A Novel Calibration Algorithm For Rimu Based On Derivative Ukf

Jianhua Cheng,Ping Liu,Yingyao Kang
DOI: https://doi.org/10.23919/ChiCC.2019.8866107
2019-01-01
Abstract:In order to improve the accuracy of redundant inertial measurement unit (RIMU), a novel calibration algorithm for RIMU based on derivative unscented Kahnan filter (DUKF) is proposed in this paper. First, the cone configuration method is proposed and the relationship between gyros and body coordinate is obtained. Then, the calibration model is built to obtain the state function and measurement function. Third, a zero-space amplification algorithm is proposed to improve the observability of the measurement function. Finally, the DUKF algorithm is proposed to estimate the biases, scale factor and installation errors which improves the calibration accuracy. Experimental results show that the proposed calibration algorithm can improve the calibration accuracy and performance of RIMU effectively, the accuracy of the system can be improved at least 2 times. Furthermore, the algorithm can be used in all kinds of RIMU based on cone configuration.
What problem does this paper attempt to address?