国防科技大学学报2025，Vol.47Issue(5)：70-77,8.DOI:10.11887/j.issn.1001-2486.25010021
神经网络拟合的激光陀螺惯组重力实时测量滤波
Real-time gravity measurement filtering of laser gyro inertial group fitted by neural network
摘要
Abstract
To overcome the inherent fixed time-delay limitation of conventional gravity filtering methods in laser gyro-based inertial navigation gravity measurement systems,a real-time gravity data processing method based on neural network-approximated FIR(finite impulse response)filtering was proposed.By fitting the FIR filter through neural network implementation,the long dependency on future data was effectively reduced,thereby effectively reducing filtering delay.Experimental results show that compared with the FIR filter,the processing delay of the proposed method is reduced by93%,and the average filtering accuracy is better than 2 mGal.This indicates that the proposed method can significantly improve the real-time performance of data processing while maintaining high accuracy,providing a solution for the real-time gravity measurement of the laser gyroscope inertial group gravity measurement system.关键词
重力测量/神经网络/FIR低通滤波/实时处理/重力数据处理Key words
gravity measurement/neural network/FIR low-pass filtering/real-time processing/gravity data processing分类
航空航天引用本文复制引用
高春峰,程嘉奕,陈迈伦,魏国,罗晖,侯承志,朱旭,马海阳,王景..神经网络拟合的激光陀螺惯组重力实时测量滤波[J].国防科技大学学报,2025,47(5):70-77,8.基金项目
国家自然科学基金资助项目(62203454) （62203454）
