Error Compensation Method for Dual Fiber Optic Gyroscope Navigation System Based on BP Neural Network

REN Hongyong; WANG Yang; MA Xiaolu; LONG Siguo; LI Yang; SUN Weidong; ZHANG Xiaoyu

Abstract：In the context of accelerated urbanization,pipe jacking engineering has emerged as a pivotal trenchless underground pipeline laying technology in the construction of urban underground infrastructure.However,traditional guiding methods encounter issues of insufficient accuracy and low efficiency in complex working conditions such as small diameters,long distances,large elevations,and large curvatures.In order to address the aforementioned issues,this paper proposes an error compensation method for a dual fiber optic gyroscope(FOG)navigation system based on a backpropagation(BP)neural network,and its validity is verified through simulation tests and experiments.The system framework,key components,and software-hardware architecture are designed.A range of optimization strategies are proposed,including signal processing optimization,elevation compensation algorithms,and curvature adaptive adjustments.These strategies are designed to enhance system adaptability and stability under long-distance,large-elevation,and high-curvature conditions.The experimental results demonstrate that the system exhibits high precision and rapid response in basic performance tests,with guidance accuracy controlled within an error range of±0.05％.The efficacy of the optimization strategies in enhancing guidance accuracy in complex scenarios is well-documented.The efficacy of the system's continuous monitoring and self-calibration mechanisms is further validated by the results of long-term stability tests.