论文部分内容阅读
分析了粒子群算法的收敛性,指出早熟是由于粒子速度降低而失去继续搜索可行解的能力.进而提出一种基于种群速度动态改变惯性权重的粒子群算法,该算法以种群粒子平均速度为信息动态改变惯性权重,避免了粒子速度过早接近0.通过5个标准测试函数的仿真实验并与其他算法相比,结果表明该算法在进化中期能很好地保持种群多样性,有效地改善算法的平均最优值和成功率.
The convergence of particle swarm optimization algorithm is analyzed and it is pointed out that precocity is the ability to search for feasible solution because of the decrease of particle velocity.A new particle swarm optimization algorithm is proposed based on dynamic change of inertia weight based on population swarm speed.The algorithm uses the average particle swarm speed as the information Dynamically changing the inertia weight and avoiding the particle velocity prematurely approaching 0. Through the simulation experiments of five standard test functions and compared with other algorithms, the results show that the algorithm can well maintain the population diversity during the evolutionary period and effectively improve the algorithm The average optimal value and success rate.