Lifecycle-based swarm optimization method for multi-objective optimization problem (MOP)

In view of the superiority of the Lifecycle-based Swarm Optimization algorithm (LSO) in benchmark functions, this paper will further study on the optimizing performance of the LSO algorithm in multi-objective optimization problem. Based on the LSO algorithm, this paper designs the LSO algorithm based on non-dominated sorting (NLSO) which has easy and lesser parameters. The NLSO algorithm divides initialization population into dominating set and non-dominated set, also adjusts dynamically the non-dominated set in the iteration, to accomplish the searching and the approximation of the Pareto optimal set. The experiments demonstrate not only effectiveness and rapidity of the NLSO algorithm, but also the NLSO algorithm outperforms other congeneric algorithms by the calculation of performance index Generational Distance (GD) and Spacing (SP).