模糊自学习控制在水电机组控制中应用研究

Research on Fuzzy Self-learning Controller Used in Hydroelectric Generating Unit

采用具有自学习能力的自适应模糊控制器来控制水电机组运行。自适应模糊控制器将模糊控制和神经网络结合,根据运行情况在线调整模糊推理规则和隶属函数,使控制系统具有自适应学习的特性。学习中学习速率和平滑因子可根据误差情况在线修改,克服了网络学习速度慢和局部最优的缺点。仿真实验表明,设计的自适应模糊控制器具有良好的鲁棒性,可有效地改善水轮发电机组系统的动、静态性能。

Conventional controller cannot get good controlling performance in the control of hydroelectric generating unit system because the system is non-linear and complicated.In the paper new fuzzy self-learning controller is designed to control hydroelectric generating unit.The designed new fuzzy controller combines neural network into fuzzy reasoning system.Fuzzy reasoning rules and member function may be trained on-line via neural network,which makes control system having the character of self-adapting.In order to accelerate learning and converging quickly,parameters may be adjusted according to error and error change.Simulation experiment shows that the designed controller has stronger robust performance and excellent static and dynamic performance.