基于多模型的自适应预测函数控制

Adaptive predictive functional control based on multiple -model

作为第三代模型预测控制算法,预测函数控制已经在线性系统控制中取得了很好的效果,但应用于非线性对象还有一定局限性.针对现实中大量存在的非线性过程,提出了基于局部模型设计的预测函数控制器加上模型切换算法的多模型自适应预测函数控制器.通过计算机仿真应用于连续发酵过程,表明基于多模型的自适应预测函数控制器比常规的预测函数控制器具有更好的控制效果.

Adaptive predictive functional control is proposed based on multiple-model. PFC (Predictive functional control), as a 3rd generation model based predictive control algorithm, has achieved great success in linear system, although there are still many problems in applying PFC for nonlinear system. As a typical nonlinear system, the continuous fermentation process can be approximated with local linear models. A PFC controller can be established on the basis of local linear models. Each controller independently calculates the manipulated variables used by a switching algorithm to calculate the final output. Simulation results confirmed that the adaptive predictive functional control based on multiple-model is superior to the traditional predictive functional control.