基于神经网络PID控制的系统非线性校正的研究

Study on nonlinear adjustment of neural network PID control system

在对BP神经网络PID控制器系统研究的基础上,提出了单神经元的自适应PSD算法。该算法兼有单神经元和自适应PSD算法的特点,简单、实时性好、自适应能力强,可用于控制过程时变、有大滞后的较复杂的对象,是一种实用价值较高的自适应控制算法。文中采用BP神经网络PID控制与单神经元PSD自适应控制两种方法对压电式微位移系统进行非线性控制,并取得了良好的效果。

Based on the study of BP neural network and PID controller, a single neuron adaptive PSD algorithm is presented. With the characters of simple, real time and adaptive, it is a kind of adaptive control algorithm with high practicality and can be used to control the complex object with the character of timevarying or large time delay. Both BP neural network PID control and single neuron PSD adaptive algorithm are applied to realize the non-linear control of piezoelectric micro displacement system. The result shows that these two algorithms are efficient.