Research On Autonomous Calibration And Combined Calibration Methods Of MEMS Inertial/magnetic Sensors

With the wide application of inertial sensor and magnetic sensor in various fields,higher requirements for sensor accuracy are put forward.As the key to eliminate the deterministic error of sensor,sensor calibration technology has gradually become a research hotspot at home and abroad with the development of demand.In order to improve the accuracy and robustness of calibration,various calibration methods of sensor error are studied in this paper.Firstly,the errors of three-axis inertial sensor and three-axis magnetic sensor are introduced,and the corresponding modeling is carried out.Under the condition of no external auxiliary equipment,the maximum likelihood estimation calibration method of accelerometer,gyroscope integral calibration method and ellipsoid fitting calibration method of magnetometer are proposed for three-axis accelerometer,three-axis gyroscope and three-axis magnetometer respectively.These methods overcome the shortcomings of traditional sensor calibration methods that need to be carried out under the condition of external auxiliary equipment,and realize the field calibration of MEMS inertial sensor and magnetometer.In addition,according to the characteristics of the sensor output changing with temperature,the third-order polynomial parameter fitting method is used to complete the parameter fitting,and then complete the temperature calibration of the sensor.Simulation experiments show the effectiveness of these calibration algorithms.Secondly,the combined calibration method of MEMS inertial and magnetic sensors is studied.For three-axis gyroscope,the cross multiplication calibration method of gyroscope assisted by magnetometer output and the quaternion update calibration method of gyroscope assisted by accelerometer output are studied.The effectiveness of the two methods is verified by simulation experiments,and the calibration effects of the two methods are compared.For the three-axis magnetometer,the accelerometer output assisted magnetometer point product calibration method and the gyroscope output assisted magnetometer least square calibration method are studied.The effectiveness of the two methods is verified by simulation experiments,and the calibration effects of the two methods are compared.In addition,the misalignment error between the inertial platform and the magnetometer platform is analyzed and modeled,and the effectiveness of these calibration algorithms is proved by simulation experiments.Finally,based on the above simulation results,the MEMS sensor hardware platform is built,and the data is collected and verified by the host computer.The experimental results show that the data of three-axis accelerometer,gyroscope and magnetometer can be calibrated through the calibration algorithm,which verifies the effectiveness and robustness of the calibration algorithm.