Multi-objective sparrow search algorithm: A novel algorithm for solving complex multi-objective optimisation problems

This study proposes a multi-objective sparrow search algorithm (MOSSA) based on a 2 crowding-distance entropy and the optimal strategy of the positions to solve complex multi-objective optimisation problems (MOPs) using the excellent performance of the sparrow search algorithm (SSA). A fast non-dominated sorting approach is incorporated to permit SSA to solve MOPs. To maintain the evenness and spread of the Pareto solution set obtained in each run, a 2 crowding-distance entropy is proposed to measure the diversity of the solution set. Modified formulas for updating positions and additional adaptive parameters can improve the global search ability of MOSSA. The position archive of the population is introduced to realise the optimal strategy of the positions. This strategy ensures that the positions of the population in each generation are optimal, which significantly improves the performance of MOSSA. MOSSA is compared with three well-known algorithms using a set of complex unconstrained test problems and a complex constrained engineering optimisation problem. This study explores the effect of the 2 crowding-distance entropy and crowding distance on maintaining diversity. The impact of the in 2 crowding-distance entropy on the performance of MOSSA is analysed. The experimental results demonstrate that MOSSA exhibits competitive performance in solving complex MOPs.