中文说明：TSP问题是一个典型的组合优化问题，也是一个NP难题，一般很难精确地求出其最优解，因而找出有效的近似解算法具有重要意义。本文针对基本遗传算法在解决TSP问题是所存在的收敛速度慢，容易“早熟”问题，提出了一种改进的交叉算子和基于种群相似度的更新策略。改进的交叉算子是通过先比较两个城市间距离再进行的交换城市序号，因此加快了收敛的速度，而基于种群的相似度更新策略则在算法的后期可以有效的防止早熟，通过对实例144进行测试，证明该算法在解决该类问题上取得了较好的效果。
English Description：
TSP is a typical combinatorial optimization problems, is also a NP problem is often difficult to accurately calculate the optimal solution, thus identifying effective approximation algorithm has important significance. In view of basic genetic algorithm in solving TSP problem the convergence is slow, easy "precocious" problem, an improved crossover operator and update strategy based on similarity of populations. Improved crossover operator is by comparing the distance between the two cities for the Exchange city number, thus accelerating the speed of convergence, based on the similarity of species updated strategy late in the algorithm can prevent premature, by instance 144 test proved that the algorithm has made good results on to solve the problem.