Abstract: This paper describes simplex hybrid genetic algorithm called SM-HGA⁺.In detail，by analyzing Neld Meld Simplex algorithm，the authors propose simplex crossover operator and K-step random simplex search operator，and respectively fuse Neld Mead Simplex algorithm and above two novel operators into the best population μP_B（t），the worst population μP_W（t） and a common population P_C（t） in a genetic algorithm in order to construct algorithm SM-HGA⁺.Neld Mead Simplex algorithm in μP_B（t） raises the computation precision；the simplex crossover operator in μP_W（t） accelerates the worst individuals evolving towards better individuals；the K-step random simplex search operator enhances the global convergence speed and genetic algorithm with big crossover probability improves global search performance of algorithm SM-HGA⁺ by population P_C（t）.The standard testing functions test and verify the correctness and efficiency of SM-HGA⁺.

摘要： 构造了单纯形混合遗传算法SM-HGA⁺。分析单纯形搜索算法，提出了单纯形交叉算子和K步随机单纯形搜索算子，并将单纯形搜索算法和这两个算子分别融入到最优微群体μP_B（t）、最差微群体μP_W（t）和普通群体P_C（t），形成SM-HGA⁺。最优微群体中的单纯搜索算法提高算法的精度；最差微群体中的单纯形交叉算子加速最差个体向优秀个体进化；普通群体中K步随机单纯性搜索提高全局搜索速度，同时在普通群体采用大交叉概率的标准遗传算法，提高全局搜索能力。遗传算法测试函数验证算法SM-HGA⁺的正确性、效率。