A correction estimation method considering sensor measurement bias was proposed to accurately estimate the lateral velocity of the tractor in autonomous semi-trailer vehicles. Based on the three-degree-of-freedom vehicle dynamics model, the measurement bias estimation model was established by expanding the front wheel angle, yaw rate bias, and lateral acceleration bias into the state variables. Taking into account the impact of dynamic model mismatch, a feedback correction mechanism based on the kinematic model was designed to adjust the estimation results of lateral velocity and yaw rate through feedback. The proposed method was implemented using the square-root cubature Kalman filter algorithm, and the observability of the estimator was analyzed. This method was verified on the joint simulation platform of TruckSim and MATLAB/Simulink. The results show that the proposed method not only eliminates the influence of measurement bias but also exhibits good convergence speed. Its performance is significantly better than estimation methods that do not consider measurement bias or only consider measurement bias.

Semi-trailer  /  Lateral velocity estimation  /  Measurement bias  /  Feedback correction  /  Square-root cubature Kalman filter