基于GM(1,2)模型的多步自调节灰色预测控制算法

Multi-step self-adjustable grey prediction control algorithm based on GM(1,2)model

基于GM(1,2)灰色模型和一种新型综合偏差,提出了基于GM(1,2)模型的多步自调节灰色预测控制算法.该算法可对系统输出偏差和预测偏差进行合成形成一个在线多步综合偏差,用此偏差代替传统最优控制算法中的偏差项,这样算法既有在线预测系统未来多步行为的功能,又能根据预测模型精度实时调节预测值在控制回路中的作用,减小了预测误差对系统的不利影响.仿真结果表明,该算法可获得较好的控制跟踪效果.

The multi-step self-adjustable grey prediction control algorithm can combine the output deviation with prediction deviation of a system to produce a on-line multi-step integrated deviation,and the deviation item in traditional optimal control algorithm is substituted by the integrated deviation.So the multi-step self-adjustable grey prediction control algorithm can not only predict the future multi-step behaviors of the system online but also adjust the function of the predicted value in control loop on-line according to the precision of the prediction model,which reduces the unfavorable effect of the predicted deviation on the system.Simulation result shows that this algorithm can achieve a good control result.