基于博弈机制的多目标粒子群优化算法

为改善多目标粒子群算法存在优化解的多样性不足和算法的收敛性问题,提出一种基于博弈机制的多目标粒子群优化算法。使用博弈机制,无需外部储备集,通过非占优排序和拥挤距离选出一部分优秀的粒子,从这些优秀的粒子中随机选择一个作为全局最优粒子,有效提升算法的收敛性和种群的多样性。算法初期使用多尺度混沌变异策略,避免算法陷入局部最优。通过与6个多目标算法在3个系列标准测试函数上进行比较,验证了该算法所得解分布性较好,能快速收敛到真实Pareto前端。

Game mechanism based multi-objective particle swarm optimization

To improve the diversity of the optimal solution and the convergence of the algorithm for multi-objective particle swarm optimization,a multi-objective particle swarm optimization algorithm based on game mechanism was proposed.The game mechanism was used,and the algorithm did not require an external reserve set,but selected some excellent particles through non-dominant sorting and using crowding distance,and randomly selected one of these excellent particles as the global optimal.Particles can effectively improve the convergence of the algorithm and the diversity of the population.Multi-scale chaotic mutation strategy was used at the beginning to avoid the algorithm falling into local optimum.By comparing with six multi-objective algorithms on three series of standard test functions,it is verified that the algorithm has better solution distribution and can quickly converge to the real Pareto front end.