Reseach On MIMU Sensors Parameter Identification And Error Compensation

In recent years, with the continuous improvement of manufacturing industry, theaccuracy of MEMS inertial device increasing rapidly.With the merit of small size, lightweight, high reliability and low power consumption, micro inertial measurement unit(MIMU) has become a central part of inertial navigation. Therefore, it is significant toimprove the accuracy of MEMS inertial devices. This paper proposed research on devicecalibration and error compensation naiming at this problem.Firstly, the development of micro inertial technology in and abroad as well as theresearch on micro inertial device error compensation technology is introduced in thispaper. Then the working principle and cause of the error are analysised.Secondly, the error source of MEMS inertial devices are analysised, data was collectedusing laboratory equipment. Then error models of both gyro and accelerometer areestablished, based on which error calibration and compensation is carried on, and later thevalidity of the error model is proved and the is compensed.Thirdly, the random error sources of MEMS inertial devices are deeply analysised.After this, this paper separated the five kinds of random noises of MEMS gyro using Allanvariance method, and calculated the random noise of each axis.Finally, by using time series analysis of random noise on the MEMS gyroscope,ARMA model was set up, and Kalman filter is conducted on the model. The randomerror coefficient before and after filtering is compared through Allan variance method,which verified the validity of the filter method in noise supression.