双重控制系统的广义预测控制

来源期刊：中南大学学报(自然科学版)2003年第4期

论文作者：李平 李刚 孟令柏

文章页码：386 - 389

关键词：双重控制;广义预测控制;胖系统;静态目标

Key words：dual-control; generalized predictive control; fat system; steady state object

摘    要：传统的双重控制方法建立在简单PID算法基础上,其性能受到一定的限制.为此,将广义预测控制算法应用于双重控制系统中,形成基于预测控制的双重控制算法.该算法继承了传统双重控制方法的控制结构和控制思想,采用主、副2个控制器进行控制.其中主控制器采用广义预测控制算法,针对快回路产生快控制作用使系统具有良好的动态特性;副控制器采用PID算法,针对慢回路产生慢控制作用以实现一定的经济指标;同时,副控制器的输出作为前馈量引入主控制器.主、副2个控制器互相协调工作,使系统既具有良好的动态特性,又符合工艺上或经济上的要求.仿真对比结果表明,该算法控制性能优良,具有较强的的抗干扰性和良好的鲁棒性,是一种较好的双重控制系统控制算法.

Abstract: The traditional dual-control system is based on simple PID algorithm and its performance is limited. A Generalized Predictive Control (GPC) based Dual-Control algorithm is proposed by introducing GPC algorithm to the Dual-Control System. The algorithm keeps the control thinking and structure of the traditional Dual-Control System, where two controllers are designed. The prime controller using GPC algorithm focuses on the rapid loop and its aim is to make good dynamical response, and the secondary controller focuses on the slow loop and PID algorithm is used to satisfy some steady state index. The output of the secondary controller is used as the feedforward input of the prime controller. The two controllers cooperate together both to make the system a good dynamical feature and to satisfy the technology or economy requirement of the system. The simulation results show that the proposed algorithm provides good control performance, strong robustness and fine anti-disturbance capability.