采用遗传算法来求解不满足先进先出原则的动态网络中的最短路径问题,并采用所提出的随机A*算法解决了利用遗传算法求解最短路径问题时的最大障碍——初始种群的产生.最后以广州市电子地图为基础随机产生了一个不满足先进先出原则的动态网络(包括20000个节点,40000条边和144个时间间隔),来对所提出的算法进行验证.试验结果表明,遗传算法适合求解非常态且不满足先进先出原则的动态网络中的路径诱导问题. The genetic algorithm is used to solve the shortest path problem in the dynamic network that does not satisfy the first in, first out and the stochastic A * algorithm is used to solve the genetic algorithm to solve the shortest path problem, the biggest obstacle - the initial population. Based on Guangzhou’s electronic map, a dynamic network (including 20000 nodes, 40,000 edges and 144 time intervals) that does not satisfy the first-in first-out principle is randomly generated to verify the proposed algorithm.The experimental results show that the genetic The algorithm is suitable for the path guidance problem in dynamic network which is not suitable to solve the first-in-first-out problem.