Path Following Control of Four‐Wheel Independent Steering Vehicles Considering Variable Adhesion Characteristics of Intermittent Ice‐Snow Roads

ABSTRACT Four‐wheel independent steering (4WIS) vehicles exhibit enhanced path following potential under nominal conditions due to their fully controllable wheel mechanisms. However, on intermittently icy and snowy roads, the time‐varying and non‐uniform adhesion characteristics at each wheel significantly challenge conventional control strategies, which are typically designed under fixed and homogeneous adhesion assumptions. To address this limitation, this paper proposes a path following control framework for 4WIS vehicles that explicitly considers variable adhesion conditions of intermittent ice‐snow roads. First, a noise‐adaptive edge particle filter‐based adhesion coefficient estimation method is developed to accurately estimate the time‐varying road adhesion coefficients at each wheel, providing precise real‐time reference data for control adaptation. Second, a Variable Constraint Model Predictive Control (VC‐MPC) strategy is introduced, which dynamically adjusts the adhesion‐related constraints and control outputs of the path following controller based on the estimated adhesion conditions. This enhances the controller's adaptability to heterogeneous adhesion variations and improves the overall tracking performance. Experimental results demonstrate that, compared to conventional approaches, the proposed method significantly improves path following accuracy and vehicle stability for 4WIS vehicles operating on intermittently icy and snowy surfaces.