光纤陀螺热致漂移误差的模糊补偿(英文)

针对光纤陀螺启动过程中的热致漂移误差问题,研究了一种模糊模型补偿方案。依据Shupe非互易性理论和Mohr加热模型试验的结论,以光纤环内侧温度和温度变化率为输入,以陀螺漂移为输出,建立了二输入一输出模糊模型。利用全温范围(-25℃～45℃)内光纤陀螺的恒温静态试验数据,基于自适应神经网络模糊推理系统的自学习功能,辨识出模糊规则库。通过实时施行模糊推理可实现光纤陀螺温度漂移的在线自动补偿。室温验证试验表明,陀螺的零偏稳定性由补偿前的0.037(°)/h提高到0.017(°)/h,陀螺启动时间由补偿前的30 min减少为2 min。

Fuzzy compensation of thermally induced bias drift in fiber optical gyro

The thermally induced bias drift in the start-up phase (i.e. start-up bias drift) of fiber optical gyro (FOG) is investigated. And a fuzzy compensation model for the start-up bias drift is presented based on Shupe's non-reciprocity theory and Mohr's heating experiment result. The model takes the temperature and temperature rate of change at the inner side of the fiber coil as inputs. Based on the self-study function of Adaptive Neuro-Fuzzy Inference System(ANFIS), the fuzzy rules of the compensation model are identified from the temperature measurement and the FOG outputs in static experiments(-25℃～45℃). According to the rules identified, the start-up bias drift can be compensated on-line by performing the fuzzy inference. The verification experiment is conducted which shows that, after compensation by the proposed model, the FOG bias stability is increased from 0.037 (°)/h to 0.017 (°)/h, and the start-up time is shortened from 30 min to 2 min.