基于小波多尺度变换的光纤陀螺振动误差分析与补偿

Error Analysis and Compensation for Fiber Optic Gyroscope under Vibration Based on Wavelet Multi-Scale Analysis

针对振动环境对光纤陀螺性能的影响,对某型号的光纤陀螺进行了线振动实验并对实验结果进行了Allan方差分析。利用小波多尺度变换提取了光纤陀螺误差模型中的各误差项,分析并验证了零漂及噪声误差与Allan方差分析误差系数中的量化噪声、角度随机游走以及零偏误差与误差系数中的零偏稳定性、速率随机游走、速率斜坡之间的对应关系。随后利用RBF神经网络对小波多尺度分析提取的零偏误差建立模型并进行了补偿。仿真结果表明,本文提出的方法有效减小了振动环境下各误差项对光纤陀螺性能的影响,Allan方差分析结果中的五个误差系数均有较大下降,其中两项误差系数下降了一个数量级及以上,极大提高了光纤陀螺在振动环境下的输出精度,对光纤陀螺在振动环境下的误差研究具有重要指导意义。

To analyze the performance of fiber optic gyroscope（FOG）under vibration environment,Wavelet multi-scale analysis and Allan variance analysis are used to analyze the FOG signal which obtained under vibration environment.Wavelet multi-scale transform is used to extract the error terms of FOG error model,the corresponding relationship between the error term and the error coefficient which obtained by Allan variance analysis is pointed out.Then RBF neural network is used to model and compensate the zero-bias error of FOG signal.The simulation results show that the proposed method can resolve the zero-drift,the noise and the zero-bias errors effectively,and can enhance the stability of FOG greatly,which has important guiding significance on the research of FOG performance in vibration environment.