混沌离散微粒群算法在矩形排样中的应用  

Application of Chaotic Discrete Particle Swarm Algorithm in Rectangle Packing

作　　者：王珊珊[1] 

机构地区：[1]合肥工业大学管理学院,合肥230009

出　　处：《计算机工程》2010年第23期174-176,共3页Computer Engineering

基　　金：国家"863"计划基金资助项目"面向机械制造业的连续多时段整体智能优化下料问题研究及系统实现"(2006AA04Z134)

摘　　要：混沌离散微粒群算法主要在离散微粒群算法框架的基础上引入混沌机制,并通过判断种群多样性处理速度的更新操作,从而解决离散微粒群算法易于陷入局部最优及搜索到解的时间较长且精度不高等缺点。将该算法应用到在定宽无限长矩形优化排样问题上,通过剩余矩形算法完成解码。实验结果表明,该算法在解决此类排样问题时可以较快地获得最优的配切方法。An improved chaotic discrete particle swarm algorithm is proposed to solve the packing problem of rectangles.This paper solves the shortcomings of the discrete particle swarm algorithm which is easy to fall into local optimum,to search the solution for a longer time and not get to higher accuracy.It decodes the result sequence by residual rectangular algorithm in the final stages.Experimental results show that this algorithm can get better with cutting rates than others.

关 键 词：离散微粒群 混沌 TENT映射 种群多样性 剩余矩形排样算法 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术]