An improved genetic algorithm for solving traveling salesman problem(TSP) is proposed. Based on the traditional genetic algorithm, the proposed algorithm introduces the greedy method into species initialization. In order to improve the optimization speed and prevent the local minimum, the improved algorithm updates the crossover probability and mutation probability adaptively according to the evolution stages and the fitness value of individuals. The heuristic crossover operator
based on greedy method is used to optimize the crossover results. The strategy of keeping the best individuals to propagate the optimal gene structure is introduced. The results of TSP example show that the improved algorithm can find the global optimal solution with high performance.