Abstract—Studies into the determining size of battery energy storage system (BESS) has become significant recently, owing to their use in a variety of complex, high performance and energy storage system applications. This paper presents a comparative study of optimization techniques between particle swarm optimization (PSO) and artificial neural network (ANN) for evaluating the optimum size of BESS in the micro-grid system. In this paper, the micro-grid system consists of two micro turbine systems, solar photovoltaic (PV) system and BESS, and it is connected to the utility grid. Simulation results show that the optimal size of BESS-based PSO approach achieves the lowest performance in achieving the optimal BESS size compared to the optimal size of BESS-based ANN. However, the optimal sizing of BESS-based ANN approach gives the fastest conversion time compared to the optimal sizing of BESS-based PSO.
Index Terms—Artificial neural network (ANN), battery energy storage system, frequency control, micro-grid, particle swarm optimization.
T. Kerdphol, K. Fuji, Y. Qudaih, and Y. Mitani are with the Kyushu Institute of Technology, Kitakyushu, Japan (e-mail: thongchartkerd@gmail.com, fuji19690708@gmail.com, yaser_qudaih@yahoo.com, mitani@ele.kyutech.ac.jp).
[PDF]
Cite:T. Kerdphol, K. Fuji, Y. Mitani, and Y. Qudaih, "Performance Comparison of Artificial Intelligence Approaches for Battery Energy Storage Size Optimization in PV Micro-grid," Journal of Clean Energy Technologies vol. 4, no. 6, pp. 389-395, 2016.
