Remote UGV Control via Practical Wireless Channels: A Model Predictive Control Approach
Authors:
inghao Cao,
Subhan Khan,
Wanchun Liu,
Yonghui Li,
Branka Vucetic
Abstract:
In addressing wireless networked control systems (WNCS) subject to unexpected packet loss and uncertainties, this paper presents a practical Model Predictive Control (MPC) based control scheme with considerations of of packet dropouts, latency, process noise and measurement noise. A discussion of the quasi-static Rayleigh fading channel is presented herein to enhance the realism of the underlying…
▽ More
In addressing wireless networked control systems (WNCS) subject to unexpected packet loss and uncertainties, this paper presents a practical Model Predictive Control (MPC) based control scheme with considerations of of packet dropouts, latency, process noise and measurement noise. A discussion of the quasi-static Rayleigh fading channel is presented herein to enhance the realism of the underlying assumption in a real-world context. To achieve a desirable performance, the proposed control scheme leverages the predictive capabilities of a direct multiple shooting MPC, employs a compensation strategy to mitigate the impact of wireless channel imperfections. Instead of feeding noisy measurements into the MPC, we employ an Extended Kalman Filter (EKF) to mitigate the influence of measurement noise and process disturbances. Finally, we implement the proposed MPC algorithm on a simulated Unmanned Ground Vehicle (UGV) and conduct a series of experiments to evaluate the performance of our control scheme across various scenarios. Through our simulation results and comparative analyses, we have substantiated the effectiveness and improvements brought about by our approach through the utilization of multiple metrics.
△ Less
Submitted 13 March, 2024;
originally announced March 2024.
