分析了求解多故障暂态稳定约束最优潮流的轨迹灵敏度法存在的问题,提出了相应的改进方案。利用轨迹灵敏度法的快速收敛特性和微分进化算法的全局搜索能力,借助用轨迹灵敏度法求解暂态稳定约束最优潮流问题过程中产生的多个逼近全局最优点的稳定和不稳定点来构成微分进化算法的初始种群,将两者结合起来构造成求解多故障TSCOPF的混合算法。在3机和10机试验系统上的计算表明,所构建的混合算法收敛性快且全局搜索能力强。
According to the analysis of existing problems in trajectory sensitivity methods for transient stability constrained optimal power flow (TSCOPF) under multi-contingency condition,an improved scheme is proposed.In order to take advantage of fast convergence of the trajectory sensitivity method and global search capability of the differential evolution(DE)method,a hybrid algorithm for TSCOPF under multi-contingency condition is formed by combining these two methods,in which many stable and unstable points approaching the global optimal point are produced during solution process of TSCOPF by the trajectory sensitivity method,and applied to initialize the DE’s first population.Calculations on 3-machine and 10-machine systems show that the proposed hybrid algorithm has good convergence and strong global search capability.