Optimal scheduling of photovoltaic and battery energy storage in distribution networks using an ameliorated sand cat swarm optimization algorithm: Economic assessment with different loading scenarios

Abstract

The rapid growth of renewable distributed generators (DGs) integrated into smart distribution grids leads to energy waste because of the restricted hosting capacity of these networks, system limit violations, inefficient resource usage, excess generation, and increasing system uncertainty. Battery energy storage (BES) systems, despite its high expense, shortened life, and complicated control strategies, provide a comprehensive solution for these challenges. Therefore, the paper contributes to developing an ameliorated version of the sand cat swarm optimization (SCSO), named ASCAO, for the optimal scheduling of the solar photovoltaic (SPV) and BES units in radial distribution networks (DNs), considering the uncertainties of the solar irradiance and the time-varying load model. The ASCAO algorithm enhances the original SCSO algorithm's effectiveness by using two efficient techniques, namely, local random sand cat strategy (LRSCS) and orthogonal learning (OL). OL is employed to generate diversity between candidate solutions and enhance the speed of the SCSO algorithm, while LRSCS enhances the exploration ability and prevents premature convergence. The embracing of the two proposed strategies improves the stability of the search strategy and ameliorates the SCSO algorithm. The effectiveness of the ASCSO algorithm is validated through cec2020 benchmark functions, and a comparison with the conventional SCSO and other efficient literature-based algorithms is conducted using various statistical analyses. The ASCSO is employed to specify the optimal placement and capacity of the SPV and SPV&BES systems into 33-bus and 69-bus DNs to mitigate energy loss, considering different loading conditions, i.e., light load (LL), normal load (NL), and heavy load (HL). The time-variant voltage-dependent load is a mix of three different load sectors, namely, residential, commercial, and industrial. The cost of energy purchased and the annual expenses including cost of energy loss, installation cost, and operation and maintenance cost are evaluated for the optimized SPV&BES units to ascertain the economic benefits of the DG technology. The findings reveal that the proposed ASCSO algorithm significantly outperforms the state-of-the-art methods in simultaneously integrating multiple SPV and SPV&BES units into DNs, allowing for the implementation in more complex systems.