Abstract: In inversion of parameters of probability integral method, the genetic algorithm is expected start the search with a series of parameters in the range, and elastic strategy is used to maintain population diversity, so that the algorithm can overcome the obstacle of local convergence, and then evolve to a global optimum solution.However, this probabilistic algorithm has the disadvantage of poor local detection capability and unstable results; it can only obtain an approximate optimal solution of the question.The pattern search is a reducing gradient algorithm which has better local detection capability and quick convergence.But the algorithm is sensitive to the initial value, and an improper choice of initial value may result in local convergence.In this paper, we propose a combinatorial algorithm combining pattern search with genetic algorithm:the genetic algorithm is used to obtain a global approximate optimal solution, and then the solution is taken as the initial value to obtain a stable and accurate global optimal solution by using the pattern search.The results show that the proposed algorithm has higher accuracy, faster convergence and better stability in inversion of parameters of probability integral method.The comprehensive performance of this algorithm is better than the genetic algorithm or the pattern search.