Data-driven event-triggered control of discrete-time nonlinear systems under mismatched input quantization and disturbance

This paper is concerned with data-driven event-triggered control of discrete-time nonlinear systems under mismatched input quantization and disturbance. A neural network method is proposed by gradient descent to estimate disturbance and unknown parameters. An improved data-driven control law employing adaptive dynamic programming strategy is supplied to compensate disturbance. A dynamic triggering rule is constructed by instantaneous and average output to reduce transmission times. Sufficient conditions are provided for resultant tracking error system to be ultimately uniformly bounded. A heat exchanger system is presented to illustrate the validity of the proposed method.