基于改进的遗传模拟退火混合算法求解TSP问题

TSP Problem Solving Based on Improved Genetic Simulated Annealing Algorithm

本文针对遗传算法（GA）早熟收敛问题就GA的交叉算予进行改进，针对模拟退火算法易陷入局部最小值的缺点．使用HFC—ADM（自适应输入阂值的分等级搜索）的SA（模拟退火算法）和改进后的GA相结合，提出了一种求解TSP问题的遗传模拟退火混合算法，并应用于求解TSP（旅行商问题）问题。实验结果表明，该算法具有比传统的GA以及基于HFC—ADM的SA具有更强的全局搜索能力和更快的收敛速度。

The paper gives a Hybrid Genetic and Simulated Annealing algorithm for TSP.The algorithm improves cross-operator of GA for genetic algorithm＇s premature convergence issues.Due to the disadwmtages that SA algorithm is vulnerable to run into minimum,the paper combines SA based on HFC-ADM with improved Genetic Algorithm.The algorithm has better capability for global searching and higher convergence speed through the TSP experiment results.