引入单纯形算子的粒子群优化算法在板形模式识别中的应用  

Application of Particle Swarm Optimization Using Simplex Method in Flatness Pattern Recognition

作　　者：刘建[1] 王益群[1] 姜万录[1] 宁淑荣[1] 

机构地区：[1]燕山大学,秦皇岛066004

出　　处：《中国机械工程》2007年第23期2877-2880,共4页China Mechanical Engineering

基　　金：国家自然科学基金资助项目(60474044);河北省自然科学基金资助项目(E2004000221)

摘　　要：将粒子群理论引入板形模糊模式识别系统,对识别后的板形进行优化,提高了板形识别的精度。粒子群算法作为一种全局优化算法,对于复杂优化问题,存在容易陷入局部极值的不足,因此,提出了粒子群和单纯形混合优化算法,将具有良好局部搜索能力的单纯形法与粒子群算法的全局搜索能力结合起来,有效地提高了板形模式识别优化的收敛速度,同时也提高了识别精度。For improving flatness pattern recognition precision, particle swarm theory was used in flatness fuzzy pattern recognition to optimize the result of pattern recognition. Particle swarm theory, a global optimization algorithm, still has the shortage of easily getting into local optimal value for complex optimization problems, so a mixed algorithm of particle swarm using simplex method was presented. Combined good local searching characteristics of simplex method with good global searching characteristics of particle swarm, the convergence speed and the precision of flatness pattern recognition is improved effectively.

关 键 词：板形 模式识别 单纯形法 粒子群 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]