Research On The Key Technology Of Accurate Calibration For MEMS Redundant Inertial Measurement Unit Under Multi-Working Conditions

Redundant inertial measurement unit(RIMU)which is designed by redundant configuration technology can improve reliability of navigation system.In the field of RIMU,MEMS-RIMU is more and more popular,because of its low cost and small volume.Due to the limitation of actual materials and manufacturing technology,the need of high precision navigation cannot be satisfied.The key technology of improving MEMS-RIMU's accuracy is calibration method.Due to the sensors in MEMS-RIMU are non-orthogonal,the calibration principle and scheme are more complex than traditional three-axis IMU.In addition,the requirements of calibration scheme are different under different wording conditions.Therefore,the systematic idea is adopted to calibrate MEMS-RIMU under different wording conditions.Based on the existing hardware equipment,it is more practical to improve the navigation accuracy of the navigation system by fully exploiting the potential accuracy of the device.This dissertation will take MEMS-RIMU as the research object,analyze the relation of the constant errors of MEMS-RIMU and the navigation accuracy,focus on the key issues of the calibration scheme of MEMS-RIMU under multi-working conditions.The specific research contents are as follows:(1)Analyzing the relation of the constant errors of MEMS-RIMU and navigation accuracy.To supplement the analysis of relation between the constant errors of MEMS-RIMU and the navigation accuracy,the measurement value of angular velocity,acceleration and navigation errors will be given by analyzing the relation between the output of each sensor and the three-axis output in body frame.According to the analysis above,it is shown that the relation between RIMU constant error and navigation performance is more complex than that of IMU.Moreover,the necessary of studying the calibration technology of MEMS-RIMU is verified.Finally,the theoretical analysis is verified by simulation experiment with four-gyro MEMSRIMU.(2)MEMS-RIMU calibration scheme based on extended Kalman filter.To solve the problems of low accuracy and poor generality caused by inaccurate model and linearized estimation algorithm in traditional RMIU calibration method under laboratory condition,a universal MEMS-RIMU calibration method based on extended Kalman filter is proposed.In this method,the coupling relationship between the constant errors is fully considered.Then universal calibration model is derived by establishing the relationship between each sensor and the body frame.In addition,the theory of observability analysis is introduced to analyze the estimation accuracy of states after the model is established.The idea of measurement amplification is proposed to improve the observability of observation model and simplify the process of calibration experiment.The MEMS-RIMU can be calibration accurately by rotating the X,Y and Z axes of the rotation platform in the same attitude.In terms of parameter estimation,to solve the problem of parameter estimation accuracy degradation caused by linearizing the nonlinear model,an extended Kalman filter algorithm is proposed to estimate the states of nonlinear model,which can improve the accuracy of parameter estimation and ensure the universality of calibration scheme.(3)MEMS-RIMU self-calibration scheme based on derivative unscented Kalman filter.Aiming at the lack of observation information for MEMS-RIMU calibration under outfield conditions,the MEMS-RIMU self-calibration scheme based on derivative unscented Kalman filter is proposed.Firstly,the accelerometer calibration model based on the difference between two norm of gravity vector and two norm of accelerometer measurement is proposed.Then a gyroscope calibration model based on the second norm of the of accelerometer output's first derivative is proposed.Finally,Combining the derivative unscented Kalman filter to estimate the parameters,MEMS-RIMU self-calibration can be completed.Moreover,the environmental adaptability and rapidity of calibration can be improved by the proposed method.(4)Backup sensor calibration scheme based on adaptive extended Kalman filter.To solve the problem of inaccurate backup sensor calibration in MEMS-RIMU due to the change of observation noise characteristics,a backup sensor calibration scheme based on adaptive extended Kalman filter is proposed.Observation noise covariance matrix can be estimated to improve the environment adaptability of the algorithm.Firstly,the nonlinear calibration model without rotation platform reference is established according to the geometric relationship between the original sensor and the backup sensors.Then,combined with the sequential variable Bayesian adaptive extended Kalman filter,the noise covariance matrix can be estimated adaptively.The environmental adaptability of the calibration method is improved.