When the conventional genetic optimization algorithm is used for reactive power optimization of the electricity dis⁃tribution network system,it has premature convergence problem and poor ability to search the local optimal solution,which may result in the optimization results that the convergence rate is slow and convergence precision is low. The adaptive dual popula⁃tion,adaptive termination conditions and new filial⁃generation generating conditions for offspring are introduced to form an im⁃proved genetic optimization algorithm,which can ensure the population diversity and avoid the premature convergence of the op⁃timization algorithm in the process of population iteration,and accelerate the search efficiency of the optimization algorithm to improve the ability of searching local optimal solution. The improved genetic optimization algorithm is compared with the conven⁃tional genetic optimization algorithm and the controlled cross genetic algorithm(CCGA)by taking IEEE57 and IEEE30 node sys⁃tems as the experiment platform,and the algorithms are tested with experimental data under same experimental conditions. The test results show that the improved genetic algorithm has the best optimization effect,faster convergence rate and higher conver⁃gence precision,can obtain the global optimal solution to make the objective function more close to the global optimal solution, and its average network loss obtained by this algorithm is lower than that of other optimization algorithms.%考虑到常规遗传优化算法进行电力配网系统的无功优化时，算法的早熟问题以及局部寻找最优解能力欠缺造成了优化结果收敛速度和收敛精度较低等问题，该文将自适应对偶种群、自适应终止条件以及全新的子代生成条件引入以形成一种改进型遗传优化算法，在种群迭代过程中保证种群的多样性以避免优化算法早熟现象，以及对优化算法的搜索效率进行加快以提高局部寻找最优解能力。通过实验IEEE57和IEEE30节点系统作为测试平台，使用常规遗传优化算法和可控交叉遗传算法（CCGA）与该文研究的改进遗传优化算法进行对比，使用相同的实验条件和实验数据进行测试。结果表明该文研究的改进遗传算法具有最好的优化效果，该算法计算得到的平均网损均低于其他优化算法，收敛精度和收敛速度更高，能够在局部最优解处跳出，距离目标函数的全局最优解更加接近。
展开▼
