无源滤波器多目标优化设计及实验仿真  

Multi-objective optimal design and experimental simulation for passive power filters

作　　者：陆秀令[1] 张松华[1] 周腊吾[2] 曹才开[1] 

机构地区：[1]湖南工学院,湖南衡阳421008 [2]湖南大学,湖南长沙410082

出　　处：《华东电力》2007年第4期29-32,共4页East China Electric Power

基　　金：湖南省自然科学基金资助项目(03YJJ5005);湖南省教育厅科学研究资助项目(05C002)

摘　　要：针对现有无源电力滤波器的设计方法中过分依赖经验和优化能力不强的情况,利用改进遗传算法的全局寻优能力,提出了一种基于遗传算法的无源电力滤波器设计方法,即从无源滤波器的成本、无功补偿和滤波效果3个目标全局优化。通过适应度函数的阈值制约以及以不同概率进行染色体选择操作,使得种群朝3个目标最佳协调点的方向进化。通过混沌算子来解决早熟收敛的问题。最后利用PSIM软件建立仿真模型,仿真实验结果说明了该方法在无源滤波器优化设计中的有效性和正确性。By using the global optimization ability of the improved genetic algorithm, a genetic algorithm-based passive power filter （PPT） design method which realizes global optimization in the aspects of PPF cost, reactive compensation effect, and filtering performance is proposed. Through threshold limitation for the fitness function and chromosome selection with different probabilities, the population is made to evolve to the best tradeoff of the three objectives. A chaotic operator was introduced to overcome premature convergence, and the simulation model was established by PSIM software. Simulation proves the design method effective and accurate.

关 键 词：改进遗传算法 多目标 优化设计 PSIM仿真 

分 类 号：TM714.3[电气工程—电力系统及自动化]