摘要
光纤陀螺是惯导系统的重要组成器件,环境温度变化会造成光纤陀螺的零偏发生漂移,从而降低测量精度。运用传统的BP神经网络进行预测易陷入局部极小值,导致补偿失败。该文采用混沌模拟退火粒子群BP神经网络的光纤陀螺零偏温度补偿模型,优化了网络参数。通过在-40~60℃的升降温实验对模型进行验证,实验结果表明,该温度补偿模型的零偏稳定性比补偿前约有70%的精度提升,与以往BP模型相比,其预测性能和补偿效果更好。
The fiber optic gyroscope is an important component of the inertial navigation system,and the ambient temperature change will cause the zero bias drift of the fiber optic gyroscope,thereby reducing the measurement accuracy.Using the traditional BP neural networks for prediction is prone to falling into local minima,resulting in compensation failure.In this paper,the zero-biased temperature compensation model of fiber optic gyroscope adopting the chaos simulated annealed particle swarm BP(CSAPSO-BP)neural network is used to optimize the network parameters.The model is verified by the temperature rise and fall experiment at-40℃~60℃.The experimental results show that the zero bias stability of the temperature compensation model is improved by about 70%compared to before compensation,and its prediction performance and compensation effect are better than previous BP models.
作者
赵深
何巍
辛璟焘
吕峥
ZHAO Shen;HE Wei;XIN Jingtao;LYU Zheng(Key Lab.of the Ministry of Education for Optoelectronic Measurement Technology and Instrument,Beijing Information Science and Technology University,Beijing 100192,China;Beijing Lab.of Optical Fiber Sensing and System,Beijing Information Science and Technology University,Beijing 100192,China)
出处
《压电与声光》
CAS
北大核心
2023年第4期589-594,共6页
Piezoelectrics & Acoustooptics
基金
国家自然科学基金资助项目(52105540)
高等学校学科创新引智计划基金资助项目(D17021)
北京市自然基金-市教委联合基金资助项目(KZ201911232044)。
关键词
光纤陀螺
温度补偿
BP神经网络
混沌理论
模拟退火粒子群
零偏
fiber optic gyroscope
temperature compensation
BP neural network
chaos theory
simulated annealing particle swarm
zero bias
作者简介
赵深(1994-),男,安徽省铜陵市人,硕士生,主要从事光纤陀螺控制与优化的研究;通信作者:何巍(1986-),男,博士生导师,主要从事光纤传感系统的研究。
