有等式约束优化问题的粒子群优化算法  

Particle swarm optimization algorithm of equation constrained optimization problem

作　　者：盛跃宾[1] 陈定昌[2] 穆森[1] 任强[1] 张朝阳[2] 

机构地区：[1]中国航天科工集团第二研究院706所,北京100854 [2]中国航天科工集团第二研究院二部,北京100854

出　　处：《计算机工程与设计》2006年第13期2412-2413,2418,共3页Computer Engineering and Design

摘　　要：目前大多数粒子群优化算法针对无约束优化问题或不等式约束优化问题,求解有等式约束优化问题的方法是把每个等式约束变成两个不等式约束,这种方法的缺点是在进化过程中粒子位置很难满足等式约束条件,影响了收敛速度和解的精度。提出了求解有等式约束优化问题的两种新粒子群优化算法,数值试验结果表明,算法是有效的。Most of current particle swarm optimization algorithms are used to solve unconstrained optimization problems or inequation constrained optimization problems. The method to solve equation constrained optimization problems is that each constrained equation is turned into two constrained inequations. But, the particles＇ positions in this method aren＇t easy to satisfy constrained equations, which reduces the speed of convergence and the precision. To solve the equation constrained optimization problems, two new particle swarm optimization algorithms are presented. The experimental results show the algorithms are effective.

关 键 词：粒子群优化 等式约束 约束优化问题 适应度函数 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]