Advanced Multi-rate Kalman Filter for Double Layer State Estimator of Electric Vehicle Based on Single Antenna GPS and Dynamic Sensors

In this paper, a double layer state estimator for electric vehicle is proposed. The first layer provides sideslip angle and yaw angle estimation. Utilizing the output the first layer, roll angle, longitudinal and lateral velocity are estimated from the second layer. The estimator is designed by using the course angle and velocity vector obtained from single antenna GPS and a novel advanced multi-rate Kalman filter. While motion control system of electric vehicle requires state estimation every 1 millisecond, the sampling time of GPS based measurement is much longer. In order to solve this problem, inter-sample residual prediction is proposed. Furthermore, by treating the combination of model uncertainties and external disturbances as extended state to be estimated, the high robustness of vehicle state estimation is achieved. The satellite information is utilized for auto-tuning of GPS measurement noise covariance matrix.