Trajectory Tracking Control of Skid-Steering Mobile Robots with Slip and Skid Compensation using Sliding-Mode Control and Deep Learning

TL;DR

This paper introduces a novel trajectory tracking method for skid-steering mobile robots that combines sliding-mode control with deep learning to compensate for slip and skid in real-time, improving outdoor navigation accuracy.

Contribution

It presents a new online slip and skid compensation technique integrating deep learning models with sliding-mode control for improved outdoor robot trajectory tracking.

Findings

Over 27% reduction in tracking errors.

Effective real-time slip and skid compensation.

Enhanced robustness in outdoor terrains.

Abstract

Compensating for slip and skid is crucial for mobile robots navigating outdoor terrains. In these challenging environments, slipping and skidding introduce uncertainties into trajectory tracking systems, potentially compromising the safety of the vehicle. Despite research in this field, having a real-world feasible online slip and skid compensation remains challenging due to the complexity of wheel-terrain interaction in outdoor environments. This paper proposes a novel trajectory tracking technique featuring real-world feasible online slip and skid compensation at the vehicle level for skid-steering mobile robots operating outdoors. The approach employs sliding-mode control to design a robust trajectory tracking system, accounting for the inherent uncertainties in this type of robot. To estimate the robot's slipping and undesired skidding and compensate for them in real-time, two previously developed deep learning models are integrated into the control-feedback loop. The main advantages of the proposed technique are that it (1) considers two slip-related parameters for the entire robot, as opposed to the conventional approach involving two slip components for each wheel along with the robot's skidding, and (2) has an online real-world feasible slip and skid compensator, reducing the tracking errors in unforeseen environments. Experimental results demonstrate a significant improvement, enhancing the trajectory tracking system's performance by over 27%.