用多种群遗传算法优化结构振动模糊控制器

Optimization of Fuzzy Controller for Structural Vibration Control Based on Multi-population Genetic Algorithm

在结构控制中模糊控制是被采用的方法之一 ,基于遗传算法的模糊系统的优化设计 ,把模糊控制和遗传算法结合起来 ,利用遗传算法的优点 ,克服了一般模糊控制设计中模糊变量的隶属度和控制规则的选取通常靠经验来获取的不足 ,使得系统的模糊控制设计更灵活方便 ,能取得更好的控制效果。采用实数和整数混合编码技术来设计遗传算法的染色体和相应算法 ,为了避免早熟和加快收敛速度 ,采用了多种群的遗传算法。为了验证优化的效果 ,对一个三层框架进行了仿真分析 ,结果显示 。

Fuzzy logic control is one of applied methods in structural control. Optimization of fuzzy controller is based on genetic algorithm that takes advantage of powerful optimization capability of GA and avoids the weakness of ordinary FLC designing in which control rules and membership functions of FLC are chosen by expert experiences. The method can improve FLC system and gets more effective control. In this paper, a genetic algorithm with real integer mixture encoding and relative operators was put forward. In order to prevent premature and quicken convergence, multi-population is considered. A simulation for a three-storey frame structure showed that the proposed method is better than the ordinary FLC.