论文部分内容阅读
针对标准粒子群算法易陷入局部最优而早熟的问题,提出了一种基于动态种群结构的粒子群算法。该算法在种群结构中引入小世界网络模型,由于网络模型的演化,使算法具有动态的种群结构,从而保持了种群的多样性。同时为了使粒子尽可能地分布在不同的搜索空间,在网络模型演化过程中考虑了结点的个体价值。为了加快算法的收敛速度,在进化后期采用全局模型粒子群算法。通过对三个经典测试函数优化问题的数值仿真并与其它方法进行比较,结果表明了算法的有效性和实用性。
Aiming at the problem that the standard particle swarm optimization is easy to fall into local optimum and precocity, a particle swarm optimization algorithm based on dynamic population structure is proposed. The algorithm introduces a small-world network model into the population structure. Due to the evolution of the network model, the algorithm has a dynamic population structure, thus maintaining the diversity of the population. At the same time, in order to make the particles distributed in different search spaces as much as possible, the individual value of nodes is considered in the evolution of network model. In order to speed up the convergence of the algorithm, a global model particle swarm optimization algorithm is adopted in the late evolutionary stage. Through numerical simulation of three classic test function optimization problems and comparison with other methods, the results show that the algorithm is effective and practical.