基于改进惯性权重的粒子群优化算法

惯性权重是粒子群算法中平衡全局搜索和局部搜索能力的重要参数,提出了一种基于改进惯性权重的粒子群优化算法。该算法在进化初期采用基于不同粒子不同维的动态自适应惯性权重策略,加快收敛速度,在进化后期采用线性递减权重策略,同时为防止陷入局优,适时引入混沌变异增加种群多样性。对5个典型测试函数的测试结果表明,NPSO在收敛速度、收敛精度、稳定性和全局搜索能力等方面比线性权重PSO(LDIWPSO)均有很大程度上的提高。

PARTICLE SWARM OPTIMISATION ALGORITHM BASED ON MODIFIED INERTIA WEIGHT

As the inertia weight is an important parameter in particle swarm optimisation to balance global search and local search,a new PSO algorithm based on modified inertia weight is proposed.In initial stage of the evolution,the algorithm uses the strategy of dynamic and self-adaptive inertia weight based on different dimensions and different particles to accelerate the convergent speed.In later stage of the evolution,it uses the strategy of linear degressive inertia weight(LDIW),at the same time introduces time...