论文部分内容阅读
提出一种多目标自适应混沌粒子群优化算法(MACPSO).首先,基于混沌序列提出一种新型动态加权方法选择全局最优粒子;然后,改进NSGA-II拥挤距离计算方法,并应用到一种严格的外部存档更新策略中;最后,针对外部存档提出一种基于世代距离的自适应变异策略.以上操作不仅提高了算法的收敛性,而且提高了Pareto最优解的均匀性.实验结果表明了所提出算法的有效性.
A new multi-objective Adaptive Chaos Particle Swarm Optimization (MACPSO) algorithm is proposed.Firstly, a new dynamic weighting method is proposed to select the global optimal particle based on chaotic sequences. Then, the calculation method of NSGA-II congestion distance is improved and applied to a Strict external archive update strategy.Finally, an adaptive mutation strategy based on distance generation is proposed for external archiving.The above operations not only improve the convergence of the algorithm, but also improve the uniformity of Pareto optimal solution.The experimental results show that The effectiveness of the proposed algorithm.