Review of recent trends in optimization techniques for

0 downloads 0 Views 706KB Size Report
May 31, 2015 - Requirements for PV–wind hybrid system optimization . ..... HYBRID2 (The Hybrid Power System Simulation Model), HOGA. (Hybrid Optimization using ... The General Algebraic Modeling System (GAMS),. Optimization of ...
Renewable and Sustainable Energy Reviews 50 (2015) 755–769

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews journal homepage: www.elsevier.com/locate/rser

Review of recent trends in optimization techniques for solar photovoltaic–wind based hybrid energy systems Sunanda Sinha, S.S. Chandel n Centre for Energy & Environmental Engineering, National Institute of Technology, Hamirpur 177005, Himachal Pradesh, India

art ic l e i nf o

a b s t r a c t

Article history: Received 14 January 2015 Received in revised form 11 April 2015 Accepted 12 May 2015 Available online 31 May 2015

An update literature review on trends in optimization techniques used for the design and development of solar photovoltaic–wind based hybrid energy systems is presented. The main objective is to identify latest promising techniques for the optimization of solar photovoltaic (PV)–wind based hybrid systems. Different techniques used by researchers for the optimization of renewable based hybrid energy systems are reviewed along with PV–wind based hybrid system sizing methodology, is presented. Optimization studies during last 2.5 decades by researchers using traditional and new generation methods are analyzed and sixteen optimization methods including hybrid algorithms are presented. The trend shows that new generation artificial intelligence algorithms are mostly used during last decade as these require less computation time and have better accuracy, good convergence in comparison to traditional methods. The study suggests using hybridization of two or more algorithms to overcome the limitations of a single algorithm. Additionally some other techniques are identified for follow up research in the design of PV–wind hybrid systems. This review will be useful for researchers to face complexity and challenges in renewable energy based hybrid system research. & 2015 Elsevier Ltd. All rights reserved.

Keywords: Hybrid energy systems Solar–wind based hybrid systems Solar photovoltaics Wind turbine Optimization techniques Hybrid algorithms

Contents 1. 2.

3.

n

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756 PV–wind hybrid system sizing methodologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 2.1. Requirements for PV–wind hybrid system optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 2.1.1. Meteorological data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 2.1.2. Load profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 2.1.3. System configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 2.1.4. Energy system model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 2.1.5. Optimization results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760 2.2. Criteria for PV–wind hybrid system optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760 2.2.1. Reliability analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760 2.2.2. Cost analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760 2.3. Modeling of hybrid system components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760 2.3.1. Modeling of photovoltaic system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 2.3.2. Modeling of wind generator system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 2.3.3. Modeling of battery system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 2.4. Conditions and steps for the execution of a hybrid optimization problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 Optimization techniques used in PV–wind based hybrid research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 3.1. Traditional approach for optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762 3.1.1. Graphical construction technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762 3.1.2. Iterative techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762 3.1.3. Probabilistic approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762

Corresponding author. Tel.: þ 91 1972 254748; fax: þ91 1972 223834. E-mail addresses: [email protected], [email protected] (S.S. Chandel).

http://dx.doi.org/10.1016/j.rser.2015.05.040 1364-0321/& 2015 Elsevier Ltd. All rights reserved.

756

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

3.1.4. Trade-off approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.5. Linear programming technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. New generation approach for optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1. Genetic Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2. Particle swarm optimization (PSO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3. Simulated annealing (SA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.4. Other new generation approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.5. Hybrid algorithm optimization techniques. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Nomenclature V ci V co Ib P f aliure Pw

ηPV

Pr Vr

α

P total CB C other Cw Id

ηw

Edef icit P load

ηm ηpc

T f ailure Ps IT Rd Rr H total H LOL Aw C total T total β¼

cut-in wind speed cut out wind speed direct normal solar radiations load which cannot be served within a time period power output from wind turbine generator. PV system efficiency rated power of a wind turbine rated speed of the wind turbine power law exponent total load cost of battery bank cost of other systems and accessories cost of wind system diffuse solar radiations efficiency of wind turbine generator and corresponding converter, energy deficit within a certain time period (t) load demand during a period module efficiency power conditioning equipment efficiency power failure time period power generated from solar PV system solar radiation on a tilted surface tilt factor for the diffused solar radiation. tilt factor for the reflected solar radiation. total hours of operation total number of hours during which loss of load (LOL) occurs total swept area total system cost total working time array efficiency temperature coefficient

1. Introduction The fossil fuel resources are becoming scarce due to ever increasing energy demand in commercial, industrial, agricultural and domestic sector. In this context, alternative energy resources like solar, wind, biomass, bio-fuel, hydro and geothermal etc. are being utilized largely to generate power in recent years. A renewable energy based hybrid system offers a better option than a single source based system in terms of cost, reliability and efficiency. One or more energy sources can be utilized in renewable energy based hybrid systems (REHS) which can work as stand alone or in a grid connected mode. Different types of hybrid system combinations are feasible depending on the need and

762 762 763 763 764 764 764 765 766 767 767

CRF ¼ Tc ¼ ηc ¼

capital recovery factor cell temperature charge efficiency (depends on the SOC and the charging current and has a value between 0.65 and 0.85) Cs ¼ cost of solar PV system ηdis ¼ discharge efficiency (generally taken equal to 1) σ ðtÞ ¼ hourly self-discharge rate depending on the battery state (taken constant at about 0.02%) i¼ interest rate C bat ¼ nominal capacity of the battery (A h). Pw ¼ power generated from wind turbine T proj ¼ project life time APV ¼ PV system area Tr ¼ reference temperature for cell efficiency C ann ¼ total annualized cost P load ¼ total load to be delivered or load demand P total ¼ total power generated from resources COE cost of energy DOD depth of discharge, LA level of autonomy LCC life cycle cost LLP/LOLP loss of load probability LOLH loss of load hours LOLR loss of load risk LPSP loss of power supply probability NPC net present cost p cumulative probability of meteorological status which corresponds to electrical energy generation q probability of failure SOC state of charge SPL system performance level UL unmet load V and V0 wind speeds at heights h and h0 (h0 is the reference height).

resource availability at a particular location, but in the present study we have focused only on PV–wind based hybrid systems as solar and wind are most promising power generating sources due to their complementary nature advantage. Wind speeds are often low in periods when the solar resource is at its best. On the other hand, the wind is often stronger in seasons when there are less solar resource. But these sources depend on climatic conditions which are unpredictable thus making the design of a hybrid system complex. In order to improve the reliability of a PV wind hybrid system other sources like fuel cell, diesel generator can also be integrated. Such a hybrid system can meet the basic power requirements in a non-electrified remote area where grid power is not available. Besides this our main research focus in the present

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

study is on PV wind based hybrid system applications in urban, rural and remote regions. However, the techniques covered are also useful for studying other types of hybrid systems. A solar–wind hybrid system consists of PV array, wind turbine, battery bank, inverter, controller, and other accessories. The schematic is shown (Fig. 1).The generated power from solar and wind energy charge the battery after meeting load demand but if generated power is less than demand then battery will supply the load as per storage capacity. The steps to be followed for the efficient design and planning of a PV–wind based hybrid system, are shown in Fig. 2. There are some constraints in the formulation and solution of the designing and optimization approach like resource availability, technology, efficiency, mathematical models and many more. But advancement in computational techniques has made it easy to deal with optimization problems by using a number of optimization and simulation techniques. A number of simulation tools HOMER (Hybrid Optimization Model of Electric Renewable), HYBRID2 (The Hybrid Power System Simulation Model), HOGA (Hybrid Optimization using Genetic Algorithm) etc. are used for optimizing, designing and performance evaluation of PV–wind

757

hybrid systems as discussed in a comprehensive review by Sinha and Chandel [1]. However, these softwares have some limitations like black box coding, different working platforms, unavailability of some of the softwares and are also not as flexible as optimization techniques which can be used as per research criteria. Sizing and optimization techniques must efficiently search for an optimum combination of parameters like system cost, system reliability, PV system size, tilt angle of PV panels, battery size, wind turbine size with hub height because, over sizing causes higher system costs and under sizing causes insufficient power supply. A number of authors have carried out detailed reviews on sizing and optimization techniques [2–10]. Summary highlights of these studies are given in Table 1. The paper is organized as follows: Section 2 provides an overview of PV–wind hybrid system sizing methodologies; Section 3 gives a literature survey on optimization techniques used in PV–wind based hybrid system research; Section 4 deals with discussion on the recent trends of optimization followed by Conclusion in Section 5.

2. PV–wind hybrid system sizing methodologies In this section sizing methodologies, requirements, criteria, mathematical modeling, conditions and execution process for a PV–wind based hybrid system are discussed. 2.1. Requirements for PV–wind hybrid system optimization Fig. 1. Schematic of a solar–wind hybrid system.

The input parameters required for a PV–wind hybrid system optimization are as follows. 2.1.1. Meteorological data Analysis of meteorological characteristics of the location has to be made for optimization process. Measured solar and wind resource data are the main inputs for PV wind hybrid based system optimization. The time series measured minute wise, hourly or daily weather data are preferable. In case measured data are not available for the location then satellite based data or estimated data can also be used for preliminary studies if accuracy is not the major consideration. 2.1.2. Load profile Yearly electric load demand profile is one of the necessary steps to design-planning and optimization of a hybrid system. It is difficult to find out and also complicated to analyze real load demand with all minute fluctuations, therefore hourly or daily averages of load demand is generally used for design-optimization purpose. But to have a real or nearest to real load variations for all the seasons is really difficult task, if it fails then system designed can be oversized or undersized. 2.1.3. System configuration After the prefeasibility studies based on weather data (e.g. wind speed, solar irradiation and temperature) and load demand the selection of proper sizing of equipment can be made. But this sizing process must be according to nature of PV and wind, e.g. if the study location have good solar potential than wind then the hybrid system must be configured with maximum share of PV system and minimum share of wind system.

Fig. 2. Basic steps for renewable energy based hybrid system design and planning.

2.1.4. Energy system model Energy system models are the mathematical models developed to represent various energy-related problems reliably. These models are used to identify and solve problems using various computing systems. The accuracy of the developed models for PV/wind system

758

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

Table 1 Summary highlights of review studies on sizing and optimization of renewable energy based hybrid systems. Reference/ year of study

Systems studied

Zhou et al. [2]

Only standalone PV–wind hybrid systems

Topics covered

 Criteria for optimizations and simulation 

modeling of photovoltaic system, wind energy system, battery storage system Software tools for hybrid solar–wind system reviewed are ○ HOMER ○ HYBRID2 ○ HOGA

Highlights

 Artificial intelligence techniques are identified to be promising which requires to be further explored

 Sizing methods reviewed are ○ ○ ○ ○ ○

Erdinc and Uzunoglu [3]

Covers all types of renewable energy based hybrid systems.

Graphic construction method Probabilistic approach Iterative technique Artificial Intelligence method Multi Objective optimization

 Software tools discussed are: HOMER, HYBRID2, Promising techniques identified are: The General Algebraic Modeling System (GAMS),  Ant colony algorithm Optimization of Renewable Intermittent Energies  Artificial immune system algorithm with Hydrogen for Autonomous Electrification  Tabu Search (ORIENTE), OptQuest, LINDO, WDILOG2, Dividing  Honey Bee Mating Algorithm Rectangles (DIRECT), Determining Optimum  Bacterial Foraging Algorithm Integration of RES (DOIRES),Simulation of  Game Theory



Photovoltaic Energy Systems (SimPhoSys), GeoSpatial Planner for Energy Investment Strategies, Hybrid methodologies require to be improved Grid-connected Renewable Hybrid Systems further Optimization (GRHYSO), H2RES Optimization techniques discussed are: ○ Genetic algorithm ○ Particle swarm optimization ○ Simulated annealing ○ Linear programming ○ Simplex algorithm ○ Neural Networks ○ Evolutionary algorithm ○ Stochastic, iterative, probabilistic, parametric and numerical approaches

Fadaee and Radzi[4]

Study covers PV–battery, PV–wind–battery and PV–wind–diesel–battery hybrid systems

 Optimization Techniques reviewed:

Luna-Rubio et al. [5]

Study includes all types of renewable energy based hybrid systems

 Reviewed hybrid system performance indicators

Khatib et al. Study includes standalone and grid [6] connected PV systems, PV–diesel generator systems, PV–wind systems, PV–wind–diesel generator systems

○ ○ ○ ○ ○ ○

Genetic Algorithm Honey Bee mating Optimization Particle Swarm Optimization Evolutionary Algorithm Artificial Intelligence Pareto-based multi-objective optimization and parallel processing

(Loss of power supply probability (LPSP), Levelized cost of energy (LCE)) ○ Hybrid energy system sizing methodologies reviewed are: ○ Probabilistic methods, ○ Iterative methods, ○ Hybrid methods (Genetic Algorithm, Artificial Intelligence) ○ Analytical methods including commercial software tools and/or numerical approximations of system component)

 Standalone PV systems size optimization methods reviewed are: ○ Intuitive methods ○ Numerical methods ○ Analytical methods ○ Other methods (Artificial Intelligence)

 GA and PSO as most useful and promising multi objective optimization methods in hybrid system design

 Study suggests hybrid optimization methodologies are superior to other methods.

 Artificial intelligence techniques have the potential to improve the process of optimization.

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

759

Table 1 (continued ) Reference/ year of study

Systems studied

Topics covered

Highlights

 Reviewed grid connected hybrid systems sizing optimization methods: ○ Intuitive methods ○ Numerical methods ○ Artificial intelligence methods

Upadhyay and Sharma [7]

Covers all types of renewable energy based hybrid systems

 Discussed design parameters, evaluation criteria 

 Hybrid optimization methodologies are

and control and energy management of hybrid energy systems Software tools reviewed are: ○ HOMER ○ HOGA ○ RETScreen ○ HYBRIDS ○ TRNSYS

recommended for hybrid system research to avoid the limitation of one methodology

 Sizing methodologies reviewed are: ○ ○ ○ ○ ○ ○

Chauhan and Saini [8]

Covers all types of renewable energy based hybrid systems

Graphic construction methods Probabilistic methods Analytical methods Iterative methods Artificial intelligence methods Hybrid methods

 Reviewed various types of integrated

 

configurations (DC coupled, AC coupled, Hybrid DC–AC coupled) and various options for energy storage technologies with system control and management Mathematical model for wind, Micro-hydro, solar, biomass gasifier energy systems are studied Reviewed various sizing methodologies used in hybrid system study ○ Artificial intelligence ○ Multi objective design ○ Iterative approach ○ Analytical method ○ Probabilistic approach ○ Graphical construction method

 Concludes artificial intelligence techniques as a 

better technique than other deterministic methods Genetic algorithm (GA), harmony search (HS), particle swarm optimization (PSO), biogeography based optimization (BBO) are the most promising Algorithms in future research

 Software tools reviewed are: ○ ○ ○ ○ ○

Bourennani et al. [9]

Focused on solar–wind–fuel cell based hybrid systems

Bhandari et al. [104]

Study covers PV–wind–diesel–battery hybrid systems

HOMER HOGA RET Screen HYBRIDS TRNSYS

Multi-Objective Optimizations (MOO) for Hybrid Energy System Design

 Reviewed energy scenario and various types of optimal design criteria

 Mathematical model for wind, solar, battery, diesel generator systems are studied

 Reviewed various sizing methodologies used in hybrid system study ○ Graphical construction method ○ Probabilistic approach ○ Iterative approach ○ Artificial intelligence

 Software based approach: HOMER

Article concludes that hybrid energy system requires more interaction between both energy and MOO research Study concludes that artificial intelligence may provide good optimization of system without extensive long term weather data

760

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

components is important in optimizing. Therefore these models must include all necessary variables which affect energy conversion. These mathematical models should use simple concepts but in many cases the complexity of a model increases or researcher may not include some important factors which may lead to inaccuracy.

2.1.5. Optimization results Optimization results must be accurate enough to avoid excess or deficit power and it is only possible if above four steps are followed. Although power generation from PV–wind system is location dependant process, yet generalization of optimization results for nearby location is also important.

2.2. Criteria for PV–wind hybrid system optimization A PV–wind hybrid system has to be optimally designed to deliver power for a particular load demand reliably. The optimum hybrid system configuration must satisfy and compromise between two main objectives: power reliability and system cost.

2.2.1. Reliability analysis The dependency on nature and unpredictability of solar and wind resources have a great impact on energy production which leads to unreliable power supply during cloudy or non windy days. A system is reliable if it is able to supply required power to the electrical load within a specific time period. A power reliability analysis is essential for system design and optimization process. There are several methods in literature to calculate the reliability of a hybrid energy system which are summarized in Table 2. 2.2.2. Cost analysis Cost analysis of a hybrid system is important for optimization so as to deliver energy at minimum cost. There are several methods used to calculate cost of a hybrid energy system like net present cost, life cycle cost and cost of energy which are summarized in Table 3. 2.3. Modeling of hybrid system components The performance of a hybrid solar–wind system is dependent on its components. The mathematical modeling equations of a PV

Table 2 Methods to determine reliability of hybrid energy systems. Sl. no.

Reliability criteria/method Definition and representation

1

Loss of power supply probability (LPSP)

Reference

Most widely used method in which probability of insufficient power supply to load demand is taken into account while [11] designing the hybrid system. LPSP is the ratio of power supply deficits to the electric load demand during a certain period n P Edef icit

¼ 1 LPSP ¼ t P n

P load

t ¼ 1

2 3

Loss of load probability (LLP/LOLP) Unmet load (UL)

Defined as the power failure time period divided by the total working time of the hybrid system

[11]

The load which cannot be served divided by the total load of a time period (normally one year) n P

[11]

P f ailure

¼ 1 UL ¼ t P n

P total

t ¼ 1

4 5

System performance level (SPL) Loss of load hours (LOLH)

6

Loss of load risk (LOLR)

7

Level of autonomy (LA)

SPL is defined as the probability of unsatisfied load

[12]

LOLH is the summation of loss of load expectation in hours over a specified time (usually one year) that the power system [13] is unable to meet load requirements due to lack of power at an instant excluding the effects of component breakdown or maintenance time LOLR is defined as the probability of the generating system failure to meet the daily electrical energy demand due to [14] deficient energy of the renewable energy sources used LOLR can be can be represent as LOLR ¼ 1  p Or LOLR ¼ q LA deals with two main parameters namely the total number of hours in which loss of load (LOL) occurs and the total [5] hours of operation. If autonomy increases system will be more reliable but same time cost will be higher LOL LA ¼ 1  HHtotal

Table 3 Cost analysis methods of hybrid energy system. Sl. no.

Cost analysis method

Definition and representation

Reference

1

Net present cost

NPC reflects cost of energy for a particular system and is defined as the ratio of the total annualized cost of the system to the annual [2] electricity delivered by the system. NPC is total present value of cash flows including initial cost of system components, replacement cost of any component within project lifetime, and cost of maintenance, i.e. investment costs plus the discounted present values of all future costs during system’s lifetime

2

Life cycle cost

3

Cost of energy

LCC deals with sum of all recurring and one-time (non-recurring) costs over the full life span or a specified period of a good, service, [2] structure, or system Includes purchase price, installation cost, operating costs, maintenance and upgrade costs, and remaining (residual or salvage) value at the end of ownership or its useful life [2] COE reflects the cost of energy or electricity generation and as the ratio of total annualized cost of the system to the annual electricity delivered by the system. Total annualized cost includes all the costs over the system’s lifetime from initial investment and capital costs, to operations and maintenance (e.g. fuel) and financing costs COE¼ total cost/energy produced

C ann NPC ¼ CRFði;T proj Þ

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

system, wind system and battery system are described in this section. 2.3.1. Modeling of photovoltaic system The power output from PV system with area Apv (m2) is given by: P ¼ I T APV ηPV

ð1Þ

where IT is expressed as I T ¼ I b Rb þ I d Rd þ ðI b þI d ÞR

ð2Þ

The PV system efficiency is expressed as

ηPV ¼ ηm ηpc ½1  βðT c  T r Þ

ð3Þ

2.3.2. Modeling of wind generator system Power output of wind turbines for a location depends on wind speed at hub height which can be calculated using power-law equation given below  α V h ¼ ð4Þ V0 h0 Total power available from wind turbine is given by: P ¼ P w Aw ηw

ð5Þ

Different wind turbines have different power output and performance curves. Therefore, the modeling equation of a wind system is strongly influenced by the power curve of the wind turbine used. Fig. 3 shows a typical wind turbine power curve characteristics and using this curve power output Pw (kW/m2) from wind generator can be calculated as follows 8 if V o V ci and V 4V co > < 0; Pw ¼

aV 3  bP r ; > :P ; r

if V ci oV o V r

ð6Þ

if V r oV o V co

where a and b are the co-efficients are given as a¼



Pr V 3r  V ci



ð7Þ

During charging process SOCðt þ 1Þ ¼ SOCðtÞ:½1  σ ðtÞ þ½I bat ðtÞ:Δt:ηc ðtÞ=C bat 

ð8Þ

2.3.3. Modeling of battery system Battery is used to store surplus generated energy, to regulate system voltage and to supply load in case of insufficient power generation occurs from the hybrid system. Battery sizing depends on maximum depth of discharge (DOD), temperature and battery life. A battery’s state of charge (SOC) is expressed as follows:

Fig. 3. Wind turbine power curve.

ð9Þ

During discharging process SOCðt þ 1Þ ¼ SOCðtÞ:½1  σ ðtÞ ½I bat ðtÞ:Δt=ηdis ðtÞ=C bat

ð10Þ

With ð1  DODÞ rSOCðtÞ r1 2.4. Conditions and steps for the execution of a hybrid optimization problem The two main conditions to be fulfilled for a good designed hybrid system are maximum power reliability and minimum cost. Total energy generated by wind and PV systems can be expressed as P total ¼ P w þ P s

ð11Þ

The total energy generated may or may not satisfy the load demand and three different situations may arise depending upon the load demand and total power generated which are shown as: i. P total ¼ P load Total energy generated matches the load demand, so there is no excess and deficiency in power. ii. P total o P load Total energy generated is less than the load demand, so there will be a deficiency in power and battery will have to supply the load. iii. P total 4 P load Total energy generated is greater than the load demand, so excess power is generated and battery will be charged. The optimum combination of a solar–wind hybrid system can make the best compromise with system cost. The economical approach is composed of minimization of the costs of individual components like PV, battery, wind system etc. This cost includes capital, maintenance and replacement cost of various components of the system. The total cost can be expressed as C total ¼ fC s þ C w þ C B þ C other gmin

V 3ci   3 V r  V 3ci

761

ð12Þ

The optimization process is shown in flow chart (Fig. 4).The first step of the optimal sizing methodology consists of providing meteorological data–electric load demand and system input data. The second step consists of mathematical equations with reliability and cost conditions. The next step is using optimization methods fulfilling the criteria of system configuration.

3. Optimization techniques used in PV–wind based hybrid research The power supplied by solar and wind resources mainly depends on meteorological conditions so it is necessary to maximize the power during sunny or windy days with properly optimally sized design. There is a continuously increasing interest in the development of hybrid energy systems using various optimal sizing techniques. The classification of optimization algorithm can be carried out in a number of ways but in this paper the algorithms are divided into two simplest categories namely traditional approaches and new generation approaches. A traditional approach follows a rigorous procedure for example linear programming, iterative techniques etc. whereas new generation approaches are genetic algorithm, particle swarm optimization etc. In this section an overview of various optimization techniques used in PV–wind hybrid systems are described.

762

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

(LEC) and life cycle unit cost (LUC) of power generation with battery bank. Borowy and Salameh [22] determined optimum size of a battery bank and PV array for a stand-alone hybrid wind–PV system and concluded that the optimum mix depends on the particular site, load profile, and the desired reliability of the hybrid system. Ashok [23] proposed an iterative method where an optimal hybrid system was obtained among different renewable energy combinations including PV and wind technology ensuring system’s reliability and minimum life cycle cost which is applicable to renewable power generation in any rural village. Iterative optimization method requires more computational efforts and usually two main parameters PV module tilt angle and wind turbine tower height are not optimized in most of the reported studies.

Fig. 4. General execution process followed in optimal sizing of hybrid systems.

3.1. Traditional approach for optimization A number of studies have been carried out using like graphical construction method, iterative techniques etc which are discussed below. 3.1.1. Graphical construction technique Borowy and Salameh [15] presented a graphical construction technique for finding optimum combination of a stand-alone PV– wind hybrid system based on long-term meteorological data. For a typical load consumption of a house in Massachusetts and a desired LPSP, the optimum configuration of battery bank and PV array was calculated based on the minimum cost of the system. Markvart [16] used graphical technique to optimally design a solar–wind hybrid power generation system using monthly-average solar and wind energy values. This is a basic and easily understandable method with no complexity, but this technique is not used currently as it is not flexible and is based on various approximations. 3.1.2. Iterative techniques An iterative method is a mathematical procedure that generates approximate solutions for problems. This is a recursive process which stops when the best configuration is reached as per specifications. In hybrid renewable energy research, iterative approach is used from beginning to design and optimize. Yang et al. [17] proposed a Hybrid Solar–Wind System Optimization (HSWSO) model, using iterative optimization technique. For power reliability and system cost, authors considered LPSP and Levelized Cost of Energy model with parameters like capacity of PV system, rated power of wind system, and capacity of the battery bank. An iterative optimization method was used by Kellogg et al. [18,19] to optimize the wind turbine size and number of PV modules. Diaf et al. [20] used a methodology to perform optimal sizing of an autonomous hybrid PV–wind system among a set of systems components, which meets the desired system reliability requirements and lowest levelized cost of energy. Prasad and Natarajan [21] used an iterative technique to determine the optimum size of solar panels, wind turbines and capacity of batteries of a hybrid system based on deficiency of power supply probability (DPSP), relative excess power generated (REPG), unutilized energy probability (UEP), life cycle cost (LEC), levelized energy cost

3.1.3. Probabilistic approach A probabilistic model is that where there are multiple possible outcomes, with varying degrees of certainty or uncertainty of occurrence. Probabilistic approach considers the effect of random variability of parameters. Probabilistic approaches enable variation and uncertainty to be quantified, mainly by using distributions instead of fixed values. This technique is also used by some researchers to solve hybrid system sizing problem. However, the probabilistic approach cannot represent the dynamic changing performance of the hybrid system which is main disadvantage of this method. Bagul et al. [24] used probabilistic approach to determine the relation between number of photovoltaic arrays and batteries to meet a given reliability if wind turbine rated capacity and load demand are known. Karaki et al. [25] used a probabilistic treatment of an autonomous solar–wind energy conversion system delivering a load. The methodology used fix an upper limit on required storage batteries and predicts the expected energy not supplied (EENS) to the hybrid system considering charging/discharging cycles of the batteries. Combination of the two separate PV and wind models were obtained and convolution theory was used. Tina et al. [26] used probabilistic approach for the long-term performance assessment of a solar–wind hybrid system for standalone and grid-linked applications and developed an analytical model. The authors modified the model developed by Karaki et al. [25] using energy index of reliability (EIR). The validity of the developed probabilistic model is shown and results are compared with those obtained from time-series simulation. 3.1.4. Trade-off approach This approach is not widely used in hybrid system sizing and not much literature is available on this method. Chedid et al. [27] used trade-off/risk method and presented a decision support technique to study the design of a hybrid solar–wind power system for grid-connected applications based on simultaneous maximization of reliability and minimization of cost. Gavanidou and Bakirtzis [28] applied the method in the design of a standalone system. The outcome of the method is not a unique “optimal” design, but a small set of robust designs and the final decision is left to the decision makers. The main disadvantage of this approach is that emission control, which has a major influence on the final trade-off curve, is not taken into account. 3.1.5. Linear programming technique The linear programming method was first developed by Leonid Kantorovich in 1939 and is a widely used technique for sizing and optimization of renewable systems. Chedid and Rehman [29] proposed an optimal design of wind–solar hybrid system using this technique to minimize the cost of electricity while meeting the load requirements in a reliable manner along with considering

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

environmental factors. Huneke et al. [30] used linear programming to obtain optimal configuration for a solar–wind–battery–diesel based power generator combination for two real off-grid energy systems in India and Colombia. The optimization results for both studies show the feasible combination of PV–battery and diesel generator. Nogueira et al. [31] used the methodology for sizing and simulation of a PV–wind–battery hybrid energy system and linear programming with minimum cost and high reliability. Lee et al. [32] formulated a Linear programming based new optimization model for hybrid power systems considering various power losses and studied three case studies. The main focus was not on minimizing the total cost of the system but on minimizing the outsourced electricity supply and electricity storage capacity. Saif et al. [33] formulated a problem of a PV–wind–diesel–battery hybrid power system as a Linear Programming model with two objectives: minimizing total cost and minimizing total CO2 emissions, while capping the Expected Unserved Energy (EUE). Nagabhushana et al. [34] used LP to calculate the sizes of the components of a PV–wind hybrid renewable energy system for three locations in Karnataka. Among all the techniques discussed above, linear programming technique is found to be best than other approaches as it improves the quality of decision. Also linear programming technique is more flexible than other methods and a wide range of problems can be solved easily. Comparison of various traditional hybrid system optimization techniques are summarized in Table 4. 3.2. New generation approach for optimization This approach is widely used now-a-days in renewable energy research to solve optimization and design problems. These are not restricted to local optimum configuration but also to determine global optimum system configuration with relative computational simplicity as compared to traditional optimization methods. 3.2.1. Genetic Algorithm Genetic Algorithm (GA) is a search process that mimics the process of natural selection and was developed by John Holland in 1960–1970 period [35,36].GA generates solutions to optimization problems using techniques inspired by natural evolution such as inheritance, mutation, selection, and crossover. GA has several advantages: it can solve problems with multiple solutions, easy to understand and can easily be transferred to existing simulations and models etc. It has some limitations like a tendency to converge towards local optima or even arbitrary points rather than the

763

global optimum of the problem, cannot assure constant optimization response times etc. A number of researches have used the application of GA for the optimal design and operation of PV–wind based hybrid energy systems. An optimal sizing of standalone PV–wind systems is proposed by Koutroulis et al. [37] using GA to select the optimal number of units with minimum cost, subject to load demand fulfillment. In another study Koutroulis et al. [38] presented a GA based optimal sizing of desalination systems by PV–wind generators as a powersupplied unit. Yang et al. [39–40] used GA to optimize the configurations of hybrid solar–wind–battery bank system where the decision variables are the number of PV modules, wind turbines and batteries, the PV slope angle and wind turbine tower height. This method was proposed for a hybrid system which supplies power for a telecommunication relay station. Bilal et al. [41] proposed an optimized sizing of a hybrid solar–wind–battery system through multi-objective genetic algorithm satisfying two principle aims of annualized cost minimization and minimization of the loss of power supply probability (LPSP). Nafeh [42] used GA to yield optimum PV wind, and battery ratings with minimum cost and power reliability. Abdullrahman and Addoweesh [43] proposed a methodology of optimum sizing of a PV–wind–diesel– battery hybrid system using GA including the effect of wind turbine parameters such as rated speed and rated power. The decision variables included in the optimization process are the PV array capacity, wind turbine number, battery bank number and diesel generator rated power. Atia and Yamada [44] adopted a two layer (main and secondary) genetic algorithm coupled with a local optimizer known as a Hybrid Genetic Algorithm (HGA) and used in designing and controlling of PV–wind–diesel systems. Authors also found that the HGA is more powerful algorithm than the conventional GA. Merei et al. [45] used GA to optimize PV–wind– diesel–battery hybrid system with three different battery technologies. A controlled elitist genetic algorithm has been applied by Abbes et al. [46] to perform a multi-objective design of PV–wind– battery hybrid system in order to find the best compromise between three objectives: life cycle cost (LCC), system embodied energy (EE) and loss of power supply probability (LPSP). Shi et al. [47] used multi-objective genetic algorithm to study technoeconomical performance of the PV–wind hybrid energy system and optimized three objectives e.g. total system cost, autonomy level, and wasted energy rate with the PV array peak power, the wind generator rated power and the rated capacitor of the battery as decisive variables. Mostofi and Shayeghi [48] used GA to solve the optimization problem of a PV–wind–hydro–fuel cell hybrid

Table 4 Comparison of various traditional approaches used for PV–wind based hybrid system optimization. Technique

Highlights

Strength

Weakness

Graphical construction technique Iterative approach

Presents graphical solution of optimization problem

Easy to understand and use

It is a recursive process which stops when the best configuration is reached as per design specifications Based on the effects of random variability of upon the performance of an system Based on a situation that involves losing one quality or aspect of something in return for gaining another quality or aspect Based on a mathematical model represented by linear relationships

Easy to understand; Tracks defects at early stages

[15–16] Some important factors (like tilt angle of PV, wind turbine hub height) are completely neglected in this technique Each phase of iteration is rigid with no overlaps [17–23]

Probabilistic approach Trade-off method

Linear programming

Reference

Easy to understand and use

It cannot represent the dynamic changing performance of the hybrid system

[24–26]

Easy to understand

Not much used in renewable energy applications

[27–28]

Best suitable for solving complex problems; simple to use; more flexible than any other methods; a wide range of problems can be solved easily

Linearity in relation of variables; assumptions of linear programming are also unrealistic: there is a change in relation between input, output gain, loss etc.

[29–34]

764

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

system and compared results with HOMER software and concluded that GA has better accuracy than HOMER. Shadmand and Balog [49] presented Multi-Objective Genetic Algorithm (MOGA) to determine the PV–wind hybrid system design, optimized by considering multiple criteria including size, cost, and availability. Tégani et al. [50] used GA for optimal sizing of PV–wind hybrid system with a lifespan of 20 years. Multi-Objectives Genetic Algorithm approach is also used by Bilal et al. [51] to design and optimize a stand-alone hybrid PV/wind/diesel/battery system minimizing the LCE and the CO2 emission for Senegal.

3.2.2. Particle swarm optimization (PSO) Particle swarm optimization was developed by Kennedy and Eberhart [52,53] based on the research of bird and fish movement behavior. The advantages of PSO are: speed of the researching is very fast; calculation in PSO is simple as compared to other methods and can be completed easily. Limitations of this optimization algorithm are that it cannot work out the problems of noncoordinate system, easily suffers from the partial optimism etc. The use of the method in the PV–wind hybrid system is just beginning and few literatures are reported till now. Basir and Sadeh [54] have taken a combination of wind, photovoltaic and tidal energy with battery source and used PSO to determine the capacity of hybrid system. The Equivalent Loss Factor (ELF) has been used as an index to evaluate the system reliability level. The paper also compared hybrid wind, solar and battery combination with wind, solar, tidal and battery combination and found the second combination as more economical. Lee and Chen [55] used an evolutionary PSO algorithm to solve wind– photovoltaic capacity coordination with the aim of maximizing the benefit–cost ratio. Kaviani et al. [56] optimized a hybrid wind– photovoltaic–fuel cell generation system over its 20 years of operation with PSO. The aim was to minimize the annual cost of the hybrid system subject to reliable supply to meet load demand. Bansal et al. [57] used Meta Particle Swarm Optimization (MPSO) to solve the PV–wind–battery hybrid system optimization problem. Using this improved PSO technique local minimum trap can be avoided thus proving it as an effective technique. Sharafi and ELMekkawy [58] studied PSO simulation based approach to tackle the multi-objective optimization problem for a hybrid system consisting of wind turbine, photovoltaic panels, diesel generator, batteries, fuel cell, electrolyzer and hydrogen tank. Pirhaghshenasvali and Asaei [59] proposed a standalone PV–wind–diesel– battery based hybrid system for Kerman, Iran and used PSO to obtain the optimal sizes for wind turbine, PV system, battery banks and the diesel generator. Borhanazad et al. [60] used MultiObjective Particle Swarm Optimization (MOPSO) in order to obtain the best configuration of the PV–wind–diesel–battery based hybrid system for three stations in Iran namely Nahavand, Rafsanjan, and Khash. Maleki and Askarzadeh [61] used four heuristic algorithms namely, particle swarm optimization (PSO), tabu search (TS), simulated annealing (SA), and harmony search (HS) for optimum sizing of a cost-effective PV–wind–fuel cell and PV– wind–battery based hybrid systems. The results show that PSO is more robust and promising than other three algorithms used by authors. A study to determine the optimum dimensions of hybrid photovoltaic systems, wind power, and storage battery bank has been carried out by Maleki et al. [62] in the remote regions of the South, North-West and North-East of Iran. Authors studied the performance of five different PSO variants and three more algorithms namely tabu search, simulated annealing and harmony search (HS). PSO, modified PSO (MPSO), PSO based on repulsion factor (PSO-RF), PSO with constriction factor (PSO-CF), and PSO with adaptive inertia weight (PSO-W) these are the variants used in the study. It has been concluded by the authors that PSO-CF is

more favorable than the other PSO variants and PV–battery based hybrid systems are suitable for most areas of the country due to the good solar radiation availability and low windy nature. 3.2.3. Simulated annealing (SA) Simulated annealing, which mimics material annealing processing, was developed by Kirkpatrick, Gelatt and Vecchi in 1983 [63]. It is a trajectory based random search technique for global optimization. The main advantage of simulated annealing is its ability to avoid being trapped in local minima. Simulated annealing is a robust and versatile technique which can deal with highly nonlinear models, chaotic and noisy data and many constraints. The main weakness of SA is that the quality of the outcome may be poor. Till now little literature has been reported using SA in this field. Ekren and Ekren [64] used simulated annealing (SA) algorithm for optimizing size of a PV–wind–battery hybrid energy system to minimize total cost. The decision variables are PV size, wind turbine rotor swept area and the battery capacity used by the authors and found that SA algorithm gives better result than the Response Surface Methodology (RSM). 3.2.4. Other new generation approaches Several artificial intelligence based/metaheuristic/heuristic/ nature inspired clever algorithms are formulated and used in recent years. Some of the techniques which are used in PV–wind hybrid system studies are discussed in this sub section. Ant colony algorithms [65–66] were initially proposed by Marco Dorigo in 1992 in his PhD thesis. The algorithm was aiming to search for an optimal path in a graph, based on the behavior of ants seeking a path between their colony and a source of food. A small renewable hydroelectric, wind and solar resources based on hybrid hydrogen storage in the northwestern region of Iran (Ardebil Province) is studied by Menshsari et al. [67]. In this paper ant colony algorithm is used by authors for improving the technical and economic performance of the hybrid system. Xu et al. [68] proposed a specific graph-based ant system to minimize the total capital cost, subject to the constraint of the LPSP for sizing of standalone hybrid wind/PV power systems. Bacterial Foraging Algorithm (BFO) [69] is inspired by the group foraging behavior of bacteria such as Escherichia coli and Myxococcus xanthus. It is the chemo taxis behavior of bacteria that will perceive chemical gradients in the environment and move toward or away from specific signals. Bazyar [70] used Bacterial Foraging Algorithm (BFA) for an optimal design of integrated wind–PV– diesel–battery system for supply of power demand in remote and rural areas of Ardebil, Iran. Result shows that hybrid wind–PV– diesel–battery is suitable from economical point of view. Artificial bee colony algorithm (ABC) is an optimization algorithm based on the intelligent foraging behavior of honey bee swarm, proposed by Karaboga and Basturk [71,72]. In ABC, the position of a food source represents a possible solution to the optimization problem and the nectar amount of a food source corresponds to the quality (fitness) of the associated solution. Nasiraghdam and Jadid [73] presented a novel multi-objective artificial bee colony algorithm to solve the distribution system reconfiguration and hybrid (photovoltaic–wind turbine–fuel cell) energy system sizing. This article also found total power loss, the total electrical energy cost, and the total emission produced by hybrid energy system and grid minimization and the voltage stability index (VSI) of distribution system maximization. To optimally size a hybrid energy system based on PV–wind–fuel cell for Rafsanjan, Iran, an efficient artificial bee swarm optimization (ABSO) algorithm is proposed by Maleki and Askarzadeh [74]. Simulation results shows that PV–wind–fuel cell is the most cost-effective hybrid system and wind– fuel cell and PV–fuel cell systems are in the other ranks. Tudu et al.

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

[75] studied the optimal sizing combinations of solar–wind–hydro– fuel cell based hybrid systems for catering the load of a remote village of Kerala, India using bee algorithm. Results show that the combination of hydro–wind–fuel cell is the most feasible hybrid energy system in relation to the net present cost for the study location. Maleki and Pourfayaz [76] evaluated the performance of different evolutionary algorithms for optimum sizing of a PV–wind–battery based hybrid system which can continuously satisfy a particular load demand with minimal total annual cost. Total seven heuristic algorithms, namely, particle swarm optimization, tabu search, simulated annealing, improved particle swarm optimization (IPSO), improved harmony search (IHS), improved harmony search-based simulated annealing (IHSBSA), and artificial bee swarm optimization (ABSO), are applied to the proposed system and the results are compared in terms of total annual cost. Authors found that ABSO shows more promising results than other six algorithm used. Bio-geography is the science of studying the behavior of species in nature against time and space and species immigration and emigration between habitats, which is a probable solution of the problem. Biogeography-based optimization (BBO) [77] optimizes a problem by maintaining a population of candidate solutions, and creating new candidate solutions by combining existing ones according to a simple formula. BBO belongs to the class of metaheuristics since it includes many variations, and since it does not make any assumptions about the problem and can therefore be applied to a wide class of problems. Kumar et al. [78] used BBO algorithm to solve the sizing problem of the hybrid PV–wind– diesel–battery system by considering economical and reliability constraints of the system to supply in the area of Jaipur, Rajasthan (India). Authors also made a comparison between HOMER software, BBO, GA, PSO, comprehensive learning particle swarm optimization (CLPSO) [79] and ensemble of mutation and crossover strategies and parameters in DE (EPSDE) algorithm [80] and shown that BBO algorithm is more rapid and give minimum cost as compared to others. Optimal design of a PV–wind–diesel– battery system is accomplished through BBO for Jaipur, India by Gupta et al [81]. For wind speed and solar radiation forecasting Back propagation trained Artificial Neural Network (BPANN) based time-series forecasting methods are implemented by authors and concluded that use of forecast data has high influence on optimal sizing algorithm performance. Geem et al. [82] proposed Harmony Search (HS) algorithm, which is inspired by the improvisation process of jazz musicians. Optimal sizing of different generation systems (PV–wind–diesel– battery, wind–diesel–battery, PV–diesel–battery, and diesel alone) are studied by Maleki and Askarzadeh [83] and found that wind– diesel–battery is the most cost-effective system for the study area Rafsanjani, Iran. Authors used discrete version of harmony search (DHS) and discrete simulated annealing (DSA) for the analysis, and found that DHS performs better than DSA. Rashedi et al. [84,85] proposed gravitational search algorithm (GSA) based on Newton’s law of gravitation and the second law of motion and superior than the traditional intelligent optimization algorithms according to optimization precision and convergence speed. Wu et al. [86] proposed an enhanced gravitational search algorithm (EGSA) to optimize the unit output and cost for largescale wind–PV–battery storage power generation in Zhangbei, China. Also to demonstrate the superiority of the proposed algorithm, a comparative analysis among EGSA, ANN and PSO were applied which shows that the unit cost of power generation obtained using EGSA is lower than the unit cost of power generation obtained using PSO and ANN. Imperialist Competitive Algorithm (ICA) is mainly inspired by imperialistic competition which is developed in 2007 by AtashpazGargari and Lucas [87]. ICA starts with an initial population. Population individuals called countries are divided into two types:

765

colonies and imperialists that all together form some empires. Imperialistic competition among these empires forms the basis of ICA. During this competition, weak empires collapse and powerful ones take possession of their colonies. Gharavi et al. [88] determines the optimal sizes of autonomous and non-autonomous PV–wind–electrolyzer–fuel cell based hybrid system with considerations for economics, reliability indices, and environmental emissions. Authors used fuzzy logic for multiobjective problem solving and ICA for optimization purposes. Ranjbar and Kouhi [89] used GA, PSO and ICA to minimize total cost of PV– wind–fuel cell based hybrid system and also studied impact of tariff change on the optimal cost of operation for Kerman province, Iran. The study deals with three different cases where thermal and electrical loads are used. Results show that hybrid systems with multiple source units perform better than single source unit and the entire algorithms have nearly same results. Tabu Search [90] originally proposed by Glover, which is an iterative procedure that starts from a random initial solution and tries to find a better solution escaping local optima. TS used in PV–wind hybrid system studies which is discussed in this paper [61–62,76,91].

3.2.5. Hybrid algorithm optimization techniques The hybrid algorithms are developed using advantages and disadvantages of two or more optimization methods. The research in this area has dramatically grown up recently. Katsigiannis et al. [91] used simulated annealing (SA) and Tabu Search (TS), for the solution of autonomous hybrid power system’s optimal sizing problem. In the study minimization of cost of energy is the main aim and the design variables are: wind turbines size, photovoltaic system size, diesel generator size, biodiesel generator size, fuel cells size, batteries size, converter size, and dispatch strategy. Authors suggested that in compared to the solutions provided by individual SA or TS methods, hybrid SA-TS improved the obtained solutions, in terms of quality and convergence. Askarzadeh [92] presented an optimum design to determine the number of PV panels, wind turbines and batteries so that the total annual cost of the hybrid system subject to some constraints is minimized. Three algorithms are merged namely: chaotic search (CS), harmony search (HS) and SA to develop a novel discrete chaotic harmony search-based simulated annealing algorithm (DCHSSA). Tutkun [93] investigated a power management for off-grid PV–wind hybrid system using GA and SVM (Support Vector Machine). In this article binary coded genetic algorithm is used to minimize the operation cost of this hybrid system and for power scheduling the SVM regression method. Khatib et al. [94] studied an optimization of hybrid PV–wind system based on loss of load probability (LLP) and system cost using hybrid iterative/ genetic algorithm. Authors used the algorithm in two parts firstly a set of possible configurations for the proposed system is determined by using the iterative part, while in second part the genetic algorithm is applied to find the optimum configuration. Dehghan et al. [95] hybridized PSO and HS algorithm to find optimal sizing of a hybrid PV–wind plant which can supply the electric load demand in a reliable manner and minimum costs. Zhou and Sun [96] proposed an improved Simulated Annealing Particle Swarm Optimization (SAPSO) algorithm to optimize a wind–solar–battery–super capacitor based hybrid system. Basic objective of this work is to minimize one-time investment and operation costs where the constraints are utilization rate and reliability of power supply. Authors also compared the proposed SAPSO algorithm with traditional PSO algorithm and the result shows that new algorithm is faster than the traditional one and effectiveness of the new hybrid algorithm is shown. Abdelhak et al. [97] proposed an optimum sizing methodology for PV–wind–battery hybrid system by using Fuzzy-Adaptive Genetic Algorithm. This Algorithm is used

766

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

Table 5 Comparison of new generation optimization techniques for PV–wind hybrid systems. Technique

Highlights

Strength

Weakness

Reference

Genetic Algorithm

Mimics process of natural evolution, like inheritance, mutation, selection, and crossover

Can solve problems with multiple solutions; easily transferable to existing simulations and models Solve problems with multiple solutions; available in MATLAB toolbox Speed of the researching is fast; calculation in PSO is simple in comparison to other methods; can be completed easily Can deal with highly nonlinear models, chaotic and noisy data and many constraints; robust and general technique; flexible with ability to approach global optimality; quite versatile as it does not rely on any restrictive properties of the model

Convergence speed is slower than other stochastic algorithms; cannot assure constant optimization response times etc.

[35–51]

Mimics bird and fish movement behavior

Cannot work out the problems of noncoordinate system; easily suffers from the partial optimism etc. There is a clear tradeoff between the quality Simulated Mimics an analogy between the way in of the solutions and the time required to annealing which a metal cools and freezes into a compute; tailoring work is required to minimum energy crystalline structure (the account for different classes of constraints annealing process) Fine-tuning of parameters of the algorithm can be rather delicate; significant effect upon the quality of the outcome Ant algorithms Inspired by the pheromone-based strategy of Algorithm has the strength in both local and Random initialization; algorithm has several parameters; parameters need to be tuned; ants foraging in nature; foraging behavior of global searches; implemented with several probabilistic approach in the local search optimization problems ants is based on finding the shortest path between source and their nests Bee-inspired Based on the intelligent foraging behavior of Algorithm has local search and global search Random initialization; algorithm has several algorithms honey bee ability; implemented with several parameters optimization problems; easy to use; available for hybridization combination with other algorithms Harmony Based on improvisation process of jazz Does not require differential gradients, thus it Complex solving process search musicians can consider discontinuous functions as well as continuous functions; can handle discrete variables as well as continuous variables; does not require initial value setting for the variables; free from divergence; ability to perform global and local search Poor in exploiting the solutions; no provision Biogeography- Biogeography is the science of studying the Fast computation time; good convergence accuracy for selecting the best members from each behavior of species in nature against time based generation; sometimes many infeasible optimization and space and species immigration and solutions are generated emigration between habitats, which is a (BBO) probable solution of the problem Based on Newton’s law of gravitation and the Good calculation accuracy; fast convergence Suffering from premature convergence Gravitational second law of motion speed problem sometime search algorithm GSA obtains the optimal solution when particles attract one another and bond together to become a large entity Complex process The ICA based on a socio-politically inspired High convergence accuracy, appropriate for Imperialist optimization of nonlinear hybrid power competition optimization strategy generation system problems with high algorithm dimensions Enhances the performance of local search; Tabu search Tabu search is a metaheuristic search fast computation method employing local search methods used for mathematical optimization Increased complexity; difficult to code Hybrid Developed by using two or more algorithms Better accuracy in results; takes less optimization computational time (in some cases); much techniques more competitive than any individual method Particle swarm optimization

to obtain the optimal number of photovoltaic panels, wind turbines and storages units ensuring the minimum cost and full availability of energy to meet load requirements. Authors concludes that fuzzy-adaptive GA is better than standard GA after comparing results of analysis using these two techniques. Mukhtaruddin et al. [98] used a hybrid Iterative-Pareto-Fuzzy (IPF) technique to obtain the best compromised solution between PV– wind–battery hybrid system that yields minimum cost and maximum reliability in a study location of Kuala Terengganu, Malaysia. Multi Criteria Decision Analysis (MCDA) optimization approach, a procedure based on both Multi Objective Genetic Algorithm (MOGA) and Multi Criteria Decision Making (MCDM) is proposed by Alsayed et al. [99] for optimal design of grid connected photovoltaic–wind hybrid system. Among all the discussed new generation algorithms in this section, GA is found to be mostly used in PV–wind hybrid system sizing

[52–62], [76] [61– 64,76]

[65–68]

[69–81],

[61–62], [76,82– 83]

[77–81],

[84–86]

[87–89]

[61– 62,76,90– 91] [91–99]

although it suffers from some shortcomings. Table 5 shows a summary of the discussed new generation optimization techniques.

4. Discussion Concise sizing methodologies for PV–wind based hybrid systems including various requirements, criteria conditions and execution process with mathematical models of PV system, wind generator and battery bank are presented in this review. Various optimization techniques have been utilized by researchers to design hybrid renewable energy systems which are also included in this study. Optimization problems of sustainable energy systems become more and more complex, especially when more number of renewable sources, are integrated together.

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

767

Table 6 Various techniques used in PV–wind based hybrid system studies during 1996– 2015. Techniques

Number of published studies Year

Graphic construction technique Iterative approach

2 7

Probabilistic method

3

Trade off approach Linear Programming Genetic algorithm

2 6 15

Particle swarm optimization Simulated annealing Bee based algorithms Ant based algorithms

11 4 4 2

Bio geography based optimization Gravitational search algorithm Tabu search Harmony search Imperialist competition algorithm Hybrid algorithms

2 1 4 4 2 9

1996 1996– 2007 1996– 2006 1992–1998 1997–2014 2006– 2015 2007–2015 2010–2015 2011–2015 2006– 2013 2013–2015 2015 2014–2015 2014–2015 2015 2009– 2015

In this study optimization methods are divided into two categories e.g. traditional and new generation methods. Traditional approaches follows a rigorous procedure and also have some drawbacks like rigid iterations, less flexibility, slow convergence speed, needs more computation time, cannot deal with dynamic changing etc. On the other hand new generation approaches are faster, flexible than traditional approaches with good convergence speed and efficient global search solutions. Both types of optimization methods are explained with their strengthsweaknesses in tabular form for better understanding. The trend shows that new generation algorithms (evolutionary-heuristic–met heuristic) are much more widely accepted and used in literature during last five years. The trend of number of papers published using various techniques during period 1996– 2015 are shown in Table 6. Table 6 shows that old techniques like graphical construction method, iterative method, probabilistic method and trade off approach were used almost a decade ago. Now-a-days these techniques are almost not used by researchers due to their drawbacks. However, in recent years, researchers are mostly using artificial intelligence (AI) based, nature inspired heuristic–met heuristic algorithms like TS, GA, SA, HS, BBO, ACO, GCA, ICA etc. for PV–wind based hybrid system optimization studies. Out of these GA is mostly used AI technique than others. For several complex problems a simple new generation algorithm might not be good enough to find the desired solution and could fail to obtain a convenient solution. This clearly shows the need for hybridization of two or more algorithms efficiently. Hybrid algorithms are found to be more reliable to use for better accuracy. The use of hybrid techniques results in improving the speed of convergence along with quality of the solutions obtained. During the period 2009–2015 various combinations of hybrid algorithms like SA-TS, GA-SVM, and SA-PSO etc. are being used in most of the studies. Although hybrid algorithms are more complex than single algorithm yet the trend in the use of hybrid algorithms in PV–wind based hybrid system research is growing as shown in Fig. 5. Further exploration of hybrid algorithms is an important future research area to face complexity and challenges of hybrid system integration and design. Some more algorithms like Cuckoo search [100], Bat algorithm [101], Firefly algorithm [102], Chaotic Ant swarm optimization [103], Cultural Algorithm [104], Memetic Algorithm [104] along

Fig. 5. Trend in the use of hybrid algorithms in PV–wind based hybrid system research 2009–2015.

with other nature inspired techniques which are used in different optimization studies, also seem promising to enrich the future research in renewable energy based hybrid energy system optimization studies [104,105].

5. Conclusion This article presents an update literature review of sixteen types of optimization techniques including hybrid algorithms used in PV–wind based hybrid energy system research and development along with PV–wind based hybrid system sizing methodology. It is found that different optimization methods used have different convergence speed, accuracy level; performance efficiency and computation speed so selection of suitable approach may change with user requirements, type of applications etc. None of the individual methods could perform better than all the other methods on all kinds of problems. However, new generation optimization approaches like artificial intelligence based, heuristic approaches are found to be more acceptable than traditional approaches because of their ability to search local and global optima, good calculation accuracy and fast convergence speed. The hybrid optimization techniques (using two or more optimization techniques together) are found best than single optimization methods. The nature inspired optimization techniques and hybrid optimization techniques will be important for further exploration in future research to face complexity and challenges of PV–wind based hybrid systems. References [1] Sinha S, Chandel SS. Review of software tools for hybrid renewable energy systems. Renewable Sustainable Energy Rev 2014;32:192–205. [2] Zhou W, Lou C, Li Z, Lu L, Yang H. Current status of research on optimum sizing of stand-alone hybrid solar–wind power generation systems. Appl Energy 2010;87:380–9. [3] Erdinc O, Uzunoglu M. Optimum design of hybrid renewable energy systems: overview of different approaches. Renewable Sustainable Energy Rev 2012;16:1412–25. [4] Fadaee M, MAM. Radzi. Multi-objective optimization of a stand-alone hybrid renewable energy system by using evolutionary algorithms: a review. Renewable Sustainable Energy Rev 2012;16:3364–9. [5] Luna-Rubio R, Trejo-Peres M, Vargas-Vazquez D, Rı´os-Moreno GJ. Optimal sizing of renewable hybrids energy systems: a review of methodologies. Sol Energy 2012;86:1077–88. [6] Khatib T, Mohamed A, Sopian K. A review of photovoltaic systems size optimization techniques. Renewable Sustainable Energy Rev 2013;22:454–65. [7] Upadhyay S, Sharma MP. A review on configurations, control and sizing methodologies of hybrid energy systems. Renewable Sustainable Energy Rev 2014;38:47–63. [8] Chauhan A, Saini RP. A review on integrated renewable energy system based power generation for stand-alone applications: configurations, storage

768

[9] [10]

[11]

[12]

[13] [14] [15]

[16] [17] [18]

[19]

[20] [21] [22]

[23] [24]

[25]

[26]

[27]

[28]

[29] [30]

[31]

[32]

[33]

[34]

[35] [36] [37]

[38]

[39]

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

options, sizing methodologies and control. Renewable Sustainable Energy Rev 2014;38:99–120. Bourennani F, Rahnamayan S, Naterer GF. Optimal design methods for hybrid renewable energy systems. Int J Green Energy 2015;12:148–59. Bhandari B, Lee KT, Lee GY, Cho YM, Ahn SH. Optimization of hybrid renewable energy power systems: a review. Int J Precis Eng Manuf Green Technol 2015;2(1):99–112. Kaldellis JK. Stand-alone and hybrid wind energy systems: technology, energy storage and applications. Woodhead Publishing Limited and CRC Press LLC; 2010. Maghraby HAM, Shwehdi MH, Al-Bassam GK. Probabilistic assessment of photovoltaic (PV) generation systems. IEEE Trans Power Syst 2002;17 (1):205–8. Shrestha GB, Goel L. A study on optimal sizing of stand-alone photovoltaic stations. IEEE Trans Energy Convers 1998;13(4):373–8. Al-Ashwal AM, Moghram IS. Proportion assessment of combined PV–wind generating systems. Renewable Energy 1997;10(1):43–51. Borowy BS, Salameh ZM. Methodology for optimally sizing the combination of a battery bank and PV array in a wind/PV hybrid system. IEEE Trans Energy Convers 1996;11(2):367–73. Markvart T. Sizing of hybrid PV–wind energy systems. Sol Energy 1996;59 (4):277–81. Yang HX, Lu L, Zhou W. A novel optimization sizing model for hybrid solar– wind power generation system. Sol Energy 2007;81(1):76–84. Kellogg W, Nehrir MH, Venkataramanan G, Gerez V. Generation unit sizing and cost analysis for stand-alone wind, photovoltaic and hybrid wind/PV systems. IEEE Trans Energy Convers 1998;13(1):70–5. Kellogg W, Nehrir MH, Venkataramanan G, Gerez V. Optimal unit sizing for a hybrid wind/photovoltaic generating system. Electr Power Syst Res 1996;39:35–8. Diaf S, Diaf D, Belhamel M, Haddadi M, Louche A. A methodology for optimal sizing of autonomous hybrid PV/wind system. Energy Policy 2007;35:5708–18. Prasad AR, Natarajan E. Optimization of integrated photovoltaic–wind power systems with battery storage. Energy 2006;31:1943–54. Borowy BS, Salameh ZM. Methodology for optimally sizing the combination of a battery bank and PV array in a wind/PV hybrid system. IEEE Trans Energy Convers 1996;11(2):367–75. Ashok S. Optimised model for community-based hybrid energy system. Renewable Energy 2007;32(7):1155–64. Bagul AD, Salameh ZM, Borowy B. Sizing of a stand-alone hybrid wind photovoltaic system using a three-event probability density approximation. Sol Energy 1996;56(4):323–35. Karaki SH, Chedid RB, Ramadan R. Probabilistic performance assessment of autonomous solar–wind energy conversion systems. IEEE Trans Energy Convers 1999;14:766–72. Tina G, Gagliano S, Raiti S. Hybrid solar/wind power system probabilistic modeling for long-term performance assessment. Sol Energy 2006;80 (5):578–88. Chedid R, Akiki H, Rahman. S. A decision support technique for the design of hybrid solar–wind power systems. IEEE Trans Energy Convers 1998;13:76–82. Gavanidou ES, Bakirtzis AG. Design of a standalone system with renewable energy sources using trade off methods. IEEE Trans Energy Convers 1992;7:42–8. Chedid R, Rahman S. Unit sizing and control of hybrid wind–solar power systems. IEEE Trans Energy Convers 1997;12(1):79–85. Huneke F, Henkel J, González JAB, Erdmann G. Optimization of hybrid offgrid energy systems by linear programming. Energy Sustainability Soc 2012;2(7):1–19. Nogueira CEC, Vidotto ML, Niedzialkoski RK, Melegari de Souza SN, Chaves LI, Edwiges T, et al. Sizing and simulation of a photovoltaic–wind energy system using batteries, applied for a small rural property located in the south of Brazil. Renewable Sustainable Energy Rev 2014;29:151–7. Lee JY, Chen CL, Chen HC. A mathematical technique for hybrid power system design with energy loss considerations. Energy Convers Manage 2014;82:301–7. Saif A, Gad Elrab K, Zeineldin HH, Kennedy S, Kirtley JL. Multi-objective capacity planning of a PV–wind–diesel–battery hybrid power system. In: IEEE international conferences; 2010. Nagabhushana AC, Jyoti R, Raju AB. Economic analysis and comparison of proposed hres for stand-alone applications at various places in Karnataka state. IEEE PES Innovative Smart Grid Technol—India 2011:380–5. Deb Kalyanmoy. Multi-objective optimization using evolutionary algorithms. vol. 2012. Chichester: John Wiley & Sons; 2001. Goldberg David Edward. Genetic algorithms in search, optimization, and machine learning. . Reading Menlo Park: Addison-Wesley; 1989. Koutroulis E, Kolokotsa D, Potirakis A, Kalaitzakis K. Methodology for optimal sizing of stand-alone photovoltaic/wind-generator systems using genetic algorithms. Sol Energy 2006;80(9):1072–88. Koutroulis E, Kolokotsa D. Design optimization of desalination systems power-supplied by PV and W/G energy sources. Desalination 2010;258(1– 3):171–81. Yang H, Zhou W, Lu L, Fang Z. Optimal sizing method for stand-alone hybrid solar–wind system with LPSP technology by using genetic algorithm. Sol Energy 2008;82(4):354–67.

[40] Yang H, Zhou W, Lou C. Optimal design and techno-economic analysis of a hybrid solar–wind power generation system. Appl Energy 2009;86(2):163–9. [41] Bilal BO, Sambou V, Ndiaye PA, Kébé CMF, Ndongo M. Optimal design of a hybrid solar–wind–battery system using the minimization of the annualized cost system and the minimization of the loss of power supply probability (LPSP). Renewable Energy 2010;35(10):2388–90. [42] AESAA Nafeh. Optimal economical sizing of a PV–wind hybrid energy system using genetic algorithm. Int J Green Energy 2011;8:25–43. [43] Abdullrahman AAS, Addoweesh KE. Optimum sizing of hybrid PV/wind/ battery/diesel system considering wind turbine parameters using genetic algorithm. In: IEEE international conference on power and energy (PECon); 2–5 December 2012. Malaysia, p. 121–26. [44] Atia R, Yamada N, Optimization of a PV–wind–diesel system using a hybrid genetic algorithm. In: IEEE electrical power and energy conference; 2012. p. 80–85. [45] Merei G, Berger C, Sauer DW. Optimization of an off-grid hybrid PV–wind– diesel system with different battery technologies using genetic algorithm. Sol Energy 2013;97:460–73. [46] Abbes D, Martinez A, Champenois G. Life cycle cost, embodied energy and loss of power supply probability for the optimal design of hybrid power systems. Math Comput Simul 2014;98:46–62. [47] Shi JH, Zhu HJ, Cao GY. Design and techno-economical optimization for stand-alone hybrid power systems with multi-objective evolutionary algorithms. Int J Energy Res 2007;31:315–28. [48] Mostofi F, Shayeghi H. Feasibility and optimal reliable design of renewable hybrid energy system for rural electrification in Iran. Int J Renewable Energy Res 2012;2(4):574–82. [49] Shadmand MB, Balog RS. Multi-objective optimization and design of photovoltaic–wind hybrid system for community smart DC micro grid. IEEE Trans Smart Grid 2014;5(5):2635–43. [50] Tégani I, Aboubou A, Ayad MY, Becherif M, Saadi R, Kraa O. Optimal sizing design and energy management of stand-alone photovoltaic/wind generator systems. Energy Procedia 2014;50:163–70. [51] Bilal BO, Nourou D, Kébé CMF, Sambou V, Ndiaye PA, Ndongo M. Multiobjective optimization of hybrid PV/wind/diesel/battery systems for decentralized application by minimizing the levelized cost of energy and the CO2 emissions. Int J Phys Sci 2015;10(5):192–203. [52] Kennedy J, Eberhart R. Particle swarm optimization. In: Proceedings of IEEE, international conference on neural networks, 1995; 4(2):1942–48. [53] Poli R, Kennedy J, Blackwell T. Particle swarm optimization. Swarm Intell 2007;1(1):33–57. [54] Bashir M, Sadeh J. Size optimization of new hybrid stand-alone renewable energy system considering a reliability index. In: Eleventh IEEE international conference in environment and electrical engineering (EEEIC); 2012. p. 989–94. [55] Lee TY, Chen CL. Wind–photovoltaic capacity coordination for a time-of-use rate industrial user. IEEE Trans Renewable Power Gener 2009;3(2):152–67. [56] Kaviani AK, Riahy GH, SHM. Kouhsari. Optimal design of a reliable hydrogen based stand-alone wind/PV generating system, considering component outages. Renewable Energy 2009;34(11):2380–90. [57] Bansal AK, Gupta RA, Kumar R. Optimization of hybrid PV/wind energy system using Meta Particle Swarm Optimization (MPSO). In: IEEE 2011 India international conference on power electronics (IICPE), 2010. p. 1–7. [58] Sharafi M, ELMekkawy TY. Multi-objective optimal design of hybrid renewable energy systems using PSO-simulation based approach. Renewable Energy 2014;68:67–79. [59] Pirhaghshenasvali M, Asaei B. Optimal modeling and sizing of a practical hybrid wind/PV/diesel generation system. In: The fifth power electronics, drive systems and technologies conference (PEDSTC 2014), Tehran, Iran; 2014. p. 506–11. [60] Borhanazad H, Mekhilef S, Ganapathy VG, Delshad MM, Mirtaheri A. Optimization of micro-grid system using MOPSO. Renewable Energy 2014;71:295–306. [61] Maleki A, Askarzadeh A. Comparative study of artificial intelligence techniques for sizing of a hydrogen-based stand-alone photovoltaic/wind hybrid system. Int J Hydrog Energy 2014;39:9973–84. [62] Maleki A, Ameri M, Keynia F. Scrutiny of multifarious particle swarm optimization for finding the optimal size of a PV/wind/battery hybrid system. Renewable Energy 2015;80:552–63. [63] Kirkpatrick S. Optimization by simulated annealing: quantitative studies. J Stat Phys 1984;34(5–6):975–86. [64] Ekren O, Ekren BY. Size optimization of a PV/wind hybrid energy conversion system with battery storage using simulated annealing. Appl Energy 2010;87:592–8. [65] Dorigo M, Gambardella LM. Ant colony system: a cooperative learning approach to the traveling salesman problem. In: IEEE transactions on evolutionary computation 199; 1.1(7):53–66. [66] Dorigo M, Birattari M, Stutzle T. Ant colony optimization. IEEE Comput Intell Mag 2006;1.4:28–39. [67] Menshsari A, Ghiamy M, Mousavi MMM, Bagal HA. Optimal design of hybrid water–wind–solar system based on hydrogen storage and evaluation of reliability index of system using ant colony algorithm. Int Res J Appl Basic Sci 2013;4(10):3582–600. [68] Xu D, Kang L, Cao B. Graph-based ant system for optimal sizing of standalone hybrid wind/PV power systems. Computational inelligence. Berlin Heidelberg: Springer; 2006. p. 1136–46.

S. Sinha, S.S. Chandel / Renewable and Sustainable Energy Reviews 50 (2015) 755–769

[69] Passino KM. Biomimicry of bacterial foraging for distributed optimization and control. IEEE Control Syst Mag 2002;22(3):52–67. [70] Bazyar R, Valipoor KH, Javadi MR, Valizade M, Kord H. Optimal design and energy management of stand-alone wind/PV/diesel/battery using bacterial foraging algorithm. In: Eighth national congress; 2010, p. 1–14. [71] Karaboga D, Basturk B. A powerful and efficient algorithm for numerical function optimization: artificial bee colony (ABC) algorithm. J Global Optim 2007;39:459–71. [72] Karaboga D, Basturk B. Artificial bee colony (ABC) optimization algorithm for solving constrained optimization problems. Foundations of fuzzy logic and soft computing. Berlin Heidelberg: Springer; 2007. p. 789–98. [73] Nasiraghdam H, Jadid S. Optimal hybrid PV/WT/FC sizing and distribution system reconfiguration using multi-objective artificial bee colony (MOABC) algorithm. Sol Energy 2012;86:3057–71. [74] Maleki A, Askarzadeh A. Artificial bee swarm optimization for optimum sizing of a stand-alone PV/WT/FC hybrid system considering LPSP concept. Sol Energy 2014;107:227–35. [75] Tudu B, Majumder S, Mandal KK, Chakraborty N. Optimal unit sizing of stand-alone renewable hybrid energy system using bees algorithm. In: IEEE international conference on energy, automation, and signal (ICEAS), 2011; pp. 1–6. [76] Maleki A, Pourfayaz F. Optimal sizing of autonomous hybrid photovoltaic/ wind/battery power system with LPSP technology by using evolutionary algorithms. Sol Energy 2015;115:471–83. [77] Simon D. Biogeography-based optimization. IEEE Trans Evol Comput 2008;12(6):702–13. [78] Kumar R, Gupta RA, Bansal AK. Economic analysis and power management of a stand-alone wind/photovoltaic hybrid energy system using biogeography based optimization algorithm. Swarm Evol Comput 2013;8:33–43. [79] Liang JJ, Qin AK, Suganthan PN, Baskar S. Comprehensive learning particle swarm optimizer for global optimization of multimodal functions. IEEE Trans Evol Comput 2006;10:281–95. [80] Mallipeddi R, Suganthan PN, Pan QK, Tasgetiren MF. Differential evolution algorithm with ensemble of parameters and mutation strategies. Appl Soft Comput 2010;11:1679–96. [81] Gupta RA, Kumar R, Bansal AK. BBO-based small autonomous hybrid power system optimization incorporating wind speed and solar radiation forecasting. Renewable Sustainable Energy Rev 2015;41:1366–75. [82] Geem ZW, Kim JH, Loganathan GV. A new heuristic optimization algorithm: harmony search. Simulation 2001;76:60–8. [83] Maleki A, Askarzadeh A. Optimal sizing of a PV–wind–diesel system with battery storage for electrification to an off-grid remote region: a case study of Rafsanjan, Iran. Sustainable Energy Technol Assess 2014;7:147–53. [84] Rashedi E, Nezamabadi-Pour H, Saryazdi S. GSA: a gravitational search algorithm. Inf Sci 2009;179(13):2232–48. [85] Rashedi E, Nezamabadi-Pour H, Saryazdi S. BGSA: binary gravitational search algorithm. Nat Comput 2010;9:727–45. [86] Wu K, Zhou H, An S, Huang T. Optimal coordinate operation control for wind–photovoltaic–battery storage power-generation units. Energy Convers Manage 2015;90:466–75.

769

[87] Atashpaz-Gargari E, Lucas C. Imperialist competitive algorithm: an algorithm for optimization inspired by imperialistic competition. In: IEEE congress on evolutionary computation, Singapore, 2007; p. 4661–4667. [88] Gharavi H, Ardehali MM, Ghanbari-Tichi S. Imperial competitive algorithm optimization of fuzzy multi-objective design of a hybrid green power system with considerations for economics, reliability, and environmental emissions. Renewable Energy 2015;78:427–37. [89] Ranjbar MR, Kouhi S. Sources response for supplying energy of a residential load in the form of on-grid hybrid systems. Electr Power Energy Syst 2015;64:635–45. [90] Glover F, Laguna M. Tabu search. US: Springer; 1999. [91] Katsigiannis YA, Georgilakis PS, Karapidakis ES. Hybrid simulated annealing —Tabu search method for optimal sizing of autonomous power systems with renewables. IEEE Trans Sustainable Energy 2012;3(3):330–8. [92] Askarzadeh A. A discrete chaotic harmony search-based simulated annealing algorithm for optimum design of PV/wind hybrid system. Sol Energy 2013;97:93–101. [93] Tutkun N. Minimization of operational cost for an off-grid renewable hybrid system to generate electricity in residential buildings through the SVM and the BCGA methods. Energy Build 2014;76:470–5. [94] Khatib T, Mohamed A, Sopian K. Optimization of a PV/wind micro-grid for rural housing electrification using a hybrid iterative/genetic algorithm: case study of Kuala Terengganu, Malaysia. Energy Build 2012;47:321–31. [95] Dehghan S, Kiani B, Kazemi A, Parizad A. Optimal sizing of a hybrid wind/PV plant considering reliability indices. World Acad Sci Eng Technol 2009;56 (32):527–35. [96] Zhou T, Sun W. Optimization of battery–super capacitor hybrid energy storage station in wind/solar generation system. IEEE Trans Sustainable Energy 2014;5(2):408–15. [97] Abdelhak BJ, Najib E, Abdelaziz H, Hnaien F. Optimum sizing of hybrid PV/ wind/battery using fuzzy-adaptive genetic algorithm in real and average battery service life. In: IEEE international symposium on power electronics, electrical drives, automation and motion; 2014. , p.871–876. [98] Mukhtaruddin RNSR Rahman HA, Hassan MY, Jamian JJ. Optimal hybrid renewable energy design in autonomous system using Iterative-ParetoFuzzy technique. Electr Power Energy Syst 2015;64:242–9. [99] Alsayed M, Cacciato M, Scarcella G, Scelba G. Design of hybrid power generation systems based on multi criteria decision analysis. Sol Energy 2014;105:548–60. [100] Yang XS, Deb S. Engineering optimization by cuckoo search. Int J Math Model Numer Optim 2010;1(4):330–43. [101] Yang XS. Bat algorithm for multi-objective optimization. Int J Bio-Inspired Comput 2011;3:267–74. [102] Łukasik S, Żak S. Firefly algorithm for continuous constrained optimization tasks. Computational collective intelligence. Semantic web, social networks and multiagent systems. Springer-Berlin-Heidelberg; 2009. p. 97–106. [103] Li Y, Wen Q, Li L, Peng H. Hybrid chaotic ant swarm optimization. Chaos, Solitons Fractals 2009;42(2):880–9. [104] Brownlee J. Clever algorithms: nature-inspired programming recipes. 1st ed.. LuLu: Jason Brownlee; 2011. [105] Yang Xin-She. Nature-inspired metaheuristic algorithms. 2nd ed. UK: Luniver Press; 2010.