Abstract: I}his paper proposed a co-evolutionary genetic algorithm by two stages based on potential field. In the first stage, each population was evolved mainly by sexual reproduction, when all populations reached evolutionary stagnate,key areas were formed by cluster analysis,and narrowing the search area can improve the efficiency of the algorithm In the second stage, each population was evolved mainly by asexual reproduction to enhance local search, and realized the directed evolution based on individuals' fitness so as to speed up the convergence rate. At the same time it proposed a concept called environmental potential field which could guide the evolution in order to make multiple populations evolve cooperatively. The experimental results show that the proposed algorithm has high quality of precision and rapid convergence rate and that it overcomes the low efficiency of traditional algorithms to some extent.

Key words: Co-evolution, Potential field, Asexual reproduction, Evolutionary direction