Tri-objective optimization and decision-making methods for stand-alone photovoltaic/battery system: A case study and techno-economic assessments

In this study, a novel accelerated convergence grey wolf optimizer (ACGWO) is developed by employing three key improvements for determining the optimal size of an stand-alone PV system for a remote place in Malaysia. The first improvement is updating the three vectors: Alpha, Beta, and Delta, which improves the exploration inclination. The w and F self-adaptive control parameters are also introduced to enhance global search capability and avoid local stagnation. A new exploration strategy and ans effective exploitation mechanism are proposed to enhance the performance of the proposed method. The improvement is the integration of the proposed ACGWO method with multi-objective optimization principle. The Maxmin function, archive evolution path (AEP), nondominated sorting, and crowding distance mechanisms are used to solve various multi-objective benchmark functions. Finally, the optimal configuration is determind for the freestanding PV system based on three competing criteria: loss of load probability, surplus energy, and total life cycle cost of the system. The proposed MOACGWO method is validated against several well-published multi-objective optimization methods to assess diversity and convergence. To address the challenge of selecting the most desirable solution from the estimated Pareto optimal set, several hybrid multi-criteria decision-making methods are implemented. The experimental results demonstrated that the proposed MOACGWO method outperforms other comparative algorithms in solving benchmark functions of ZDT and DTLZ groups based on several statistical metrics, while the selected optimum configuration consists of 240 PV modules (10 in series and 24 in parallel) and 60 battery storage units at zero LLP, 12.46 (KWh) of Pdump, and 60677.3 ($) of LCC. Based on these findings, he results show that the proposed MOACGWO approach is capable of effectively managing intricate multi-objective optimization problems and can be applied for handling other practical optimization problems too. The MATLAB code for the MOACGWO method is available at: https://github.com/hussein198812/MOACGWO-