在分析遗传算法基本理论的基础上，针对泵站优化运行问题的特性，从控制参数、选择方式、交叉机理和种群设置等方面进行改进，建立求解高维整型离散决策变量的改进遗传算法。以淮安四站为例进行计算，比较并分析改进遗传算法和标准遗传算法在泵站优化运行中的适用性，为泵站优化运行算法的设计和应用提供参考。结果表明：同时改进适应度和种群的混合改进遗传算法——FFGA&TPGA，精度高，计算速度快，具有收敛一致性，可用于复杂泵站系统的实时优化调度计算。

According to basic theory of genetic algorithm and the characteristics of optimal operation for pumping station, established improved genetic algorithm for solving high-dimensional discrete integer decision variables through improving the control parameter s, crossover mechanism and population. The No.4 Huaian Pumping Station is taken as a study case to compare and analyze the genetic algorithm and improved genetic algorithm applicability to optimal operation for pumping station, which can provide reference for the the design and application of the pumping algorithm. Calculations showed: mix advanced genetic algorithm——FFGA&TPGA, which improved the fitness and population, had high precision, efficiency and uniform convergence. Thus, it had advantages in solving large and complex problems about optimal operation of pumping station.

江苏省水利科技项目（2012041）；江苏省普通高校研究生科研创新计划项目（CXZZ13_0256）