Abstract: For the problem of local optimum and the performance of stability，an adaptive genetic algorithm based on co-evolution is proposed.Adaptive probability of mutation is applied in local evolution and the capability of local search on the top floor is enhanced.On the bottom floor，co-evolution algorithm is used for solving the premature convergence among sub-populations，and adaptive genetic manipulation is used to protect outstanding individuals in sub-populations.The experiments demonstrate that this algorithm can increase the convergent speed and it has the ability of searching an optimum solution.

Key words: adaptive genetic algorithm, co-evolution, convergent speed

摘要： 针对传统遗传算法易于陷入局部最优解，性能不稳定的问题，提出了一种基于协同进化的自适应遗传算法（CEAGA）。在协同进化的两层框架模型的基础上，引入一个自适应的变异策略，改进了协同进化遗传算法中的局部进化操作，加强了在上层中的局部搜索;在下层，在种群之间采用协同进化算法，克服未成熟收敛，在种群内部进化中引入自适应遗传操作，保护种群中的优秀个体。实验验证CEAGA既具有很快的收敛速度，又具有很好的全局搜索性能。

关键词: 自适应遗传算法, 协同进化, 收敛速度