Vehicle state parameter estimation based on improved adaptive observer

To accurately obtain the state parameters during vehicle driving, this paper designs an adaptive nonlinear observer for vehicle state parameters to mitigate the influence of measurement noise within the framework of extended Kalman filtering. Based on the vehicle nonlinear three degree of freedom, a dead zone observer was designed to accurately estimate the sideslip angle, yaw rate, and longitudinal velocity of the vehicle in answer to the problem of significant noise in lateral acceleration. According to the output variable error, determine the dead zone boundary of the observed state variables at each time based on observation errors, limiting the estimation error within the dead zone boundary and improving noise suppression performance. The validity of this method is validated through a variety of different simulation conditions and by comparison to the EKF algorithm. Results demonstrate that the improved algorithm significantly enhances the estimation precision compared to Extended Kalman Filter algorithm.