Hydrogen economy of the fuel cell hybrid power system optimized by air flow control to mitigate the effect of the uncertainty about available renewable power and load dynamics

Hydrogen Economy of the Fuel Cell Hybrid Power System optimized by air flow control to mitigate the effect of the uncertainty about available renewable power and load dynamics Nicu Bizon 1) Faculty of Electronics, Communication and Computers, University of Pitesti, 1 Targu din Vale, 110040, Pitesti, Romania 2) University Politehnica of Bucharest, 313 Splaiul Independentei, 060042, Bucharest, Romania [email protected]om, ORCID: 0000-0001-9311-7598 Jose Manuel Lopez-Guede Systems Engineering and Automatic Control Department, Faculty of Engineering Vitoria-Gasteiz, University of the Basque Country (UPV/EHU), Vitoria, Spain, [email protected], ORCID: 00000002-5310-1601 Erol Kurt Gazi University, Technology Faculty, Department of Electrical and Electronics Engineering, 06500 Teknikokullar, Ankara, Turkey, [email protected], ORCID: 0000-0002-3615-6926 1 2 Nomenclature AirFr AV EMS ESS FuelFr FC Air Flow rate Average value Energy Management Strategy Energy Storage System Fuel Flow rate Fuel cell This is the accepted manuscript of the article that appeared in final form in Energy Conversion and Management 179 : 152-165 (2019) , which has been published in final form at https://doi.org/10.1016/j.enconman.2018.10.058. © 2018 Elsevier under CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) GES HPS LF MV PEMFC RTO RES sFF Global Extremum Seeking Hybrid Power System Load Following Mean Value Proton Exchange Membrane Fuel Real-Time Optimization Renewable Energies Source Static Feed-Forward 3 Abstract: A new Energy Management Strategy to reduce the hydrogen consumption is proposed 4 for Hybrid Power Systems based on Proton Exchange Membrane Fuel Cell system used as backup 5 source. The Energy Management Strategy uses a Load Following control loop of requested load 6 demand on DC bus and an optimization control loop to improve the fuel economy based on the 7 Global Extremum Seeking algorithm applied to the air flow rate. The performance of proposed 8 strategy is compared to the one obtained with the Static Feed-Forward strategy considering three 9 case studies for the optimization function used in different scenarios for power flow on DC bus 10 (variable or constant load demand, without or with variable renewable energy power). 11 Keywords: Proton Exchange Membrane Fuel Cell; Hydrogen Economy; Power Variability 12 Mitigation; Air Flow Control; Global Extremum Seeking; Load Following 13 14 1. INTRODUCTION15 In the last years Renewable Energies Sources (RESs) have experienced a rapid development [1] and 16 these (especially wind and solar energy [2]) are now recognized as important energy sources for 17 micro-grids [3]. Furthermore, worldwide laws have been adopted to encourage the use of renewable 18 energies [4,5]. RESs used in smart grids is important for eco-friendly development in all countries 19 [6,7]. 20 In addition to the most used renewable energies (solar and wind energy), other resources are used as 21 well (such as geothermal, biomass and biogas energies) along with backup sources (such as the 22 Proton Exchange Membrane Fuel Cell (PEMFC) or the diesel generator) and Energy Storage 23 Systems (ESSs) in order to mitigate the variability of the RES power flow [8,9]. A Hybrid Power 24 System (HPS) always uses two or more energy sources to sustain the power demand based on 25 appropriate Energy Management Strategy (EMS) designed to manage the power flows on DC bus 26 [10]. If for each region the RES are known and chosen according to the requirements of the power 27 demand, the use of fossil fuels will be considerably reduced [7-9,11]. 28 RESs are very attractive for the investors but the RES power flow (PRES) and load demand (Pload) 29 are variable and difficult to be predicted [12]. So, HPS optimal operation must be ensured with an 30 appropriate EMS design [10,13]. The RES power flow depends on weather conditions and load 31 demand on many factors, so ESSs store the excess energy or supply the needed power due to lack of 32 RES power on DC bus [14,15]. The EMS will motorize the charging ESS status in order to 33 guarantee the HPS operation under high dynamic load profiles [16,17]. The power flow requested 34 on DC bus (PDCreq) must follow the power flow Pload - PRES using the Load Following (LF) control 35 [15,18,19]. During light load stages, when Pload < PRES, the excess of power PRES - Pload will be 36 stored or used to supply an electrolyzer [16,17]. Many optimization techniques of the HPS 37 operation were proposed in the literature [20-24] using different algorithms [25], including the 38 Global Extremum Seeking (GES) algorithm [26-30]. EMSs must be implemented using Real-Time 39 Optimization (RTO) loops [6,15,16,31]. 40 In this paper the RTO1 strategy is proposed based on LF control of the boost converter and GES 41 optimization of the Air Flow rate (AirFr) of the PEMFC system, which is used here as 42 environmental-friendly backup source. 43 In the last decade, the environmental-friendly backup sources such as PEMFC systems are preferred 44 to non-renewable energy backup sources such as diesel generators due to some disadvantages of the 45 last: high price for maintenance and the pollutant emissions [32]. Furthermore, the hydrogen that is 46 supplied to the PEMFC could be generated renewably using an electrolyzer [33-35], as it was 47 mentioned above. 48 When two or more energy sources and loads are involved in the HPS topology, the EMS is required 49 to be appropriately designed based on state-flow diagrams and RTO algorithms. Different EMSs 50 have been proposed in the literature that can ensure the load demand and safe operation of the HPS 51 [36-44]. 52 The RTO1-based EMS has supplementary only the RTO control loop compared to the Static Feed-53 Forward (sFF) strategy [44], which is most know and already commercially implemented. Note that 54 both RTO1 and sFF strategies use the LF control loop based on PDCreq = Pload - PRES and the control 55 of the Fuel Flow rate (FuelFr) based on FC current. The RTO control loop uses the GES algorithm 56 to find the global maximum of the optimization function due to high performance (high accuracy, 57 speed, and searching resolution) reported in [25-29]. The topological differences between other 58 RTO strategies proposed in the literature are mentioned in Table 1. 59 60 Table 1. The topological differences between the RTO strategies 61 62 The mitigation of the uncertainty of available power from renewable sources and the variability of 63 load demand is analyzed in this paper based on LF control of the boost converter. The hydrogen 64 economy is obtained based on optimized control applied to air flow regulator. The performance of 65 the RTO1-based EMS is validated using different scenarios for power flow on DC bus (variable or 66 constant load demand, without or with variable renewable energy power). 67 Thus, the structure of the paper is as follows. EMSs are presented and compared in Section 2. The 68 proposed HPS is modeled in Section 3 and the RTO1-based EMS is presented in Section 4. The 69 results are presented and discussed in Section 5. Last section concludes the paper. 70 71 2. ENERGY MANAGEMENT STRATEGIES FOR HYBRID POWER SYSTEM 72 The necessity for EMSs is not only for standalone hybrid system but also for systems connected to 73 the power grid. For standalone HPSs, EMSs ensure the continuous power supply of the load 74 regardless of the conditions, the use of all renewable resources to the fullest reducing production 75 costs to minimize the cost of energy and increasing the system reliability. In addition, in the case of 76 HPSs connected to the power grid, EMSs ensure the control of energy flow from the grid to 77 consumers and vice versa (from the hybrid system to the grid), in addition to measure parameters to 78 minimize the costs [45,46]. 79 There are in the literature a number of review papers about energy management strategies 80 [5,6,10,17,26,38,39,47]. For example, in [47] there are presented the principal configurations, usual 81 sizing methodologies and main EMSs used for HPSs, which can be classified in centralized (Figure 82 1), distributed (Figure 2) and hybrid (Figure 3) EMSs. The main goals for any strategy are to meet 83 the demands of consumers, extract the maximum amount of energy from each renewable source, 84 minimize the cost of energy, and reduce the number of loading and unloading cycles for ESSs. 85 It can be observed for centralized EMSs (Figure 1) that two communications are present for each 86 component of the system: one to transfer data from the center to the components and another for the 87 reverse flux. So, centralized EMSs are not effective in smart micro-grids, because future topologies 88 are always subject to changes, resulting in low stability of the system [48]. Also, when a system has 89 many components, this is not a cost efficient solution due to the high number of communications 90 channels between components. However, centralized methods cannot be yet replaced by other new 91 methods proposed in the literature of distributed or hybrid types [49]. 92 In distributed EMSs (Figure 2) each component has its own local controller and each component 93 provides and uses an appropriate set of measurements (currents and voltages), fact that implies 94 more correct decisions taken by the controller. The advantage of distributed strategies is the 95 increased stability in comparison with the centralized ones and the ability to protect data from the 96 components [50]. 97 In comparison with distributed EMSs (where components provides/uses a set of measurements 98 without sharing information with other components of the system), hybrid EMSs (Figure 3) allow 99 their components to share information between them, resulting in high flexibility and stability of 100 these micro-grids. The advantage of these strategies it is its use in the future smart grids due to the 101 fact that new elements can be added without affecting the EMS [51]. 102 EMSs must be designed considering the topology of the HPSs [52-54]. An algorithm to predict 103 sizes of hybrid sources and an EMS to share power between the PEMFC system and ESS is 104 proposed in [52] for automotive applications. A review regarding modeling, energy flow 105 management and control strategies of RES HPSs is presented in [53]. The issues related to HPSs 106 configurations and energy management using PV panels, wind turbines, hydro-power stations and 107 fuel cells are revised and some solutions are proposed to overcome these problems. So, in the next 108 section, the proposed HPS is presented in the frame of this classification proposed for HPSs. 109 110 Figure 1. Centralized EMSs. Figure 4. The DC bus HPS topology. Figure 2. Distributed EMSs. Figure 5. The AC bus HPS topology. Figure 3. Hybrid EMSs. Figure 6. The hybrid bus HPS topology. 111 3. PROPOSED PEMFC HYBRID POWER SYSTEM TOPOLOGY 112 The classification of HPSs can be made depending on the type of common bus used [53-58]. The 113 most known HPS topologies are the DC bus HPS topology (Figure 4), the AC bus HPS topology 114 (Figure 5) and hybrid bus HPS topology (Figure 6). In each topology, the DC or AC bus is used to 115 connect the energy sources and the load (and the grid if the HPS topology is grid connected) using 116 appropriate power converters. The EMSs control the power flow of the available energy sources to 117 ensure the load demand based on the power flow balance on common bus [57]. The advantages and 118 disadvantages of the DC bus topology compared with the AC bus topology are revised in [58]. 119 Anyway, a disadvantage of all topologies is clearly represented by the increased number of 120 conversion stages, but this could be reduced by the hybrid bus topology that uses bidirectional 121 converters [54-56]. However, the EMSs of this hybrid bus topology are more complex due to the 122 energy management of both buses and the control of bidirectional converter between them [57,58]. 123 A bidirectional hybrid bus HPS topology is proposed in [58] to simplify the energy management 124 between the DC and AC buses based on bidirectional DC-AC converters (Figure 6). 125 The HPS topology considered in this study is reduced to a DC bus modeling the other parts of the 126 HPS with an equivalent DC load (Figure 7). So, the inverter system (grid-connected or not) is 127 modeled on DC bus with an equivalent load with dynamic profile as in reality. Choosing a specific 128 HPS topology depends on several factors such as the RES variability, load demand dynamics and 129 economic constraints. 130 The RTO1-based HPS topology is obtained with the switch on the GES position, as it is shown in 131 Figure 7, when the GES optimization control is applied to AirFr regulator and the FuelFr regulator 132 is controlled by the FC current. The sFF strategy (used as reference to compare the results obtained 133 with the RTO1 strategy) will be obtained if the switch is moved to the sFF position and both AirFr 134 and FuelFr variables are controlled by the FC current [34]. The LF control is applied to boost 135 converter in both RTO1 and sFF strategies. The 6 kW / 45 V PEMFC model from the library of 136 Matlab-Simulink (with the FC time constant set to 0.1 s) is used. The FC voltage is boosted to 137 VDC  VDC(ref) = 200 V using a boost converter. The boost controller is of hysteretic type, having as 138 inputs the FC current (IFC) and the reference current (IrefLF). The reference current IrefLF is generated 139 by LF control block, as will be detailed in next section. 140 141 Figure 8. The LF control block. 142 Figure 7. The RTO1-based HPS topology. 143 Figure 9. The GES control block. 144 4. MODELING OF HYBRID POWER SYSTEM 145 The power flow balance on DC bus HPS shown in Figure 7 is given by (1): 146 CDC  udc  dudc/dt =  boost  pFC+ pESS - pLoad (1) where CDC is the capacitor on DC bus (which is used to filter the voltage on DC bus, udc),  boost is 147 the energy efficiency of the boost converter (set to 95%) and pFC, pESS , and pLoad are the FC net 148 power, the ESS power, and the load demand requested on DC bus respectively. 149 The average value (AV) of power flow balance of (1) is given by (2): 150 0 =  boost PFC(AV)+ PESS(AV) - PLoad(AV) (2) where  boost is the energy efficiency of the boost converter, which was set at 95%. 151 So, since that the LF control of the power requested on DC bus (PDCreq(AV)  PDC(AV) = PLoad(AV) – 152 PRES(AV)) is implemented for the FC system, the FC power will follow the average value of needed 153 energy on DC bus (PDCreq(AV)): 154 PFC(AV)= PDCreq(AV) /  boost (3) The LF reference current will be: 155 Iref(LF) =IFC(AV)= PDCreq(AV) / (VFC(AV  boost) (4) Thus, the AV of the ESS power will be almost zero (considering (1-4)): 156 PESS(AV)  0 (5) Consequently, the ESS size can be reduced at minimum. 157 The LF control block (Figure 8) implements (3), with PDCreq(AV) obtained by Mean Value (MV) of 158 the pDCreq power given by Eq. (6), but other filtering techniques could be used as well [38]: 159 pDCreq = pload (6) The GES control block (Figure 9) implements the relationships (6) [6]: 160   12 ,y f v v , N Ny y k y (6a) h f h N f y y y       , HPF N f y y y , l BPF l HPF BPF y y y       (6b) , sin( ) DM BPF d d y y s s t     , (6c) DM Int yy  (6d) 1 , d MV MV BPF d G y y y dt T    (6e) Md yG (6f) 1 1 1 , Int sd p k y k      (6g) 22Md p k y s   (6h) 3md p A s (6i)   1 2 3refGES Np I k p p p    , (6j) 161 where (6a) to (6f) represent the optimization function, the input normalization gain (kNy), the high-162 pass filter (HPF) and the band-pass filter (BPF), the demodulation, the integration, the computing of 163 the dither gain (Gd) based on AV of the ybpf signal, and the signal that will modulate the dither [6]. 164 The components p1, p2, and p3 of the searching signal (p) are given by (6g) to (6i), and the reference 165 current IrefGES by (6j), where and kNp is the output normalization gain [6]. 166 The switch selects the reference currents for Iref(O2) as IrefGES and Iref(H2) as IFC in the RTO1, while 167 Iref(H2) = Iref(O2) = IFC in sFF strategy. 168 The normalization gains are kNy= 1/YMax and kNp= IFC(rated) / 2, where IFC(rated) and YMax are the 169 nominal value of the FC current and the estimation of maximum value of the optimization function 170 respectively. The tuning parameters are designed based on [59,60]. Apart from other optimization 171 techniques proposed in the literature [26-33,61-66], the optimization function f used here is defined 172 to increase the FC system energy efficiency and reduce the total fuel consumption by maximizing 173 (7a): 174 ( , , , ) net FCnet fuel eff Load k P k Fuel f x AirFr FuelFr P    (7a) subject to dynamics given by Eq. [7b]: (7b) 6. CONCLUSION 306 In this last section, the main findings of this study are summarized as follows: 307  Both strategies (sFF and RTO1) use the LF control of the FC boost converter so that the FC 308 power ensures the load demand on DC bus, operating the battery in charge sustained mode; 309  The RTO1-based EMS proposed here has better performance compared with that of sFF 310 strategy by adding only a optimization loop of the AirFr regulator, which is implemented in 311 real-time by using the GES algorithm; 312  The performance indicators such as the FC energy efficiency, the fuel efficiency, and the 313 fuel economy could increase up to 2.65 %, 10.35 W/lpm, and 11.8 l for the HPS under 8 kW 314 load; 315  The fuel economy could be increased 2-times using kfuel=25 in the optimization function 316 defined as a mix relationship of the FC net power and the fuel efficiency (see Table 5 for 317 kfuel=0 and kfuel=25); 318  The fuel economy is obtained on the full range of load demand, constant or variable, by 319 using the RTO1 strategy with kfuel0 (see Table 5 and Table 7); 320  The effect of variability of the load demand dynamics and uncertainty of available power 321 from renewable sources can be mitigated by LF control of the FC power generated, which 322 will follow the power demand requested on DC bus; 323 The next work will be focused on operation of the HPS with variable RES power and dynamic load 324 profile that will use an electrolyzer supplied with energy in excess (PRES(AV) - Pload) in order to 325 maintain the charge sustained mode for the battery and at least the performances reported in this 326 paper. 327 328 ACKNOWLEDGEMENTS 329 This work was supported by Research Center “Modeling and Simulation of the Systems and 330 Processes” based on grants of the Ministry of National Education and Scientific Research, 331 CNCS/CCCDI-UEFISCDI within PNCDI III “Experimental validation of a propulsion system with 332 hydrogen fuel cell for a light vehicle - Mobility with Hydrogen Demonstrator”, 53PED, ID: PN-III 333 P2-2.1-PED-2016-1223, and within RDI Program for Space Technology and Advanced Research - 334 STAR, project number 167/2017: “Concept Development of an Energy Storage Unit Using High 335 Temperature Superconducting Coil for Spacecraft Power Systems (SMESinSpace). 336 337 REFERENCES 338 [1] Perera ATD, Attalage RA, Perera KKCK, Dassanayake VPC. Designing standalone hybrid 339 energy systems minimizing initial investment, life cycle cost and pollutant emission. Energy 340 2013;54(0):220–30. 341 [2] Bandara K, Sweet T, Ekanayake J. Photovoltaic applications for off-grid electrification using 342 novel multi-level inverter technology with energy storage. Renew Energy 2011;37(1):82–8. 343 [3] Moghavvemi M, Ismail MS, Murali B, Yang SS, Attaran A, Moghavvemi S. Development and 344 optimization of a PV/diesel hybrid supply system for remote controlled commercial large scale 345 FM transmitters. Energy Convers Manag 2013;75:542–51. 346 [4] Balcombe P, Rigby D, Azapagic A. Motivations and barriers associated with adopting 347 microgeneration energy technologies in the UK. Renew Sustain Energy Rev 2013;22(0):655–348 66. 349 [5] Meyar-Naimi H, Vaez-Zadeh S. Sustainable development based energy policy making 350 frameworks, a critical review. Energy Policy 2012;43(0):351–61. 351 [6] Bizon N, Thounthong P. Real-time strategies to optimize the fueling of the fuel cell hybrid 352 power source: A review of issues, challenges and a new approach. Renew Sustain Energy Rev 353 2018;91:1089–1102. 354 [7] Ismail MS, Moghavvemi M, Mahlia TMI. Energy trends in Palestinian territories of West Bank 355 and Gaza Strip: possibilities for reducing the reliance on external energy sources. Renew 356 Sustain Energy Rev 2013;28(0):117–29. 357 [8] Hoicka CE, Rowlands IH. Solar and wind resource complementarity: advancing options for 358 renewable electricity integration in Ontario, Canada. Renew Energy 2011;36(1):97–107. 359 [9] Mekhilef S, Faramarzi SZ, Saidur R, Salam Z. The application of solar technologies for 360 sustainable development of agricultural sector. Renew Sustain Energy Rev 2013;18(0):583–94.] 361 [10] Olatomiwa L, Mekhilef S, Ismail MS, Moghavvemi M. Energy management strategies in 362 hybrid renewable energy systems: A review. Renew Sustain Energy Rev 2016; 62:821–835. 363 [11] Gupta SK, Purohit P. Renewable energy certificate mechanism in India: a preliminary 364 assessment. Renew Sustain Energy Rev 2013;22(0):380–92. 365 [12] Vergara-Dietrich JD, et al. Advanced chance-constrained predictive control for the efficient 366 energy management of renewable power systems, J Process Control (2017), 367 https://doi.org/10.1016/j.jprocont.2017.11.003] 368 [13] Bizon N. Optimal Operation of Fuel Cell / Wind Turbine Hybrid Power System under 369 Turbulent Wind and Variable Load. Appl Energy 2018;212:196-209 370 https://doi.org/10.1016/j.apenergy.2017.12.034 371 [14] Cristian N, Bizon N, Alexandru A. Design of hybrid power systems using HOMER 372 simulator for different renewable energy sources. ECAI - 9th Edition of International 373 Conference on Electronics, Computers and Artificial Intelligence. Targoviste, ROMÂNIA, 29 374 June -01 July, 2017 10.1109/ECAI.2017.8166507 375 [15] Bizon N. Real-time optimization strategy for fuel cell hybrid power sources with load-376 following control of the fuel or air flow. Energy Convers Manage 2018;157:13–27; 377 https://doi.org/10.1016/j.enconman.2017.11.084 378 [16] Bizon N, Mazare AG, Ionescu LM, Enescu FM. Optimization of the Proton Exchange 379 Membrane Fuel Cell Hybrid Power System for Residential Buildings. Energy Convers Manage 380 2018;163:22–37; 381 [17] Chauhan A, Saini RP. A review on Integrated Renewable Energy System based power 382 generation for stand-alone applications: configurations, storage options, sizing methodologies 383 and control. Renew Sustain Energy Rev 2014;38:99–120. 384 [18] Bizon N, Radut M, Oproescu M. Energy control strategies for the Fuel Cell Hybrid Power 385 Source under unknown load profile. Energy 2015;86:31-41, 386 http://dx.doi.org/10.1016/j.energy.2015.03.118 387 [19] Bizon N. Load-following Mode Control of a Standalone Renewable/Fuel Cell Hybrid Power 388 Source. Energy Convers Manage 2014;77:763–772 389 http://dx.doi.org/10.1016/j.enconman.2013.10.035 390 [20] Trifkovic M, Sheikhzadeh M, Nigim K, Daoutidis P. Modeling and control of a renewable 391 hybrid energy system with hydrogen storage. IEEE Trans Control Syst Technol 2014;22(1): 169 392 - 179. https://ieeexplore.ieee.org/document/6477101/ 393 [21] Bizon N. Nonlinear control of fuel cell hybrid power sources: Part II –Current control. Appl 394 Energ 88(7) (2011) 2574–2591 http://dx.doi.org/10.1016/j.apenergy.2011.01.044 395 [22] Hatti M, Meharrar A, Tioursi M. Power management strategy in the alternative energy 396 photovoltaic/PEM fuel cell hybrid system. Renew Sustain Energy Rev 2011;15(9):5104–10. 397 [23] Bizon N. Tracking the maximum efficiency point for the FC system based on extremum 398 seeking scheme to control the air flow. Appl Energ 2014;129:147–157 399 http://dx.doi.org/10.1016/j.apenergy.2014.05.002 400 [24] Uzunoglu M, Onar OC, Alam MS. Modeling, control and simulation of a PV/FC/UC based 401 hybrid power generation system for stand-alone applications. Renew Energy 2009;34(3):509–402 20. 403 [25] Sikkabut S, Mungporn P, Ekkaravarodome C, Bizon N, et. al. Control of High-Energy High-404 Power Densities Storage Devices by Li-ion Battery and Supercapacitor for Fuel 405 Cell/Photovoltaic Hybrid Power Plant for Autonomous System Applications. IEEE T Ind Appl 406 2016 52(5):4395-4407, 10.1109/TIA.2016.2581138. 407 [26] Bizon N. Global Maximum Power Point Tracking (GMPPT) of Photovoltaic array using the 408 Extremum Seeking Control (ESC): A review and a new GMPPT ESC scheme. Renew Sustain 409 Energy Rev 2016;57:524–539 doi:10.1016/j.rser.2015.12.22 410 [27] Bizon N. Global Maximum Power Point Tracking based on new Extremum Seeking Control 411 scheme. Prog Photovoltaics: Research and Applications 2016;24(5):600-22. DOI: 412 10.1002/pip.2700. 413 [28] Bizon N. Searching of the Extreme Points on Photovoltaic Patterns using a new 414 Asymptotic Perturbed Extremum Seeking Control scheme. Energy Convers Manage 415 2017;144:286–302; https://doi.org/10.1016/j.enconman.2017.04.065 15 July 2017. 416 [29] Bizon N. Energy optimization of Fuel Cell System by using Global Extremum Seeking 417 algorithm. Appl Energ 2017;206:458-474; https://doi.org/10.1016/j.apenergy.2017.08.097 418 [30] Bizon N. Global Extremum Seeking Control of the Power Generated by a 419 Photovoltaic Array under Partially Shaded Conditions. Energy Convers Manage 2016;109; 420 71-85 (February 2016), http://dx.doi.org/10.1016/j.enconman.2015.11.046 421 [31] Dagdougui H, Minciardi R, Ouammi A, Robba M, Sacile R. A dynamic decision model for 422 the real-time control of hybrid renewable energy production systems. IEEE Syst J 423 2010;4(3):323–33. http://dx.doi.org/10.1109/JSYST.2010.2059150. 424 [32] Valverde L, Ali D, and Gordo R. A technical evaluation of Wind-Hydrogen (WH) 425 demonstration projects in Europe. In: Proceedings of power engineering, energy and electrical 426 drives (POWERENG) 2013 10.1109/PowerEng.2013.6635764 427 [33] Dufo-López R, Bernal-Agustín JL, Contreras J. Optimization of control strategies for stand-428 alone renewable energy systems with hydrogen storage. Renew Energy 2007;32(7):1102–26. 429 [34] Vasallo MJ, Bravo JM, Andújar JM. Optimal sizing for UPS systems based on batteries 430 and/or fuel cell. Appl Energy 2013;105:170–81. 431 [35] Segura F, Durán E, Andújar JM. Design, building and testing of a standalone fuel cell hybrid 432 system. J Power Sources 2009;193(1):276–84. 433 [36] Semaoui S, Hadj Arab a, Bacha S, Azoui B. The new strategy of energy management for a 434 photovoltaic system without extra intended for remote-housing. Sol Energy 2013;94:71–85. 435 [37] Bizon N, Oproescu M, Raceanu M. Efficient Energy Control Strategies for a Standalone 436 Renewable/Fuel Cell Hybrid Power Source. Energy Convers Manage 2015;90:93-110 437 doi:10.1016/j.enconman.2014.11.002. 438 [38] Bizon N, Tabatabaei NM, Blaabjerg F, Kurt E (Ed.), Energy Harvesting and Energy 439 Efficiency: Technology, Methods and Applications, Springer Verlag London Limited, 2017. 440 [39] Bizon N, Tabatabaei NM, Shayeghi H (Ed.), Analysis, Control and Optimal Operations in 441 Hybrid Power Systems - Advanced Techniques and Applications for Linear and Nonlinear 442 Systems, Springer UK, 2013. 443 [40] Bizon N, Culcer M, Iliescu, Mazare AG, Laurentiu IM, Beloiu R. Real-time strategy to 444 optimize the Fuel Flow rate of Fuel Cell Hybrid Power Source under variable load cycle. ECAI 445 - 9th Edition of International Conference on Electronics, Computers and Artificial Intelligence. 446 Targoviste, ROMÂNIA, 2017 10.1109/ECAI.2017.8166513 447 [41] Bizon N, Culcer M, Oproescu M, Iana G, Laurentiu IM, Mazare AG, Iliescu M. Real-time 448 strategy to optimize the Airflow rate of Fuel Cell Hybrid Power Source under variable load 449 cycle, The 22rd International Conference on Applied Electronics - APPEL 2017, Pilsen, 2017 450 10.23919/AE.2017.8053577 451 [42] Mehne HH. Evaluation of Parallelism in Ant Colony Optimization Method for Numerical 452 Solution of Optimal Control Problems. Journal of Electrical Engineering, Electronics, Control 453 and Computer Science - JEEECCS 2015 1(2):15-20; 454 http://jeeeccs.net/index.php/journal/article/view/17 455 [43] Hoarca IC, Raducu M. On the micro-inverter performance based on three MPPT controllers. 456 Journal of Electrical Engineering, Electronics, Control and Computer Science - JEEECCS 2015 457 1(1):7-14; available on http://jeeeccs.net/index.php/journal/article/view/9 458 [44] Pukrushpan JT, Stefanopoulou AG, Peng H. Control of fuel cell power systems. New York: 459 Springer; 2004. 460 [45] Valenciaga F. and Pablo F. Puleston and Pedro E. Battaiotto. Supervisor Control for a 461 Stand-Alone Hybrid Generation System Using Wind and Photovoltaic Energy. IEEE Trans 462 Energy Convers 2005;20(2):398-405. 463 [46] Motaz Amer M, Naamane A, M’Sirdi KN. Optimization of Hybrid Renewable Energy 464 Systems (HRES) Using PSO for Cost Reduction. Energy Procedia 2013;42:318-327. 465 [47] Chauhan A, Saini RP. A review on integrated renewable energy system based power 466 generation for stand-alone applications: configurations, storage options, sizing methodologies 467 and control. Renew Sustain Energy Rev 2014;38:99–120 468 [48] Yang S, Tan S, Xu J-X. Consensus Based Approach for Economic Dispatch Problem in a 469 Smart Grid. IEEE Transactions on Power Systems 2013;28(4):4416 – 4426. 470 [49] Binetti G, Naso D, Turchiano B, Davoudi A, Frank L. Lewis.Consensus-based Approach for 471 the Economic Dispatch Problem. IFAC Proceedings Volumes 2014;47(3):3140-3145. 472 [50] He Y, Venkatesh B, Guan L. Optimal scheduling for charging and discharging of electric 473 vehicles. IEEE Trans Smart Grid 2012;3(3):1095–1105. 474 [51] Guo Y, Xiong J, Xu S, Su W. Two-Stage Economic Operation of Microgrid-Like Electric 475 Vehicle Parking Deck. IEEE Transactions on Smart Grid 2016;7(3):1703 – 1712. 476 [52] Bendjedia B, Rizoug N, Boukhnifer M, Bouchafaa F. Hybrid Fuel Cell/Battery Source 477 Sizing and Energy Management for Automotive Applications. IFAC PapersOnLine 2017; 478 50(1):4745–475. 479 [53] Nehrir MH, Wang C, Strunz K, Aki H, Ramakumar R, Bing J, et al. A review of hybrid 480 renewable/alternative energy systems for electric power generation: configurations, control, and 481 applications. IEEE Trans Sustain Energy 2011;2(4):392–403. 482 [54] Bizon N. Energy efficiency for the multiport power converters architectures of series and 483 parallel hybrid power source type used in plug-in/V2G fuel cell vehicles. Appl Energ 484 2013;102:726-734 http://dx.doi.org/10.1016/j.apenergy.2012.08.021 485 [55] Bizon N. Energy Efficiency of Multiport Power Converters used in Plug-In/V2G Fuel Cell 486 Vehicles, Appl Energ 2012;96: 431-443 http://dx.doi.org/10.1016/j.apenergy.2012.02.075 487 [56] Mihaescu M. Applications of multiport converters. Journal of Electrical Engineering, 488 Electronics, Control and Computer Science - JEEECCS 2016 2(1):13-18; available on 489 http://jeeeccs.net/index.php/journal/article/view/25 490 [57] Tabatabaei NM, Bizon N, Aghbolaghi AJ, Blaabjerg F (Ed.), Fundamentals and 491 Contemporary Issues of Reactive Power Control in AC Power Systems, Springer Verlag 492 London Limited, 2017. 493 [58] Chauhan A, Saini RP. A review on Integrated Renewable Energy System based power 494 generation for stand-alone applications: configurations, storage options, sizing methodologies 495 and control. Renew Sustain Energy Rev 2014;38:99–120. 496 [59] Bizon N, Kurt E. Performance Analysis of Tracking of the Global Extreme on Multimodal 497 Patterns using the Asymptotic Perturbed Extremum Seeking Control Scheme. Int J Hydrogen 498 Energ 2017; 42(28):17645-54, http://dx.doi.org/10.1016/j.ijhydene.2016.11.173 499 [60] Bizon N, Thounthong P, Raducu M, Constantinescu LM. Designing and Modelling of the 500 Asymptotic Perturbed Extremum Seeking Control Scheme for Tracking the Global Extreme. Int 501 J Hydrogen Energ 2017; 42(28):17632-44, http://dx.doi.org/10.1016/j.ijhydene.2017.01.086 502 [61] opezuede M, Ramos-Hernanz A, arra aga M, Garmendia A, Ionescu VM. Study on 503 the influence of Lambda parameter on several performance indexes in Dynamic Matrix Control. 504 Journal of Electrical Engineering, Electronics, Control and Computer Science - JEEECCS 2015 505 2(1):1-8; available on https://jeeeccs.net/index.php/journal/article/view/15 506 [62] Erdinç O, Vural B, Uzunoglu M. A wavelet-fuzzy logic based energy management strategy 507 for a fuel cell/battery/ultra-capacitor hybrid vehicular power system, J Power Sources 508 2009;194:369-80. 509 [63] Naciri M, Aggour M, Ahmed WA. Wind energy storage by pumped hydro station / Pages: 510 32-42 Journal of Energy Systems (JES) 2017;1(1):32-42, 511 http://dergipark.gov.tr/jes/issue/30882/329315 512 [64] Uğurlu A, ökçöl C. A case study of PV-wind-diesel-battery hybrid system. Journal of 513 Energy Systems (JES) 2017;1(4):138-147, http://dergipark.gov.tr/jes/issue/33528/348335 514 [65] Santos-Herrero JM, Lopez-Guede JM, Ivan Flores I. A Short review on the use of renewable 515 energies and model predictive control in buildings. Journal of Energy Systems (JES) 2017;1(3) 516 112-119, http://dergipark.gov.tr/jes/issue/32837/346653 517 [66] Ramos-Paja CA, Spagnuolo G, Petrone G, Emilio Mamarelis M. A perturbation strategy for 518 fuel consumption minimization in polymer electrolyte membrane fuel cells: Analysis, Design 519 and FPGA implementation. Appl Energ 2014;119:21-32. 520 521 522 Table 1. The topological differences between the RTO strategies No. Iref(Boost) Iref(Air) Iref(Fuel) RTO strategy Ref 0 ILF IFC IFC sFF [44] 1 ILF IGES+IFC IFC RTO1 [41] 2 ILF IFC IGES+IFC RTO2 [40] 3 IGES ILF IFC RTO3 [29] 4 IGES IFC ILF RTO4 [15] Table 2. The sFF strategy applied to FCHPS at different Pload Pload PFCnet0 sys0 Fueleff0 FuelT0 [kW] [W] [%] [W/lpm] [l] 2 1942 93.26 137.3 34.02 3 2884 91.85 129.5 56.3 4 3786 90.43 121.6 74.88 5 4650 88.75 113.4 98.6 6 5467 86.89 104.7 125.58 7 6229 84.78 95.16 158.34 8 6912 82.3 83.75 176 Table 3. The RTO1 strategy applied to FCHPS at different Pload Pload PFCnet1 sys1 Fueleff1 FuelT1 [kW] [W] [%] [W/lpm] [l] 2 2009 93.53 136.3 35.24 3 2960 92.27 128.8 56.43 4 3868 90.96 122.3 74.75 5 4751 89.36 115.2 98.22 6 5595 87.58 108.2 124.2 7 6504 85.69 101.37 154 8 7240 84.95 94.1 164.2 Table 4. The gaps in FC energy efficiency, fuel efficiency, and fuel economy for the RTO1 strategy compared to the sFF strategy Pload sys Fueleff FuelT [kW] [%] [W/lpm] [l] 2 0.27 -1 1.22 3 0.42 -0.7 0.13 4 0.53 0.7 -0.13 5 0.61 1.8 -0.38 6 0.69 3.5 -1.38 7 0.91 6.21 -4.34 8 2.65 10.35 -11.8 Table(s) Figure 13. The gap in total fuel consumption for Pload constant and different kfuel Figure 14. The gap in total fuel consumption for Pload variable and different kfuel Figure 15. The behavior of RTO1 strategy with PRES=0, Pload(AV)=4kW, and knet=0.5 W-1 and kfuel=50 W-1 lpm Figure 16. The behavior of gaps in performance indicators for the RTO1 strategy compared to sFF strategy Figure 17. The behavior of the RTO1-based HPS with PRES variable, Pload=6kW, and knet=0.5 W-1 and kfuel=50 W-1 lpm Figure 18. The behavior of RTO1-based HPS with variable profiles for PRES and Pload, and knet=0.5 W-1 and kfuel=50 W-1 lpm