Abstract: At present, the dynamic analysis and control of distributed-drive electric vehicles are based on deterministic dynamic models. However, the random perturbation of system parameters, along with uncertainties in state variables and sensor information, causes control systems based on deterministic models to struggle in adapting to complex road conditions. To address this, the influence of random disturbances such as wheel load fluctuations and non-uniform distribution of road surface adhesion coefficients in the wheel-ground interaction process is considered, and a wheel-ground uncertainty model is established. Dynamics simulation experiments are conducted under both hard and soft road conditions using the wheel-ground uncertainty model. In-depth analysis of the experimental results leads to the proposal of a "Tension-Relaxation Theory " between the wheel-ground interaction constraints and wheel-end control constraints. Using parasitic power, traction efficiency, and additional yaw moment as evaluation indices, statistical disturbance analysis is performed in conjunction with the wheel-ground uncertainty model. This further clarifies the inherent relationship between the wheel-ground contact state and control modes, ultimately leading to the proposal of a mode-switching drive control mechanism. Multi-condition simulation results indicate that, compared to a single control mode, the mode-switching drive control mechanism effectively reduces parasitic power and additional yaw moment while maintaining stable and high traction efficiency, thereby verifying the correctness and effectiveness of the proposed theoretical method.

Key words: distributed drive electric vehicle, tension-relaxation theory, parasitic power, torque control, speed control