Hopfield网络求解旅行商问题有效性的研究  

Study on the Efficiency of Solving TSP Based on Hopfield Network

作　　者：赵建明[1] 姚念民[2] 

机构地区：[1]福建师范大学福清分校,福建350000 [2]哈尔滨工程大学计算机科学与技术学院,哈尔滨150001

出　　处：《科学技术与工程》2009年第2期437-440,共4页Science Technology and Engineering

基　　金：国家自然科学基金(60503055);黑龙江省博士后科研启动金(323630217)资助

摘　　要：对Hopfield网络求解旅行商问题的有效性进行了研究,实验结果表明使用Hopfield网络求解TSP问题时存在两难问题,提高解有效性概率导致解的路径长度优化能力降低,而提高解的路径长度优化能力导致解有效性概率降低。The efficiency of solving the TSP problem based on CHNN Hopfield is studied. A lot of experiments of using Hopfield neural network to solve the TSP problem are done and the results are analyzed. It shows that using CHNN Hopfield to solve TSP problems will run into a dilemma that raising the probability of getting the right answer will reduce the optimization ability of TSP＇s path length and increasing the optimization ability of TSP＇s path length will reduce the probability of getting the right answer.

关 键 词：HOPFIELD网络 旅行商问题 有效性概率 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]