基于遗传算法无源电力滤波器的优化设计

针对无源电力滤波器的参数选择比较困难,普通设计方法中过分依赖经验,优化能力不强的问题,研究了一种基于遗传算法的无源电力滤波器设计方法。利用遗传算法全局寻优的特点,对无源滤波器成本、无功补偿和滤波效果3个指标进行全局优化,并通过改变遗传交叉和变异概率来克服算法早熟收敛的情况。同时开发出界面友好,多种参数自选的自动设计软件,可帮助设计人员方便地进行无源电力滤波器的设计。某大型电机的实际滤波结果表明,所提出的设计方法正确有效,所开发的软件实用性强。

Optimization Design of Passive Power Filter Based on Genetic Algorithms

Conventional approaches for passive power filters design depend on experiences too much and lack of ability for optimization because design of passive power filters is comparative hard. A new design approach for passive power filters based on genetic algorithms was adopted. It macroscopically optimized for the cost, harmonic suppression and the effect of filtering of passive power filter. Changing the probability of inheritance and variation improve the forwardness of genetic algorithms. And auto design software was given, which had the good interface and many self-help parameters. It can help engineer design passive power filters conveniently. The simulation and experiment indicates that this approach is correct and efficiency, and the software are worth generalizing.