Abstract: Distributed drive electric vehicles feature independently controllable wheel-end torque, enabling highly maneuverable steering modes such as compass steering and significantly enhancing the vehicle's ability to navigate narrow spaces. To address the inherent trade-off between motion accuracy and tire wear in compass steering scenarios, this paper proposes a control strategy based on non-cooperative game theory that simultaneously considers displacement error and tire wear. The compass steering motion is classified into two representative types, and a three-degree-of-freedom vehicle dynamics model is established to analyze the underlying mechanisms. A hierarchical control architecture is developed, consisting of a strategy optimization layer and an execution control layer, with the sum of squared tire forces on the pivot wheel and the total squared slip velocities of non-pivot wheels defined as the control objectives. The dynamic states of individual wheels under different compass steering types are further examined. A decoupling method for lateral and longitudinal tire forces is realized based on the mapping between wheel speed and tire force, and the corresponding optimization objectives and constraints are formulated. Leveraging a non-cooperative game framework, a Nash equilibrium-based optimization algorithm is designed to compute the optimal front wheel steering angles and wheel speeds under various test scenarios. A PI controller is then applied to ensure accurate wheel speed tracking. Hardware-in-the-loop(HiL) test results show that, compared with the minimum-displacement-error strategy, the proposed method maintains similarly low displacement errors across six test scenarios, while significantly reducing tire wear. The results validate the proposed strategy's high motion accuracy, strong practical applicability, and robust performance.

Key words: compass steering, dynamic mechanism, differential control, nash equilibrium, distributed drive electric vehicles