Torque Vectoring Control for Remote-control Driving of Differentialdrive Mobile Robots

This paper proposes a torque vectoring control algorithm for remote control considering the dynamic characteristics of differential-drive mobile robots (DDMRs). We utilized remote control techniques to address the challenging control issues associated with parking and reversing maneuvers when a vehicle is mounted in front of DDMRs Initially, we converted inputs from the remote controller into control state variables by leveraging the structural similarities between rear-wheel steering vehicles and the DDMR driving mechanisms. Furthermore, even when considering the dynamic characteristics, we represented the control state variables solely as tire longitudinal velocities. In doing so, we introduced disturbances representing tire longitudinal forces, which are challenging to estimate but crucial for control, to demonstrate robustness. Considering dynamic effects, including tire slip and caster wheel influence, which become increasingly significant with increasing weight and speed, we formulated the dynamics of the planar model. Finally, we conducted dynamic modeling using actual DDMR parameters and validated control stability through MATLAB simulations.