Dynamic Modeling and Disturbance-Observer-Enhanced Control for Mecanum-Wheeled Vehicles Under Load and Noise Disturbance

This paper investigates the dynamic modeling and robust control of a Mecanum-wheeled vehicle (MWV) under load disturbances and measurement noise. The system is modeled as a cascaded state-space representation, where the motor transfer function (PWM input → torque output) and the vehicle transfer function (torque input → vehicle speed output) are combined. The PWM-induced motor delay is linearized, and the complete dynamic model is derived using Lagrangian mechanics, addressing the limitations of conventional models that are incomplete and unable to decouple control signals from disturbance signals. For the developed model, a robust stability controller is designed by integrating Internal Model Control (IMC) with a Disturbance Observer (DOB), enhancing real-time disturbance rejection. Open-loop experiments validate the model’s accuracy, showing a Dynamic Time Warping (DTW) error of 0.2662 m, significantly lower than the 0.3198 m observed in traditional models. In closed-loop simulations, under load disturbances ( T L = 0.1 to T L = − 0.7 ) and Gaussian noise (power: 0.0001–0.00005), the proposed IMC + DOB controller achieves 97.6% faster stabilization than IMC and 98.3% faster than PID, demonstrating superior convergence speed, robustness, and disturbance rejection. This study provides a novel control strategy that effectively handles non-square system dynamics while mitigating external disturbances in real time. The proposed framework enhances trajectory tracking accuracy and stability, with potential applications in autonomous robotics and vehicular systems.