一种基于双偏差的动态预测控制算法

A Dynamic Predictive Algorithm Based on Double Errors

针对大惯性大迟延对象控制特性过渡过程时间长、响应特性差的特点 ,文中提出了一种新的预测控制算法。该算法采用单步动态预测模型 ,并在算法中引入了模型与对象、参考轨迹和预测输出的双向偏差以进行优化 ,在不增加算法复杂性的基础上 ,提出了一种基于双偏差的动态预测控制算法。并与传统 PID控制方式比较 ,其动态品质大大提高。

It is well known that the industrial process with large time delay is hard to be controlled. Since the controlled process with large inertia and time delay has long transient time and poor response characteristics, a new predictive control algorithm is presented in this paper to improve control performance. The single step dynamic predictive model is utilized,furthermore,double deviations have been introduced in the algorithm,one is the difference between model and controlled process,the other is the difference between reference locus and predictive output. Finally,this algorithm is compared with traditional PID controller, it has shown that the dynamic quality is improved. The simulation result is satisfactory.