Cat Swarm Optimization for Sizing Photovoltaic-Battery based Stand-Alone System

Abstract

Stand-Alone Photovoltaic (SAPV) systems have been a popular mode of electricity generation especially in remote areas which deprived of grid electricity. However, one of the main challenges in the development of such systems is optimal system sizing. An undersized system causes failure of meeting the load demand. Thus, this paper presents Cat Swarm Optimization (CSO) for optimal sizing of an SAPV system. A single objective function was formulated for the sizing optimization problem in which models of system components, i.e. PV module, battery, charge controller and inverter, each represented in integer codes, were transcribed as decision variables whereas the Loss of Power Supply Probability (LPSP) was set as the objective function that needs to be minimized. Prior to the optimization, an Iterative Sizing Algorithm (ISA) was initially developed to produce the optimal sizing solution using non-computational intelligence approach. The solution from ISA served as benchmark for CSO. Later, CSO was used to select the optimal model of PV module, battery, charge controller and inverter such that the LPSP was minimized. The results showed that CSO was able to yield similar LPSP as recommended by ISA. Besides having computation time of approximately 3.65 times lower than ISA, CSO was also found to be superior by producing lower LPSP when compared with Evolutionary Programming (EP).