基于适应度空间距离评估选取的多目标粒子群算法在电网无功优化中的应用  

Application of Multi-Objective Particle Swarm Optimization Algorithm in Power System Reactive Power Optimization Based on Evaluation of Distance in Fitness Space

作　　者：娄素华[1] 吴耀武[1] 熊信银[1] 

机构地区：[1]电力安全与高效湖北省重点实验室(华中科技大学),湖北省武汉市430074

出　　处：《电网技术》2007年第19期41-46,共6页Power System Technology

摘　　要：提出了一种基于适应度空间距离评估选取最优解的多目标粒子群算法。该方法避免了目前多目标优化求解方法中权重选择的难题,保证了寻优方向的多向性,可以获得多目标优化问题的Pareto解集。将该算法应用于网损最小、静态电压稳定裕度最大为目标的多目标无功优化问题,算例表明在有效性和最优性等方面均有良好表现。A multi-objective reactive power optimization algorithm, which chooses solution by means of evaluating distance of fitness space, is proposed. Using this algorithm, the puzzle of weight selection in the solution of current multi-objective optimization algorithms can be avoided and the multidirectional in the searching process can be ensured, thus the Pareto optimal solution set of multi-objective optimization problem can be achieved. Applying the proposed algorithm to multi-objective reactive power optimization and taking minimized network loss and maximized static voltage stability margin as objective function, the results show that the computation complexity is reduced, the convergence precision is improved.

关 键 词：多目标 无功优化 适应度空间距离 评估向量选取 多目标粒子群算法 

分 类 号：TM712[电气工程—电力系统及自动化]