Stabilization Study Of Networked Systems With Quantization

The study of quantization errors in digital control systems has been an important area of research since digital controllers have been employed in feedback systems. Nowadays, many control systems are remotely implemented via communication channels with limited bandwidth, which we will refer to as remote control systems. In such systems, the communication link is commonly shared by different application processes, and a natural issue is to minimize the quantity of information needed to be transmitted while achieving a certain closed-loop performance. In such remote control systems, where the plant and the controller are physically distributed, the sensor information (sent to the controller) and the control signal (feed back to the plant) are connected via a communication network. In other words, the feedback information is exchanged through a shared communication channel among the control system components (sensors, controller, actuator, etc.). Thus, it is natural to suppose that both control and measurement signals are quantized before transmission. A simple classical approach to analysis of quantization effects is to treat the quantization error as uncertainty or nonlinearity and bound it using a sector bound. Meanwhile, the robustness analysis tools, such as absolute stability theory (see [10] and [11]), can be applied to study the quantization effect. Further, control parameters can be optimized by minimizing the quantization effect, which is called the sector bound method in the quantized control field. Hence, it is important to investigate networked control systems with input and output quantization by sector bound method in theory and practice.The main research results of this paper are as follows:In chapter1, the theoretical significance of this thesis is introduced. And the research situation at home and abroad is recalled.In chapter2, the dynamic output-feedback control problem is considered for linear networked control system involving in signal quantization and data packet dropout. The state of the controlled system is unavailable and the measured output signal and the controlled output signal are logarithmic quantized before being communicated. Because of the limited bandwidth of the channels, such packet dropouts can occur stochastically in the communication channels from the sensor to the controller and from the controller to the actuator. Then, the closed-loop NCS is modeled as a discrete-time switched system with four subsystems. By using the sector bound approach and the average dwell-time method, sufficient conditions for the exponential stability of the closed-loop NCS are presented in terms of nonlinear matrix inequalities, and the dynamic output feedback controller can be obtained by solving a nonlinear minimization problem.In the third chapter, the stabilization problem the networked control systems with random packet loss and logarithmic quantization. With the given packet loss rate and the quantization error parameters, using the Lyapunov functional method, the dynamic output feedback controller is designed in the term of linear matrix inequalities.In the fourth chapter, the stabilization problem for linear networked systems with packet dropout and time delay is invested. The networked control system is modeled as a switching system with time-delay by the delay input method. By using the switching system bound approach and the piecewise Lyapunov method, sufficient conditions for the exponential stability of the closed-loop NCS are presented in terms of nonlinear matrix inequalities, and the dynamic output feedback controller can be obtained by solving linear matrix inequalities (LMIs).In chapter5, the main results of the paper are concluded, and some research directions in future are proposed.