Aug 2026
Robust Slip-Compensated Control of a Wheeled Mobile Manipulator via Vision-Based Estimation and Deep Reinforcement Learning
The proposed framework combines dynamic modeling, vision-based slip measurement, multilayer perceptron (MLP)–based real-time slip estimation, compensation of nonlinear wheel–ground interaction effects, and an adaptive PID controller whose gains are tuned online by deep reinforcement learning.
S. K. Saraswat, B. A. Zneid, Merwa Alhadrawi et al.
· Journal of Vibration Enginee... · 0 citations