：组合差分进化算法CoDE是一新的具有竞争力的算法，但收敛速度和寻优性能仍有待改进。为解决上述问题，提出对

组合差分进化算法CoDE从生成策略和控制参数两个方面进行改进，提出了两种改进的CoDE版本MCoDE和MCoDE—P，并

利用6个典型的测试甬数对改进性能进行检验。结果表明结合了最好个体信息的MCoDE方法能够改善CoDE的寻优性能，

而采用控制参数扩展的MCoDE—P方法却难以达到期望的效果。-: Combined differential evolution algorithm CoDE is a new competitive algorithm, but convergence speed and optimization performance needs to be improved. In order to solve the above problem, the pair differential evolution algorithm CoDE generation strategy and control parameters from the two aspects of improvement, we propose two improved versions MCoDE CoDE and MCoDE-P, and using six typical test Yong Number on improving performance tested. The results show that combines the best individual information MCoDE method can improve CoDE the optimization performance, while the use of the control parameter expansion MCoDE-P method is difficult to achieve the desired effect.