Enhanced Disturbance Observer-Based Sliding Mode Control for Precise Trajectory Tracking of Liquid-Carrying Robots on Uncertain Terrains

Abstract

This paper investigates a high-precision trajectory tracking control strategy for differential-drive tanker robots subject to the combined effects of liquid sloshing dynamics and mixed-terrain disturbances. A key challenge in this application arises from the contrasting characteristics of different payload regimes: while a full tank presents maximum total mass and inertia, a half-filled tank (50% partial-load) possesses an expanded free-surface area that facilitates larger wave oscillations (exhibiting peak transient amplitudes near -0.28" m" , compared to approximately -0.22" m" under full load). To mitigate these state-dependent internal hydrodynamic forces alongside stochastic surface resistance, a control architecture integrating Sliding Mode Control with a Nonlinear Disturbance Observer (DOB-SMC) is formulated. Within this scheme, the observer estimates and compensates for the lumped disturbance composite in real time, thereby allowing a reduction in the discontinuous switching gain to alleviate control chattering. Simulation results along a multi-zone circular trajectory indicate that the DOB-SMC framework consistently attenuates tracking discrepancies across both load conditions, achieving Root Mean Square Error (RMSE) reductions of 42.69% (from 0.0061" m" to 0.0035" m" ) under full load and 33.32% (from 0.0052" m" to 0.0035" m" ) at partial load relative to standalone SMC. Overall, the proposed approach demonstrates practical feasibility for improving trajectory tracking precision and maintaining bounded transient responses during liquid transport over uneven terrains.