中国电机工程学报2012，Vol.32Issue(28)：90-96,7.
基于外部存档的并行遗传算法在水轮机调速器参数优化中的应用
Optimization of Hydro-turbine Governor Parameters Based on Parallel Genetic Algorithm With Exterior Archive Method
唐正茂 1马士虎 2解德1
作者信息
- 1. 华中科技大学船舶与海洋工程学院,湖北省武汉市430074
- 2. 武汉第二船舶设计研究院,湖北省武汉市430064
- 折叠
摘要
Abstract
The way of tuning proportional- integral-derivative (PID) parameters for hydro-turbine governor is usually based on empirical formula. However, it is difficult to reach the optimum parameters. A novel tuning strategy based on multi-objective genetic algorithm and multi-attribute decision-making (MADM) method was proposed, which aimed at three controller performances simultaneously, which are accurate setpoint tracking, disturbance attenuation, and robust stability. The parallel multi-population genetic algorithm, which improved the intercommunion and migration among each population, and the concordance evolutionary modes, was applied to reach the global optimization. Two exterior archives were taken as containers. The one kept the estimated Pareto optimal set for MADM as possible as the algorithm searched. The other used niched technology and the distance between the two individuals in vector space to adjust the unevenly Pareto front set. An evenly distributed Pareto optimal subset was formed for migration guidance. So the multi-population could search Pareto front in a coordinated fashion. In the subsequent stage, a multi-attribute decision making approach was adopted to find the "best" compromising one among Pareto optimal solutions to the best of algorithm's ability. Simulation result indicates that the performance of hydro-turbine governor could be improved by the proposed method and the method is simple and feasible.
关键词
水轮机调速器/参数优化/并行遗传算法/外部档案/多目标优化/多属性决策Key words
hydro-turbine governor/parameter optimization/parallel genetic algorithm/exterior archive/multi-objective optimization/multi-attribute decision making分类
信息技术与安全科学引用本文复制引用
唐正茂,马士虎,解德..基于外部存档的并行遗传算法在水轮机调速器参数优化中的应用[J].中国电机工程学报,2012,32(28):90-96,7.
