基于粒子群算法和内点算法无功优化研究.PDF

第 38 卷 第 13 期 电力系统保护与控制 Vol.38 No.13 2010 年 7 月 1 日 Power System Protection and Control Jul. 1, 2010 基于粒子群算法与内点算法的无功优化研究 张 锋，段余平，邱 军，冯小琴 (武钢股份能源动力总厂，湖北 武汉 430080) 摘要：提出了基于改进粒子群算法和预测-校正内点法的解耦无功优化算法。通过引入时代因子和邻近变异策略，同时采用 分段处理方法对粒子群算法进行改进。运用预测-校正算法替代原-对偶内点，使得在内点法寻优过程中的迭代步长加大，同 时避免寻优过程中振荡的出现。将改进粒子群算法和预测-校正内点算法分别用于无功优化的离散优化和连续优化子问题。 将所提出的方法应用于IEEE30节点和IEEE118节点的系统。算例表明：与采用传统粒子群算法和原-对偶内点算法的混合无 功优化相比，提出的方法在计算速度和优化效果方面都具有明显的优势。 关键词：无功优化；粒子群算法；内点法；预测-校正内点算法 Research on reactive power flow based on particle swarm optimization and interior point method ZHANG Feng ，DUAN Yu-ping ，QIU Jun，FENG Xiao-qin (Energy and Power Plant, Wuhan Iron Steel Group Corp (WISCO)，Wuhan 430080 ，China) Abstract: Decoupling reactive power optimization method is proposed. By introducing the time factor and the neighboring mutation strategies, while using segmentation approach to improve particle swarm optimization. The use of predictor-corrector algorithm makes the iteration step length longer while avoiding oscillating. The improved particle swarm optimization and prediction-correct interior point algorithm are used to discrete optimization and continuous optimization sub-problems of reactive power optimization respectively. The proposed formulation is applied to an equivalent of an area of the IEEE 30 and 118 buses test system and the results illustrate that the proposed method is effective in calculation speed and optimization result. Key words: reactive power flow ；particle swarm optimization ；interior point method ；predictor-corrector interior point method 中图分类号： TM715 文献标识码：A 文章编号： 1674-3415(2010)13-0011-06