Research On Calibration Method Of All Parameters Of MEMS Inertial Measurement Unit

With its advantages of low cost,small size and easy mass production,MEMS inertial measurement unit is becoming the core information source of small strapdown inertial navigation system and heading and attitude measurement system at present.It plays an increasingly important role in both military and civilian fields.With the advancement of the micro-electromechanical industry and manufacturing processes,MEMS inertial measurement units have a tendency to gradually replace other low-and medium-precision IMUs.However,similar to other types of inertial measurement units,MEMS inertial measurement units also face the problem of periodic calibration.Even for low-precision MEMS-IMUs,it is more important to compensate for the measurement accuracy based on their error parameters.An indispensable means to improve its performance.In this paper,the existing MEMS inertial measurement unit is used for full error parameter analysis and calibration compensation.In this paper,the error characteristics of the MEMS inertial measurement unit are analyzed from the two aspects of deterministic error and random error,and the generation mechanism and identification method of each error are studied.For the three deterministic errors,a positive and negative rate experiment and a static multi-position experiment were designed for calibration;random noise was analyzed using the Allan variance method,and five random noises were estimated.In view of the defects that the accuracy of the test turntable restricts the calibration efficiency of the device and the calibration process is too cumbersome,the systemlevel calibration theory is studied for the calibration and compensation method of the MEMS inertial measurement unit.The system-level calibration method is based on the introduction of high-precision main inertial group data as a reference,the Kalman filter is reduced to 12 dimensions,and the calibration of the first-order error of the gyroscope g value is added.To a certain extent,it accelerates the filter convergence speed and shortens the calibration time.Finally,the article compares the IMU compensation results of the discrete and systemlevel calibration methods through experiments based on low-precision three-axis turntables.Through comparative experiments,it can be proved that for low-precision turntables,the main principle based on Kalman filter The inertial group matching calibration method has higher calibration accuracy,shorter time,and better compensation effect,which can meet the actual calibration needs of low and medium precision inertial measurement units.