【作者】 张蕾； 陈月辉； 杨波；

【机构】 济南大学网络中心；

【摘要】 <正> 1 引言粒子群优化(Particle Swarrn Optimization,PSO)算法是一种新兴的全局优化技术。PSO算法同遗传算法类似,是一种基于迭代的优化工具。系统初始化为一组随机解,通过粒子在解空间追随最优的粒子搜索最优值。PSO算法简单易实现,但易陷入局部极小值点。一些改进的PSO仍存在着比较复杂、不便使用等缺点。作为一种简单、有效的随机全局优化算法,PSO算法是一种有着潜在竞更多还原

【Abstract】 This paper proposes an artificial neural network (ANN) optimal algorithm based on parallel particle swarm optimization(PSO). A stochastic global optimization technique is used in the proposed algorithm. The connection weights of ANN optimized by the position vectors of each particle improves the ability of solving the pattern classification problems. The algorithm is successfully applied to pattern classification problems, for example the intrusion detection problems. This algorithm not only can improve the classification accuracy, but also can speed up the convergence process ,while getting the better acceleration. The experiment results show that parallel PSO is a potentially robust optimal algorithm.更多还原