-
PGC调制技术主要有外调制和内调制两种。外调制中的相位调制可通过调制干涉仪中由光纤缠绕在压电陶瓷环[20]上构成的相位调制器来完成,可以使输入光信号的相位按照一定规律发生变化。内调制技术不需要将调制信号加在调制器上,而是将其直接加载在激光器上,然后通过非平衡干涉仪实现相位生成载波,内调制不需要在调制干涉仪中添加任何器件,对于实现系统的全光化起着重要作用,但是由于内调制需要改变激光器内部的振荡参量,因此要求使用的光源可调频。图 1为内调制型马赫-曾德尔(Mach-Zehnder)光纤干涉仪的基本结构。图中, laser为光源,coupler为耦合器,SMF(single-mode fiber)为单模光纤,signal arm为信号臂,reference arm为参考臂,PD(photodetector)为光电探测器。
Figure 1. Basic structure of internal modulation type Mach-Zehnder optical fiber interference sensor
上述Mach-Zehnder光纤干涉仪结构中,经光电探测器PD响应输出的干涉信号I可表示为[19]:
$ I = A + B{\rm{cos}}[C{\rm{cos}}({\mathit{\omega }_0}t) + \mathit{\varphi }(t)] $
(1) 式中,A是与光纤干涉仪的输入光强、偏振器等的插入损耗有关的直流项;B=kA(k < 1,为干涉条纹的相干度);C表示调制深度;cos(ω0t)为载波信号,ω0为载波信号的角频率;φ(t)为待测信号,可以表示为:
$ \mathit{\varphi }\left( t \right) = D{\rm{cos}}({\mathit{\omega }_s}t) + \mathit{\varphi }(t) $
(2) 式中,D表示待测信号的幅值;ωs为待测信号的角频率;φ(t)为因受外界环境影响而产生的初始相位。将(1)式按照贝塞尔(Bessel)函数展开可以得到:
$ \begin{array}{c} I = A + B\left\{ {[{{\rm{J}}_0}(C) + 2\sum\limits_{k = 1}^\infty {{{\left( { - 1} \right)}^k}{{\rm{J}}_{2k}}\left( C \right) \times } {\rm{ }}} \right.\\ {\rm{cos}}(2k{\mathit{\omega }_0}t)]{\rm{cos}}[\mathit{\varphi }\left( t \right)] - \left[ {2\sum\limits_{k = 0}^\infty {{{\left( { - 1} \right)}^k}{{\rm{J}}_{2k + 1}}(C) \times } {\rm{ }}} \right.\\ \left. {\left. {{\rm{cos}}[\left( {2k + 1} \right){\mathit{\omega }_0}t]]{\rm{sin}}[\mathit{\varphi }(t)} \right]} \right\} \end{array} $
(3) 式中,Jk(C)表示调制深度C的k阶Bessel函数。
-
传统的PGC解调算法主要包括DCM算法和arctan算法两种,其中DCM解调原理框图如图 2所示。图中,⊗表示乘法器,LPF(low pass filter)表示低通滤波器,DIFF(differentiator)表示微分器,∫表示积分器,HPF(high pass filter)表示高通滤波器。
由图 2可知,干涉信号I分别与角频率为ω0、幅度为G和角频率为2ω0、幅度为H的信号发生混频,再通过低通滤波器LPF 1和LPF 2,然后通过LPF 1的信号与经过LPF 2并微分运算后的信号相乘,同时通过LPF 2的信号与经过LPF 1并微分运算后的信号相乘,即可实现两路信号的微分交叉相乘,再对输出的两路信号进行差分和积分运算,经高通滤波后输出的信号最终可表示为[21]:
$ I = {B^2}GH{{\rm{J}}_1}(C){{\rm{J}}_2}(C)\mathit{\varphi }(t) $
(4) 通过上式可以看出,系统最终输出的信号与待测信号之间呈线性关系,由此便可实现对待测信号φ(t)的解调。
另一种传统的PGC解调算法为arctan算法,其基本原理如图 3所示。
由图 3可以看出,输入的干涉信号I通过系统,分别与一倍频和二倍频信号发生混频,混频之后分别通过了低通滤波器LPF 1与LPF 2,输出的两路信号进行相除操作后可以得到[16]:
$ \frac{{{I_1}}}{{{I_2}}} = \frac{{G{{\rm{J}}_1}\left( C \right)}}{{H{{\rm{J}}_2}(C)}}{\rm{tan}}\left[ {\mathit{\varphi }(t)} \right] $
(5) 令式中G=H,并调整C值为2.63rad,此时J1(C)= J2(C),通过对上式进行简化便可得到tan$\left[ {\mathit{\varphi }(t)} \right] $的值,进而进行反正切运算,即可得到φ(t)的值,从而完成对待测信号的解调。
-
对于PGC-DCM算法来说,从(4)式所示的解调信号中可以发现,式中含有与调制深度C有关的贝塞尔函数项,这对于解调的准确性以及系统的稳定性都有很大的影响。为了保证输出信号的稳定,通常将调制深度C的值设定为2.37rad,从而使得J1(C)J2(C)取得极大值,此时解调信号受调制深度C的影响相对较小,但是由于外界环境的影响,C值会产生一定的漂移现象,这使得解调信号的幅度仍然会发生变化,影响了系统整体的稳定性。为得到解调信号的值,需要进行如下运算[21]:
$ \mathit{\varphi }\left( t \right) = {\rm{ }}\frac{{{B^2}GH{{\rm{J}}_1}(C){{\rm{J}}_2}(C)D{\rm{cos}}({\mathit{\omega }_\rm s}t)}}{{{\rm{ }}{B^2}GH{{\rm{J}}_1}(C){{\rm{J}}_2}(C)}} $
(6) 当C值发生漂移时,产生的信号会发生一定程度的失真,此时的信号可表示为:
$ \begin{array}{c} {\mathit{\varphi }^\prime }\left( t \right) = {\rm{ }}\frac{{{B^2}GH{{\rm{J}}_1}(C \pm \Delta C){{\rm{J}}_2}(C \pm \Delta C)D{\rm{cos}}({\mathit{\omega }_\rm s}t)}}{{{B^2}GH{{\rm{J}}_1}(C){{\rm{J}}_2}(C)}} = \\ {D^\prime }{\rm{cos}}({\mathit{\omega }_\rm s}t) \end{array} $
(7) 而对于PGC-arctan算法,最终获得的解调结果如(5)式所示,解调的信号仍然受到调制深度C的影响,为了减小失真,需要将C值设定为2.63rad以保证J1(C)=J2(C),从而使得在后续进行反正切运算时解调结果能够保持准确。当由于外界环境扰动导致C的值发生漂移而偏离2.63rad时,此时式中J1(C)≠J2(C),正切函数的系数偏离了1,此时若对信号直接进行反正切运算,输出的解调信号同样会出现波形失真现象。因此,当外界影响造成调制光频漂移以及调制电压的不稳定时,会导致C值发生漂移,进而影响系统解调信号的稳定性,这些对于调制深度的限定条件都在不同程度上大大降低了系统的灵活性。
-
为了解决传统PGC算法中对于调制深度依赖的问题,提出了一种改进的解调算法,其主要的解调原理和系统框图如图 4所示。
图 4表明,干涉信号I分别与角频率为ω0、幅度为G和角频率为2ω0、幅度为H的信号发生混频后,经过低通滤波器LPF 1和LPF 2,然后通过LPF 2且经过微分处理后的信号与通过LPF 1的信号进行相除运算,同时通过LPF 1且经过微分处理后的信号与LPF 2的信号进行相除运算,即实现两路信号的微分交叉相除,然后再将交叉相除的两路信号进行相乘操作来消去含有调制深度C的贝塞尔函数项,并对输出信号进行开方和积分运算,最终便可实现对待测信号的解调。干涉信号I与基频和二倍频信号分别混频并通过低通滤波器后的信号可分别表示为:
$ {I_1} = - BG{{\rm{J}}_1}(C){\rm{sin}}\left[ {\mathit{\varphi }(t)} \right] $
(8) $ {I_2} = - BG{{\rm{J}}_2}(C){\rm{cos}}\left[ {\mathit{\varphi }(t)} \right] $
(9) 两路信号分别进行微分运算后得到:
$ \frac{{{\rm{d}}{I_1}}}{{{\rm{d}}t}} = - BG{{\rm{J}}_1}(C){\rm{cos}}\left[ {\mathit{\varphi }(t)} \right]{\mathit{\varphi }^\prime }(t) $
(10) $ \frac{{{\rm{d}}{I_2}}}{{{\rm{d}}t}} = - BG{{\rm{J}}_2}(C){\rm{sin}}\left[ {\mathit{\varphi }(t)} \right]{\mathit{\varphi }^\prime }(t) $
(11) 当G=H时,分别进行交叉相除运算后的两路信号可表示为:
$ \frac{{{\rm{d}}{I_1}/{\rm{d}}t}}{{{I_2}}} = \frac{{{{\rm{J}}_1}(C)}}{{{{\rm{J}}_2}(C)}}{\mathit{\varphi }^\prime }(t) $
(12) $ \frac{{{\rm{d}}{I_2}/{\rm{d}}t}}{{{I_1}}} = \frac{{{{\rm{J}}_2}(C)}}{{{{\rm{J}}_1}(C)}}{\mathit{\varphi }^\prime }(t) $
(13) 为方便后续处理,将(12)式和(13)式分别进行绝对值运算后, 再将两式相乘以达到消去贝塞尔函数项的目的,经上述运算后的式子如下:
$ \left| {\left( {\frac{{{\rm{d}}{I_1}/{\rm{d}}t}}{{{I_2}}}} \right)} \right| \times {\rm{ }}\left( {\frac{{{\rm{d}}{I_2}/{\rm{d}}t}}{{{I_1}}}} \right) = {[{\mathit{\varphi }^\prime }(t)]^2} $
(14) 根据系统流程图,再对上式进行开方运算后得到:
$ \sqrt {{{\left[ {{\mathit{\varphi }^\prime }\left( t \right)} \right]}^2}} = {\mathit{\varphi }^\prime }(t) $
(15) 最后通过积分运算以及高通滤波处理来滤除低频干扰之后便可实现对待测信号的解调:
$ \mathit{\varphi }\left( t \right) = D{\rm{cos}}({\mathit{\omega }_{_\rm s}}t) $
(16) 从最终的解调结果中可以看出,输出的解调信号中只包含待测信号,消除了与调制深度C有关的贝塞尔函数项,使得解调结果摆脱了因C值漂移而带来的影响,提高了系统的稳定性。
光纤干涉传感器相位生成载波解调算法研究
Research on improvement of phase generated carrier demodulation algorithm for fiber optic interferometric sensor
-
摘要: 为了解决传统的相位生成载波解调算法中由调制深度漂移引起的解调结果失真现象,采用微分交叉相除的信号解调方法进行了相关的理论分析及仿真验证,得到了一种不受调制深度限制的高性能相位生成载波解调方案。结果表明,在采用不同幅值和频率的待测信号进行仿真时,改进算法的解调性能始终十分优异;且当调制深度的值为典型值2.63rad和2.37rad以及非典型值1.5rad和3.0rad时,使用该算法得到的解调信号都没有发生失真;同时,当调制深度在0.5rad~3.5rad范围内变化时,与传统解调算法相比,改进算法中解调信号的幅值始终与待测信号保持一致, 且高次谐波分量始终非常小。该研究解决了传统解调算法中调制深度变化带来的失真现象,为光纤干涉型传感系统的解调方案提供了参考。Abstract: In order to solve the distortion of the demodulated signal caused by the drift of modulation depth in the traditional phase generation carrier demodulation algorithms, the signal demodulation method of differential cross division was adopted. Related theoretical analysis and simulation verification were carried out, and a high-performance phase-generating carrier demodulation scheme that was not limited by the modulation depth was obtained. The results show that the demodulation performance of the improved algorithm is always excellent when the signals under test of different amplitudes and frequencies are used for simulation. When the values of modulation depth are respectively typical values such as 2.63rad and 2.37rad and atypical values such as 1.5rad and 3.0rad, the demodulated signals obtained by improved algorithm has no distortion. At the same time, when the modulation depth changes in the range of 0.5rad~3.5rad, compared with traditional demodulation algorithms, the amplitude of the demodulated signal in the improved algorithm is always consistent with the signal under test and the high-order harmonic components are always very small. This research solves the distortion phenomenon caused by the modulation depth change in traditional demodulation algorithms, and provides a reference for the demodulation scheme of the optical fiber interferometric sensor system.
-
Key words:
- sensor technique /
- signal demodulation /
- differential cross division /
- modulation depth
-
-
[1] LEE B H, KIM Y H, PARK K S, et al. Interferometric fiber optic sensors[J]. Sensors, 2012, 12(3): 2467-2486. doi: 10.3390/s120302467 [2] LI Y Sh, ZHOU B, LI X Y. Phase modulation and demodulation of interferometric fiber-optic hydrophone using phase generated carrier techniques[J]. Journal of Transducer Technology, 2004, 23(2): 14-17(in Chinese). [3] CAO J N, LI X Y, ZHANG L K, et al. Dynamic range analysis of Mach-Zehnder fiber optic interferometer using PGC homodyne detection scheme[J]. Journal of Harbin Engineering University, 1998, 19(5): 81-87(in Chinese). [4] YIN J, LIU T, JIANG J, et al. Assembly-free-based fiber-optic micro-Michelson interferometer for high temperature sensing[J]. IEEE Photonics Technology Letters, 2016, 28(6): 625-628. doi: 10.1109/LPT.2015.2503276 [5] ZHANG R, JIANG Sh, YAN Q Zh, et al. All-fiber perimeter alarm system based on Mach-Zehnder interference[J]. Laser Technology, 2013, 37(3): 334-337(in Chinese). [6] DANDRIDGE A, TVETEN A B, GIALLORENZI T G. Homodyne demodulation scheme for fiber optic sensors using phase generated ca-rrier[J]. IEEE Journal of Quantum Electronics, 1982, 18(10): 1647-1653. doi: 10.1109/JQE.1982.1071416 [7] WANG Z F, HU Y M, MENG Z, et al. Pseudo working-point control measurement scheme for acoustic sensitivity of interferometric fiber-optic hydrophones[J]. Chinese Optics Letters, 2008, 6(5): 381-383. doi: 10.3788/COL20080605.0381 [8] NI M, HU Y M, MENG Zh, et al. Dynamic range of fiber optic hydrophone using digitized phase generated carrier[J]. Laser & Opto-electronics Progress, 2005, 42(2): 33-37(in Chinese). [9] WANG L, LI Y Q, ZHANG L X, et al. Research progress of phase demodulation in ϕ-OTDR system[J]. Laser Technology, 2019, 43(1): 69-74(in Chinese). [10] CRANCH G A, NASH P J, KIRKENDALL C K. Large-scale remotely interrogated arrays of fiber-optic interferometric sensors for underwater acoustic applications[J]. IEEE Sensors Journal, 2003, 3(1): 19-30. doi: 10.1109/JSEN.2003.810102 [11] BO L H, LIAO Y B, ZHANG M, et al. The improvement on PGC demodulation method based on optical fiber interferometer sensors[J]. Acta Photonica Sinica, 2005, 34(9): 1324-1327(in Chin-ese). [12] MA L, LIU Y, LI Y, et al. Analysis of frequency drift effects in the phase-generated carrier method[J]. Acta Photonica Sinica, 2013, 42(1): 34-37(in Chinese). doi: 10.3788/gzxb20134201.0034 [13] WANG G Q, XU T W, LI F, et al. PGC demodulation technique with high stability and low harmonic distortion[J]. IEEE Photonics Technology Letters, 2012, 24(23): 2093-2096. doi: 10.1109/LPT.2012.2220129 [14] HUANG S C, HUANG Y F, HWANG F H. An improved sensitivity normalization technique of PGC demodulation with low minimum phase detection sensitivity using laser modulation to generate carrier signal[J]. Sensors & Actuators, 2013, A191(2): 1-10. [15] ZHANG S, ZHANG A L, PAN H G, et al. Eliminating light intensity disturbance with reference compensation in interferometers[J]. IEEE Photonics Technology Letters, 2015, 27(17): 1888-1891. doi: 10.1109/LPT.2015.2444421 [16] HE J, WANG L, LI F, et al. An ameliorated phase generated carrier demodulation algorithm with low harmonic distortion and high stability[J]. Journal of Lightwave Technology, 2010, 28(22): 3258-3265. [17] LI Y, SU X X, LIU Y, et al. A PGC demodulation method based on the fundamental frequency mixing[J]. Journal of Optoelectronics·Laser, 2012, 23(5): 933-938(in Chinese). [18] ZHANG A L, WANG Y, GONG M J, et al. An improved algorithm of PGC demodulation method based on fundamental frequency mixing[J]. Acta Photonica Sinica, 2014, 43(2): 0206003(in Chin-ese). doi: 10.3788/gzxb20144302.0206003 [19] SUN W, YU M, CHANG T Y, et al. Research and improvement based in PGC demodulation method[J]. Acta Photonica Sinica, 2018, 47(8): 806004 (in Chinese). doi: 10.3788/gzxb20184708.0806004 [20] CHEN D S, HUANG X D, WANG H B, et al. PGC modulation based on PZT for fiber interferometric sensor[C]// Symposium on Piezoelectricity, Acoustic Waves, and Device Applications (SPAWDA). Shanghai, China: IEEE, 2012: 148-150. [21] NI M. Investigation of the key technologies of fiber optic hydrophone[D]. Beijing: Chinese Academy of Sciences, 2003: 65-68(in Ch-inese).