Research On Stability Of A Scalar Nonlinear Quantitative Feedback System Based On Event Triggering

Recently,Networked Control Systems and event-triggered control have caught a great deal of attention.This dissertation considers to stabilize a scalar continuous-time nonlinear system based on the event-triggered control.Meanwhile,stability is expected to be maintained at as low as possible feedback bit rate.Notice that the obtained stabilizing bit rate conditions are only up to some inherent system parameters.Compared with the traditional periodic sampling method,a lower bit rate is required by the proposed event-triggered strategy.Thus,the bandwidth of the feedback network is saved.The main contributions can be summarized as below:1)Based on the model-based continuous time event-triggered control scheme,we consider to stabilize the scalar continuous time nonlinear system with bounded network delay,noise and model uncertainty.We design the exponential convergence triggering condition and appropriate updating rules.Each event is triggered and only 1 bit is utilized to represent the sign of the estimation error.Extra state information is actually extracted from the receiving time instants under the proposed method.Furthermore,a lower occupied bit rate than the periodic sampling strategy is obtained while guaranteeing the desired input-to-state stability of the system.The stabilizing bit rate conditions are only determined by the parameter of Lipschitz condition,the upper bounds of the network delay and the model uncertainty,but are not concerned with the system state or quantized state at all.In particular,the proposed method is also efficient for the linear system.Simulations are done to verify the effectiveness of the achieved stabilizing bit rate conditions.2)We propose the model-based periodic event-triggered control strategy to stabilize a scalar continuous time nonlinear system with bounded network delay and noise.Without ACK,the updating of nominal models at the sensor and at the controller are asynchronous.Four different scenarios of the nonlinear systems are investigated.Compared with the periodic sampling method,our scheme can make full use of the received data packets,particularly their receiving time instant information,which yields a lower occupied bit rate than the periodic sampling strategy in order to guarantee the desired input-to-state stability of the system.Note that our stabilizing bit rate conditions depend only on the parameter of Lipschitz condition,the upper bound of the network delay,the number of quantization bits,and the sampling period.The bounded noise will not deteriorate the upper bound of the stabilizing bit rate conditions based on our proposed strategy.