Abstract: Modular vehicles eliminate mechanical constraints between wheels and achieve independent steering of all wheels, greatly enhancing the flexibility of chassis control. However, under curved conditions, each steering wheel is dynamically independent, and the vertical loads of each wheel exhibit time-varying differences, posing challenges for the design of the controller. This paper proposes a linear parameter varying(LPV) modeling method for full wheels independent steering vehicles based on tensor product model(TP model) to address the problems of constraints on scheduling parameter selection, conservatism of system convex hull, and lack of unified tight convex hull construction. The TP model can use higher-order singular value decomposition(HOSVD) to select vertices of the convex hull for the original time-varying system and construct a tight convex hull of the system, resulting in a polytopic LPV(PLPV) model. A robust gain scheduling controller is designed by the parallel distributed compensation(PDC) framework to solve the problem of trajectory tracking caused by different speeds and time-varying vertical loads on each wheel. At the same time, in the derivation of the trajectory tracking model, the reference heading angle of the controller is corrected using the vehicle's side-slip angle, and a robust controller with side slip compensation(RLPV-SC) is derived, which further improves the accuracy of the trajectory tracking. Finally, the effectiveness of the proposed control algorithm is verified through Carsim-Simulink simulation.

Key words: TP model, HOSVD, FWIS, trajectory tracking