Abstract: An improved differential evolution algorithm is proposed to solve the constrained optimization problems. The members of the initial set are generated based on good-point-set method, which keeps the diversity of population. During the evolution, mutation strategies and crossover operations are adaptively selected based on the proportion of feasible solution in the population. It enhances the exploitation ability and the exploration ability of algorithm. Several well-known Benchmark problems are used to verify the algorithm. The experiment results show that the proposed algorithm is effective to deal with different constrained optimization problems.