Robust Path-Tracking Control of Networked Autonomous Vehicles for Multiple Source Uncertainties

The control of networked autonomous vehicles is susceptible to the inevitable multiple source uncertainties, including vehicle dynamic parameter deviations, frame dropouts, and network delays. To enhance the path-tracking accuracy and yaw stability, a robust control strategy with rapid error convergence has been developed. By progressively incorporating uncertainties arising from vehicle dynamics and network conditions into a polytopic linear parameter-varying model, the proposed method formulates a comprehensive description of all possible vehicle operating conditions under uncertainties through a finite set of vertices. Subsequently, a novel asymptotically stable controller is designed using Lyapunov-based linear matrix inequalities, incorporating stability, optimality, and constraint considerations. Combined with convergence speed optimization, both robustness and optimality in path-tracking control (PTC) can be guaranteed in the presence of these uncertainties. The simulation and experimental results have verified that the proposed method can effectively address the impact of multiple source uncertainties and is feasible for real-time path tracking control. Compared with existing methods, the proposed strategy significantly enhances path-tracking performance, achieving improvements of 25.58% in lateral accuracy and 15.79% in heading accuracy, along with better yaw stability.