-
MZI-BDB传感器包括两个部分:SMF和ECF。传感器的结构示意图、ECF在电子显微镜下的横切面图以及偏心的分布示意图如图4所示。ECF的纤芯的偏心距离(d1)是30.02 µm;纤芯和包层的折射率分别是1.468和1.457。ECF的两端通过型号80S的光纤熔接机(Fujikura Inc., 日本)与SMF错位融合而成。
图 4 (a) MZI-BDB传感器设计示意图; (b) 偏心光纤在电子显微镜下的横切面图
Figure 4. (a) Schematic diagram of the MZI-BDB sensor design; (b) Microscope image of the cross-sections of the eccentric core fiber
熔接前准备工作如下:将光纤熔接机设置成“Manual”手动模式,光纤熔接类型选择“SS”,持续放电电流设置成“−23 bit”,持续放电时间选择350 ms。首先将需要熔接的单模光纤通过FC/PC接口连接至扫频光源,由于光功率计采用端面接收,因此将ECF一端利用FC/PC接口直接和光功率计相连,保证在进行熔接过程中的实时监测;利用光纤熔接机手动模式调节ECF与第一段SMF在X轴、Y轴相对位置,当光功率计监测光功率值最大时表示SMF与ECF纤芯完全对准,此后继续采用小步进模式进行调节两芯相对位置,当光功率计显示数值衰减至约50%时,代表SMF与ECF纤芯错位约4 μm,此时进行熔接,并且完毕后将ECF切割3 cm。将切割后的ECF另一端与另一段SMF在光纤熔接机中进行错芯熔接,SMF的另一段连接至光谱仪,通过观察光谱仪中干涉条纹对比度变化的方式进行调节。通过光纤熔接机的调节,光谱仪上的干涉条纹对比度会出现由最小到最大,再到最小,再逐渐变大这种周期性变化。这是由于光纤熔接机在初始状态下将ECF和SMF中心纤芯完全对准,此时干涉对比度最小。当调节相对位置时,SMF纤芯进入ECF偏心纤芯区域,SMF纤芯与ECF偏心纤芯产生错位,干涉对比度增强,当错位约4 μm时干涉对比度最大;当继续调节两芯相对位置时,SMF纤芯与ECF的偏心纤芯错位距离减少,干涉对比度减弱;继续调节时,两芯的错位距离再次增加,干涉对比强度开始增强;当SMF纤芯逐渐远离ECF偏心纤芯,错位距离减少,干涉对比度随之变小。根据以上变化规律,将SMF纤芯与ECF的纤芯在干涉对比度最强时进行错位熔接。
-
MZI-BDB传感器工作原理如下:传感器被封装于软硅树脂片内,通过绑带固定于膝关节处,随着受试者关节的运动,传感器被诱导弯曲。当扫频激光光源的光信号从MZI-BDB传感器中SMF和ECF的第一个熔接点透射时,ECF的包层模被激发。在第二个熔接点处,ECF的基模和包层模耦合进入第二个SMF的纤芯。如果基模和包层模的透射光信号存在相位差,则第二次耦合会产生干涉,并且在该结构中激发更多的高阶包层模。
在关节弯曲检测过程中,MZI-BDB传感器的弯曲模拟示意图如图5所示,S1是SMF和ECF的第1熔接面,S2是ECF与SMF的第2熔接面,S1和S2的夹角如公式(1)所示:
$$\theta \approx \frac{{{L_{{{cl}}}}}}{R} = \frac{{{L_{co}}}}{{R + d}}$$ (1) 式中:R为弯曲曲率半径;d为偏心纤芯和中性面的距离;Lcl和Lco分别为包层模和基模长度。公式(1)可以转换成:
$${L_{{{co}}}} = \left( {1 + d \cdot C} \right) \cdot {L_{cl}}$$ (2) 式中:C为膝关节弯曲曲率,代表膝关节在运动过程中的弯曲程度。弯曲曲率越大,表示膝关节的弯曲程度越大。光信号从ECF透射后,基模和包层模之间的相位差γ如公式(3)所示:
$$\gamma = \frac{{2\pi }}{\lambda }\left( {{n_{co}}{L_{co}} - {n_{cl,i}}{L_{cl}}} \right)$$ (3) 式中:nco和ncl,i分别为基模和第i阶包层模的有效折射率;λ为扫频激光光源波长。当相位差满足
$\gamma = \left( {2k + 1} \right) \cdot \pi $ ,谐振波长如公式(4)所示:$${\lambda _d} = \frac{{2({n_{co}}{L_{co}} - {n_{cl,i}}{L_{cl}})}}{{2k + 1}}$$ (4) 传感器被固定于膝关节处,其芯内中性面垂直于肌肉表面。当受试者正常行走时,膝关节的运动会诱导MZI-BDB传感器发生弯曲。此时基模的nco和Lco会发生明显的变化,然而包层模的ncl,i和Lcl改变量比基模的改变量要小。因此,基模与包层模之间的光程差(
${n_{co}}{L_{co}} - {n_{cl,i}}{L_c}_l$ )发生变化,导致干涉谐振波长发生漂移。具体来说,当膝关节运动,诱导传感器在0°方向发生弯曲(正向弯曲),基模与包层模之间的光程差(${n_{co}}{L_{co}} - {n_{cl,i}}{L_c}_l$ )增大,从而导致干涉谐振波长增加;当膝关节诱导传感器向180°方向弯曲(负向弯曲),基模与包层模之间的光程差(${n_{co}}{L_{co}} - {n_{cl,i}}{L_c}_l$ )减小,从而导致干涉谐振波长减小。 -
检测MZI-BDB传感器特性的装置示意图如图6所示,该装置由扫频光源、偏振控制器、光纤旋转夹具、金属板、千分尺以及光谱仪(YOKOGAWA, 日本)构成。扫频光源输出光信号通过单模光纤传输至MZI-BDB传感器,其透射光信号输入光谱仪。其中,偏振控制器确保光信号在该过程中偏振态保持不变;光纤旋转夹具用于调整MZI-BDB传感器的弯曲方向;传感器被固定于金属板上,放置砝码(3 g)确保传感器在测试过程中始终紧贴于金属板;千分尺用于调节被测传感器的弯曲曲率。按照2.1节所述步骤熔接后,将固定于载玻片上的MZI-BDB传感器摆放状态进行标记,此时载玻片平面为90°方向和270°方向所在平面,垂直于载玻片平面为0°方向和180°方向所在平面,传感器的传感方位基于以上标记确定。
图 6 MZI-BDB传感器特性检测装置示意图
Figure 6. Schematic diagram of MZI-BDB sensor characteristic detection device
调节光纤旋转夹具,使得MZI-BDB传感器分别处于0°、90°和180°弯曲方向下。随着弯曲曲率的调节,通过光谱仪检测出传感器在不同弯曲方向和弯曲曲率下透射光的谐振波长变化情况。该传感器研制长度为35.1 mm,千分尺调节MZI-BDB传感器的曲率步长和范围分别是0.1 m−1、0~2 m−1。图7中,在0°弯曲方向下(正向弯曲),MZI-BDB传感器的弯曲灵敏度为5.29 nm/m−1,分辨率为0.11 m−1,线性拟合相关系数R2为0.989;在180°弯曲方向下(负向弯曲),MZI-BDB传感器的弯曲灵敏度为−3.11 nm/m−1,分辨率为0.12 m−1,线性拟合相关系数R2为0.995。理想情况下在90°弯曲方向下,MZI-BDB传感器的弯曲灵敏度应为0,然而实验测量结果是−0.14 nm/m−1,通过分析认为是测量过程中弯曲方向的误差所导致。将传感器水平放置在加温板上,温度调节范围为20~90 ℃,所测MZI-BDB传感器的温度灵敏度为0.043 nm/℃,相关系数R2为0.998,该结果表明传感器对温度不敏感,具备避免在检测过程中受温度交叉影响的特性。
Research on knee joint curvature detection system based on fiber optic MZI-BDB curvature sensor
-
摘要: 在人体运动监测的过程中,膝关节运动信息在中老年慢性疾病的诊断和康复评估方面具有重要意义。研制了一种基于光纤马赫曾德尔干涉曲率传感器(Mach-Zehnder interferometer-based directional bending,MZI-BDB)的膝关节弯曲检测系统。该系统由MZI-BDB传感器、扫频激光光源、光纤耦合器、光纤隔离器、光电探测器及信号处理系统构成。MZI-BDB传感器由偏心光纤和单模光纤错位熔接而成,封装于软硅树脂片内,通过绑带固定于膝关节处。当膝关节屈曲和伸展时,诱导MZI-BDB传感器发生弯曲,传感器内透射光信号干涉场的模场状态发生变化,谐振波长发生漂移,从而对膝关节的弯曲方向和曲率进行监测。MZI-BDB传感器在正向和负向弯曲的测量角度范围为0°~90°;在正向弯曲方向上灵敏度和分辨率分别为5.29 nm/m−1和0.11 m−1;在负向弯曲方向上的灵敏度和分辨率分别为−3.11 nm/m−1和0.12 m−1。实验测试MZI-BDB传感器温度敏感度为0.043 nm/℃,该结果显示传感器对温度的不敏感特性。光电编码器与MZI-BDB传感器同时进行数据的传感采集。实验结果表明:该检测系统和光电编码器验证平台在准确度和响应度上具有一致性。Abstract: In the monitoring process of human motion and posture, knee movement information was of great significance in the diagnosis and rehabilitation evaluation of chronic diseases in the middle and old age. The knee directional bending measurement device using a Mach-Zehnder interferometer-based directional bending (MZI-BDB) sensor was presented in this paper. The system consisted of MZI-BDB sensor, swept laser light source, optical fiber coupler, optical fiber isolator, photodetector and signal processing system. MZI-BDB sensor was fabricated by fusion-splicing a section of eccentric core fiber (ECF) between two single-mode fibers (SMF) with core-offset. It was encapsulated in a soft silicone sheet and fixed to the knee joint by a band. When the knee joint was flexed and extended, the MZI-BDB sensor was led to bend, causing the modal interferes of the sensor changes and the resonant wavelength shifted, so as to monitor the bending direction and curvature of the knee joint. The range of measurement of MZI-BDB sensor was 0°-90° at the positive and negative bendings respectively. At the positive bending (bending direction of 0°), the proposed MZI-BDB sensor sensitivity and resolution were 5.29 nm/m−1 and 0.11 m−1 respectively. The sensitivity and resolution were −3.11 nm/m−1 and 0.12 m−1 at the negative bending (bending direction of 180°). The temperature sensitivity of the proposed MZI-BDB was 0.043 nm/℃, which had minimal effect on the experiments. Furthermore, the photoelectric encoder and MZI-BDB sensor were used for data acquisition simultaneously. The experimental results show that the detection system and the photoelectric encoder verification platform are consistent in accuracy and responsiveness.
-
Key words:
- optical fiber sensing /
- eccentric core optical fiber /
- curvature sensor /
- gait analysis
-
-
[1] Nathalie C, Bairc W, Susan M, et al. Prediagnostic markers of idiopathic Parkinson’s disease: Gait, visuospatial ability and executive function [J]. Gait & Posture, 2019, 68: 500-505. [2] Rossom S, Smith C, Thelen D, et al. Knee joint loading in healthy adults during functional exercises: Implications for rehabilitation guidelines [J]. J Orthop Sport Phys, 2018, 48(3): 162-173. doi: 10.2519/jospt.2018.7459 [3] Mohamed A, Ren J, Sharaf A, et al. Optimal path planning for unmanned ground vehicles using potential field method and optimal control method [J]. International Journal of Vehicle Performance, 2018, 4(1): 1-14. doi: 10.1504/IJVP.2018.088780 [4] Song L M, Huang H Z, Chen Y, et al. Fast measurement of human body posture based on three-dimensional optical information [J]. Infrared and Laser Engineering, 2020, 49(6): 20200079. (in Chinese) doi: 10.3788/irla.7_2020-0079 [5] Ciechanowicz K, Stolarz M, Glabet G, et al. Biomechanical image of the knee motion in patients with chronic anterior instability of the knee joint before and after Kinesio Taping [J]. J Back Musculoskelet, 2019, 33(7): 1-9. [6] Bloomfield, Aaron R, Fennema, et al. Proposal and validation of a knee measurement system for patients with osteoarthritis [J]. IEEE T Bio-med Eng, 2019, 66(2): 319-326. doi: 10.1109/TBME.2018.2837620 [7] Zhu G Y, Tian X D, Du D F, et al. Medical image analysis of knee joint lipoma arborescens and arthroscopic treatment [J]. Comput Med Imag Grap, 2018, 1(3): 66-72. [8] Lin G, Battistelli G, Chisci L, et al. Distributed joint sensor registration and multitarget tracking via sensor network [J]. IEEE T Aero Elec Sys, 2019, 99: 1-1. [9] Bokhman E, Filatov Y, Pavel A, et al. Research of dynamic goniometer system for direction measurements [J]. Opt Eng, 2019, 58(9): 094106. [10] Jun K, Lee D W, Lee K, et al. Feature extraction using an RNN autoencoder for skeleton-based abnormal gait recognition [J]. IEEE Access, 2020, 8: 19196-19207. doi: 10.1109/ACCESS.2020.2967845 [11] Roetenberg D, Höller C, Mattmüller K, et al. Comparison of a low-cost miniature inertial sensor module and a fiber-optic gyroscope for clinical balance and gait assessments [J]. J Healthc Eng, 2019, 2(3): 1-11. [12] Bernini R, Persichetti G, Catalano E, et al. Refractive index sensing by Brillouin scattering in side-polished optical fibers [J]. Opt Lett, 2018, 43(10): 2280-2283. doi: 10.1364/OL.43.002280 [13] Zhang C B, Ning T G, Li J, et al. Etching twin core fiber for the temperature-independent refractive index sensing [J]. Joumal of Optics, 2018, 20(4): 045802. doi: 10.1088/2040-8986/aab0e5 [14] Pant S, Umesh S, Asokan S. Knee angle measurement device using fiber Bragg grating sensor [J]. IEEE Sens J, 2018, 18(24): 10034-10040. doi: 10.1109/JSEN.2018.2875564 [15] Woyessa G, Fasano A, Markos C, et al. Low loss polycarbonate polymer optical fiber for high temperature FBG humidity sensing [J]. IEEE Sens J, 2017, 29(7): 575-578. [16] Sun B, Fang F, Zhang Z X, et al. High-sensitivity and low-temperature magnetic field sensor based on tapered two-mode fiber interference [J]. Opt Lett, 2018, 43(6): 1311-1314. doi: 10.1364/OL.43.001311 [17] Huang B, Shu X W. Highly sensitive torsion sensor with femtosecond laser-induced low birefringence single-mode fiber based Sagnac interferometer [J]. Opt Express, 2018, 26(4): 4563-4571. doi: 10.1364/OE.26.004563