The Study Of Generalized Predictive Control In Delta Domain

The discrete sampled systems have been developing rapidly with the development of computer technology. But when sampling rates are high, most traditional digital signal processing and control algorithms are inherently ill-conditioned. Considerably recent progress toward ameliorating such as ill-conditioned problems has been made by the use of delta operator, rather than the conventional shift operator, to represent the dynamics of sampled data. And in processing high sampling data system, discrete time model using Delta operator approaches corresponding continuous time model, which is the remarkable advantage of the Delta operator. On the other hand, generalized predictive control, which is known as a new type of control approach, has been applied successful in industry. The merger of Delta operator theory and generalized predictive control is a new theme, which is significant for practical control systems, especially for high speed sampled systems.Based on the methods of Delta operator theory and the generalized predictive control, this paper studies the Delta domain generalized predictive control. Some initially control theory of Delta domain generalized predictive control are investigated as follows.The main contents are stated as follows:1. Emulated-based DGPC (Delta-operator Generalized Predictive Control) with state feedback is proposed and emulator-based DGPC with state feedback are discussed. The structure of state feedback generalized predictive control in delta domain is a dynamic state variable feedback control using the estimated state variable feedback control and the ability of rejection undetectable disturbance is increased and the robustness of the controlled system is proved.2. The major weakness of GPC is the lack of general stability guarantees. Instead of SIORHC/CRHPC and SGPC for general stability guarantee, Infinite horizon generalized predictive control in Delta-domain are proposed. We adopt infinite costing horizon which is known to lead stability but it has its own implementation difficulties. In order to improve the stability of the controller, a practical means of implementingδ? GPC∞with infinite costing horizon with inequality constraints is presented.3. For the large scale computing burden of receding horizon optimization, Delta-operator GPC Based on Linear Matrix Inequalities is proposed. With the help of LMI approach, a output feedback controller is obtained. Then the structure of stabilation controller and the existence condition of stable controller are presented.