چكيده به لاتين
Nowadays, with the development of networked control systems, event-triggered control strategies have received considerable attention due to their advantages over the time-triggered control frame work. On the other hand, the problem of controlling uncertain nonlinear systems is a hot topic due to its variety of applications in various areas; and among the existing nonlinear control techniques, the dynamic surface control method has attracted considerable attention due to the advantages of simplicity of the control law and its ability in controlling the nonlinear systems with mismatched uncertainties. This thesis investigates the event-triggered dynamic surface control of strict-feedback uncertain nonlinear networked systems. The controller is designed for both the ideal networks and networks with transmission time delays. In each case, some appropriate error signals and Lyapunov candidate functions are firstly defined, and the virtual control signals and the control input are de signed such that the time derivative of the Lyapunov functions becomes negative definite. By including a set of low-pass filters in the design procedure, the proposed method does not involve model differentiation and does not face the mathematical difficulties of the "explosion of terms" problem. By imposing a constraint on the change value of the control input than its latest transmitted value through the network, an event-triggering mechanism is constructed, which reduces the data trans mission in the network and saves limited communication bandwidth. Then, the closed-loop error dynamics is derived and rewritten as a linear dynamics with some perturbation terms by using the Lipschitz property of the nonlinear functions. Finally, the ultimate error boundedness and quadratic tracking objectives are addressed in the framework of convex optimization. The absence of the Zeno behaviour, which is an important feature of any event-triggered methodology, is also proved. The proposed method, which is firstly presented for single-input single-output strict-feedback nonlinear systems, is also extended to multi-input multi-output strict-feedback nonlinear systems. The validity of the theoretical results is illustrated by applying the proposed method to a remotely operated underwater vehicle and a vehicle platoon via simulation as well as an experimental implementation on a magnetic levitation system.
