带自适应变异的量子粒子群优化算法

Quantum particle swarm optimization algorithm with adaptive mutation

提出了一种带有自适应变异的量子粒子群优化(AMQPSO)算法,利用粒子群的适应度方差和空间位置聚集度来发现粒子群陷入局部寻优时,对当前每个粒子经历过的最好位置进行自适应变异以实现全局寻优。通过对典型函数的测试以及与量子粒子群优化(QPSO)算法和自适应粒子群优化(AMPSO)算法的比较,说明AMQPSO算法增强了全局搜索的性能,优于其他算法。

A Quantum Particle Swarm Optimization Algorithm with Adaptive Mutation（AMQPSO） is given.When the proposed algorithm is found to sink into the local optimization by fitness variance and space position aggregation degree,a new adaptive mutation operator is implemented at the best position of each particle at first so as to realize global optimization.The experiments show that the AMQPSO is better than QPSO and the AMPSO in global optimization.