纸浆浓度的模糊神经网络自适应PID控制  

Fuzzy Neural Nework Adaptive PID Control of Pulp Consistency

作　　者：吴新生[1] 

机构地区：[1]广东科学技术职业学院广州学院,广州510640

出　　处：《计算机测量与控制》2013年第11期2969-2971,共3页Computer Measurement &Control

基　　金：广东省自然科学基金资助项目(8451064007000003)

摘　　要：纸浆浓度的控制是一个大纯滞后和建立精确数学模型困难的系统,使用常规的PID控制器很难保持最优的控制效果;为了得到最优的控制效果,设计了纸浆浓度的模糊神经网络PID控制系统;系统采用模糊神经网络根据控制过程的纸浆浓度偏差和偏差变化率来实时和在线地整定PID控制器的参数,以适应纸浆浓度控制系统的环境参数变化的要求;仿真结果表明:系统的控制模型在对象增益k=3,惯性时间T=2,滞后时间τ=4时,模糊神经网络整定PID控制器得到的控制参数为Kp=0.07,Ki=0.03,Kd=0.017。仿真结果表明:模糊神经网络PID控制器在响应快速性、调节平稳性及抗干扰能力等控制性能均优于常规PID控制器和模糊PID控制器;这为纸浆浓度的最优控制提供了一种新的控制方法。It is difficult to obtain the optimal control effects with conventional PID controller because the control system of pulp consis tency is a large and pure time delay system and an accurate mathematical control model is difficult to be obtained. To solve the problem, the PID controller based on fuzzy neural network was adopted as a regulator in the pulp consistency control system. The parameters of PID con troller can be online set in real time considering error and error changing rate so as to suit the requirements of pulp consistency control sys tem. As the parameters of mathematical control model were set as gain k= 3, inertia time T= 2, delay time r： 4, the parameters of PID were tuned by fuzzy neural network and obtained as followed： Kp = 0.07, Ki = O. 03, Kd = 0. 017. Simulation results show that the PID pulp consistency control system based on fuzzy neural network is more excellent than conventional PID controller and fuzzy PID controller in aspects of response rapidity, regulation stability and antiinterference ability. It is a new optimal method for control of pulp consistency.

关 键 词：模糊神经网络 PID控制器 纸浆浓度 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]