Loading...
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 48122 (05)
- University: Sharif University of Technology
- Department: Electrical Engineering
- Advisor(s): Haeri, Mohammad
- Abstract:
- The focus of this thesis is on robust control of hybrid systems using tube-based model predictive control. Hybrid systems are systems that contain both continuous and discrete components. The hybrid nature of these systems can stem from inherent switching behavior that occurs in many dynamical or digitally controlled systems, or from approximation of nonlinear systems. The ability of these systems to model a vast amount of physical phenomena has been attracted a lot of attention from control society. Although much effort have been made on the control of “nominal” hybrid systems, robust stabilization of these systems still seems to be immature. One of the most successful approaches toward robust control of constrained systems is tube-based MPC, which have been widely studied for linear and “continuous” nonlinear systems. However, utilization of this method for hybrid systems seems to be inadequate and further research is needed in this area. This thesis tries to extend tube-based model predictive control approach toward piecewise-affine systems, by means of noble use of set invariance theories. This leads to new formulation of model predictive control problem, which ensures stability of close-loop system in the presence of disturbance. Furthermore, a new dual-mode control method is proposed to expand the classes of PWA systems under study. In addition, the abovementioned methods are incorporated in order to guarantee robust tracking of piecewise constant set points. The performance of close-loop system with use of controllers attained via proposed methods, have been investigated by means of many illustrative examples
- Keywords:
- Hybrid System ; Dual Mode Controllers ; Robust Control ; Tube-Based Model Predictive Control ; Constrained Piecewise-Affine Systems ; Dual-Mode Control ; Set Invariance Theories