论文部分内容阅读
ＤＭＣ（动态矩阵控制）是预测控制的典型算法之一，它具有建模简单，计算量少，鲁棒性强等特点。但是，现有的各种ＤＭＣ算法，在任一时刻的误差信息均是采用该时刻的实际输出与上一时刻的一步模型输出预测的差值，并以此补偿基于模型的预测，当模型失配较为严重时，其控制品质不太理想。为此，利用预测误差的历史数据建立误差预测模型，通过对误差的预测来修正基于非参数模型的过程预测输出，从而得到了具有误差预测修正功能的ＤＭＣ算法。仿真结果表明，该算法能够较好地克服建模误差的影响，提高了控制品质。
DMC (Dynamic Matrix Control) is one of the typical algorithms for predictive control. It has the characteristics of simple modeling, less computation and strong robustness. However, in current DMC algorithms, the error information at any time is calculated by using the difference between the actual output at that moment and the one-step model output prediction at the previous moment to compensate for the model-based prediction. When the model is mismatched More serious, its control quality is not ideal. Therefore, the error prediction model is established by using the historical data of the prediction error, and the prediction output of the process based on the non-parametric model is corrected through the prediction of the error. Thus, the DMC algorithm with error prediction and correction function is obtained. The simulation results show that the algorithm can better overcome the influence of modeling error and improve the control quality.