一类非自衡系统的多变量预测函数控制

PREDICTIVE FUNCTIONAL CONTROL OF MULTIVARIABLE NON-SELF-REGULATING SYSTEM

针对过程控制中大量存在的多变量非自衡系统 ,研究了一类非自衡系统的多变量预测函数控制问题。常规预测函数控制主要针对渐进稳定系统 ,不能直接控制非自衡系统。根据含有一阶积分环节的非自衡系统的开环阶跃响应在某一时刻后可视为直线的特点 ,改进了常规预测函数控制算法 ,提出了一种可直接控制多变量非自衡系统的预测函数控制改进算法。该改进算法基于简单的开环阶跃响应模型 ,克服了非自衡系统预测模型截断误差的影响 ,具有模型易于获取、算法简单的特点。仿真结果表明 。

The problem of predictive functional control of multivariable non-self-regulating systems was considered for the non-self-regulating systems universally exist in the process control. The general PFC algorithm is mainly designed for asymptotic stable plants and mostly cannot be applied in non-self-regulating systems. Based on the characteristic that the open-loop step response for a non-self-regulating plant with a first-order integrating link can be regarded as a beeline after certain time, an existing PFC algorithm was improved, which could be directly applied in multivariable non-self-regulating systems. The simple open-loop step response for a non-self-regulating plant was adopted in the improved algorithm, truncated errors of non-self-regulating system model can be avoided by use of this algorithm, and this algorithm also has the strongpoint such as the model is easy to get and the method is very simple. The results of simulations show this algorithm is practicable for multivariable non-self-regulating system control.