Key points are not available for this paper at this time.
The use of artificial intelligence (AI) – based tools in the optimization of renewable energy (RE) systems is increasing. These tools could even be more useful to developing countries like Cameroon with abundant RE resources, yet low rural electrification rate. However, the optimization of these energy systems especially in hybrid forms is still a challenge. This paper uses an AI-based Particle Swarm Optimization (PSO) and Differential Evolution (DE) for the design and optimization of a stand-alone hybrid solar PV – hydro- battery power system. These algorithms were developed using the MATLAB software. The proposed smart algorithms ensure that the load is met at a minimum levelized cost of energy (LCOE) and acceptable loss of power supply probability (LPSP). After simulation, DE gave an optimum LPSP of 0. 0499 and optimum LCOE of 0. 06192 /kWh after the 19th iteration under set operational limits while PSO gave an optimum LPSP of 0. 0492 and optimum LCOE of 0. 06358 /kWh after the 40th iteration. The optimal net present value (NPC) obtained from the PSO and DE were USD 96, 175. 26 and USD 93, 958. 07 respectively. While DE gave a lesser LCOE than PSO, the LPSP obtained using the PSO technique was smaller, signifying more system reliability. The optimum system size of DE showed the least LCOE with the proposed capacities of 1 kW PV, 33. 96 kW hydropower and zero battery. The optimized system ensures a proper power management within the hybrid system. An appraisal of the two algorithms showed that the DE tool is accurate and a better option than PSO in terms of cost and speed of convergence. Further statistical analysis revealed that PSO was more robust. The optimal cost function obtained from both algorithms is acceptable for rural electrification projects.
Iweh et al. (Wed,) studied this question.