State Estimation for Traction Control of Dual-Motor Electric Vehicles

To address inaccurate longitudinal speed acquisition, difficult road adhesion identification, and insufficient reliability of state inputs for traction control in dual-motor electric vehicles under low-adhesion, adhesion-transition, and drive-slip conditions, this paper proposes a state estimation method oriented to traction control. Four-wheel speeds, inertial measurement unit (IMU) signals, and vehicle dynamics are fused to establish a layered longitudinal speed estimation structure, including slip-confidence evaluation, inertial correction, kinematic and dynamic fusion, and multi-mode weight decision. Standard road adhesion curves, fuzzy inference, and recursive correction are further combined to estimate the peak adhesion coefficient and the optimal slip ratio online. CarSim/Simulink co-simulation results show that the root mean square errors of the proposed speed estimation method are 0.1226, 0.1728, 0.1070, and 0.0322 m/s under comprehensive driving, acceleration slip, emergency braking, and high-speed steering conditions, respectively. Under an adhesion-transition condition, the peak adhesion coefficient and optimal slip ratio can be updated rapidly with road changes. Application results suggest that the estimated states can provide useful inputs for front–rear axle traction coordination under the investigated low-adhesion conditions.