Roll estimation algorithm based on Sage-Husa adaptive Kalman filtering with rotation criteria

作者：Wang Jiawei, Qi Keyu, Xu Guotai, Qian Rongzhao, Jie Yan · 年份：2017 · DOI：10.1109/icemi.2017.8265931 · 被引用次数：3 · 研究领域：Target Tracking and Data Fusion in Sensor Networks、Inertial Sensor and Navigation、Astronomical Observations and Instrumentation

According to the problem of increasing error caused by irresistible measuring noise in traditional EKF method within trajectory, a new solution of roll estimation for correction fuze, enlightened by axial output of gyro can be treated as rotation compensation for system noise, using SHAKF with rotation criteria is proposed. Firstly, based on the simulating comparison of estimation precision between traditional EKF and SHAKF methods, the result indicates that the roll measuring error of the new solution conspicuously lower than that of EKF's, the mean value of measuring error is 0.26deg and the variance of that is 0.97deg, that means the adaptivity offilter can follow the innovation to make estimated state convergence and eventually decreases the absolute error of roll angle. Furthermore, the new SHAKF algorithm is also verified by in-lab testing with MEMS three-axis turntable under an actually varying rotation setting, and the result shows that the estimation error always below 4.2degs in dynamic range changing from 30r/s to 1r/s.