摘要
针对水轮发电机组系统具有时变非线性,传统的控制方法很难达到最优控制的特性,提出采用基于RBF神经网络和遗传算法的自适应模糊控制器来控制水轮发电机组运行.模糊控制器的比例因子、模糊推理规则和隶属函数由遗传算法在线寻优.由RBF神经网络进行被控对象的动态特性模型辨识,以评价模糊控制器控制性能.仿真实验表明,控制效果良好,特别在变工况和扰动情况下优于最优PID控制.
Considering the nonliner and much varying feather of hydraulic turbine system in operation, conventional hydraulic turbine PID governor cannot control effectively; a new self-tuning fuzzy governor based on RBF neural networks and genetic algorithms is designed. The parameters and rules of fuzzy controller are optimized based on GA in operating. Dynamic identification model of hydraulic turbine system is designed based on the RBF neural networks to appraise the the adaptive fuzzy controller is better controlling performance of fuzzy controller. Simulation results show that than the traditional PID control strategy.
出处
《武汉大学学报（工学版）》
CAS
CSCD
北大核心
2005年第4期9-12,16,共5页
Engineering Journal of Wuhan University
关键词
水轮机
模糊控制
遗传算法
RBF神经网络
hydraulic turbine
fuzzy control
genetic algorithms
RBF neural networks
