基于改进型能量函数和瞬态混沌神经网络的TSP问题研究  

Study on the TSP Based on an Improved Energy Function and Transiently Chaotic Neural Network Model

作　　者：严晨[1] 王直杰[1] 

机构地区：[1]东华大学信息科学与技术学院自动化系,上海201620

出　　处：《系统仿真学报》2006年第5期1402-1405,共4页Journal of System Simulation

摘　　要：针对传统神经网络在搜索NP类问题的解时易陷于局部最优点的不足,提出了一种基于改进型能量函数(IEF)和瞬态混沌神经网络(TCNN)的优化模型,将此应用于旅行商问题(TSP)的求解,并和传统神经网络优化方法进行了比较。仿真研究结果表明,该论文所提出的方法在解的可行性以及全局最优解的获取能力方面都有很大优势,收敛速度和准确度也令人满意。The solutions to the NP problems searched by the conventional neural networks such as Hopfield neural networks tend to be trapped into local optimal. To overcome this deficiency, a new optimization model which is composed of an improved energy function （1EF） and transiently chaotic neural network model （TCNN） was proposed. A number of simulation experiments were carried out to solve the traveling salesman problem （TSP）. Compared with the conventional methods, the simulation results show that the new method has a strong ability to access a feasible and global optimal solution. Moreover, the speed of convergence and accuracy are also satisfying.

关 键 词：旅行商问题 优化 混沌神经网络 能量函数 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]