基于BP神经网络的PID控制器

A Neural Network PID Controller Based on BP

介绍了一种基于神经网络自学习PID控制器,该控制器能通过自学习不断进行适应性控制,以保证系统的输出符合实际应用的要求, 其主要特点是采用线性预测模型来近似确定控制参数,进而进行神经网络控制,仿真结果表明该方法有较好的效果。

A Neural Network PID Controller Based on BP is presented in the paper. The system can continue adapt ability-control by self-learning to make sure that the systemic output meets the practical requirements, so it has some application value. Its main characteristic is that it adopts linear predicting modelto confirm the control parameter approximately, then controls the object by use of the neural network. The simulation in the paper also indicates the method has better control effect than others.