A Prediction-Based Approach Realizing Finite-Time Convergence of Networked Control Systems

This brief presents a networked model predictive control (MPC) strategy for the finite-time convergence of networked control systems with two-channel random communication constraints. Network-induced delays and packet dropouts in the backward and forward channels are transformed into the equivalent delays in each channel, which are actively compensated in the controller and actuator, respectively. Future control inputs are obtained by a finite-time MPC strategy, and the upper bound of the convergence time is given by theoretical analysis. Finally, the proposed scheme is verified by the numerical simulation of a mass-spring-damper system.

Automated summary

This brief introduces a networked model predictive control (MPC) strategy for the finite-time convergence of networked control systems with two-channel random communication constraints. It transforms network-induced delays and packet dropouts into equivalent delays in each channel, which are actively compensated in the controller and actuator. By using a finite-time MPC strategy, future control inputs are obtained, and the upper bound of the convergence time is analytically determined. The proposed scheme is verified through a numerical simulation of a mass-spring-damper system.