GPC design technique based on MQFT for MIMO uncertain system

According to the properties of the generalized predictive control theory and quantitative feedback theory, this paper presents two combined robust control methods. The first method uses MQFT to conquer plant uncertainty and stabilize the system. And the multivariable GPC (MGPC) controls. the system which has been preparatorily controlled by MQFT. Two closed loops are used in this design approach. Any kind of MQFT and any kind of MGPC can be used to realize this control technique. This method also includes the advantages of GPC and QFT. The second method is different from the first one because the single-input single-output (SISO) GPC is used in the outer loop. As the first method, using MGPC, it needs more calculation time and memory. But its application area is rather wide and its performance is better than the latter. In GPC and MGPC, the model is identified with the system input and output online. If some of the specifications are changed, the MQFT controller needn't be redesigned in the proposed methods since GPC can adapt to the changes. They can also be used to control unstable plants. Finally, the simulation shows that the integration of GPC and QFT have better performance than if only one of them is used.