Consolidated Control Architecture for Mecanum-Wheeled Mobile Robots Using SMO-MPC-PID-Fuzzy Hybridization

Abstract

This paper presents the first systematic integration of sliding mode observer (SMO), model predictive control (MPC), Kalman-enhanced PID, and adaptive fuzzy logic for Mecanum-wheeled mobile robots (MWMRs), addressing the critical gap in unified control architectures for omnidirectional mobile systems. The key innovations include: (1) a novel adaptive weighting mechanism β(k) for optimal MPC-PID fusion based on real-time tracking dynamics, (2) fuzzy rule-based parameter adaptation specifically designed for MWMR omnidirectional characteristics, and (3) enhanced SMO with wheel-slippage compensation capability. Mathematical modeling incorporates detailed four-wheel dynamics with motor characteristics, followed by systematic integration design ensuring global stability. Comprehensive MATLAB/Simulink simulations across circular, square, figure-eight, S-curve, and infinity-shaped trajectories demonstrate superior performance: 37.2% improvement in tracking accuracy (RMSE: 0.032 m vs 0.087 m for PID), 25.8% reduction in energy consumption, 47.6% faster settling time (0.43 s vs 0.82 s), and enhanced robustness with disturbance rejection factors of 0.873–0.945 under external forces, parameter uncertainties (±20%), and measurement noise. Validation conducted in MATLAB/Simulink R2024b environment with real robot parameters (4.5 kg, 245.6 × 300 × 515 mm). Key limitations include validation restricted to simulation, performance degradation above 2 m/s velocities, and requirement for real-time computational resources (4.92 ms per iteration). This integration provides a systematic framework for high-precision MWMR control in industrial applications, with immediate applicability to material handling, warehouse automation, and service robotics requiring reliable omnidirectional navigation.