基于退火策略的混沌神经网络及其在TSP中的应用  

Chaotic Neural Network Based on Annealing Strategy and Their Application to TSP

作　　者：翁妙凤[1] 高晶[2] 

机构地区：[1]华东船舶工业学院计算机系,江苏镇江212003 [2]沈阳师范大学数学系,辽宁沈阳110034

出　　处：《小型微型计算机系统》2002年第5期574-576,共3页Journal of Chinese Computer Systems

摘　　要：本文主要研究混沌模拟退火神经网络 (CSAN)在求解 TSP中的应用 .我们采用了四种 CSAN模型 ,分别将它们对 15、2 0、5 0个城市的 TSP求解结果比较 ,并研究其模型参数的设置对 TSP优化解的影响 .仿真结果表明 ,CSAN比HNN具有更丰富和更为灵活的动力学特性 。In this paper,We mainly do researches on using chaotic neural network based on simulated annealing(CSAN) to solve TSP. We'll introduce four chaotic neural network (CSAN)models and use them on TSP of 15?20?50 cities,then compare with conclusion.TSP resultions about the effect of each papamater of CSAN are summed up.simulation result have shown that CSAN has richer and and more flexible dynamics rather than HNN like only with point attractors, so that it can be expected to have higher ablity of searching for globally optimal or near optimal solutions.

关 键 词：退火策略 混沌神经网络 TSP 旅行商问题 目标函数 

分 类 号：O22[理学—运筹学与控制论] TP183[理学—数学]