差分算法是一种新型的进化算法,它易与其他算法混合,适应性强,鲁棒性好。本文将演化博弈论的思想引入到差分算法中,对差分算法的改进提出了一种新的思路。提出了增加种群多样性的双种群烟花博弈差分进化算法DGDE,同时对原差分算法参数的设置进行了讨论,通过仿真实验进行验证,得出其算法参数的设置。 Differential algorithm is a new evolutionary algorithm, which is easy to mix with other algorithms, adaptable and robust. In this paper, the evolutionary game theory is introduced into the differential algorithm, and a new idea is proposed for the improvement of the differential algorithm. This paper proposes a two population fireworks game evolutionary algorithm (DGDE) based on population diversity. At the same time, the parameters of the original differential algorithm are discussed. The simulation results show that the algorithm parameters are set up.