No abstract

The position and azimuth determining system (PADS) chiefly consists of an inertial navigation system and odometer. In order to improve the performance of the PADS, a backtracking scheme is proposed for odometer calibration and odometer-aided rapid alignment of PADS with a single pre-known landmark. Based on a backtracking scheme, a newly derived backward filter is presented to accomplish odometer calibration, fine alignment and positioning simultaneously by using the recorded forward data. Compared with normal methods, the proposed method can calibrate the odometer with only one pre-known landmark and lessens the environment restraints. Both the simulation and the experiment results show that the proposed method can calibrate the odometer with only one landmark and achieve good alignment and positioning accuracy in a short period of time.

Micro-electro-mechanical systems (MEMS) inertial measurement devices tend to be widely used in inertial navigation systems and have quickly emerged on the market due to their characteristics of low cost, high reliability and small size. Calibration is the most effective way to remove the deterministic error of an inertial reference unit (IRU), which in this paper consists of three orthogonally mounted MEMS gyros. However, common testing methods in the lab cannot predict the corresponding errors precisely when the turntable’s working condition is restricted. In this paper, the turntable can only provide a relatively small rotation angle. Moreover, the errors must be compensated exactly because of the great effect caused by the high angular velocity of the craft. To deal with this question, a new method is proposed to evaluate the MEMS IRU’s performance. In the calibration procedure, a one-axis table that can rotate a limited angle in the form of a sine function is utilized to provide the MEMS IRU’s angular velocity. A new algorithm based on Fourier series is designed to calculate the misalignment and scale factor errors. The proposed method is tested in a set of experiments, and the calibration results are compared to a traditional calibration method performed under normal working conditions to verify their correctness. In addition, a verification test in the given rotation speed is implemented for further demonstration.

No abstract