Dynamic Energy Management of Hybrid Energy Storage ... - IEEE Xplore

150

IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 30, NO. 1, MARCH 2015

Dynamic Energy Management of Hybrid Energy Storage System With High-Gain PV Converter Narsa Reddy Tummuru, Student Member, IEEE, Mahesh K. Mishra, Senior Member, IEEE, and S. Srinivas, Member, IEEE

Abstract—In this paper, fast acting dc-link voltage-based energy management schemes are proposed for a hybrid energy storage system fed by solar photovoltaic (PV) energy. Using the proposed control schemes, quick fluctuations of load are supplied by the supercapacitors and the average load demand is controlled by the batteries. Fast dc-link voltage, effective energy management, and reduced current stress on battery are the main features achieved from the proposed control schemes. The effectiveness of the proposed control schemes are compared with the unified cascaded control. Small-signal control gains are formulated to design the voltage and current loops of the proposed energy management schemes. Detailed stability analysis is also presented to find the boundary values of compensator gains. In addition, a high-gain PV converter is proposed for extraction of maximum power from the solar panels. High voltage gain, reduced reverse recovery of diodes, and less duty cycle operation are the key features obtained from the proposed high-gain converter. The validity of the proposed energy management schemes with high-gain converter is verified by the detailed simulation and experimental studies. Index Terms—Battery, dynamic energy management, photovoltaic (PV) system, small-signal models, supercapacitor.

I. INTRODUCTION HE POWER converter-based microgrids have low rotational inertia compared to that of utility grid; as a result of this, system becomes unstable under extreme operating conditions. By adding energy storage devices in the microgrid, the equivalent inertia introduced enhances the dynamic stability of the system against the changes in the load or changes in the renewable power production due to the variations in the atmospheric conditions [1]. Microgrids have two modes of operation, namely, grid-interactive and islanding modes [2]. During islanding mode, the main objective of the storage is to maintain the energy balance. During grid-interactive mode, the aim is to prevent the propagation of the renewable source intermittency and load fluctuations to the utility grid [3]. Energy storage of a single type cannot perform these jobs efficiently. The intermittent nature of renewable energy sources like photovoltaic (PV) demands usage of storage with high energy density [4]. However, applications like portable electronic devices,

T

Manuscript received November 5, 2013; revised May 1, 2014; accepted August 29, 2014. Date of publication September 26, 2014; date of current version February 16, 2015. This work was supported by the Department of Science and Technology, India, under Project Grant DST/TM/SERI/2k10/47(G). Paper no. TEC-00655-2013. The authors are with the Department of Electrical Engineering, Indian Institute of Technology Madras, Chennai 600036, India (e-mail: narasaiitm@ gmail.com; [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TEC.2014.2357076

telecommunication systems, and electric vehicles have highpulse power requirement and relatively low average power requirement. The high-pulse power requirement of load, thus, demands storage with high power density. Therefore, a hybrid energy storage system (HESS) that contains both high energy density storage battery and high power density storage supercapacitor are necessary to meet the aforementioned requirements [5]–[8]. On the PV source side, a dc–dc converter with a high voltage gain is necessary for converting the low PV panel voltage into the high dc-link voltage [9], [10]. There are many high-gain converter topologies reported in the literature [9]–[15]. The work proposed in [11] and [10] uses voltage multiplier cells (VMCs) that comprise of diodes and capacitors to increase the steadystate voltage gain. Achieving high gain requires more number of components that result in reduced converter efficiency. In [13], the phase-shifted full-bridge can achieve a high step-up gain by increasing the turn’s ratio of the transformer. However, more input electrolytic capacitors are required to suppress the large input current ripple and the output-diode voltage stress is much higher than the output voltage. The passive lossless three-level boost converter [9] is another alternative for achieving higher gains. However, the voltage stress or current stress of the power devices increases. Compared to the converter topologies mentioned before, the proposed converter has the following features. 1) It provides higher voltage gains. 2) It does not employ VMCs and, hence, is less complex. 3) It avoids the converter operation at extreme duty cycles. 4) Though the converter switching stress are more, low-dutycycle operation results in moderate efficiency. A bidirectional dc–dc converter at the dc link enables power flow into and from the energy storage elements. Various control strategies have been proposed for bidirectional dc–dc converters that include use of neural networks to fuzzy logic [16], optimal control [17], and model-predictive control (MPC) [18]. In [19], fuzzy controlled battery/supercapacitor energy system for electric vehicle was reported. In [20], rule-based control methods are proposed and these are verified by simulation studies only. The main drawback of the method proposed in [18] is that it is computationally intensive, since it is based on the classic MPC. Classic MPC relies on a discrete model of the control system and a cost function. The work relating to detailed dynamics of the dc-link voltage against the changes in the load conditions or changes in the renewable power generation has to be addressed. It is well known that any change in the load/renewable power affects the dc-link voltage directly. The sudden decrease/increase of load/renewable power would result in an increase in the

0885-8969 © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications standards/publications/rights/index.html for more information.

TUMMURU et al.: DYNAMIC ENERGY MANAGEMENT OF HYBRID ENERGY STORAGE SYSTEM WITH HIGH-GAIN PV CONVERTER

Fig. 1.

151

HESS with the proposed high-gain PV converter.

A. High-Gain PV DC–DC Converter To integrate the low-voltage PV panels to the distribution system, the output voltage of the intermediate dc–dc converter should be high enough to generate the required dc-link voltage [9]. Hence, a dc–dc converter with a high-voltage gain is necessary. The high-gain PV converter is essentially used for achieving higher voltage gains; as a result, the magnitude of the current drawn from PV side is high, and hence, the converter operates most of the times in continuous current mode (CCM). Upon application of volt–second balance on inductors Ls1 , Ls2 , and L1 (see Fig. 1) during CCM operation, the following equations are obtained: Fig. 2.

vpv − vc1 (1 − D) = 0 2 vc1 D + (vc1 − vo )(1 − D) = 0.

2 vpv D +

Steady-state voltage gain versus duty cycle.

dc-link voltage above the reference value, whereas a sudden increase/decrease in load/renewable power would reduce the dc-link voltage below its reference value. The multifunctional operation of microgrid inverter requires variation of the dc-link voltage within the prescribed limits [21]. Keeping the above perspective and issues, a fast acting dc-link voltage-based energy management schemes for HESS are proposed in this paper. The main features obtained from the proposed control schemes are tight dc-link voltage regulation, effective power management at the dc-link, allows current limits for both the battery and the supercapacitor, less current stress on the battery pack, and more importantly the control schemes are computationally less intensive. II. MODELING AND CONTROL OF THE PROPOSED SYSTEM The envisaged system consists of PV modules, batteries, and supercapacitors connected to the dc grid through the intermediate dc–dc converter power stages as shown in Fig. 1. A high-gain dc–dc converter topology is proposed in this paper to extract the maximum power from the PV panels. The bidirectional buckboost dc–dc converter topologies are used to control the power flow between batteries, supercapacitors, and the dc load.

(1) (2)

After simplification of (1) and (2), the steady-state voltage gain of the proposed converter (Gccm ) can be written as Gccm =

Vo 1 + 3D = Vpv (1 − D)2

(3)

where D, Vo , and Vpv are the duty cycle, average output voltage, and PV voltage, respectively. The steady-state voltage gain (Gccm ) of the proposed high-gain PV converter is plotted against duty cycle with inductor and capacitor parasitic elements (rsL 1 , rsL 2 , rL 1 = 0.1 Ω and rc1 , rco = 0.15 Ω). The voltage gain of this converter is compared with the some of the recently reported converters [9], [11], [14], [15] in literature as shown in Fig. 2. It is clear from Fig. 2 that, for a given duty cycle, the proposed high-gain converter topology provides higher gain when compared to other topologies. B. Tracking Performance of High-Gain Converter The small-signal control to inductor current gain [GiD (s)] of the high-gain PV converter topology is obtained as GiD (s) =

Δipv (s) = (Z1 − Z2 )X + Z(SI − A)−1 f ΔD(s)

(4)

152

Fig. 3.


Bode plot of G i D (s) with D as the parameter. Fig. 5. Energy management schemes for battery control. (a) Control scheme-I. (b) Control scheme-II (proposed in this paper).

C. Design of Energy Storage Elements The design of the supercapacitor is based on the energy that has to be supplied or absorbed during dynamic conditions. The design of capacitance of the supercapacitor pack is considered for constant power mode and the maximum useful energy that the supercapacitor pack can provide is given as follows: Fig. 4.

Variable reference current perturbed MPPT algorithm.

Esc =

Csc (Vm2 ax − Vm2 in ) . 2

(5)

Upon simplification of (5), the capacitance of the supercapacitor pack (Csc ) can be derived as follows: where Z, A, B, Δipv , ΔD, and X are the average output, system, input matrices, perturbed PV current, perturbed duty cycle, and steady-state solution, respectively. f = (A1 − A2 )X + (B1 − B2 )Vpv and subscripts 1 and 2 represent ON and OFF status of switch. The tracking performance of inductor current (iL s1 ) variations are investigated at different solar insolation levels. The changes in the solar insolation directly affect the duty cycle of the high-gain converter. Hence, the dynamics of the inductor current (iL s1 ) is investigated against the duty cycle. Irrespective of variations in the duty cycle (0.16 dB) as seen from Fig. 3. The variable reference current perturbed MPPT algorithm used in this paper is illustrated in Fig. 4. This algorithm generates the variable reference current perturbation (Δi∗pv ) based on the rate of change of PV array power (Δppv ). The Δppv is given as an input to the PID controller as shown in the Fig. 4. For every change of PV array power, the PID controller generates a variable reference current perturbation, and therefore, the perturbation is directly related to Δppv . The perturb and observe (P&O) algorithm generates reference PV current (i∗pv ) using Δi∗pv as one of the input. In Fig. 4, Ns and ie are the number of delayed samples and error in PV current, respectively.

Csc =

ηVm2 ax

2E sc ϑ 2 1 − 100

(6)

where Vm ax , Vm in , ϑ, and η are the maximum, minimum supercapacitor pack voltages, discharge voltage ratio, and efficiency of the supercapacitor pack, respectively. III. ENERGY MANAGEMENT SCHEMES FOR HESS The energy management schemes, in general, should provide several operational objectives depending on the PV power availability (ppv ), batteries, and supercapacitors state of charge (SoC) status. Fast acting dc-link voltage-based energy management schemes are proposed in this paper for battery and supercapacitor control. A. Battery Energy Management Scheme The battery energy management scheme is mainly used to regulate the power balance in the dc grid. Many control schemes are reported in the literature to control the battery current in various applications. Among them, a unified current-mode control [22] for hybrid vehicles and for renewable power system [23] is reported and is shown in Fig. 5(a) (called as control scheme-I in this paper). The main drawbacks of this scheme


153

where wb is the weighting factor of the battery control. The key difference between control scheme-I and the scheme proposed here is that the inductor current is programmed such that it can be controlled to follow the desired reference waveform, which is generated by the outer voltage loop within a switching cycle using (9). Therefore, the control scheme-II for battery control does not require any current controller. Fig. 6.

Proposed energy management scheme for supercapacitor control.

B. Supercapacitor Energy Management Scheme

Fig. 7. Small-signal average model of battery converter with control scheme-II (proposed).

are complicated compensator design and low efficiency at light-load conditions. To overcome these issues, a fast acting dc-link voltage-based energy management scheme is proposed for battery current control. The proposed scheme mainly consists of battery SoC estimation, battery reference current generation, and generation of control variable (δB ) as illustrated in the Fig. 5(b). The dynamics of the inductor current during buckand boost-mode operation of bidirectional converter are given as (refer Fig. 1) vdc − vB diB vdc = δB (t) − [1 − δB (t)] dt LB LB

(7)

diB vB vdc − vB = δB (t) − [1 − δB (t)] . dt LB LB

(8)

The bidirectional converter switches are operated in complementary mode. Therefore, it is sufficient to find out the control law in the boost mode of operation. Upon simplification of (8), the control variable (δB ) is derived as δB (t) =

vdc − vB LB diB . + vdc vdc dt

disc + vdc (1 − δsc ) dt dvdc vdc = isc (1 − δsc ) + Cdsc . dt RL ,eq vsc = Lsc

(10)

(12) (13)

The dynamics of the supercapacitor error current (ie =isc,ref − isc ) can be written as isc =

die + (wsc − 1)ie + isc,ref . dt

(14)

Using (13), the supercapacitor current can be written as isc = δsc isc + Cdsc

(9)

In (9), the first part is the feedforward term, which improves the dc-link voltage regulation against battery voltage changes. The second part is the predictive term and the coefficients of these parts are not constant. These coefficients are varying with the dc-link voltage, and hence, the dynamics of the dc-link voltage is faster. The reference current of the battery converter (iB ,ref ) is generated by a simple voltage compensator [Gcv (z)] in the outer dc-link voltage loop. This Gcv (z) is responsible for regulation of the dc-link voltage. In order to further improve the transient response of the dc-link voltage, a term proportional to the voltage error, i.e., wb (vdc,ref − vdc ) is added in the outer voltage loop iB ,ref = iB ,r + wb (vdc,ref − vdc )

The supercapacitor energy management scheme plays a vital role for achieving fast dc-link voltage dynamics. The proposed scheme mainly contains the generation of supercapacitor reference current (isc,ref ) and fast acting current control loop. A moving average filter (MAF) is used to calculate the average load demand. The supercapacitor current control is obtained by a fast acting direct current feedback and generation of control variable (δsc ) based on error current dynamics as illustrated in Fig. 6. The control law of the supercapacitor ensures that it quickly responds to sudden changes in the dc-link voltage. The supercapacitor reference current (isc,ref = iscl,ref + G (vdc,ref − vdc )) using MAF is obtained as 1 t o +T iLoad (t) dt + G (vdc,ref − vdc ) isc,ref = iLoad (t) − T to (11) where G is the gain of the dc-link voltage dynamics estimator (shown in Fig. 6). The average state model of the supercapacitor converter (shown in Fig. 1) during boost mode of operation is given as

dvsc vdc − . dt RL ,eq

(15)

Using (14) and (15), the control variable is obtained as Cdsc dvsc isc,ref − io (wsc − 1) 1 die − + ie + isc isc isc dt isc dt (16) where wsc , Ksc , and RL ,eq are the weighting factor for supercapacitor control, direct supercapacitor current feedback gain, and equivalent load resistance, respectively. The main feature of this controller compared to the conventional one is that the gain values of the proportional and derivative coefficients of supercapacitor error currents are not constant. These gains are varying with respect to the supercapacitor current. The variation in the controller gains shows the adaptiveness of the proposed supercapacitor control law. δsc =

154


Fig. 8. Stability of control scheme-II (proposed) with K c as a parameter. (a) Current loop gain Nyquist plot. (b) Current loop gain Bode plot. (c) Overall loop gain Nyquist plot.

TABLE I SYSTEM PARAMETERS

IV. DESIGN AND ANALYSIS OF BATTERY AND SUPERCAPACITOR CONTROL LOOPS The main objective of the battery and supercapacitor control schemes is to balance the power flow at the dc link by regulating the voltage at the dc link effectively. The dynamics of the dclink voltage is mainly dependent on changes in the equivalent battery, supercapacitor, and PV currents at the dc link. Therefore, the dc-link voltage dynamics can be written in terms of these currents as Δvdc = Gv i B Δi B + Gv i s c Δi sc + Gv i p v Δi pv

(17)

where Δi B , Δi sc , and Δi pv are equivalent dc-link currents of battery, supercapacitor, and high-gain PV converter, respectively. Gv i B , Gv i s c , and Gv i p v are the small-signal impedances of battery, supercapacitor, and high-gain PV converter, respectively. The following are the important gains that are used to design the voltage [Gcv (z)] and current [Gci (z)] compensators 1) the duty-cycle-to-dc-link voltage gain [Gv δ (z)], carries the information needed to determine the type of voltage feedback compensation, 2) the duty-cycle-to-inductor current gain [Giδ (z)] is needed to determine the current controller structure, 3) audio susceptibility [Gv v B (z)], and 4) input admittance [Giv (z)]. The structure of voltage compensator used in this work for battery control is given as follows: Gcv (z) =

Kv z(z − α) (z − 1) (z − β)

(18)

where Kv , α, and β are the voltage compensator gain and voltage compensator control parameters, respectively. The small-signal model of the battery converter with control scheme-II is shown in Fig. 7. In the similar lines, the smallsignal model of supercapacitor along with its control can be derived. The inner loop current gain [Li (z)], outer loop voltage gain [Lv (z)], and overall loop gain [Lo (z)] can be written as follows: Li (z) = Giδ (z) Ri Gci (z) Fm

PV module parameters Rated power Open circuit voltage (V o c ) Maximum power voltage (V m p ) Short circuit current (I s c ) Maximum power current (I m p ) High-gain converter parameters Supercapacitor pack parameters Terminal voltage (V s c ) Max. peak current rate (I p ) Capacitance/pack (C s c ) Max. continuous current (I m c ) Battery pack specifications Ah Capacity Terminal voltage (V B ) No. of batteries in series

Values 240 W 37.5 V 31.1 V 8.3 A 7.72 A L s 1 = 1 mH, L s 2 = 0.5 mH, L 1 = 0.1 mH, C o = C 1 = 220 μF Values 16.2 V 200 A 58 F 19 A Values 14 Ah 12 V 4

Lv (z) = Gv δ (z) Gcv (z) [1 + Gci (z)] Fm Lo (z) = Li (z) + Lv (z)

(19)

where Gci =Kc Gi , Kc , Gi , Ri , and Fm are the overall current control gain, current loop gain constant, current-control gain, current sensor scaling factor, and modulator gain, respectively. A. Stability Analysis of Voltage and Current Control Loops The detailed stability analysis of the voltage and current loops of supercapacitor are investigated at different compensator gains and some of the loop gain frequency response plots are presented in Fig. 8 with Kc as a parameter. In general, for the system to be stable, its loop gain Nyquist plot envelop should be on the right side of the critical point (CP) (−1, 0). However, in the event of disturbances in the dc-link voltage, the loop gain plot trajectory will move towards or away from the CP. By proper selection of compensator parameters, the overall loop gain trajectory should be within the desired stability bounds. Fig. 8(a) and (b) shows


155

Fig. 9. Performance of HESS. (a) Control scheme-I with w b = 1 and w sc = 0.1. (b) Control scheme-I with w b = 0.1 and w sc = 0.5. (c) Control scheme-II (proposed) with w b = 0.1 and w sc = 0.5.

the Nyquist and Bode plots of current loop gain [Li (z)] with gain (Kc ) as the parameter. It can be observed from these figures that the current loop of the supercapacitor control is stable with sufficient stability margins (P M > 45◦ , GM > 6 dB) as Kc is varied in range of 5 to 50. Fig. 8(c) shows the Nyquist plot of the overall loop gain [Lo (z)] with Kc varied from 5 to 50. It is clear from Fig. 8 that even the variation in the compensator gain will ensure the stability of the system. V. RESULTS AND DISCUSSION A. Simulation Results Detailed simulation studies are carried out to verify the validity of the control scheme-II for battery and supercapacitor energy management with high-gain PV converter as a MPP extractor. The system parameters used for simulation study are presented in Table I. The steady state and dynamic performance of control scheme-I and the control scheme proposed in this paper are investigated with different weighting factors of battery (wb ) and supercapacitor (wsc ) control under the following cases: 1) sudden changes of load power (pl ) and 2) pulsed load conditions. All these investigations are carried out at one solar insolation level. 1) Performance of HESS With Control Scheme-I: The performance of the control scheme-I against the sudden changes of load power (pl ) at the dc link is illustrated in the Fig. 9(a) and (b) with different weighting factors (wb , wsc ). The battery management control scheme-I regulates the dc-link voltage by balancing the powers at the dc link against the sudden changes of load. Fig. 9(a) shows the performance of the HESS with

high weighting factor of battery. For this case, the supercapacitor pack only supplies/absorbs (psc ) quick changes in the load power and the average load power is supplied/absorbed by the batteries (pB ). During the time t = 0 to 0.45 sec, the PV source (ppv = 240 W) supplies the load (200 W) and the excess power (40 W) is used to charge the batteries (iB ) and supercapacitors (isc ) as shown in the Fig. 9(a). The corresponding changes in the dc-link voltage (vdc ), supercapacitor voltage (vsc ), and battery voltage (vB ) are presented in the Fig. 9(a). The performance of the HESS with high weighting factor (wb = 0.1, wsc = 0.5) for the supercapacitor is shown in the Fig. 9(b). The dynamics of the dc-link voltage (vdc ) with different weighing factors are shown in the Fig. 9(a) and (b). It can be concluded from these figures that the control scheme-I is not able to regulate the dclink voltage effectively when weighting factors are wb = 0.1 and wsc = 0.5. 2) Performance of HESS With Control Scheme-II (Proposed): The investigations on the control scheme-II for battery and supercapacitor control are carried out with the same operating conditions (pl , ppv , wb , wsc ) considered for the control scheme-I. The performance of the control scheme-II against the same changes of load power (pl ) at the dc link is now illustrated in Fig. 9(c) with least weighting factor for battery pack control (wb = 0.1). For this case, the supercapacitor pack (psc ) only supplies/absorbs sudden changes in the load power. The changes in the dc-link voltage (vdc ), supercapacitor voltage (vsc ), and battery voltage (vB ) are shown in Fig. 9(c). The moment when the load power is changed, then the supercapacitor (isc ) supplies/absorbs the sudden changes of load and the average load power is supplied/absorbed by the batteries (iB )

156

Fig. 10. Performance of the control scheme-I under pulsed loads. (a) Load, battery, and supercapacitor currents. (b) Supercapacitor and battery voltages. (c) DC-link voltage. (d) Load, battery, PV, and supercapacitor powers.


Fig. 11. Performance of the control scheme-II (proposed) under pulsed loads. (a) Load, battery, and supercapacitor currents. (b) Supercapacitor and battery voltages. (c) DC-link voltage. (d) Load, battery, PV, and supercapacitor powers.

as illustrated in the Fig. 9(c). It can be observed from these figures that, irrespective of weighting factors/load changes, the control scheme-II regulates the dc-link voltage effectively compared with the control scheme-I. Further, the control schemeII improves the life cycles by reducing sudden current stress on battery [18]. 3) Performance of Control Schemes Under Pulsed Loads: The steady state and dynamic performance of the control scheme-I and control scheme-II with pulsed load conditions are illustrated in Figs. 10 and 11, respectively. It is clear from these figures that the control scheme-II has superior performance in terms of dc-link voltage dynamics (vdc ) and battery stress (in terms of current (iB )) compared with the control scheme-I. B. Experimental Results Experimental results are presented in order to validate the performance of control scheme-I and the control scheme proposed in this paper. A photograph of developed experimental setup in the laboratory is shown in Fig. 12. It consists of supply from the PV panel, battery and supercapacitor packs, high-gain converter, bidirectional converters and dc load as labeled in the figure.

Fig. 12.

Experimental setup of HESS with high-gain converter.

The dynamic performance of the control scheme-I and control scheme-II are validated by step variation of load from full load to half load (20 to 40 Ω) and from half load to full load (40 to 20 Ω). The maximum power tracking capability of the high-gain converter at one insulation level is shown in Fig. 13. The power


Fig. 13.

157

Peak power tracking performance of PV converter.

which is available from the PV panel is 65 W at this insolation level. The measured efficiency of the high-gain converter at this operating point is 92.3%. 1) Performance of Control Scheme-I: Figs. 14 and 15 show the steady state and dynamic performance of control scheme-I with different weighting factors. Fig. 14(a) shows the steadystate waveforms of the dc-link voltage (vdc ), battery current (iB ), and supercapacitor current (isc ). Fig. 14(b) and (c) shows the dynamic response of HESS at different weighting factors (wb , wsc ). The load demands 320 W under steady-state condition, of which 60 W is supplied by the PV system and remaining 260 W is supplied by the battery pack as observed from Fig. 14(a). The load power changes from 320 to 160 W at t = t1 . As a result, 60 W is supplied by the PV system and 100 W is supplied from the battery pack to balance the power as observed from Fig. 14(b). The load power changes from 160 to 320 W at t = t2 , the increase in the load will reduce the dc-link voltage and settles to the reference value (80 V) in 2 s. After load change, the PV system supplies 60 W and remaining 260 W is supplied by the battery pack as observed from Fig. 14(c). The dynamic response of various parameters is shown in Table II. Fig. 15 shows the dynamic response of control scheme-I with pulsed type of load. Fig. 15(a) shows the dynamics of vdc and iB at wb = 0.1 and wsc = 0.5. In this case, scheme-I does not allowing the battery pack to share the load demand effectively. This is mainly due to the low weighting factor for the battery pack control, and therefore, it affects the regulation capability of the dc-link voltage against the pulse load disturbances. Moreover, the sudden change in the battery current leads to more stress on the battery pack and it affects the battery life cycles [18]. Fig. 15 (b) shows the dynamic performance of the control scheme-I with wb = 1 and wsc = 0.5. In this case, sufficient weight is given to the battery pack current control to share the load demand, and hence, the dc-link voltage regulation is achieved. However, the large variations in the dc-link voltage effects the battery pack performance [18]. 2) Performance of Control Scheme-II (Proposed): The steady state and dynamic response of control scheme-II are illustrated in Fig. 16(a)–(c) with different weighting factors. Under steady-state condition, 160 W of load power is shared by the PV system (60 W) and the battery pack (100 W) as shown in Fig. 16(a). When the battery pack weighting factor wb = 0.1, the battery pack control does not allow battery pack to share

Fig. 14. Experimental results of control scheme-I. (a) Steady-state waveforms. (b) Dynamic performance with w b = 0.5 and w sc = 0.5. (c) Dynamic performance with w b = 1 and w sc = 0.5.

the load fully. As a result, the supercapacitor pack supplies the change in load demand for the short time, and therefore, effective regulation of the dc-link voltage is possible even at wb = 0.1 and wsc = 0.5 as shown in Fig. 16(b). In contrast, the control scheme-I fails to regulate the dc-link voltage under the same weighting factors [see Fig. 15(a)]. The performance of the control scheme-II under pulsed load is shown in Fig. 16(c). The load demand of 160 W is supplied

158


TABLE II COMPARISON OF STEADY STATE AND TRANSIENT PERFORMANCE OF CONTROL SCHEME-I AND CONTROL SCHEME-II (PROPOSED) For w b =1 and w s c =0.5

Weighting factors Control schemes Parameters Slew rate (A/s or V/s) Settling time (s) % Over/under shoot Steady-state error (V, A)

Control scheme-I vd c 5.7 3.5 37.5 0.01

iB 0.85 3.52 0.01 0.04

is c −0.21 1.1 5 0.03

For w b =0.1 and w s c =0.5

Control scheme-II il 30 – – –

vd c 0.001 0.005 0.01 0.001

iB 0.05 4.5 Nil 0.002

is c −0.92 0.05 8 0.008

Control scheme-I il 30 – – –

vd c 22 < 10 21 40

iB 27 20 17 2.5

is c 0.1 15 5 0.8

Control scheme-II il 30 – – –

vd c 0.06 22 0.07 0.06

iB 0.08 27 0.06 0.03

is c −1.2 – 15 0.01

il 30 – – –

Fig. 15. Experimental results of control scheme-I with pulse type of load. (a) With w b = 0.1 and w sc = 0.5. (b) With w b = 1 and w sc = 0.5.

by the battery (100 W) and the PV system (60 W) up to t = t1 . At t = t1 , the load demand increases to 320 W. As a result, the sudden change in the load demand is supplied by supercapacitor pack and 260 W of average load demand is supplied by battery pack as illustrated in Fig. 16(c). At t = t2 , the load power decreases to 160 W. As a result, the sudden decrease in load power is absorbed by the supercapacitor pack. The battery pack supplies the average demand of 100 W. It is clear from these figures that the fast acting dc-link voltage and less current stress on the battery are the two key features achieved by the control schemeII compared with the control scheme-I (see Figs. 14 and 15). The detailed steady state and transient performance of control

Fig. 16. Experimental results of control scheme-II (proposed). (a) Steadystate performance. (b) Dynamic performance with w b = 0.1 and w sc = 0.5. (c) With pulsed load (w b = 1 and w sc = 0.5).


scheme-I and control scheme-II at different weighting factors for battery back and supercapacitor control are shown in the Table II. From the Table II, it can be seen that the proposed control scheme provides better dc-link voltage regulation capability for the same weighting factors over control scheme-I. Moreover, the current stress on the battery pack is significantly reduced in the control scheme-II. Therefore, the energy management scheme proposed in this paper is suitable for effective power management in the hybrid energy storage environment. VI. CONCLUSION A fast acting dc-link voltage-based energy management schemes are proposed for HESS. The battery and supercapacitor units act as the energy storage devices and solar PV as the renewable energy source. By using the proposed schemes, it is shown that, the quick fluctuations of load are supplied by the supercapacitor pack and the average load demand is taken care by the batteries. The proposed control schemes provide fast dc-link voltage regulation, effective power management, and maintains current limits for both the battery and the supercapacitor packs. Also, the schemes are computationally less intensive. All the above features with reduced current stress on the batteries make this control schemes more suitable for HESS applications. The effectiveness of HESS with proposed converter is verified by detailed simulation and experimental studies. REFERENCES [1] F. Inthamoussou, J. Pegueroles-Queralt, and F. Bianchi, “Control of a supercapacitor energy storage system for microgrid applications,” IEEE Trans. Energy Convers., vol. 28, no. 3, pp. 690–697, Sep. 2013. [2] R. Kamel, A. Chaouachi, and K. Nagasaka, “Three control strategies to improve the microgrid transient dynamic response during isolated mode: A comparative study,” IEEE Trans. Ind. Electron., vol. 60, no. 4, pp. 1314–1322, Apr. 2013. [3] P. Thounthong, “Model based-energy control of a solar power plant with a supercapacitor for grid-independent applications,” IEEE Trans. Energy Convers., vol. 26, no. 4, pp. 1210–1218, Dec. 2011. [4] T. Senjyu, M. Datta, A. Yona, and C.-H. Kim, “A control method for small utility connected large PV system to reduce frequency deviation using a minimal-order observer,” IEEE Trans. Energy Convers., vol. 24, no. 2, pp. 520–528, Jun. 2009. [5] A. Gee, F. Robinson, and R. Dunn, “Analysis of battery lifetime extension in a small-scale wind-energy system using supercapacitors,” IEEE Trans. Energy Convers., vol. 28, no. 1, pp. 24–33, Mar. 2013. [6] N. Rizoug, P. Bartholomeus, and P. Le Moigne, “Study of the ageing process of a supercapacitor module using direct method of characterization,” IEEE Trans. Energy Convers., vol. 27, no. 2, pp. 220–228, Jun. 2012. [7] A. Hajizadeh, M. Golkar, and A. Feliachi, “Voltage control and active power management of hybrid fuel-cell/energy-storage power conversion system under unbalanced voltage sag conditions,” IEEE Trans. Energy Convers., vol. 25, no. 4, pp. 1195–1208, Dec. 2010. [8] M. Uzunoglu and M. Alam, “Dynamic modeling, design, and simulation of a combined PEM fuel cell and ultracapacitor system for stand-alone residential applications,” IEEE Trans. Energy Convers., vol. 21, no. 3, pp. 767–775, Sep. 2006. [9] W. Li and X. He, “Review of nonisolated high-step-up dc-dc converters in photovoltaic grid-connected applications,” IEEE Trans. Ind. Electron., vol. 58, no. 4, pp. 1239–1250, Apr. 2011.

159

[10] F. Tofoli, D. de Souza Oliveira, R. Torrico-Bascope, and Y. Alcazar, “Novel nonisolated high-voltage gain dc-dc converters based on 3SSC and VMC,” IEEE Trans. Power Electron., vol. 27, no. 9, pp. 3897–3907, Sep. 2012. [11] G. Silveira, F. Tofoli, L. Bezerra, and R. Torrico-Bascope, “A nonisolated dc-dc boost converter with high voltage gain and balanced output voltage,” IEEE Trans. Ind. Electron., vol. 61, no. 12, pp. 6739–6746, Dec. 2014. [12] X. Hu and C. Gong, “A high voltage gain dc-dc converter integrating coupled-inductor and diode capacitor techniques,” IEEE Trans. Power Electron., vol. 29, no. 2, pp. 789–800, Feb. 2014. [13] Z. Guo, D. sha, X. Liao, and J. Luo, “Input-series-output-parallel phase shift full bridge derived dc-dc converters with auxiliary LC networks to achieve wide zero voltage switching range,” IEEE Trans. Power Electron., vol. 29, no. 10, pp. 5081–5086, Oct. 2014. [14] S.-K. Changchien, T.-J. Liang, J.-F. Chen, and L.-S. Yang, “Novel high step-up dc-dc converter for fuel cell energy conversion system,” IEEE Trans. Ind. Electron., vol. 57, no. 6, pp. 2007–2017, Jun. 2010. [15] N. Tummuru, Mahesh K. Mishra, and S. Srinivas, “Integration of PV/battery hybrid energy conversion system to the grid with power quality improvement features,” in Proc. IEEE Int. Conf. Ind. Technol., Jan. 2013, pp. 1751–1756. [16] A. Mohamed, V. Salehi, and O. Mohammed, “Real-time energy management algorithm for mitigation of pulse loads in hybrid microgrids,” IEEE Trans. Smart Grid, vol. 3, no. 4, pp. 1911–1922, Dec. 2012. [17] S. Teleke, M. Baran, S. Bhattacharya, and A. Huang, “Optimal control of battery energy storage for wind farm dispatching,” IEEE Trans. Energy Convers., vol. 25, no. 3, pp. 787–794, Sep. 2010. [18] B. Hredzak, V. Agelidis, and M. Jang, “A model predictive control system for a hybrid battery-ultracapacitor power source,” IEEE Trans. Power Electron., vol. 29, no. 3, pp. 1469–1479, Mar. 2014. [19] A. Melero-Perez, W. Gao, and J. Fernandez-Lozano, “Fuzzy logic energy management strategy for fuel cell/ultracapacitor/battery hybrid vehicle with multiple-input dc-dc converter,” in Proc. IEEE Vehicle Power Propul. Conf., Dec. 2009, pp. 199–206. [20] S. Teleke, M. Baran, S. Bhattacharya, and A. Huang, “Rule-based control of battery energy storage for dispatching intermittent renewable sources,” IEEE Trans. Sustain. Energy, vol. 1, no. 3, pp. 117–124, Oct. 2010. [21] L. Xu and D. Chen, “Control and operation of a dc microgrid with variable generation and energy storage,” IEEE Trans. Power Del., vol. 26, no. 4, pp. 2513–2522, Oct. 2011. [22] L. Solero, A. Lidozzi, and J. Pomilio, “Design of multiple-input power converter for hybrid vehicles,” IEEE Trans. Power Electron., vol. 20, no. 5, pp. 1007–1016, Sep. 2005. [23] A. Haruni, M. Negnevitsky, M. Haque, and A. Gargoom, “A novel operation and control strategy for a standalone hybrid renewable power system,” IEEE Trans. Sustain. Energy, vol. 4, no. 2, pp. 402–413, Apr. 2013.

Narsa Reddy Tummuru (S’12) received the Bachelor’s degree from Jawaharlal Nehru Technological University, Hyderabad, India, in 2002, and the Master’s of Technology degree from the Indian Institute of Technology, Delhi, India, in 2006. He is currently working toward the Ph.D. degree with the Indian Institute of Technology Madras, Chennai, India. His research interests include power electronic converter applications in microgrid and renewable energy systems, power quality, and control of switchmode power converters.

160


Mahesh K. Mishra (S’00–M’02–SM’10) received the B.Tech. degree in electrical engineering from the College of Technology, Pantnagar, India, in 1991; the M.E. degree in electrical engineering from the University of Roorkee, Roorkee, India, in 1993; and the Ph.D. degree in electrical engineering from the Indian Institute of Technology, Kanpur, India, in 2002. He has teaching and research experience of about 23 years. For about ten years, he was with the Electrical Engineering Department, Visvesvaraya National Institute of Technology, Nagpur, India. He is currently a Professor in the Electrical Engineering Department, Indian Institute of Technology Madras, Chennai, India. His research interests include the areas of power distribution systems, power electronic applications in microgrid, and renewable energy systems. Dr. Mahesh is Life Member of the Indian Society of Technical Education.

S. Srinivas (M’12) received the Bachelor’s degree in electrical engineering from the University College of Engineering, Osmania University, Hyderabad, India, in 1997, and the Master’s degree in electrical engineering with specialization in electrical machines and industrial drives and the Ph.D. degree in electrical engineering from the National Institute of Technology, Warangal, India, in 2002 and 2008, respectively. From 1997 to 2008, he was with the Faculty of Electrical Engineering, National Institute of Technology Warangal. Since 2008, he has been with the Department of Electrical Engineering, Indian Institute of Technology Madras, Chennai, India, where he is currently an Associate Professor. His research interests include multilevel inverters, dc and ac drives, and power electronic applications in renewable energy systems and distributed energy systems. Dr. Srinivas received the Best paper Award at the 2011 IEEE Power Electronic and Drive Systems and Technologies Conference held in Tehran, Iran.