Online Error Estimation of Inertial Measurement Unit in Transfer Alignment Problem

Document Type : Research Article

Authors

1 Department of Aerospace Engineering, Amirkabir University of Technology, Tehran, IRAN.

2 Department of aerospace Engineering Amirkabir University of Technology, Tehran,Iran

3 Department of aerospace Engineering Amirkabir University of Technology, Tehran, Iran

Abstract

The presence of bias in low-cost inertial sensors has always been a fundamental problem in the estimation of Euler angles. The transfer alignment technique involves using two sets of sensors: a set of low-cost sensors (which may have significant bias) and a set of high-accuracy sensors (which provide reliable data). By comparing the outputs of these two sets, the biases in the low-cost sensors can be estimated and corrected. To achieve this goal, it is necessary to obtain the difference between the master and slave frames. The “master frame” typically refers to the coordinate system defined by the high-accuracy sensors, while the “slave frame” refers to the coordinate system defined by the low-cost sensors. In this article, the method of small rotation angles is used to calculate this difference between the two frames. This article proposes a new alignment algorithm that is both fast and accurate. The biases of accelerometers and gyroscopes can be rapidly estimated and compensated for. The alignment discussed in this article is conducted during motion, which is more complex than stationary mode estimation. The desired maneuver in this study is a constant acceleration maneuver, where the biases of both the accelerometer and gyroscope are well estimated. For other maneuvers, gyroscope biases may not be accurately estimated, while accelerometer biases are generally well estimated. The estimated biases for accelerometers and gyroscopes are often noisy due to errors in the inertial sensors; therefore, preprocessing filtering is also performed to estimate the biases smoothly.

Keywords

Main Subjects

Files

Share

How to cite

Statistics