论文部分内容阅读
粒子群算法在求解多目标问题时具有收敛速度快、计算代价小等优点,尤其在局部最优搜索上的速度优势而被众多研究者所青睐.本文在粒子群算法的基础上提出一种新的分组策略即将分组分解融合到多目标粒子群算法中以提高邻域局部搜索的速度.该算法根据个体到权重矢量的距离大小以及各个聚合函数值进行最佳的分组匹配,并动态利用粒子群算法来增强局部搜索能力从而得到Pareto最优解集.在仿真实验中,将该算法应用于ZDT和DTLZ测试函数中进行性能测试,并与NSGAII、M OPSO、M OEA/D和REVA算法进行比较.实验结果表明,与其他四种算法相比,该算法的收敛性能更优,分布性能更好,所获Pareto最优解集的质量更高.
Particle swarm optimization (PSO) has many advantages such as fast convergence speed, small computational cost, especially in the local optimal search speed, which is favored by many researchers in solving multi-objective problems.In this paper, a new kind of particle swarm optimization The grouping strategy is to converge the packet decomposition into the multi-objective particle swarm optimization to improve the local search speed.The algorithm performs the best group matching according to the distance between the individual and the weight vector and each aggregation function value, and dynamically uses the particle swarm Algorithm to enhance the local search ability to get the Pareto optimal solution set.In the simulation experiment, this algorithm is applied to the ZDT and DTLZ test functions for performance testing and compared with the NSGAII, M OPSO, M OEA / D and REVA algorithms The experimental results show that compared with the other four algorithms, the proposed algorithm has better convergence performance and better distribution performance, and the quality of the Pareto optimal solution set obtained is higher.