Changing Dimensional Feedback Correction Method of INS/GPS Integrated Navigation System Based on Lever Arm Estimation

Abstract

In traditional INS/GPS integrated navigation system, whether a error state can be feedback is determined by its observability, however, in that there is no precise motion reference, it is difficult to judge whether the system is observable, therefore, which error states can be feedback is hard to determine. According to the vehicle’s maneuver’s effect on error states’ observability, a changing dimensional feedback correction method based on lever arm estimation error minimization is proposed. In the beginning phase of Kalman filtering, part of error states feedback correction was executed on a 18 dimensional error sates vector including lever arm error, this scheme runs until the difference between lever arm estimation and it’s measurement less than the setting threshold, all error states feedback correction was executed with a 15 dimensional error states model. The flying test shows that the estimation result with all error states feedback is prevail to that of part of error states feedback correction after lever arm estimation error converges. The proposed method is of great value in engineering application.