| In recent years,the measurement of angular velocity and acceleration of moving objects in high speed environments has become increasingly important in both civil and defense applications.Highly dynamic MEMS-IMU is one of the effective means to achieve the above measurements.However,it requires further error compensation of the devices,because the errors generated in the manufacturing process of the devices,which in turn affects the measurement accuracy in high speed environments.At the same time,the existing calibration process also suffers from poor automation and low calibration efficiency.Therefore,it is important to study how to improve the accuracy of the highly dynamic MEMS-IMU and the efficiency of the calibration process.This paper focuses on the highly dynamic MEMS-IMU error compensation technology and automatic calibration system,the main research content is as follows:(1)Firstly,aiming at the problem that the higher-order coupling error terms generated in the high speed environment can not be effectively characterized when the conventianal model is used to compensate for the highly dynamic MEMS-IMU,an all-parameter error model for the high dynamic MEMS-IMU is established by analyzing the operating principle and error mechanism of the high dynamic MEMS-IMU,and an error model optimization method based on significance analysis is proposed to filter the error terms in the allparameter error model.The model optimization method is proposed to filter the error terms in the full-parameter error model.After experimental validation,the number of error terms in the optimized model was reduced from 42 to 28,a reduction of 33%,and the optimized model had the same significant regression effect compared with the full-parameter model.(2)Secondly,to address the problem of low accuracy of highly dynamic MEMS-IMU compensated by conventional error compensation methods,an improved generalized recursive least-squares algorithm is proposed by analyzing the limited measurement data utilized by the traditional method for estimating the model coefficients and the presence of colored noise due to the high speed environment.The algorithm takes the results of conventional calibration as initial values,filters the noise first,and then performs recursive estimation using global data.After verification,the accuracy of MEMS gyroscope is improved by 90.6% on average and that of MEMS accelerometer is improved by 13.8% on average compared with the traditional calibration method.(3)Thirdly,to address the problems of poor automation such as time-consuming,laborintensive and manual dependence of traditional calibration,the requirements of the automatic calibration system are analyzed,the general architecture of the automatic calibration system is proposed,the highly dynamic MEMS-IMU automatic calibration system is designed,and the test environment is built to verify the functions in the system.The experiments has proved that the automatic calibration system can automate the calibration process and shorten the time required from 2 h to 30 min,compared with the traditional calibration,which improves the calibration efficiency.Meanwhile,the system has good stability,and the output accuracy of the highly dynamic MEMS-IMU can meet the index requirements after the automatic calibration compensation. |
