基于速度交流的多种群多目标粒子群算法研究

为提高多目标优化算法的收敛精度和搜索性能,提出一种基于速度交流的多种群多目标粒子群算法。算法引入速度交流机制,将种群划分为多个子种群以实现速度信息共享,改善粒子单一搜索模式,提高算法的全局搜索能力。采用混沌映射优化惯性权重,提高粒子搜索遍历性和全局性,为降低算法在运行后期陷入局部最优Pareto前沿的可能性,对各个子种群执行不同的变异操作。将算法与NSGA-Ⅱ、SPEA2、AbYSS、MOPSO、SMPSO和GWASF-GA先进多目标优化算法进行对比,实验结果表明:该算法得到的解集具有更好的收敛性和分布性。

Research on Multi-population Multi-objective Particle Swarm Optimization Algorithm Based on Velocity Communication

In order to improve the convergence precision and search performance of multi-objective optimization algorithms, a multi-population of multi-objective particle swarm optimization algorithm based on velocity communication is proposed. The algorithm introduces the speed communication mechanism, divides the population into multiple sub-populations to achieve speed information sharing, improves the particle single search mode, and enhances the global search ability of the algorithm. Chaos mapping is used to optimize the inertia weight, and the particle search ergodicity and globality are improved. In order to reduce the possibility that the algorithm falls into the local optimal Pareto frontier in the late stage of operation, different mutation operations are performed on each sub-population. The algorithm is compared with NSGA-Ⅱ, SPEA2, AbYSS, MOPSO, SMPSO and GWASF-GA state-of-the-art multi-objective optimization algorithms. Experimental results show that the solution set obtained by this algorithm has better convergence and distribution.