Observer-based event-triggered model-free adaptive sliding mode predictive control

Abstract

For simple-input and simple-output (SISO) discrete-time nonlinear systems, an observer-based event-triggered model-free adaptive sliding mode predictive control technique (EMFASPC) is put forth in this study. The estimate of pseudo partial derivatives (PPD) and the transmission of I/O data are both carried out periodically at the time of event triggering to conserve network resources. A unified framework of event-triggered modelfree adaptive control with an adaptive observer and an event-triggered PPD estimation method is constructed based on the equivalent data model after compact format dynamic linearization (CFDL). The controller part adopts integral sliding mode control (SMC) combined with a rolling optimization idea of model predictive control (MPC) to predict the expected trajectory of the sliding mode state and generate the optimal control input. According to the relationship among the system tracking error, current measurement data, and the previous trigger time output, the event trigger condition is set to determine the next event trigger time, which reduces the unnecessary transmission on the premise of system stability. The stability performance of the closed-loop system is analyzed by the Lyapunov method. Finally, numerical simulation and the shell-and-tube heat exchanger control system simulation are carried out to verify that the proposed algorithm has good robustness and tracking accuracy under the limited bandwidth and computing resources.

Keywords

• adaptive observer
• compact format dynamic linearization
• event-triggered
• sliding mode predictive control
• model-free adaptive control