Multisource Coordination Energy Management Strategy ... - IEEE Xplore

296

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 67, NO. 1, JANUARY 2018

Multisource Coordination Energy Management Strategy Based on SOC Consensus for a PEMFC–Battery–Supercapacitor Hybrid Tramway Ying Han

, Student Member, IEEE, Qi Li , Senior Member, IEEE, Tianhong Wang, Student Member, IEEE, Weirong Chen, Senior Member, IEEE, and Lei Ma, Member, IEEE

Abstract—For the sake of coordinating multiple energy sources appropriately from power demand and guarantee stage of charge (SOC) consensus of the energy storage systems in different operation conditions, a multisource coordination energy management strategy based on self-convergence droop control is proposed for a large-scale and high-power hybrid tramway. A hybrid powertrain configuration that includes multiple proton exchange membrane fuel cell systems, batteries, and supercapacitors is designed for a 100% low-floor light rail vehicle (LF-LRV) tramway. According to the hybrid system model of LF-LRV tramway developed with commercial equipment, this proposed multisource coordination energy management strategy is assessed with a real driving cycle of tramway. The results obtained from RT-LAB platform testify that the proposed strategy is capable of coordinating multiple energy sources, guaranteeing the SOC consensus and improving the efficiency of overall tramway. Index Terms—Energy management system, hybrid tramway, multi-source coordination, proton exchange membrane fuel cell, self-convergence droop control, stage of charge consensus.

I. INTRODUCTION T PRESENT, there is a significant motivation to accelerate the utilization of renewable energy and promote the industrial conversion toward a sustainable transportation [1]. As one of the most clean energy sources, fuel cells which convert chemical energy of the fuel into electricity are characterized by high energy conversion efficiency and almost zero emissions for transportation applications [2], [3]. In order to reduce dependence on fossil fuels and promote hydrogen economy development, a proton exchange membrane fuel cell (PEMFC)

A

Manuscript received March 10, 2017; revised June 17, 2017 and August 12, 2017; accepted August 22, 2017. Date of publication August 30, 2017; date of current version January 15, 2018. This work was supported in part by the National Natural Science Foundation of China under Grant 61473238 and Grant 51407146 and in part by the Sichuan Provincial Youth Science and Technology Fund under Grant 2015JQ0016. The review of this paper was coordinated by the Guest Editors of the VPPC 2016 Special Section. (Corresponding author: Qi Li.) The authors are with the School of Electrical Engineering, Southwest Jiaotong University, Chengdu 610031, China (e-mail: [email protected]. edu.cn; [email protected]; [email protected]; wrchen@swjtu. edu.cn; [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/TVT.2017.2747135

is considered as one of the most main candidates to develop the locomotives and the tramways [4], [5]. Contrasted with the catenary-electric and the diesel-electric types, the tramways and the locomotives powered by the PEMFC and the energy storage system (ESS) display expansive application potential [6], [7]. In spite of the PEMFC shows great energy capability under stable operation, the dynamic performance and the lifetime of PEMFC from being utilized in large-scale and high-power transportation applications may be dramatically influenced on transient peak power demand limitation and the fast power demand variations [8], [9]. In addition, the energy from regenerative braking of a vehicle is unable to be recovered by the PEMFC. Therefore, the PEMFC is never utilized alone to satisfy the demand power, and the ESS serving as auxiliary energy source is required to compensate the transient output power and recover the braking energy [10]. Normally, the battery has higher specific energy than the supercapacitor (SC) as the ESS and thus is able to supply extra power for a longer period of time. The SC that possesses a higher specific power than the battery shows a longer lifetime in terms of number of charge/discharge cycles, and enhances the overall vehicle dynamic response [11]. The task of the energy management is to coordinate the power flow between the PEMFC and the ESS. Pablo Garc´ıa et al. [12], [13] have proposed an energy management system based on the equivalent consumption minimization strategy to minimize the fuel consumption, and also have fulfilled a comparison to choose an appropriate control method for the FC/battery/SC vehicles. The hybrid system of tramway proposed in these literatures gives the inspiration of the design of hybrid topology for this paper. Souleman Njoya Motapon et al. [14], [15] have presented a H2 -consumption-minimization energy management strategy for a more electric aircraft hybrid power system, and also have achieved a comparative analysis of different energy management schemes for this hybrid power system. Liangfei Xu et al. [16], [17] have fulfilled a pontryagin’s minimal energy management strategy to optimize the operation cost for a fuel cell/battery city bus, and also have used an adaptive supervisory control strategy to achieve the road conditions for this kind of city bus. Wu et al. [18] have presented an adaptive control strategy based on parameter estimations to manage the power sharing

0018-9545 © 2017 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.

HAN et al.: MULTISOURCE COORDINATION ENERGY MANAGEMENT STRATEGY BASED ON SOC CONSENSUS

in this hybrid power system and satisfy the system constraints. Changjun Xie et al. [19] have implemented a power management to achieve the hydrogen consumption minimization for the fuel cell/battery hybrid powertrain. Torreglosa et al. [20] have presented an equivalent consumption minimization strategy for a hybrid tramway powered by the fuel cell and the battery. Luis M. Fernandez et al. [21] have proposed a hybrid tramway based on the fuel cell and the battery, and have presented a state machine control to offer power demand. Erdinc et al. [22] have proposed a load sharing algorithm based fuzzy logic control for a fuel cell/SC hybrid system. Vural et al. [23] have developed an energy management strategy based on wavelet transform and fuzzy logic control for a fuel cell/SC hybrid systems. Qi Li et al. [6], [24] have presented a power sharing strategy and a state machine strategy for PEMFC/battery/SC hybrid tramway. The energy management strategies proposed in the above literatures give lots of light on improving and optimizing the multi-source energy management strategy for this paper. Nevertheless, the hybrid tramway consisted of multiple motor and trailer units is powered by multiple PEMFCs and ESSs based on the requirement of the power level and the mounting space. In order to achieve the power balance between the load power demand and the energy sources, the power flows are managed by the energy management strategy which decides the power distribution between the PEMFCs and the ESSs. Therefore, a proper energy management strategy that fulfill the power distribution are important for the hybrid propulsion system of the tramway. The energy management strategies mentioned above have been verified effectively to achieve power assignment for small or medium power level of hybrid system. However, the existing energy management strategies are not considered carefully to manage multiple motor and trailer units of hybrid tramway for large-scale and high-power transportation applications. In particular, multiple PEMFCs and ESSs on the motor and trailer units will try to realize the requested voltages by injecting or absorbing active power. In the face of such situation, traditional energy management strategies will be embarrassed to guarantee the DC bus voltage stabilization, and reduce the circulating currents which flow in the power sources and the converters [25]–[28]. Moreover, the proposed energy strategies have not adequately considered the state of charge (SOC) consensus of ESSs in different operation conditions. If one certain ESS is overused, the life cycle of the ESS is cut down and the efficiency of whole hybrid system is also decreased [26]–[28]. To mitigate these problem, according to the natural characteristics of the energy sources and fulfilling high efficiency without degrading the mechanism performance, a suitable energy management strategy based on self-convergence droop control approach which should coordinate multiple motor and trailer units, guarantee the SOC consensus of ESSs and prolong the lifetime of the hybrid system is quite essential for satisfying the load requests from a large-scale and high-power tramway. In this paper, a hybrid propulsion system which consists of two PEMFC systems, batteries and SCs is designed for 100% Low-floor Light Rail Vehicle (LF-LRV) tramway, and then a multi-source coordination energy management strategy based

Fig. 1.

100% LF-LRV hybrid tramway.

Fig. 2.

Structure of hybrid tramway powered by PEMFC-battery-SC.

297

on self-convergence droop control is proposed to coordinate the power demand to each power source appropriately and guarantee SOC consensus of batteries and SCs for the energy management system. Furthermore, according to the hybrid system model of tramway which is developed with commercial equipment, this proposed energy management strategy is assessed with a real driving cycle on RT-LAB which is useful engineering design platform for the large-scale and high-power applications. Finally, the comparisons of performance testing with other control strategies are fulfilled to certify the rationality and validity of the proposed method. This paper is organized as follows. Section II is dedicated to the configuration of hybrid LF-LRV tramway. Modeling of hybrid power system of tramway is brief explained in Section III. The proposed energy management strategy is introduced in Section IV. Section V details results and discussions. Finally, the main conclusions are drawn in Section VI. II. CONFIGURATION OF HYBRID LF-LRV TRAMWAY A 100% LF-LRV hybrid tramway composed of two motor units and one trailer unit is being developed without grid connection by Chinese manufacturer of Tangshan Railway Vehicle Co. Ltd and Southwest Jiaotong University as shown in Fig. 1. The proposed PEMFC-battery-SC tramway configuration including two PEMFC systems (No.1 and No.2), two SCs (No.1 and No.2), two batteries (No.1 and No.2), four bidirectional DC/DC converters, two unidirectional DC/DC converters, auxiliary service module and braking resistor is shown in Fig. 2 and the detailed description of the tramway configuration can be found in [7], [24].

298


SimPowerSystems Toolbox of Simulink. The output voltage and SOC of the battery and the SC can be obtained as follows [16], [20], [35]–[38] Ubat = E − Rb ibat = E0 − K

Q it Q − it

+ Ab exp (−B · it) − P olres · i∗ − Rb ibat QT − Rsc isc CT 1 t SOCB = 100 · 1 − ibat (t)dt Q 0

t Qinit − 0 isc (t)dt SOCSC = 100 · QT Usc =

Fig. 3.

Diagram of PEMFC-battery-SC hybrid system power flow.

III. MODELING OF HYBRID SYSTEM FOR THE TRAMWAY A. Modeling of PEMFC Power Unit With Ballard 150 kW HD6 Module A PEMFC power unit that composes of a PEMFC stack module, an air delivery module, and a cooling module is considered as the core of the hybrid power system for LF-LRV tramway [7], [24]. This designed PEMFC power unit is setup in the clean energy laboratory of Southwest Jiaotong University. A Ballard HD6 Module (FC-velocity) is utilized as the PEMFC stack module rated at 150 kW gross power. It contains the auxiliary components for air delivery and humidification, hydrogen recirculation and purge, and water-cooling circulation. The schematic diagram of PEMFC power unit with HD6 Module is shown in Fig. 4. A model of PEMFC power unit with the testing data of 150 kW HD6 Module is developed. The output voltage equation of PEMFC is expressed as [29]–[34] Ufc = Eo c − Uact − Uohm ic with

⎧ −44.43 ⎪ ⎪ E = K oc c Ee + (T − 298) ⎪ ⎪ zF ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ RT 1/2 ⎪ ⎨ ln PH 2 PO 2 + zF ⎪

⎪ ⎪ ⎪ ifc 1 ⎪ ⎪ · N Anom ln Uact = ⎪ ⎪ τs + 1 i0 ⎪ ⎪ ⎪ ⎪ ⎩ Uohm ic = Rinternal · ifc

(1)

(2)

The detailed description about the model, parameters and experimental verification of the PEMFC power unit can be found in [7], [24]. B. Modeling of Energy Store Systems (ESSs) In this paper, the Li-ion batteries and the SCs are adopted both for providing a portion of the base load together with the PEMFC systems and capturing the braking energy. The behaviors of the batteries and the SCs are respectively represented by the modified Shepherd curve-fitting model and the Stern model presented in [35]–[38], which are available in the

(3) (4) (5) (6)

The detailed description about the model, parameters and experimental verification of the battery and the SC can be found in [10], [16], [20], [24]. In addition, The PEMFC, battery and SC present a lower terminal voltage than the DC voltage necessary to feed the traction inverter. In this work, two unidirectional boost DC/DC converter have been developed by using the two-quadrant chopper models included in SimPowerSystems of Simulink [10], [16], [17]. In these models, the converters are represented by their average value equivalent, which are composed of the controlled current source at the DC bus side and the controlled voltage source at the PEMFC side. The average value models are less time-consuming by controlled voltage and current sources and all the converter dynamics are maintained, which make these models attractive as larger sampling time can be used [17]. And also, two bidirectional converters are respectively utilized to the batteries and the SCs with boost operation if discharging and buck operation if charging. These converter models also been developed by using the two-quadrant chopper models included in SimPowerSystems [16], [17]. The batteries and the SCs are located on the low voltage side, and the high voltage side is connected to the 750VDC bus. IV. MULTI-SOURCE COORDINATION ENERGY MANAGEMENT STRATEGY BASED ON SELF-CONVERGENCE DROOP CONTROL The PEMFC systems (No.1 and No.2), the batteries and the SCs (No.1 and No.2) on two motor units and one trailer unit undergo different operation conditions to drive the tramway and recover the braking energy. During the operation of hybrid tramway, the energy management system is responsible for distributing the power and guaranteeing the SOC consensus of ESSs. While one certain ESS is overused, the lifetime of the ESS is shorten and the efficiency of whole hybrid system is also decreased. Thus, the SOC balancing and injected or output power equalization should be realized in both charging and discharging processes [25]–[28]. In this work, a self-convergence droop control method is implemented to coordinate and manage multiple ESSs, which is responsible for deciding the reference voltage command Udcref i of DC/DC converter i (i = 1, 2, . . . , n) to match the load demand from a driving cycle of tramway. This characteristic is utilized to convert the output power of DC/DC


Fig. 4.

299

Schematic of PEMFC power unit with 150 kW HD6 module.

converter i into the reference voltage command Udcref i . The conventional droop curve which represents the discharging process in the first quadrant is capable of being extended to acquire the double-quadrant droop curve that represents the charging process. The reference ESS voltage droop can be represented as charging Udcref − λi Po ci (7) Udcref i = Udcref − λi Po di discharging where, Udcref i is the reference voltage of DC/DC converter i which is the same of all the converters, λi is the droop coefficient, Udcref is the reference of DC bus voltage, Po ci and Po di is the charging and discharging power of DC/DC converter i, respectively. For sake of satisfying the requirement of SOC balancing, λi in the charging and discharging processes is expressed as charging λc SOCim (8) λi = m discharging λd / SOCi where, λc and λd are the droop coefficients for charging and discharging processes if SOCi equals 100%, m is the exponent of SOC, which is involved to regulate the speed of SOC balancing. The double-quadrant droop characteristic of reference voltage versus injected/output power is shown in Fig. 5. In the charging process, the injected power (absolute value) with higher SOC of ESS is lower, and the injected power (absolute value) with lower SOC of ESS is higher. In the discharging process, with higher SOC the output power is higher, and with lower SOC the output power is lower. Specially, the idle ESS which is fully discharged is restarted and λi for each ESS is set to be proportional to SOCm in the charging process. If the ESS reaches its maximum acceptable SOC during charging, it is disconnected to wait for the next discharging process. The idle ESS which is fully charged is restarted and λi is set to be inversely proportional to the SOCm in the discharging process. If the ESS reaches its minimum SOC, it is disconnected to wait for the next charging process [26], [28]. For sake of avoiding the injected or output power is out of the maximum acceptable power range, a saturation limiter is utilized [28]. Particularly, if the injected or output power

Fig. 5. Double-quadrant droop characteristic of reference voltage versus injected/output power.

exceeds the upper restriction Po ci or Po di , it is restricted to the maximum power Pm c or Pm d , and then the objective can be obtained by involving a power-controlled loop. In the starting process, the hybrid system works in the voltage-controlled loop and the self-convergence droop control approach is implemented. While the injected or output power is out of the upper restriction Po ci or Po di , the operation loop of the corresponding converter is switched from the voltage-controlled loop to the power-controlled loop. Simultaneously, the reference value of the power-controlled loop is set to be the maximum acceptable value Pm c or Pm d . If the injected or output power is lower than the upper restriction, the operation loop of the related converter is switched back from the power-controlled loop to the voltage-controlled loop. For distributing the reference power reasonably, three operation modes are defined to dispatch the reference power signals for the PEMFC systems through adjusting two unidirectional DC/DC converters, and the charging and discharging processes of the batteries and the SCs are based on the selfconvergence droop control strategy through regulating four bidirectional DC/DC converters. The specification of the operation modes is according to the range [SOCBm in , SOCBm ax ] of the

300

Fig. 6.


Flow chart of the operation principle for hybrid system.

batteries SOCB and the range [SOCSCm in , SOCSCm ax ] of the SCs SOCSC , which depends on the designer’s knowledge about the power sources and traction devices constraints, and practical aspects regarding the dynamic behavior of the hybrid propulsion system. Hence, three operation modes of the PEMFC systems are defined as follows. Mode 1 (SOCB i ≥ SOCBm ax || SOCSC i ≥ SOCSCm ax ): Pfcref i = ⎧ ⎨ Pfcm ax Pd /n ⎩ Pfcm in

if Pd ≥ nPfcm ax + Paux if nPfcm ax > Pd > nPfcm in + Paux if nPfcm in + Paux ≥ Pd

Mode 2 (SOCBm in < SOCB < SOCSC i < SOCSCm ax ): Pfcref i = ⎧ Pfcm ax ⎪ ⎪ ⎪ ⎨ P /n d ⎪ P fcopt ⎪ ⎪ ⎩ Pfcm in

(9)

i

< SOCBm ax || SOCSCm in

if Pd ≥ nPfcm ax + Paux Pfcm ax > Pd > nPfcopt + Paux (10) if nPfcopt + Paux ≥ Pd > nPfcm in + Paux if nPfcm in + Paux ≥ Pd

Mode 3 (SOCB Pfcref i = ⎧ Pfcm ax ⎪ ⎪ ⎨ Pd /n Pfcopt ⎪ ⎪ ⎩ Pfcm in + Pchar

i

≤ SOCBm in || SOCSC

i

≤ SOCSCm in ):

if Pd ≥ nPfcm ax + Paux if nPfcm ax > Pd > nPfcopt + Paux if nPfcopt + Paux ≥ Pd > nPfcm in + Paux if nPfcm in + Paux ≥ Pd (11)

where, Pd is the demand power, Pfcref i denotes the reference power of i–th PEMFC system, Pfcm in and Pfcm ax are the lower and upper restrictions of PEMFC system, Paux is the auxiliary services power of tramway, Pchar is the minimum charging power of battery and SC, Pfcopt is the optimal power of PEMFC system. The flow chart of the above operation principle is shown in Fig. 6. As one operation mode is triggered, the PEMFC systems and the ESSs obtain their own power commands sets according to a amount of imported or exported power, and the SOC consensus of batteries and SCs is guaranteed by using the selfconvergence droop control method. The required power demand from a driving cycle is able to be satisfied accurately for the individual power source by using the proposed energy management strategy, and this method has the advantage of being able to enforce the SOC balance of each ESS while adopting the


Fig. 7.

301

Energy management scheme with the control loops.

double-quadrant droop curve. And also, the SOCB i and SOCSC i are ensured within an appropriate range as the PEMFC systems offer power flow in a suitable power changing rate supplied. Hence, the high efficiency of the whole systems of the tramway are realized and the lifetime of the each power sources is prolonged. The control loops of proposed energy management strategy are shown in Fig. 7. V. RESULTS AND DISCUSSIONS In this study, a real round-trip driving cycle of LF-LRV tramway from Turkey as shown in Fig. 8 is used to evaluate the feasibility of the multi-source coordination energy management strategy based on self-convergence droop control. For sizing the hybrid system, the maximum power of two PEMFC systems ought to be higher than the average power demanded during the driving cycle for the sake of avoiding excessive SOC drop of the ESSs. In Fig. 8, the average power of driving requirement is 160.2 kW so that the PEMFC rated power is chosen the value of 300 kW to prevent great demand from the ESSs. Because of the peak power requirement is 716.8 kW, it needs a 416.8 kW peak power complementally from the ESSs. In addition, due to the average power and the peak power of regenerative braking is 129.9 kW and 495.7 kW, the SC is adopted to enhance the dynamic response of the hybrid system. Hence, a hybrid system taken into account for the LF-LRV tramway consists of two PEMFC systems (300 kW), two Li-ion batteries (40 Ah) and two SC banks (45 F). All of the initial conditions and control

Fig. 8.

Driving cycle of LF-LRV tramway in Turkey.

parameters are shown in Table I. Specially, the SOC of No.1 and No.2 batteries and No.1 and No.2 SCs set as different initial values to verify the validity of the proposed energy management strategy. The initial SOCB1 and SOCB2 values of batteries are 75% and 55%, the initial SOCSC1 and SOCSC2 values of SCs are 76% and 48%, respectively. RT-LAB is a new kind of engineering design application platform based on the MATLAB/Simulink model exploited by Canada Opal-RT Technologies. The mathematical MATLAB/Simulink model of dynamic system can be applied directly

302

Fig. 9.


RT-LAB real-time simulation platform. TABLE I INITIAL CONDITIONS AND CONTROL PARAMETERS Parameters

Values

Parameters

Values

U d c re f mB mS C P fc m in SOC B 1 SOC B 2 SOC S C 1 SOC S C 2

750 V 6 4 10 kW 75% 55% 76% 48%

λc λd n

0.03 0.001 2 150 kW 40% 85% 20% 95%

P fc m a x SOC B m in SOC B m a x SOC S C m in SOC S C m a x

to control field, test filed and other related fields. Furthermore, the design of project through the real-time simulation, rapid prototyping and hardware in the loop can all be realized on this platform. In this work, the experimental platform is set up based on the RT-LAB as shown in Fig. 9 to verify the effectiveness and accuracy of the theoretical analysis. The experimental platform is composed of the RT-LAB simulator OP5600, DSP TMS320F28335, and the computer as a real-time control interface. Considering the expensive experimental cost and the testing safety of high power level equipment, the simulation model of large-scale hybrid system for the tramway as shown in Fig. 3 is implemented in RT-LAB real-time simulation platform to validate the proposed energy management strategy. As the results of the hybrid system by using the proposed multi-source coordination energy management strategy, the output power, voltage, and current of two PEMFC systems are illustrated in Fig. 10, the power, voltage, current and SOC of two batteries are shown in Figs. 11 and 12, the power, voltage, current and SOC of two SCs are presented in Figs. 13 and 14, respectively. During the driving cycle, all the dynamic output performance of No.1 and No.2 PEMFC systems are similar because of the sharing concept for prolonging the lifetime, and two PEMFC systems just increases or decreases the output power according to high accelerations or brakings as shown in Fig. 10. The output power of two batteries and two SCs alternates between negative and positive according to charging and discharging process. Two batteries help supply a portion of the positive low frequency components of demand power to relieve

Fig. 10.

Output power, voltage, and current of PEMFC systems.

the stress of the PEMFC systems, and absorb the slow-variation negative portion as shown in Fig. 11. In particular, the SOC of No.1 and No.2 batteries with the proposed strategy are gradually balanced during charging and discharging processes as shown in Fig. 12. Furthermore, two SCs afford the transient power demand which is unable to be satisfied with the PEMFC dynamic response during sudden acceleration or braking as shown in Fig. 13. Their rapid response is able to achieve the transient power demand, and supply or absorb the power that either the PEMFC or the battery can not provide or absorb. Similarly, the balanced SOC of No.1 and No.2 SCs with the proposed strategy are gradually achieved as shown in Fig. 14. The overall performance of different strategies based on the comparison criteria are summarized in Table II. The comparisons with a fuzzy logic control (FLC) [7], an equivalent consumption minimization strategy (ECMS) [11], and the proposed energy management strategy (MCS-DDC) are implemented under the same the driving cycle and hybrid topology. The performance criteria are the average efficiency of tramway, the average value of SOCB and SOCSC , the end values of SOCB 1 , SOCB 2 , SOCSC 1 , and SOCSC 2 , and total energy dissipated in the braking resistor. In Table II, the best value


Fig. 14. Fig. 11.

Output power, voltage, and current of batteries.

Fig. 12.

Waveforms of batteries SOC.

Fig. 13.

Output power, voltage, and current of SCs.

303

Waveforms of SCs SOC.

TABLE II RESULTS OF DIFFERENT STRATEGIES DURING THE DRIVING CYCLE

realized is denoted in italic and bold. The average efficiency of tramway is calculated as follows

dt t PHT

(12) ηHT = t Pfcs dt + t Pbat dt + t Psc dt + t Pbr dt where, PHT is the traction power of the hybrid tramway, Pfcs is total output power of the PEMFC systems including the power demanded by the auxiliary components, Pbr is the power dissipated in the braking resistor, Pbat is the total batteries output power, Psc is the total SCs output power. Compared with other strategies, the MCS-DDC provides better tramway average efficiency (50.40%), the equivalent hydrogen consumptions (836.40 g), the average SOC of the batteries and SCs (67.39% and 60.61%, respectively) and the energy dissipation (0.83 kWh). Furthermore, the highest SOC end values of No.2 battery and No.2 SC (66.91% and 58.70%, respectively) are achieved with the MCS-DDC to guarantee the SOC consensus of two batteries and two SCs. Nevertheless, the highest SOC end values of No.1 battery and No.1 SC (74.54% and 73.07%) are realized with the ECMS at the expense of neglecting the SOC consensus of two batteries and two SCs. To conclude, due to employ the self-convergence droop control approach, the

304


MCS-DDC which supplies the highest tramway average efficiency and the average SOC of the ESSs, and the lowest equivalent hydrogen consumptions and energy dissipation, is more suitable for the hybrid tramway application. As a matter of fact, the MCS-DDC effectively achieves to coordinate and manage multiple motor units and trailer units (multiple power sources) by satisfying the power demand based on the characteristics of each power source, enhancing the efficiency of overall tramway hybrid system, and balancing the SOC of each ESS. VI. CONCLUSIONS In this paper, the hybrid propulsion system of the LF-LRV tramway on two motor units and one trailer unit is developed with two PEMFC systems, two batteries and two SCs. In order to coordinate power demand to multiple energy sources properly and guarantee SOC consensus of ESSs in different operation conditions, the multi-source coordination energy management strategy with self-convergence droop control is proposed for a large-scale and high-power hybrid tramway. This self-convergence droop control approach is capable of limiting the circulating currents, balancing the SOC of each ESS, and modifying the power injections or absorptions to fulfill the regulated voltages. Three operation modes are defined to distribute the reference power reasonably for two PEMFC systems, two batteries and two SCs through adjusting the unidirectional and bidirectional DC/DC converters. The results obtained from RT-LAB platform by using the real driving cycle verify that the proposed strategy is capable to warrant the steady operation of different power sources, and the SOC of ESSs are gradually balanced during charging and discharging processes. The comparisons with other strategies demonstrate that the highest tramway average efficiency and the average SOC of the ESSs, and the lowest energy dissipation are achieved by the proposed strategy. Hence, the proposed multi-source coordination strategy can coordinate different power sources, guarantee the SOC consensus, improve the efficiency of overall tramway, extend the operating life cycle of the hybrid system, and also will offer a novel method for the advanced energy management system of large-scale and highpower hybrid tramway. ACKNOWLEDGMENT The authors would like to thank the reviewers for their helpful suggestions. REFERENCES [1] O. Erdinc and M. Uzunoglu, “Recent trends in PEM fuel cell-powered hybrid systems: Investigation of application areas, design architectures and energy management approaches,” Renew. Sustain. Energy Rev., vol. 14, no. 9, pp. 2874–2884, Dec. 2010. [2] L. Cao, K. H. Loo, and Y. M. Lai, “Systematic derivation of a family of output-impedance shaping methods for power converters—A case study using fuel cell-battery-powered single-phase inverter system,” IEEE Trans. Power Electron., vol. 30, no. 10, pp. 5854–5869, Oct. 2015. [3] L. Cao, K. H. Loo, and Y. M. Lai, “Frequency-adaptive filtering of lowfrequency harmonic current in fuel cell power conditioning systems,” IEEE Trans. Power Electron., vol. 30, no. 4, pp. 1966–1978, Apr. 2015.

[4] F. Peng, W. Chen, Z. Liu, Q. Li, and C. Dai, “System development of 150 kW electric locomotive powered by individual PEM fuel cell stack,” Int. J. Hydrogen Energy, vol. 39, pp. 13886–13893, 2014. [5] A. R. Miller, K. S. Hess, D. L. Barnes, and T. L. Erickson, “System design of a large fuel cell hybrid locomotive,” J. Power Sources, vol. 173, pp. 935–942, 2007. [6] Q. Li, W. Chen, Z. Liu, M. Li, and L. Ma, “Development of energy management system based on a power sharing strategy for a fuel cell-batterysupercapacitor hybrid tramway,” J. Power Sources, vol. 279, pp. 267–280, 2015. [7] L. Hong et al., “A nonlinear control strategy for fuel delivery in PEM fuel cells considering nitrogen permeation,” Int. J. Hydrogen Energy, vol. 42, no. 2, pp. 1565–1576, Jan. 2017. [8] J. Liu et al., “PEM fuel cell air-feed system observer design for automotive applications: An adaptive numerical differentiation approach,” Int. J. Hydrogen Energy, vol. 39, no. 30, pp. 17210–17221, Oct. 2014. [9] J. Liu, W. Luo, X. Yang, and L. Wu, “Robust model-based fault diagnosis for PEM fuel cell air-feed system,” IEEE Trans. Ind. Electron., vol. 63, no. 5, pp. 3261–3269, May 2016. [10] J. Jia, G. Wang, Y. T. Cham, Y. Wang, and M. Han, “Electrical characteristic study of a hybrid PEMFC and ultracapacitor system,” IEEE Trans. Ind. Electron., vol. 57, no. 6, pp. 1945–1953, Jun. 2010. [11] W. Greenwell and A. Vahidi, “Predictive control of voltage and current in a fuel cell–ultracapacitor hybrid,” IEEE Trans. Ind. Electron., vol. 57, no. 6, pp. 1954–1963, Jun. 2010. [12] P. Garc´ıa, J. P. Torreglosa, L. M. Fernández, and F. Jurado, “Control strategies for high-power electric vehicles powered by hydrogen fuel cell, battery and supercapacitor,” Expert Syst. Appl., vol. 40, pp. 4791–4804, 2013. [13] P. Garc´ıa, J. P. Torreglosa, L. M. Fernández, and F. Jurado, “Viability study of a FC-battery-SC tramway controlled by equivalent consumption minimization strategy,” Int. J. Hydrogen Energy, vol. 37, pp. 9368–9382, 2012. [14] S. N. Motapon, L.-A. Dessaint, and K. Al-Haddad, “A comparative study of energy management schemes for a fuel-cell hybrid emergency power system of more-electric aircraft,” IEEE Trans. Ind. Electron., vol. 61, no. 4, pp. 1320–1334, Mar. 2014. [15] S. N. Motapon, L.-A. Dessaint, and K. Al-Haddad, “A robust H2consumption-minimization-based energy management strategy for a fuel cell hybrid emergency power system of more electric aircraft,” IEEE Trans. Ind. Electron., vol. 61, no. 11, pp. 6148–6156, Nov. 2014. [16] L. Xu, M. Ouyang, J. Li, F. Yang, L. Lu, and J. Hua, “Application of Pontryagin’s minimal principle to the energy management strategy of plugin fuel cell electric vehicles,” Int. J. Hydrogen Energy, vol. 38, pp. 10104– 10115, 2013. [17] L. Xu, J. Li, J. Hua, X. Li, and M. Ouyang, “Adaptive supervisory control strategy of a fuel cell/battery-powered city bus,” J. Power Sources, vol. 194, pp. 360–368, 2009. [18] C. Wu, J. Chen, C. Cu, and Z. Liu, “Real-time adaptive control of a fuel cell-battery hybrid power system with guaranteed stability,” IEEE Trans. Control Syst. Technol., vol. 25, no. 4, pp. 1394–1405, Jul. 2017. [19] C. Xie, J. M. Ogden, S. Quan, and Q. Chen, “Optimal power management for fuel cell–battery full hybrid powertrain on a test station,” J. Electr. Power Energy Syst., vol. 53, pp. 307–20, 2013. [20] J. P. Torreglosa, F. Jurado, P. Garc´ıa, and L. M. Fernández, “Hybrid fuel cell and battery tramway control based on an equivalent consumption minimization strategy,” Control Eng. Pract., vol. 19, pp. 1182–1194, 2011. [21] L. M. Fernandez, P. Garcia, C. A. Garcia, and F. Jurado, “Hybrid electric system based on fuel cell and battery and integrating a single dc/dc converter for a tramway,” Energy Convers. Manage., vol. 52, pp. 2183–2192, 2011. [22] O. Erdinc, B. Vural, M. Uzunoglu, and Y. Ates, “Modeling and analysis of an FC/UC hybrid vehicular power system using a wavelet-fuzzy logic based load sharing and control algorithm,” Int. J. Hydrogen Energy, vol. 34, pp. 5223–5233, 2009. [23] B. Vural et al., “Fuel cell and ultra-capacitor hybridization: A prototype test bench based analysis of different energy management strategies for vehicular applications,” Int. J. Hydrogen Energy, no. 35, pp. 11161–11171, 2010. [24] Q. Li, H. Yang, Y. Han, M. Li, and W. Chen, “A state machine strategy based on droop control for an energy management system of PEMFC-battery-supercapacitor hybrid tramway%,” Int. J. Hydrogen Energy, vol. 41, pp. 16148–16159, 2016.


[25] T. L. Vandoorn, B. Meersman, J. D. M. De Kooning, and L. Vandevelde, “Analogy between conventional grid control and islanded microgrid control based on a global DC-link voltage droop,” IEEE Trans. Power Del., vol. 27, no. 3, pp. 278–286, Jul. 2012. [26] C. Jin, P. Wang, J. Xiao, Y. Tang, and F. Choo, “Implementation of hierarchical control in DC microgrids,” IEEE Trans. Ind. Electron., vol. 61, no. 8, pp. 4033–4042, Aug. 2014. [27] J. Xiao, P. Wang, L. Setyawan, and Q. Xu, “Multi-level energy management system for real-time scheduling of DC microgrids with multiple slack terminals,” IEEE Trans. Energy Convers., vol. 31, no. 1, pp. 392– 400, Mar. 2016. [28] X. Lu, K. Sun, J. M. Guerrero, J. C. Vasquez, and L. Huang, “Doublequadrant state-of-charge-based droop control method for distributed energy storage systems in autonomous DC microgrids,” IEEE Trans. Smart Grid, vol. 6, no. 1, pp. 147–157, Jan. 2015. [29] Q. Li, W. Chen, Y. Wang, S. Liu, and J. Jia, “Parameter identification for PEM fuel cell mechanism model based on effective informed adaptive particle swarm optimization,” IEEE Trans. Ind. Electron., vol. 58, no. 6, pp. 2410–2419, Jun. 2011. [30] Y.-X. Wang, F.-F. Qin, K. Ou, and Y.-B. Kim, “Temperature control for a polymer electrolyte membrane fuel cell by using fuzzy rule,” IEEE Trans. Energy Convers., vol. 31, no. 2, pp. 667–675, Jan. 2016. [31] J. T. Pukrushpan, A. G. Stefanopoulou, and H. Peng, “Control of fuel cell breathing,” IEEE Control Syst. Mag., vol. 24, no. 2, pp. 30–46, Apr. 2004. [32] C. Ramos-Paja, C. Bordons, A. Romero, R. Giral, and L. Mart´ınezSalamero, “Minimum fuel consumption strategy for PEM fuel cells,” IEEE Trans. Ind. Electron., vol. 56, no. 3, pp. 685–696, Mar. 2009. [33] A. Arce, A. J. del Real, C. Bordons, and D. R. Ramirez, “Real-time implementation of a constrained MPC for efficient airflow control in a PEM fuel cell,” IEEE Trans. Ind. Electron., vol. 57, no. 6, pp. 1892–1905, Jun. 2010. [34] Q. Li, W. Chen, Z. Liu, J. Huang, and L. Ma, “Net power control based on linear matrix inequality for proton exchange membrane fuel cell system,” IEEE Trans. Energy Convers., vol. 29, no. 1, pp. 1–8, Mar. 2014. [35] T. C. Yang, “Initial study of using rechargeable batteries in wind power generation with variable speed induction generators,” IET Renew. Power Gener., vol. 2, pp. 89–101, 2008. [36] G. Sikha, R. E. White, and B. N. Popov, “A mathematical model for a lithium-ion battery/electrochemical capacitor hybrid system,” J. Electrochem. Soc., vol. 152, no. 8, pp. 1682–1693, 2005. [37] K. B. Oldham, “A Gouy-Chapman-Stern model of the double layer at a (metal)/(ionic liquid) interface,” J. Electroanal. Chem., vol. 613, no. 2, pp. 131–138, Feb. 2008. [38] N. Xu and J. Riley, “Nonlinear analysis of a classical system: The doublelayer capacitor,” Electrochem. Commun., vol. 13, pp. 1077–1081, 2011.

Ying Han (S’15) received the B.S. degree in electrical engineering, in 2013, from Southwest Jiaotong University, Chengdu, China, where he is currently working toward the Ph.D. degree in the School of Electrical Engineering. His research interests include hybrid system modeling and energy management strategy, and microgird stability control.

305

Qi Li (M’12–SM’15) received the B.S. and Ph.D. degrees from the Electrical Engineering School, Southwest Jiaotong University, Chengdu, China, in 2006 and 2011, respectively. From 2009 and 2011, he conducted research as a Visiting Scholar in the School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. He is currenty a Professor in the School of Electrical Engineering, Southwest Jiaotong University. His research interests include fuel cell locomotives optimal control, energy management strategy of hybrid system, and power system stability and control.

Tianhong Wang (S’17) received the B.S. degree in electrical engineering, in 2016, from Southwest Jiaotong University, Chengdu, China, where he is currently working toward the Ph.D. degree in the School of Electrical Engineering. His research interests include fuel cell modeling and optimal control, dc/dc converter control, and energy management strategy of a hybrid system.

Weirong Chen (M’99–SM’16) received the B.S. and M.S. degrees in electronic engineering from Electronic Science and Technology University, Chengdu, China, in 1985 and 1988, respectively, and the Ph.D. degree in power system and its automation from Southwest Jiaotong University, Chengdu, China, in 1998. In 1999, he was a Senior Visiting Scholar at Brunel University, London, U.K. He is currently a Professor in the School of Electrical Engineering, Southwest Jiaotong University. He has published more than 120 refereed journals and conference papers, six books, and is the holder of more than 40 Chinese patents. His research interests include renewable energy and its applications, fuel cell locomotive technology, and power system control. He an IET Fellow.

Lei Ma (M’08) received the B.S. degree in automatic control from Chongqing University, Chongqing, China, in 1993, the M.S degree in electrical engineering from Southwest Jiaotong University, Chengdu, China, in 1996, and the Doktor Ingenieur degree from Ruhr-University Bochum, Bochum, Germany, in 2006. He is currently a Professor in the School of Electrical Engineering, Southwest Jiaotong University. His research interests include control theory and its applications in electromachanical systems, robot control, and fuel cell control.