摘要
为了减小光纤陀螺(FOG)的温度漂移误差以提高导航精度,从FOG温度误差机理出发,利用Bagging集成学习技术泛化能力强的优点,采用Python语言将分类回归树(CART)集成,建立了FOG温度漂移误差的CART-Bagging模型。最后,利用某型FOG温控试验的实测数据,使用此模型进行补偿实验;同时与单反向传播(BP)神经网络模型和CART模型的补偿效果进行对比分析。结果表明:相比于单BP神经网络模型和CART模型,使用CART-Bagging模型补偿后,FOG温度漂移误差减少了59%以上,具有更好的补偿效果。
In order to reduce the temperature drift error of fiber-optic gyroscope(FOG)so as to improve the navigation precision,start from FOG temperature drift error mechanism,using the advantage of strong generalization ability of Bagging ensemble learning technology.the classification and regression tree(CART)is integrated with Python language,CART-Bagging model is established.Finally,this model is used for compensation experiment by using the measured data of a certain type of FOG temperature control experiment.The compensation effect is compared with the single back propagation(BP)neural network model and the CART model and analyzed.The results show that temperature drift error of FOG is reduced by more than 59%after compensated by CART-Bagging model,compared with the single BP neural network model and the CART model,it has better compensation effect.
作者
毛宁
许江宁
何泓洋
吴苗
MAO Ning;XU Jiangning;HE Hongyang;WU Miao(College of Electrical Engineering,Naval University of Engineering,Wuhan 430033,China)
出处
《传感器与微系统》
CSCD
北大核心
2023年第6期43-46,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(41804076)
湖北省自然科学基金资助项目(2018CFB544)。
关键词
光纤陀螺
温度补偿
PYTHON语言
集成学习
fiber-optic gyroscope(FOG)
temperature compensation
Python language
ensemble learning
