We present a twist compensated, high accuracy and dynamic fiber optic shape sensing based on phase demodulation in Optical Frequency Domain Reflectometry (OFDR) by using multiple single core fiber based sensor (MFS). A dynamic strain sensing is realized by tracking the optical phase in OFDR and combining with the phase de-hopping filtering algorithm, and the sensing spatial resolution reaches 45 μm. In addition, in order to eliminate the influence of external twist and fluctuation of inherent spin on the shape reconstruction results, we propose an external twist compensation method and inherent spin rate calibration method, respectively. Finally, we use a circle segment method to reconstruct a 3D shape of MFS. The experimental results show that the reconstruction accuracies by the proposed external twist compensation and inherent spin rate calibration methods increase over 18 times and 20 times than those without these two methods, respectively. At the same time, comparing with the traditional cross-correlation-based method, we find that the proposed phase demodulation method has a similar reconstruction accuracy, the maximum reconstruction error is 0.61 %, whereas the shape reconstruction speed is improved by nearly 10 times. This is of great significance for the application of FOSS, which can be used for dynamic shape sensing such as intelligent soft robots, surgical robot and etc.

中文翻译：

---

基于光频域反射计相位解调的扭转补偿高精度动态光纤形状传感

我们通过使用多个基于单芯光纤的传感器 (MFS)，提出了一种基于光频域反射计 (OFDR) 中相位解调的扭曲补偿、高精度和动态光纤形状感测。通过OFDR中的光学相位跟踪并结合相位去跳频滤波算法实现动态应变传感，传感空间分辨率达到45 μm。此外，为了消除外部扭曲和固有自旋波动对形状重构结果的影响，我们分别提出了外部扭曲补偿方法和固有自旋速率校准方法。最后，我们使用圆分段方法重建 MFS 的 3D 形状。实验结果表明，所提出的外部扭曲补偿和固有旋转速率校准方法的重建精度比没有这两种方法的重建精度分别提高了18倍和20倍。同时，与传统的基于互相关的方法相比，我们发现所提出的相位解调方法具有相似的重构精度，最大重构误差为0.61%，而形状重构速度提高了近10倍。这对于FOSS的应用具有重要意义，可用于智能软体机器人、手术机器人等动态形状传感。