Stimation; (d) the relative error of accelerometer bias estimation. of accelerometer bias estimation.As can be noticed in Figure 4a, amongst the attitude errors, the relative yaw error could be the largest. 5. Conclusions yaw error reaches 5 `without covariance transformation. The integrated The relative The benefit with covariance transformation method is that yaw error of 0.2`. navigation result of thecovariance transformation features a less relative it establishes the As transformation relationship of theposition error is 12 m, devoid of involving the n-frame and also the shown in Figure 4b, the relative integrated navigation filter covariance transformation. G-frame. It fundamentally solves the problem of transformation shows Aligeron manufacturer superior stability along with a integrated navigation outcome with covariance filter overshoot and error discontinuity. caused by the adjust in navigation 8 m. As enhancing navigation accuracy when crossing the smaller relative position error of frame, shown in Figure 4c,d, the maximum bias error from the polar region. In addition to this, the covariance transformation technique doesn’t alter h, gyroscope with and with no covariance transformation reached 0.001 h and 0.02 the current navigation algorithm. The resultsthe the flight experimentand with no covariance respectively. The maximum bias error of of accelerometer, with and semi-physical simulation show that the covariance transformation system is effective at any latitude. As transformation, reached 0.1 ug and 25 ug, respectively. the latitude increases, the horizontal component on the earth’s rotational angular velocity decreases, resulting in weaker observability on the yaw angle error. The error fluctuation five. Conclusions caused by the frame switching will increase. Within this case, the covariance transformation The benefit in the covariance transformation approach is the fact that it establishes the process nevertheless makes a smooth transformation by way of the integrated navigation filter. transformation relationship from the integrated navigation filter in between the n-frame and Gframe. It fundamentally solves the problem of curation, Y.Z.; formal analysis, Y.Z.; investigaAuthor Contributions: Conceptualization, L.W.; data filter overshoot and error discontinuity. triggered tion,by the change in navigation frame, improving navigation draft, Y.Z.;when crossing the polar C.G.; methodology, L.W.; supervision, G.W.; writing–original accuracy writing–review and region. Besides this, the covariance transformation approach does not adjust the existing editing, L.W. All authors have study and agreed for the published version in the manuscript. navigation algorithm. The results on the flight experiment and semi-physical simulation show Funding: This study was funded by the National Nature Science Foundation of China, grant that the covariance transformation technique is powerful at any latitude. Because the latitude number 62003360, and also the Basic Research Project of College of Sophisticated Interdisciplinary Research, increases, the horizontal component on the earth’s rotational angular velocity decreases, grant quantity ZDJC19-07. resulting in weaker observability on the yaw angle error. The error fluctuation caused by the frame switching will improve. Within this case, the covariance transformation approach still makes a smooth transformation through the integrated navigation filter.Author Contributions: Conceptualization, L.W.; information curation, Y.Z.; formal evaluation, Y.Z.; investigation, C.G.; methodology, L.W.; supervision, G.W.; writing–original d.
