一种求解约束优化问题的微粒群算法

约束保持法是目前求解约束问题时处理约束的主要方法之一，该方法的思想是确保进化过程中所有粒子始终在可行域范围内。本文借鉴复合形法的思想，提出一种求解约束优化问题的新方法。当粒子超出可行域范围时，通过反射、扩张、收缩等操作，为粒子重新产生一个可行位置。通过对标准函数仿真实验表明，该算法实现原理简单，而且能得到较优的解。

Particle Swarm Optimization for Solving Constrained Optimization Problems

The constraint-preserving method is one main method to handle constraints for solving constrained optimization problems. Its idea is to guarantee all particles within the feasible region in the iteration procedure. A new method is proposed to solve constrained optimization problem by particle swarm optimization by borrowing the idea of complex shape method. When the particle escapes from the feasible region, a new feasible position will be reproduced for this particle by reflecting, extending or reversing. Experiments on standard functions show that the princi- ple of the algorithm this paper presented is simple, and can get better results.