基于神经网络的自适应PID控制器设计

The Design of Adaptive PID Controller Based on Neural Network

为了提高PID控制器参数根据实际误差自动调整能力,提出了基于神经网络的自适应PID控制器的设计。详细分析了神经网络自适应控制原理,设计了通过神经网络对PID控制器参数自动调整的方案。利用Matlab仿真工具,对自适应PID控制器进行了仿真测试,给出了神经网络训练参数,通过大量数据的模拟测试,结果表明基于神经网络的自适应PID控制器能够根据控制误差自动调整PID参数,对输入的阶跃信号响应平稳,综合性能大大提高。

To improve the auto-adjust ability to practical error,it raised the design of adaptive PID controller based on neural network.It specifically analyzed the adaptive principle,and designed the auto-adjust implementation scheme of PID controller parameters by neural network.Using Matlab as a simulation tool,the adaptive PID controller was emulation tested and the neural network training parameter was given.By the simulation test of mass data,it shows that the adaptive PID controller which based on neural network can automatically adjust PID parameter by controlling error,and has stable response to step signal.Its comprehensive performance is greatly improved.