Spatiotemporal Calibration of Dual-Antenna GNSS and Odometry Based on Kalman Filter

Dual-antenna Global Navigation Satellite System (GNSS) has the advantage of independently measuring absolute attitude information without relying on Inertial Measurement Unit (IMU). In order to utilize this advantage to assist in the rapid absolute initialization of odometry, it is necessary to study the spatiotemporal calibration between dual-antenna GNSS and odometry. Considering that existing research mainly focuses on the calibration of single antenna GNSS/IMU and odometry, and frequently neglects the time-delay between sensors, so in order to make up for the gap in the calibration of dual-antenna GNSS and odometry, this paper proposes a spatiotemporal calibration method based on Kalman filter. The method comprehensively considers motion constraints among multiple vectors during vehicle movement and the time-delay between sensors, employs the feedback correction strategy of indirect filtering to solve the challenge of estimating large installation misalignment angle, and realizes joint estimation of time-delay and spatial extrinsic parameters. The results of simulation and field tests show that the proposed method can accurately achieve the spatiotemporal calibration and has significant advantages over existing methods. The time-delay estimation accuracy is smaller than the minimum data output interval, and the estimation errors of rotation and translation extrinsic parameters are about 0.1° and 0.01 m, respectively.