多维异步随机扰动的粒子群优化算法

Particle swarm optimization algorithm with multidimensional asynchronism and stochastic disturbance

针对粒子群优化算法存在易陷入局部最优和在多维空间中搜索效率降低的问题,结合惯性权重凹函数递减策略,提出了随机扰动和多维异步策略。该策略不仅能提高算法的全局搜索能力,而且还能改善维数的束缚。通过对四个典型基准函数的实验表明,该改进算法能够兼顾局部和全局搜索,使得搜索达优率得到较大提高,所得结果精度较高。

Particle swarm optimization has the disadvantages of being easily trapped into a local optimal solution and searching with lower efficiency in multi-dimensional space. With reference to the strategy of concave function to the inertia weight, the authors proposed a method of multidimensional asynchronism and stochastic disturbance to improve the ability to search for global optimum as well as solve the limitation of dimensionality problem. The experimental results of four classic benchmark functions show that the algorithm can keep the balance between the global search and local search, which effectively improves the success probability of searching with higher precision.