What is it about?
The diagram of the iterative systematic calculation adopted in this paper is shown in Fig.1. The method is organized as follows: First, IMU raw data (accelerometer and gyro readouts) is accumulated during the rotation sequence. Second, IMU raw data is corrected with IMU estimation parameters. Then, the correction data is processed by unified mathematical framework, to get acceleration and angular rate measurements. Based on the measurements, Least Square Method (LSM) is utilized to estimate the residual error of IMU parameters. Finally, if the estimation results meet the conditions, such as iteration number is less than 7 or the estimation residual error of accelerometer bias is less than 1ug, the calibration procedure is finished. If not, the iterative calculation should continue to be performed.
Photo by Aaron Burden on Unsplash
Why is it important?
Using the method, one can estimate the Inertial Measurement Unit (IMU) parameters both at manufacturer's facilities and in the field Since a unified mathematical framework is designed, its’ not necessary to keep two IMU axes lying approximately in the horizontal plane, or rebuilding the measurement matrixes for the different rotation test sequences The observability analysis of the method in 19 specified IMU positions is done in this paper, and a fully observable factor is obtained. The fully observable factor is that in the 0th static position, z axis and any one of the other two axes couldn’t be lied in the horizontal plane at the same time. The influence of gravity disturbance is analyzed. Only the estimation of accelerometer scale factor error will be influenced by gravity anomaly, and about 1ppm estimation error will be produced by the gravity anomaly of 1ug. An iterative calculation method is designed to get an accurate calibration result.
Read the Original
This page is a summary of: Systematic calibration method based on acceleration and angular rate measurements for fiber-optic gyro SINS, Review of Scientific Instruments, January 2021, American Institute of Physics, DOI: 10.1063/5.0023674.
You can read the full text:
The following have contributed to this page