Model-free observer-based voltage control of non-holonomic mobile robot using reinforcement learning

This paper presents a trajectory tracking control strategy for non-holonomic mobile robots with unavailable orientation angle measurements. An observer is developed to reconstruct the trigonometric terms required by the controller, while the Fourier-series-based reinforcement learning scheme performs online estimation of unknown nonlinear terms, and the controller compensates for bounded external disturbances. The nonlinear system is transformed into lower dimensional subsystems, and controller design is simplified using a virtual auxiliary error function. A dynamic model of the mobile robot driven by permanent-magnet DC motors is considered, where the controller generates the required motor voltage signals. By integrating reinforcement learning with Fourier-series expansion, a model-free control scheme is introduced that achieves accurate trajectory tracking with fewer tuning parameters and lower computational cost than the considered neural-network-based benchmark. The stability of the observer and controller is theoretically analyzed through separate Lyapunov-based analyses. Simulation results demonstrate improved tracking performance, smoother control signals, competitive control effort, and computational efficiency compared with neural-network-based and robust voltage controllers.