After a short - term load forecasting method based analogous and linear extrapolation is proposed , the load forecast and the priority of equipment action are led into static reactive power optimization . the aim function is constructed for the practical situation of power system . on the basis of traditional genetic algorithm the fitness function and the holding of population diversity are improved 在提出基于相似日和线性外推的短期负荷预测新方法的基础上,将负荷预测和设备动作优先级引入静态无功优化中,并结合电网实际情况,构造了实用的目标函数,对遗传算法的适应度函数和群体多样性的保持进行了改进,采用邻域搜索算子增加遗传算法的局部寻优能力。