Self-Calibration for Land Navigation Using Inertial Sensors and Odometer: Observability Analysis

Abstract

Determination of frame misalignment is a common problem for any real system combining information from more than one sensor. By way of observability analysis, this paper theoretically addresses the feasibility of automatic determination of inertial sensors/odometer misalignment for land navigation applications. The case is also considered without the aid of an uncalibrated odometer. Specifically, global observability analysis is performed for the nonlinear target system augmented with the unknown parameters to be calibrated. We obtain moderate sufficient conditions for successful self-calibration as well as guidelines for selecting feasible calibration trajectory, which lays a good foundation for calibration design and implementation that follow.