基于RBF神经网络的智能PID控制算法

Intelligent PID Control Based on RBF Neural Network

针对工业中的纯滞后现象提出一种智能控制方法,在常规的PID控制器中引入Smith预估器,对纯滞后时间τ所产生的特性进行预估补偿,同时利用RBF神经网络对PID的参数进行自整定,在一定程度上增强了系统的鲁棒性和稳定性。实验证明,此控制方法对纯滞后工业对象进行控制可以得到良好的效果。

Presents an intelligent controller against the phenomenon of time delay in industry.Smith predictor will be introduced into the conventional PID controller,in order to give compensation for the characteristics which produced by the time delay parameter τ,while use of RBF neural network to adjusts the PID parameters on line,to a certain extent,enhance the system＇s robustness and stability.The experiment has shown that use this intelligent controller can get satisfactory control performance.