This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/authorsrights
13
Embed
Comparative study of artificial intelligence-based building thermal control methods - Application of fuzzy, adaptive neuro-fuzzy inference system, and artificial neural network
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies areencouraged to visit:
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 9 9 7 3e9 9 8 4
http://dx.doi.org/10.1016/j.ijhydene.2014.04.1470360-3199/Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Author's personal copy
electrolyzer produces hydrogen by the excess electrical energy
of the PV and wind sources. The hydrogen can then be used to
supply an FCwhich is considered as a secondary power source
when the demand is high.
For the better understanding of the different aspects of
hydrogen-based hybrid systems, thereby to efficiently utilize
PV/wind/FC systems, various investigations have been
developed. In hybrid systems, appropriate sizing is one of the
most important issues that results in having a cost-effective
energy system. Literature study indicates that there are
many attempts based on probabilistic, analytical and heuristic
methods for optimal sizing of hybrid systems. Diaf et al. [18]
have optimized hybrid system size based on loss of power
supply probability (LPSP) and the levelized cost of energy
(LCE). Borowy and Salameh [19] have introduced loss of load
probability (LLP) concept for finding the optimal size of the PV/
wind hybrid system. Shrestha and Goel [20] have presented a
methodology for optimal sizing based on energy generation
simulation. Maghraby et al. [21] have used the desired system
performance level (SPL) requirement to select the number of
PVs and batteries. Energy balance has been used for design of
hybrid PV/wind systems [22]. Prasad and Natarajan [23] have
presented a methodology for optimization of PV/wind system
based on deficiency of power supply probability (DPSP), rela-
tive excess power generated (REPG), unutilized energy proba-
bility (UEP), life cycle cost (LCC), levelized energy cost (LEC)
and life cycle unit cost (LUC) of power generation with battery
bank. Nonlinear programming [24] and HOMER [25] are other
algorithms used for optimal design of hybrid systems. Heu-
ristic algorithms such as genetic algorithm (GA) [26,27],
(a)Wind Generator
PV Panel
Battery
+
DC/DC
Load
DC bus
DC/DC
AC/DC
DC/AC
(b)Wind Generator
PV Panel
Fuel cell
+
DC/DC
Load
DC bus
DC/DC
AC/DC
DC/AC
H2 Tanks
Electrolyzer
Fig. 1 e Schematic of the hybrid systems. (a) PV/wind/battery-based hybrid system and (b) PV/wind/FC-based hybrid
system.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 9 9 7 3e9 9 8 49974
tween the generated and demanded powers (DP) for the six
optimized systems and Fig. 11 shows the storage level of the
batteries and hydrogen tanks. For example, consider the PV/
WT/FC system. In this case, negative sign of DPmeans that the
generated power of PV and WT systems can not satisfy the
Fig. 5 e Break down of the total annual cost. (a) PV/wind/
FC; (b) PV/FC and (c) wind/FC.
Fig. 6 e Break down of the total annual cost. (a) PV/wind/
battery; (b) PV/battery and (c) wind/battery.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 9 9 7 3e9 9 8 4 9981
Author's personal copy
load. So, the FC starts to work and meets the remaining load
demand. For instance, at 9th hour,DP is negative (Fig. 10(a)). At
this time, form Fig. 11(a), it is seen that the storage level of
hydrogen tank is positive meaning that the storage device can
Fig. 7 e Convergence process of the algorithms for finding
the optimum size. (a) PV/wind/FC; (b) PV/wind/battery.
Fig. 8 e Convergence process of the algorithms for finding
the optimum size. (a) PV/FC; (b) PV/battery.
Fig. 9 e Convergence process of the algorithms for finding
the optimum size. (a) wind/FC; (b) wind/battery.
Table 4 e Comparison of the algorithms in terms of thecomputational cost (second).
Hybrid system Index Algorithm
PSO TS SA HS
PV/wind/FC Mean 0.192 0.047 0.082 0.082
Min 0.156 0.031 0.047 0.062
Max 0.250 0.094 0.140 0.109
PV/FC Mean 0.185 0.039 0.07 0.082
Min 0.172 0.016 0.062 0.047
Max 0.218 0.094 0.094 0.125
Wind/FC Mean 0.172 0.027 0.072 0.078
Min 0.156 0.016 0.047 0.062
Max 0.218 0.062 0.094 0.109
PV/wind/battery Mean 0.663 0.051 0.070 0.077
Min 0.546 0.031 0.062 0.062
Max 0.78 0.078 0.094 0.140
PV/battery Mean 0.706 0.047 0.065 0.071
Min 0.515 0.031 0.047 0.062
Max 0.967 0.078 0.094 0.094
Wind/battery Mean 0.698 0.057 0.066 0.074
Min 0.624 0.031 0.062 0.062
Max 0.796 0.12 0.094 0.109
Author's personal copy
meet the deficit power. Simulation results show that the
optimized hybrid systems will be able to supply the load
demand.
In practice, most of the remote regions have electrical
power typically generated by diesel generators. This system is
difficult to maintain, very inefficient, and subject to frequent
outages. To protect against power outages due to mechanical
breakdowns, redundant generators are frequently in place,
resulting in higher capital costs. Maintenance work is most
often performed by skilled workers from outside of the re-
gions. Fuel must also be transported to the regions. All of this
leads to having an expensive generation system. For this aim
as well as in response to concerns about climate change, en-
ergy independence and economic stimulus, development of
renewable energy in worldwide has been encouraged by
government policy. As a result, according to the above-
mentioned reasons and the difficulties of transforming the
electrical energy to the remote regions, the renewable sources
can be promising alternatives.
Conclusion
This paper studies the economic aspects of PV/wind/FC and
PV/wind/battery systems and performance of different heu-
ristic optimization techniques to optimally size these sys-
tems. It is found that hybrid systems with battery storage are
economically better choice for producing electrical power
than hybrid systems with hydrogen-based storage systems.
With improvement in the efficiency of both FC and electro-
lyzer, FC/electrolyzer storage system can be economically
competitive in the future.
To optimally size the hybrid systems, different heuristic
techniques are applied to find the optimum number of each
component. From the optimization viewpoint, it is found that
PSO yields more promising results than TS, SA, and HS in
terms of the total annual cost.
Acknowledgment
The financial support of the Graduate University of Advanced
Technology is greatly acknowledged.
r e f e r e n c e s
[1] Ahmed NA, Miyatake M, Al-Othman A. Power fluctuationssuppression of stand-alone hybrid generation combiningsolar photovoltaic/wind turbine and fuel cell systems. EnergyConvers Manag 2008;49:2711e9.
[2] Calderon M, Calderon A, Ramiro A, Gonzalez J. Automaticmanagement of energy flows of a stand-alone renewable
Fig. 10 e The difference between the generated and
Fig. 11 e Storage level. (a) Hydrogen tanks and (b) batteries.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 9 9 7 3e9 9 8 4 9983
Author's personal copy
energy supply with hydrogen support. Int J Hydrogen Energy2010;35:2226e35.
[3] Lagorse J, Simoes MG, Miraoui A, Costerg P. Energy costanalysis of a solar-hydrogen hybrid energy system for stand-alone applications. Int J Hydrogen Energy 2008;33:2871e9.
[4] Mohamed FA, Koivo HN. System modelling and onlineoptimal management of microGrid using mesh adaptivedirect search. Int J Electr Power Energy Syst 2010;32:398e407.
[5] Nehrir MH, Wang C. Modeling and control of fuel cells:distributed generation applications, vol. 41. John Wiley &Sons; 2009.
[6] Nelson D, Nehrir M, Wang C. Unit sizing and cost analysis ofstand-alone hybrid wind/PV/fuel cell power generationsystems. Renew Energy 2006;31:1641e56.
[7] Onar O, Uzunoglu M, Alam M. Modeling, control andsimulation of an autonomous wind turbine/photovoltaic/fuelcell/ultra-capacitor hybrid power system. J Power Sources2008;185:1273e83.
[8] Uzunoglu M, Onar O, Alam M. Modeling, control andsimulation of a PV/FC/UC based hybrid power generationsystem for stand-alone applications. Renew Energy2009;34:509e20.
[9] Valdes R, Rodrıguez L, Lucio J. Procedure for optimal designof hydrogen production plants with reserve storage and astand-alone photovoltaic power system. Int J HydrogenEnergy 2012;37:4018e25.
[10] Zervas P, Sarimveis H, Palyvos J, Markatos N. Model-basedoptimal control of a hybrid power generation systemconsisting of photovoltaic arrays and fuel cells. J PowerSources 2008;181:327e38.
[11] Gopalakrishnan K, Khaitan SK, Kalogirou S. Soft computingin green and renewable energy systems, vol. 269. Springer-Verlag New York Inc; 2011.
[12] Khaitan SK, Raju M. Dynamics of hydrogen powered CAESbased gas turbine plant using sodium alanate storagesystem. Int J Hydrogen Energy 2012;37(24):18904e14.
[13] Khaitan SK, Raju M. Discharge dynamics of coupled fuel celland metal hydride hydrogen storage bed for small windhybrid systems. Int J Hydrogen Energy 2012;37(3):2344e52.
[14] Raju M, Khaitan S. Charging dynamics of metal hydridehydrogen storage bed for small wind hybrid systems. Int JHydrogen Energy 2011;36(17):10797e807.
[15] Raju M, Khaitan SK. System simulation of compressedhydrogen storage based residential wind hybrid powersystems. J Power Sources 2012;210:303e20.
[16] Gopalakrishnan K, Khaitan SK. Use of hydrogen fromrenewable energy source for powering hot-mix asphaltplant. Int J Traffic Transp Eng 2012;2(1):30e43.
[17] Chavez-Ramırez A, Vallejo-Becerra V, Cruz J, Ornelas R,Orozco G, Munoz-Guerrero R, et al. A hybrid power plant(SolareWindeHydrogen) model based in artificialintelligence for a remote-housing application in Mexico. Int JHydrogen Energy 2013;38:2641e55.
[18] Diaf S, Diaf D, Belhamel M, Haddadi M, Louche A. Amethodology for optimal sizing of autonomous hybrid PV/wind system. Energy Policy 2007;35:5708e18.
[19] Borowy BS, Salameh ZM. Methodology for optimally sizingthe combination of a battery bank and PV array in a wind/PVhybrid system. Energy Convers IEEE Transac 1996;11:367e75.
[20] Shrestha G, Goel L. A study on optimal sizing of stand-alonephotovoltaic stations. Energy Convers IEEE Transac1998;13:373e8.
[21] Maghraby H, Shwehdi M, Al-Bassam GK. Probabilisticassessment of photovoltaic (PV) generation systems. PowerSyst IEEE Transac 2002;17:205e8.
[22] Kellogg W, Nehrir M, Venkataramanan G, Gerez V.Generation unit sizing and cost analysis for stand-alone
wind, photovoltaic, and hybrid wind/PV systems. EnergyConvers IEEE Transac 1998;13:70e5.
[23] Prasad AR, Natarajan E. Optimization of integratedphotovoltaicewind power generation systems with batterystorage. Energy 2006;31:1943e54.
[24] Roy S. Optimal planning of wind energy conversion systemsover an energy scenario. Energy Convers IEEE Transac1997;12:248e54.
[25] Rohani A, Mazlumi K, Kord H. Modeling of a hybrid powersystem for economic analysis and environmental impact inHOMER. In: Electrical engineering (ICEE), 2010 18th Iranianconference on; 2010. pp. 819e23.
[26] Koutroulis E, Kolokotsa D, Potirakis A, Kalaitzakis K.Methodology for optimal sizing of stand-alone photovoltaic/wind-generator systems using genetic algorithms. Sol Energy2006;80:1072e88.
[27] Zhang B, Yang Y, Gan L. Dynamic control of wind/photovoltaic hybrid power systems based on anadvanced particle swarm optimization. In: Industrialtechnology, 2008. ICIT 2008. IEEE international conferenceon; 2008. pp. 1e6.
[28] Askarzadeh A. A discrete chaotic harmony search-basedsimulated annealing algorithm for optimum design of PV/wind hybrid system. Sol Energy 2013;97:93e101.
[29] Askarzadeh A. Developing a discrete harmony searchalgorithm for size optimization of windephotovoltaic hybridenergy system. Sol Energy 2013;98:190e5.
[30] Luna-Rubio R, Trejo-Perea M, Vargas-Vazquez D, Rıos-Moreno G. Optimal sizing of renewable hybrids energysystems: a review of methodologies. Sol Energy2012;86:1077e88.
[31] Garcia RS, Weisser D. A windediesel system with hydrogenstorage: joint optimisation of design and dispatch. RenewEnergy 2006;31:2296e320.
[32] El-Shatter TF, Eskander MN, El-Hagry MT. Energy flow andmanagement of a hybrid wind/PV/fuel cell generationsystem. Energy Convers Manag 2006;47:1264e80.
[33] Khan M, Iqbal M. Dynamic modeling and simulation of asmall windefuel cell hybrid energy system. Renew Energy2005;30:421e39.
[34] Khan M, Iqbal M. Pre-feasibility study of stand-alone hybridenergy systems for applications in Newfoundland. RenewEnergy 2005;30:835e54.
[35] Hakimi S, Tafreshi SM, Kashefi A. Unit sizing of a stand-alone hybrid power system using particle swarmoptimization (PSO). In: Automation and logistics, 2007 IEEEinternational conference on; 2007. pp. 3107e12.
[36] Hakimi S, Moghaddas-Tafreshi S. Optimal sizing of a stand-alone hybrid power system via particle swarm optimizationfor Kahnouj area in south-east of Iran. Renew Energy2009;34:1855e62.
[37] Eberhart R, Kennedy J. A new optimizer using particle swarmtheory. In: Micro machine and human science, 1995. MHS’95,Proc sixth int symposium; 1995. pp. 39e43.
[38] Glover F. Future paths for integer programming and links toartificial intelligence. Comput Oper Res 1986;13:533e49.
[39] Kirkpatrick S, Vecchi M. Optimization by simulatedannealing. Science 1983;220:671e80.
[40] �Cerny V. Thermodynamical approach to the travelingsalesman problem: an efficient simulation algorithm. JOptim Theory Appl 1985;45:41e51.
[41] Geem ZW, Kim JH, Loganathan G. A new heuristicoptimization algorithm: harmony search. Simulation2001;76:60e8.
[42] Abido M. Optimal power flow using tabu search algorithm.Electr Power Comp Syst 2002;30:469e83.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 9 9 7 3e9 9 8 49984