一种基于球隙迁移的改进粒子群优化算法

An Improved Particle Swarm Optimization Algorithm Based on Sphere-gap Transferring

针对粒子群算法易早熟、陷入局部最优、求解精度不高等缺陷,提出了一种基于球隙迁移的改进粒子群算法。改进后的算法重新定义了粒子的位置、速度以及操作;引入了交换序,并用贪心算法的思想以节省搜索时间、提高算法收敛速度;引入了球隙迁移算法,克服了算法易陷入局部最优的缺陷;引入了基于松弛操作的思想的扰动机制。实验结果表明改进后的算法是可行的、有效的。

In order to overcome the shortcoming of the standard particle swarm optimization algorithm such as premature, bad convergence and accuracy, an improved PSO based on sphere-gap transferring is proposed. The al- gorithm draws on the thinking of the greedy algorithm to initialize the particle swarm, and redefines the particle po- sition, velocity, operating. It also introduces sphere-gap transferring and fluctuation operator to escape the current local minimum. The numerical results show that the improved algorithm is feasible and effective.