CN115811357A

CN115811357A - Optical fiber end event positioning method for OTDR (optical time Domain reflectometer) sampling signals

Info

Publication number: CN115811357A
Application number: CN202211646149.8A
Authority: CN
Inventors: 李凌; 辜嘉; 李文超; 宋凯旋; 欧巧凤
Original assignee: Suzhou Zhongke Advanced Technology Research Institute Co Ltd
Current assignee: Suzhou Zhongke Advanced Technology Research Institute Co Ltd
Priority date: 2022-12-21
Filing date: 2022-12-21
Publication date: 2023-03-17

Abstract

The invention relates to a method for positioning an optical fiber end event of an OTDR sampling signal, which comprises the following steps: denoising an original signal by using wavelet filtering; carrying out sudden change energy extraction on the noise-reduced signal, and detecting the maximum peak value of the sampling signal in the weighted signal; the difference value of the upper envelope and the lower envelope of the signal is obtained by morphology and is used as a noise intensity signal, the noise intensities of the left side and the right side of the maximum peak value are calculated, whether tail noise exists in the signal or not is judged, and the position of an optical fiber ending event is accurately detected. The invention can effectively detect the end event, and has reliable result, high precision and simple calculation.

Description

Method for positioning optical fiber end event of OTDR (optical time Domain reflectometer) sampling signal

Technical Field

The invention relates to a method for positioning an optical fiber end event of an OTDR sampling signal.

Background

An OTDR (optical time-domain reflectometer) is an instrument that uses the back rayleigh scattered light signal generated when pulsed light is transmitted in an optical fiber to characterize the transmission characteristics of the optical fiber. OTDR, a non-destructive fiber measurement technique, can measure the length of an optical fiber, the transmission attenuation of an optical fiber, and fault location. In the OTDR test curve, the types of events included are mainly non-reflection events, and fiber ends, such as fiber splices, fusion splices, bends, breaks, etc., which correspond to losses at various locations in the fiber.

For the event analysis of the OTDR curve, the common methods include a two-point method, a least square method, a wavelet denoising method, and the like. The basic idea is to slope the OTDR backscattering curve and then locate the event on the slope curve. However, many invalid peaks and valleys are generated in the slope curve along with the difference of the signal-to-noise ratio, which affects the positioning of normal events, especially non-reflection events, which are difficult to distinguish by the slope curve alone, and the valid interval of the signal is difficult to clearly detect.

The fiber-ending event may have small reflection peaks or be overwhelmed by noise or the input is some incomplete data segments, and the above method is not robust. The fiber end event is the most critical event and can provide a reference for other parameter detection. Therefore, it is necessary to provide a method for detecting an end event in a complex situation.

Disclosure of Invention

In view of this, it is necessary to provide a method for positioning an optical fiber termination event of an OTDR sampling signal, which can solve the problems of difficult positioning of an effective signal of the OTDR sampling signal, susceptibility to influence of different signal-to-noise ratios, unstable accuracy, and the like.

The invention provides a method for positioning an optical fiber end event of an OTDR sampling signal, which comprises the following steps: s1, denoising an original signal by using wavelet filtering; s2, carrying out mutation energy extraction on the noise-reduced signal, and detecting the maximum peak value of the sampling signal in the weighted signal; and S3, solving a difference value of upper and lower envelopes of the signal by using morphology to serve as a noise intensity signal, calculating the noise intensities of the left side and the right side of the maximum peak value, judging whether tail noise exists in the signal or not, and accurately detecting the position of an optical fiber ending event.

Specifically, the step S1 includes:

5-layer decomposition is carried out on the signals by adopting 'sym8' in a Symlets wavelet family, then a 'heursure' threshold selection method is adopted to determine a threshold, the signals are processed by adopting a soft threshold method, and the signals are not scaled.

Specifically, the step S2 includes:

s21, performing wavelet decomposition on the denoised signal, and reconstructing wavelet coefficients of each layer into high-frequency components to obtain signal mutation energy and SE; extracting a signal pulse SP; fusing the signal mutation energy sum and signal pulse by adopting a maximum value method to obtain a first weight signal;

step S22, multiple weighting of the signals:

step S23, multiplying the weight signal by the original signal to obtain a signal after adaptive weighting;

in step S24, the maximum peak of the sample signal is detected in the weighted signal.

Specifically, the step S21 includes:

6-layer haar wavelet decomposition is carried out on the denoised signals, the wavelet coefficient of each layer is reconstructed into high-frequency components, absolute values of the high-frequency components are taken and added, and signal mutation energy and SE are obtained; extracting a signal pulse SP by utilizing white top hat conversion; fusing signal mutation energy and signal pulse by adopting a maximum value method to obtain a first weight signal: s = max (SE, SP).

Specifically, the step S22 includes:

calculating a sliding window gradient weight signal of the signal: GD = mean (Data (i-WLen: i)) -mean (Data (i: i + WLen));

calculating a position Gaussian weight signal: w = exp (- (x-u) · 2./(sigma. × sigma))' + base; wherein: base is a reference weight value of 0.1; u is the expectation of a gaussian function, i.e. the prior attenuation decibel 10 of the end event; the sigma is 2000, and the probability range of 0.9974 is shown as 6000 data sampling points before and after.

Specifically, step S3 includes:

step S31, solving upper and lower envelopes of a signal by using morphology, and taking a difference value between an upper envelope EU and a lower envelope ED as a noise signal intensity estimation value;

step S32, calculating the noise intensity of the left side and the right side of the maximum peak value, and judging whether tail noise exists in the signal or not by combining the low-frequency amplitude characteristic;

and step S33, if tail noise exists, screening a plurality of pulses before and after the maximum peak position in the filtered signal, and accurately detecting the position of the optical fiber end event.

Specifically, the step S31 includes:

and (3) solving upper and lower envelopes of the signal by using morphology, wherein the difference value of the upper envelope EU and the lower envelope ED is used as a noise signal intensity estimated value: EU = GC (GO (Data), ED = GO (GC (Data, strel)), where strel denotes a one-dimensional flat structure element with a size of 201, GO denotes an open operation, and GC denotes a closed operation.

Specifically, the step S33 further includes:

if there is no tail noise, the cue signal is incomplete.

The method uses various auxiliary signals to carry out self-adaptive weighting on the original signal to highlight the OTDR signal tail peak value, and has the characteristics of high accuracy and strong generalization; extracting signal noise energy by a morphological method to identify whether the signal has tail noise, and identifying the OTDR signal which is abnormally ended with high accuracy; and the end event is accurately detected by screening around the maximum mutation energy, so that the accuracy is high. The invention has small integral operand and is suitable for being deployed in handheld optical fiber detection equipment.

Drawings

Fig. 1 is a flowchart of a method for locating an optical fiber termination event of an OTDR sampled signal according to the present invention;

FIG. 2 is a diagram illustrating a normal end event detection result according to an embodiment of the present invention;

fig. 3 is a schematic diagram of a recognition result of a tailless noise signal according to an embodiment of the present invention;

fig. 4 is a schematic diagram of a detection result of an end event with a small pulse amplitude according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of the end event detection result with a very low SNR and a small pulse amplitude according to an embodiment of the present invention;

FIG. 6 is a diagram illustrating a detection result of an end event of a dense reflection burst according to an embodiment of the present invention;

fig. 7 is a schematic diagram illustrating a detection result of an end event of a large burst according to an embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

In the embodiment, a test sampling signal of a domestic OTDR is adopted for identification, the length of a measured optical fiber is different from 1.25km to 80km, a plurality of attenuation connectors and optical fiber connectors with different sizes are added in the middle, the sampling interval is different from 5ns to 500ns, the signal acquisition time is 15s, 30s and 45s, two optical pulses with wavelengths of 1310nm and 1550nm are adopted, the test event is an end event, and the test signal set comprises a complete test signal containing tail noise and a signal segment without tail noise. The implementation of the method of the embodiment is realized by Matlab programming.

Please refer to fig. 1, which is a flowchart illustrating an operation of a method for positioning an end-of-fiber event of an OTDR sampling signal according to a preferred embodiment of the present invention. Please refer to FIGS. 2-7:

and S1, denoising the original signal by using wavelet filtering. The method specifically comprises the following steps:

calling a wden function in Matlab to realize one-dimensional signal wavelet filtering, performing 5-layer decomposition on a signal by adopting 'sym8' in a Symlets wavelet family, then determining a threshold by adopting a 'soursure' threshold selection method, processing the signal by adopting a soft threshold method, and not scaling the signal. The specific implementation code is as follows:

Data＝wden(Data,'heursure','s','one',5,'sym8')

and S2, carrying out sudden change energy extraction on the noise-reduced signal, and detecting the maximum peak value of the sampling signal in the weighted signal. The method specifically comprises the following steps:

s21, performing 6-layer haar wavelet decomposition on the denoised signals, reconstructing wavelet coefficients of each layer into high-frequency components, taking absolute values of the high-frequency components, and adding the absolute values to obtain signal mutation energy and SE; extracting a signal pulse SP by utilizing white top hat conversion; fusing the signal mutation energy sum and the signal pulse by adopting a maximum value method to obtain a first weight signal: s = max (SE, SP).

Step S22, multiple weighting of the signals:

calculating position Gaussian weight signals: w = exp (- (x-u) · 2./(sigma. × sigma))' + base; wherein: base is a reference weight value of 0.1; u is the expectation of a gaussian function, i.e. the a priori attenuation in decibels 10 of the end event; the sigma is taken as 2000, and represents 6000 data sampling points before and after the probability range of 0.9974.

And S23, multiplying the weight signal by the original signal to obtain a signal subjected to self-adaptive weighting.

DataWeighted＝Data.*S.*GD.*W。

In step S24, the maximum peak value of the sample signal is detected in the weighted signal.

[pks,locs,w1,p1]

＝findpeaks(DataWeight,'MinPeakHeight',0.01,'MinPeakProminence',0.0

1)；

And screening and removing the peak signals with the pulse amplitude smaller than the noise amplitude from the returned peak signals.

Step S3, detecting the position of the optical fiber end event: the difference value of the upper envelope and the lower envelope of the signal is obtained by morphology and is used as a noise intensity signal, the noise intensity of the left side and the right side of the maximum peak value is calculated, and whether tail noise exists in the signal or not is judged by combining with the low-frequency amplitude characteristic; and if tail noise exists, screening a plurality of pulses before and after the maximum peak position in the filtered signal, and accurately detecting the position of the optical fiber ending event. If there is no tail noise, the cue signal is incomplete. Specifically, the method comprises the following steps:

step S31, calculating noise intensity: and (3) solving upper and lower envelopes of the signal by using morphology, wherein the difference value of the upper envelope EU and the lower envelope ED is used as a noise signal intensity estimated value: EU = GC (GO (Data), ED = GO (GC (Data, strel)), where strel denotes a one-dimensional flat structure element of size 201, GO denotes an open operation, and GC denotes a closed operation;

and step S33, if tail noise exists, screening a plurality of pulses before and after the maximum peak position in the filtered signal, and accurately detecting the position of the optical fiber end event. If there is no tail noise, the cue signal is incomplete.

Although the present invention has been described with reference to the presently preferred embodiments, it will be understood by those skilled in the art that the foregoing description is illustrative only and is not intended to limit the scope of the invention, as claimed.

Claims

1. A method for locating an end-of-fiber event of an OTDR sampled signal, the method comprising:

s1, denoising an original signal by using wavelet filtering;

s2, carrying out mutation energy extraction on the noise-reduced signal, and detecting the maximum peak value of the sampling signal in the weighted signal;

and S3, solving a difference value of upper and lower envelopes of the signal by using morphology to serve as a noise intensity signal, calculating the noise intensities of the left side and the right side of the maximum peak value, judging whether tail noise exists in the signal or not, and accurately detecting the position of an optical fiber ending event.

2. A method for locating an end-of-fiber event of an OTDR sampled signal as in claim 1, wherein said step S1 comprises:

3. A method for locating an end-of-fiber event of an OTDR sampled signal according to claim 2, characterized in that said step S2 comprises:

step S21, performing wavelet decomposition on the denoised signal, and reconstructing wavelet coefficients of each layer into high-frequency components to obtain signal mutation energy and SE; extracting a signal pulse SP; fusing the signal mutation energy sum and signal pulse by adopting a maximum value method to obtain a first weight signal;

step S22, multiple weighting of the signals:

4. A method for locating an end-of-fiber event of an OTDR sampled signal according to claim 3, characterized in that said step S21 includes:

5. The method for locating fiber end events of an OTDR sampled signal in accordance with claim 4, wherein said step S22 includes:

calculating position Gaussian weight signals: w = exp (- (x-u) · 2./(sigma. × sigma))' + base; wherein: base is a reference weight value of 0.1; u is the expectation of a gaussian function, i.e. the prior attenuation decibel 10 of the end event; the sigma is taken as 2000, and represents 6000 data sampling points before and after the probability range of 0.9974.

6. A method for locating an end-of-fiber event of an OTDR sampled signal according to claim 5, characterized in that step S3 comprises:

step S31, solving upper and lower envelopes of the signal by using morphology, wherein a difference value between an upper envelope EU and a lower envelope ED is used as a noise signal intensity estimated value;

7. A method for locating an end-of-fiber event of an OTDR sampled signal according to claim 6, wherein said step S31 includes:

and (3) solving upper and lower envelopes of the signal by using morphology, wherein the difference value of the upper envelope EU and the lower envelope ED is used as a noise signal intensity estimated value: EU = GC (GO (Data, strel)), ED = GO (GC (Data, strel)), where strel denotes a one-dimensional flat structural element with a size of 201, GO denotes an open operation, and GC denotes a closed operation.

8. The method for locating an end-of-fiber event of an OTDR sampled signal of claim 7, wherein said step S33 further includes:

if there is no tail noise, the cue signal is incomplete.