Physics-Based Hybrid Control of Mobile Robot Drives with Adaptive Neural Network Compensation
Open Access
Abstract
This paper proposes a physically based hybrid architecture for controlling mobile robot drives. It combines a model-based controller, an adaptive neural network compensator for residual dynamics, and a Lyapunov-based stability supervision mechanism. Unlike existing hybrid control approaches, the proposed architecture implements a structured injection of neural network correction directly into the physical drive model with a controlled Lyapunov-based adaptation constraint. A mathematical model of the electromechanical drive of a differential mobile platform is developed, taking into account electrical and mechanical dynamics, wheel-to-surface contact interaction, and the system’s energy characteristics. Numerical simulation results demonstrate that the hybrid approach improves tracking accuracy, improves transient response, and ensures stable operation of the control system under parametric uncertainty, adhesion changes, and external disturbances. The proposed architecture maintains the physical interpretability of the model while simultaneously enhancing the system’s adaptability. The obtained results confirm the effectiveness of the developed method and its potential for application in control systems for mobile robotic platforms.Keywords
Funding Information
- Committee of Science of the Ministry of Science and Higher Education of the Republic of Kazakhstan (BR27195331)
This publication has 45 references indexed in Scilit:
- Adaptive Longitudinal Slip Compensation for Wheeled Mobile Robots Using Velocity RatioIFAC-PapersOnLine, 2025
- Hybrid Stable Adaptive Control Approach for Navigation of Nonholonomic Wheeled Mobile Robots on Slippery Surfaces With ObstaclesInternational Journal of Control, Automation, and Systems, 2025
- Advancements and Challenges in Mobile Robot Navigation: A Comprehensive Review of Algorithms and Potential for Self-Learning ApproachesJournal of Intelligent & Robotic Systems, 2024
- Nonlinear Disturbance Observer Incorporated Model Predictive Strategy for Wheeled Mobile Robot’s Trajectory Tracking ControlInternational Journal of Control, Automation, and Systems, 2024
- Physics‐Informed Neural Networks to Model and Control Robots: A Theoretical and Experimental InvestigationAdvanced Intelligent Systems, 2024
- An Optimal Robust Trajectory Tracking Control Strategy for the Wheeled Mobile RobotInternational Journal of Control, Automation, and Systems, 2024
- Adaptive learning control of robot manipulators via incremental hybrid neural networkNeurocomputing, 2023
- Physics-informed Neural Networks-based Model Predictive Control for Multi-link ManipulatorsIFAC-PapersOnLine, 2022
- An adapted Burckhardt tire model for off-road vehicle applicationsJournal of Terramechanics, 2022
- Model-based observers for tire/road contact friction prediction2007