چكيده به لاتين
This dissertation addresses the problem of designing a robust tube-based model predictive control for the linear parameter-varying systems. In this class of systems, the state space matrices are affine functions of the time-varying scheduling parameter that indicates the nonlinearities of the orginal system and its behavior in the future is usually unknown and uncertain. Using the robust predictive control method, the effects of uncertainty in the future evolutions of the scheduling signal, efficiency, stability and robustness in the presence of external disturbances can be involved in the controller design procedure. Therefor, in the first step of this dissertation, the problem of designing a robust tube-based model predictive control is investigated by considering the future behavior of the scheduling signal. In order to increase the domain of attraction of the closed-loop system, the tube cross-sections are parameterized such that the volume and geometric shape of each cross-section are considered as decision variables in the optimization problem. Therefore, the predictive controller has the maximum flexibility and degree of freedom in designing tubes and in determining the optimal control signals; this reduces the conservatism in computing the control inputs. Therefore, at each sampling instant, the tube parameters and the optimal control inputs are obtained online by solving a constrained optimization problem. On the other hand, in many of control applications, the closed-loop system is configurated as Networked Control System (NCS), in which to reduce the utilization of the communication and computational resources, the aperiodic control algorithms can be used. In self-triggered control, at every sampling instant, the controller provides both the next sampling instant, as well as the inputs that are applied to the system until the next sampling instant. Therefore, in the second step of this dissertation, a robust self-triggered control algorithm for constrained linear parameter varying systems is proposed. In order to reduce the number of samplings and control updates without deteriorating the control performance, at very triggering instant, the proposed algorithm determines the next update time and the applied control values simultaneously. In the following, the analysis of recursive feasibility and proving the closed-loop stability are addressed. The simulations are performed on several numerical examples and practical systems. The simulation results show the satisfactory performance of the proposed methods in comparison with other methods presented in the literatures, as well as time-triggered strategies.
