神经网络PID复合智能控制参数自整定研究

针对具有时滞一阶非最小相位系统的被控对象,提出了一种基于神经网络的PID复合智能控制设计方案,推导出复合PID控制器控制参数的整定方法.考虑到被控对象的不确定性、慢时变等特点,以神经网络元构造基于自动化复合PID控制器,解决了基于自动化复合PID控制参数在线调整的问题,使复合PID控制器适用范围更广泛.对闭环控制系统模拟结果表明,该系统控制参数整定方法简单,具有非常好的控制效果.

Research on self-setting control parameters of PID composite intelligent controller based on neural network

In light of a controlled object of non\|least phase system with first order time\|lag, a design scheme of PID composite intelligent control based on neural network has been presented. And a method for setting control parameters of a PID composite controller has been derived. Considering the characters of the controlled object such as uncertainty, slow time\|varying etc., an automatic composite PID based controller is constructed with neural network elements. Therefore, the problem of on\|line setting control parameters based on automatic composite PID is solved so as to widen the scope of suitable application of the composite PID controller. The simulation results for closed loop control system show that the system is simple in setting method of control parameters and good in control effectiveness.