Research on Information Fusion Technology of MEMS Gyro Array

Abstract The measurement accuracy of MEMS gyroscope is relatively low. Starting from the software level, multi-sensor information fusion technology array (MGA) used to reduce drift gyroscope. Firstly, a signal acquisition and processing system for communicating with ADXRS810 based on field-programmable gate (FPGA) designed. Secondly, collected data preprocessed. long-term trend terms are obtained by ensemble empirical mode decomposition (EEMD) linear function fitting, filtered out obtain smooth normal signal. Then, model time series analysis built error performed Allan variance. Finally, AR (1) model, moving horizon optimization Kalman filter method array. experiment shows that when domain length N=3, bias instability 1, 2, 3 4 decreases 9.716 o/h, 8.5682 13.484 o/h 26.414 1.2922 0.61147 1.4184 1.6964 respectively, average noise coefficient more than 85%. Compared ordinary method, has been significantly improved.