Fiber core on-line rapid test method and system

Abstract

The invention discloses a fiber core on-line rapid test method and system, and relates to the technical field of communication, and the method comprises the steps of optical signal data collection, edge calculation primary processing, data receiving and integration, principal component analysis dimension reduction, optical signal attenuation prediction model establishment, fault point positioning and decision support. The system is composed of an online OTDR test unit, a cloud platform and a resource visualization system, and all the parts cooperate to complete a test process. Compared with the prior art, real-time online monitoring of the fiber core state can be realized, and the monitoring timeliness is improved; the data processing speed is high, and the test result acquisition period is shortened; the state of the fiber core can be mastered more accurately through multi-dimensional data processing; the fault positioning precision is high, and the repair time is saved; the resource visualization system provides comprehensive visualization display and decision support, and the scientificity and effectiveness of optical fiber communication network operation and maintenance are improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a fiber core online rapid testing method and system. Background Art

[0002] As the second physical network of the power system, the power communication network undertakes the important responsibility of ensuring the safe and stable operation of the power grid and achieving efficient dispatching and management. As the core physical transmission medium of the power communication network, the power communication optical cable carries important services such as power grid relay protection, stability control, automation and video monitoring. It is the basis for the informationization of power grid dispatching, operation and management, and is also an important means to ensure the safe and stable operation of the power grid. Therefore, its stability and reliability are directly related to the safety and efficiency of the entire power system, and play an important role in promoting the green and low-carbon transformation of the supply and demand sides of the power system.

[0003] The power communication optical cable network is becoming increasingly large and complex. Taking the Dongying Power Supply Company as an example, the total length of the existing power communication optical cables exceeds 5,000 kilometers, with a total of more than 650 optical cables, containing a total of more than 15,000 fiber cores. According to the specifications, the communication profession should inspect all the spare fiber cores every year to understand the availability, electrical corrosion, joint boxes and attenuation of the optical cable cores. The conventional detection method is that the inspectors carry bulky test equipment to the station and use handheld test equipment to perform single-core detection of the optical cable. The time for testing the optical cable core at a single station is 1.5-2 hours. Therefore, for such a large optical cable network, the traditional manual inspection and testing methods can no longer meet the operation and maintenance needs, resulting in huge operation and maintenance pressure and low operation and maintenance efficiency. A fast detection system is urgently needed to solve the pressure of optical cable network operation and maintenance.

[0004] The traditional manual operation and maintenance mode is difficult to guarantee the reliability of optical cable operation and maintenance data. The traditional optical cable testing process relies on manual operation and input of test results, which has problems such as low test efficiency, human factors, and low accuracy. The increasingly complex optical cable network topology structure further increases the difficulty of operation and maintenance. With the development of smart grids, the network topology structure of optical cables is becoming more and more complex, which puts higher requirements on the operation and maintenance management of power communication optical cables. Especially in emergencies such as extreme weather and natural disasters, the failure risk of optical cables is significantly increased, the monitoring demand for high-risk sites is increasing, and the demand for the network topology structure of optical cables and the health prediction of optical fibers is more urgent. The traditional operation and maintenance mode lacks the visualization resources of the optical cable network topology structure, and it is difficult to locate, respond to and handle faults in time. At the same time, the level of intelligent operation and maintenance is low, and the optical fiber status cannot be monitored in real time, resulting in untimely fault discovery and positioning, which seriously affects the safe and stable operation of the power grid. Therefore, it has become a top priority to improve the intelligent level of optical cable operation and maintenance and realize reliable operation and maintenance of optical cable core monitoring. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for online fast testing of fiber cores, which collects optical signal data through an online OTDR test unit, and analyzes transmission characteristics in combination with optical propagation theory to locate fault points through edge computing, cloud platform processing, principal component analysis, and establishment of an attenuation prediction model. This is accomplished collaboratively by an overall architecture including a test unit, a cloud platform, and a resource visualization system. The above problems have been solved.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A fiber core online rapid testing method, characterized by comprising the following steps:

[0008] S1: optical signal data acquisition;

[0009] S2: Edge computing performs preliminary processing;

[0010] S3: data reception and integration;

[0011] S4: principal component analysis and data dimensionality reduction;

[0012] S5: Establish an optical signal attenuation prediction model;

[0013] S6: Fault point location and decision support.

[0014] Step S1 of optical signal data collection specifically includes: using an online OTDR test unit to collect an optical signal data sequence {I n}, whose sequence is represented as:

[0015]

[0016] Where t represents time, Δt represents the acquisition time interval, T is the total acquisition time, I n is the nth data point in the optical signal data sequence collected within the time interval [0, T] according to the collection time interval Δt.

[0017] Step S2 edge computing performs preliminary processing, including data filtering and feature extraction;

[0018] The data filtering is specifically: using a finite impulse response FIR filter, whose filter coefficient sequence is {h k}, the length is M, the filtered optical signal data sequence {I′ n Through convolution operation, we get:

[0019]

[0020] Where I′ nis the nth data point in the optical signal data sequence after filtering, and its filter coefficient sequence is {h k}, M is the filter coefficient sequence {h k} length, I n-k is the filtered data point I′ in the original collected optical signal data sequence that is currently to be calculated n The relevant nk-th data point.

[0021] Step S2 edge computing performs preliminary processing, including data filtering and feature extraction;

[0022] The feature extraction is specifically: extracting features from the filtered optical signal data and calculating the average intensity of the optical signal and intensity variance σ 2 , the formula is:

[0023]

[0024] in is the average intensity of the optical signal, and the filtered optical signal data is obtained from the M-1th data point to the last data point, that is, The data points are summed and averaged, which can roughly reflect the overall intensity level of the light signal during this period of time; σ 2 is the optical signal intensity variance, which is used to measure the dispersion of the optical signal intensity relative to the average intensity within a certain time range. n With average intensity The square sum of the differences is obtained by corresponding averaging, which can be used to further analyze the stability and other characteristics of the optical signal; N is the number of data points involved in calculating the average intensity and variance, and its value is T is the total acquisition duration, Δt is the acquisition time interval, and M is the length of the filter coefficient sequence.

[0025] Step S3 data reception and integration specifically includes: using the cloud platform to receive the characteristic value data sequence from the online OTDR test unit, including the average intensity and the intensity variance σ 2 , these data are arranged in chronological order to form time series data, which can be expressed in vector form:

[0026]

[0027] in The average intensity of the optical signal received by the cloud platform from the online OTDR test unit at time t i The specific value of σ, which forms a time series data about the average intensity over time;2 (t i ) is the optical signal strength variance received by the cloud platform from the online OTDR test unit at time t i The specific value of forms a time series data about the intensity variance as time goes by; X is a two-dimensional data matrix formed by arranging the eigenvalue data such as the average intensity and intensity variance of the optical signal received by the cloud platform at different times in time sequence; L is the length of the received data time series, that is, the number of groups of eigenvalue data received by the cloud platform at different times, which determines the number of rows of the data matrix X.

[0028] Step S4: principal component analysis and data dimensionality reduction are as follows: principal component analysis is used for data dimensionality reduction:

[0029]

[0030] Among them, C is the covariance matrix of the data matrix X, which is an important basis for data processing operations such as principal component analysis. is the vector mean of X, L is the length of the received data time series, that is, the number of rows of the data matrix X, and T is the total duration of the acquisition;

[0031] The principal component analysis is performed on the two dimensions of the received average intensity and intensity variance to calculate the covariance matrix C of the data matrix X; is the mean vector of X, solving for the eigenvalue λ of the covariance matrix C i and the corresponding eigenvector v i (i=1,2), by selecting the eigenvectors corresponding to the main eigenvalues ​​to form the projection matrix P, the original data is reduced in dimension to obtain the reduced-dimensional data Y:

[0032] Y=P T X

[0033] Where Y is the data after dimensionality reduction by principal component analysis, which is obtained by multiplying the original data matrix X with the transpose of the projection matrix P consisting of the eigenvectors corresponding to the main features. P is the projection matrix, which consists of the eigenvectors corresponding to the selected main eigenvalues. These eigenvectors are obtained by solving the eigenvalues ​​and eigenvectors of the covariance matrix C.

[0034] Step S5 establishes the optical signal attenuation prediction model specifically by using a nonlinear regression model to establish a prediction model based on the optical signal attenuation:

[0035] A(t)=β0+β1t+β2t 2 +…+β m t m +γ1z1(t)+γ2z2(t)+…+γ n z n(t)

[0036] Where A(t) is the function of the optical signal attenuation over time, β i (i=0,1,…,m) are regression coefficients in the optical signal attenuation prediction model. The values ​​of these coefficients are determined by fitting the data using the least squares method, thereby determining the specific optical signal attenuation prediction model. j (j = 1, 2, ..., n) is the regression coefficient associated with the vector z(t) composed of other factors affecting the optical signal attenuation in the optical signal attenuation prediction model. j (t) is the function of the jth factor in the vector z(t) composed of other factors that affect the attenuation of the optical signal changing with time;

[0037] The actual observed optical signal attenuation data series {A k (t k )} and the corresponding predicted value is {A′ k )t k )}, set the error square sum function to:

[0038]

[0039] where E)β0,β1,…,β m ,γ1,…,γ n ) is the error sum of squares function, A k (t k ) is the kth data point in the actually observed optical signal attenuation data sequence, A′ k (t k ) is the corresponding predicted value, K is the length of the optical signal attenuation data sequence actually observed, that is, the number of attenuation data points measured at different times;

[0040] By solving the system of partial derivative equations that minimizes E:

[0041]

[0042]

[0043] The optimal regression coefficient is obtained to determine the optical signal attenuation prediction model.

[0044] Step S6 of fault location and decision support is as follows: using the established prediction model and the optical signal data received in real time, analyzing the transmission characteristics of the optical signal in the optical cable, and establishing the following relationship based on the light propagation theory:

[0045] I(x)=I0e -αx +R(x)

[0046] Where I(x) is the function of the change of optical signal intensity with transmission distance, where x is the transmission distance of the optical signal in the optical cable, I0 is the initial intensity of the optical signal, that is, the intensity value of the optical signal when it enters the optical cable and starts to transmit, α is the attenuation coefficient of the optical signal in the optical cable, and R(x) is the reflected signal intensity function of the optical signal during the transmission process;

[0047] By fitting the collected optical signal data, the specific form of the attenuation coefficient α and the reflected signal intensity function R(x) is determined, thereby analyzing the transmission characteristics of the optical signal in the optical cable;

[0048] When an optical cable fails, the transmission characteristics of the optical signal will change significantly. At this time, the optical signal strength will drop sharply and the reflected signal strength will suddenly increase. By monitoring the changes in the optical signal strength and the reflected signal strength, combined with the above optical signal transmission characteristics analysis, when the fault judgment condition x is met, f , the fault point location can be determined:

[0049] x f = argmax{x|I(x)≤I0-ΔI th and R(x)≥R0+ΔR th}

[0050] where x f is the distance from the starting point to the fault point, I0 is the initial intensity of the optical signal, ΔI th is the threshold value of the optical signal intensity drop, R0 is the reference value of the reflected signal intensity under normal circumstances, ΔR th Threshold value added for reflected signals;

[0051] Based on the fault point location and the analysis results of optical signal transmission characteristics, decision support is provided for fault detection and repair.

[0052] A fiber core online rapid testing system, characterized in that it comprises an online OTDR testing unit, a cloud platform and a resource visualization system, characterized in that the online OTDR testing unit implements the steps of the method described in any one of claims 1 to 9 when executed with the cloud platform.

[0053] The resource visualization system provides visualization of optical fiber status and optical cable resources, including the optical fiber direction, connection points, occupancy of each port, optical fiber connection status, fault detection information, and distinguishes the quality status of optical cables based on GIS maps and optical cable quality coloring functions.

[0054] First, the optical signal data sequence is collected within the set time interval through the online OTDR test unit to complete the optical signal data collection step. Then, edge computing is used to perform preliminary processing on the collected data, including data filtering using a finite impulse response FIR filter, and extracting features such as average intensity and intensity variance from the filtered data. After that, the cloud platform receives the eigenvalue data sequence from the test unit and arranges it in chronological order to form time series data, and then performs dimensionality reduction processing on the received data in two dimensions of average intensity and intensity variance through principal component analysis. Subsequently, a nonlinear regression model is used to establish a prediction model based on the attenuation of the optical signal, and the optimal regression coefficient is determined by solving the group of partial derivative equations that minimize the error sum of squares function. Finally, the established prediction model is combined with the real-time received optical signal data, and the transmission characteristics of the optical signal in the optical cable are analyzed according to the light propagation theory. When a fault occurs in the optical cable, the fault point location is determined according to the fault judgment conditions by monitoring the changes in the optical signal intensity and the reflected signal intensity, and based on this, decision support is provided for troubleshooting and repair. At the same time, the entire process is completed by the online OTDR test unit, the cloud platform, and the resource visualization system that provides functions such as optical fiber and optical cable resource visualization. Each part is executed according to the corresponding steps to realize the workflow of the entire fiber core online rapid test.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] This technology can continuously collect optical signal data through the online OTDR test unit to achieve real-time online monitoring of the fiber core status. There is no need to interrupt the normal operation of the optical fiber, and the actual working state of the optical fiber can be grasped at all times, which greatly improves the timeliness of monitoring and can detect potential problems in the first place.

[0057] With the help of edge computing for preliminary processing and a series of processes such as data reception and integration on the cloud platform, data processing speed is fast. The conversion from optical signal acquisition to effective data for analysis can be completed in a short time, which greatly shortens the cycle of obtaining test results compared with traditional methods, allowing operation and maintenance personnel to respond more quickly.

[0058] It not only collects optical signal data, but also processes and reduces multi-dimensional data such as average intensity and intensity variance through principal component analysis, so as to more comprehensively and deeply explore data features. Compared with traditional single-dimensional or simple data processing test methods, it can more accurately grasp the true state of the fiber core and reduce misjudgment.

[0059] In the fault location link, based on the established optical signal attenuation prediction model and the detailed analysis of optical signal transmission characteristics, combined with the light propagation theory, the fault point can be determined more accurately. Compared with the traditional fault location method that relies on experience or rough troubleshooting, the accuracy has been significantly improved, which can effectively reduce the scope of troubleshooting and save fault repair time.

[0060] The supporting resource visualization system provides a comprehensive visualization of the fiber status and optical cable resources, including fiber directions, connection points, port occupancy, and many other details. Operation and maintenance personnel can intuitively understand the overall situation, providing a strong basis for formulating reasonable operation and maintenance strategies and troubleshooting plans, which is an intuitive auxiliary decision-making function lacking in traditional technologies.

[0061] By integrating the collected and analyzed data, scientific and accurate decision support can be provided for troubleshooting and repair. No longer based on experience alone, but based on actual data and accurate analysis results, operation and maintenance decisions are more reasonable and efficient, improving the scientificity and effectiveness of the entire optical fiber communication network operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a diagram of a fiber core online rapid testing system of the present invention;

[0063] Figure 2 This is a working diagram of a fiber core online rapid testing system of the present invention;

[0064] Figure 3 This is a flow chart of a method for online rapid testing of a fiber core according to the present invention; DETAILED DESCRIPTION

[0065] The technical solutions in the embodiments of the present invention will be fully described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0066] like Figure 1-3 As shown, a fiber core online rapid testing method is characterized by comprising the following steps:

[0067] S1: optical signal data acquisition;

[0068] S2: Edge computing performs preliminary processing;

[0069] S3: data reception and integration;

[0070] S4: principal component analysis and data dimensionality reduction;

[0071] S5: Establish an optical signal attenuation prediction model;

[0072] S6: Fault point location and decision support.

[0073] Step S1 of optical signal data collection specifically includes: using an online OTDR test unit to collect an optical signal data sequence {I n}, whose sequence is represented as:

[0074]

[0075] Where t represents time, Δt represents the acquisition time interval, T is the total acquisition time, I n is the nth data point in the optical signal data sequence collected within the time interval [0, T] according to the collection time interval Δt.

[0076] Step S2 edge computing performs preliminary processing, including data filtering and feature extraction;

[0077] The data filtering is specifically: using a finite impulse response FIR filter, whose filter coefficient sequence is {h k}, the length is M, the filtered optical signal data sequence {I′ n Through convolution operation, we get:

[0078]

[0079] Where I′ n is the nth data point in the optical signal data sequence after filtering, and its filter coefficient sequence is {h k}, M is the filter coefficient sequence {h k} length, I n-k is the filtered data point I′ in the original collected optical signal data sequence that is currently to be calculated n The relevant nk-th data point.

[0080] Step S2 edge computing performs preliminary processing, including data filtering and feature extraction;

[0081] The feature extraction is specifically: extracting features from the filtered optical signal data and calculating the average intensity of the optical signal and the intensity variance σ 2 , the formula is:

[0082]

[0083]

[0084] in is the average intensity of the optical signal, σ 2 is the optical signal intensity variance, N is the number of data points involved in calculating the average intensity and variance, and its value is T is the total acquisition duration, Δt is the acquisition time interval, and M is the length of the filter coefficient sequence.

[0085] Step S3 data reception and integration specifically includes: using the cloud platform to receive the characteristic value data sequence from the online OTDR test unit, including the average intensity and the intensity variance σ 2 , these data are arranged in chronological order to form time series data, which can be expressed in vector form:

[0086]

[0087] in The average intensity of the optical signal received by the cloud platform from the online OTDR test unit at time t i The specific value of σ 2 (t i ) is the optical signal strength variance received by the cloud platform from the online OTDR test unit at time t i , X is the two-dimensional data matrix formed by arranging the characteristic value data such as the average intensity and intensity variance of the optical signal received by the cloud platform at different times in chronological order, and L is the length of the received data time series.

[0088] Step S4: principal component analysis and data dimensionality reduction are as follows: principal component analysis is used for data dimensionality reduction:

[0089]

[0090] Among them, C is the covariance matrix of the data matrix X, which is an important basis for data processing operations such as principal component analysis. is the vector mean of X, L is the length of the received data time series, that is, the number of rows of the data matrix X, and T is the total duration of the acquisition;

[0091] The principal component analysis is performed on the two dimensions of the received average intensity and intensity variance to calculate the covariance matrix C of the data matrix X; is the mean vector of X, solving for the eigenvalue λ of the covariance matrix C i and the corresponding eigenvector v i (i=1,2), by selecting the eigenvectors corresponding to the main eigenvalues ​​to form the projection matrix P, the original data is reduced in dimension to obtain the reduced-dimensional data Y:

[0092] Y=P T X

[0093] Where Y is the data after dimensionality reduction by principal component analysis, which is obtained by multiplying the original data matrix X with the transpose of the projection matrix P consisting of the eigenvectors corresponding to the main features. P is the projection matrix, which consists of the eigenvectors corresponding to the selected main eigenvalues. These eigenvectors are obtained by solving the eigenvalues ​​and eigenvectors of the covariance matrix C.

[0094] Step S5 establishes the optical signal attenuation prediction model specifically by using a nonlinear regression model to establish a prediction model based on the optical signal attenuation:

[0095] A(t)=β0+β1t+β2t 2 +…+β m t m +γ1z1(t)+γ2z2(t)+…+γ n z n (t

[0096] Where A(t) is the function of the optical signal attenuation over time, β i (i=0,1,…,m) are regression coefficients in the optical signal attenuation prediction model. The values ​​of these coefficients are determined by fitting the data using the least squares method, thereby determining the specific optical signal attenuation prediction model. j (j = 1, 2, ..., n is the regression coefficient associated with the vector z(t) composed of other factors affecting the optical signal attenuation in the optical signal attenuation prediction model, z j (t) is the function of the jth factor in the vector z(t) composed of other factors that affect the attenuation of the optical signal changing with time;

[0097] The actual observed optical signal attenuation data series {A k (t k )} and the corresponding predicted value is {A′ k (t k )}, set the error sum of squares function to:

[0098]

[0099] where E(β0,β1,…,β m ,γ1,…,γ n ) is the error sum of squares function, A k (t k ) is the kth data point in the actually observed optical signal attenuation data sequence, A′ k (t k ) is the corresponding predicted value, K is the length of the optical signal attenuation data sequence actually observed, that is, the number of attenuation data points measured at different times;

[0100] By solving the system of partial derivative equations that minimizes E:

[0101]

[0102] The optimal regression coefficient is obtained to determine the optical signal attenuation prediction model.

[0103] Step S6 of fault location and decision support is as follows: using the established prediction model and the optical signal data received in real time, analyzing the transmission characteristics of the optical signal in the optical cable, and establishing the following relationship based on the light propagation theory:

[0104] I(x)=I0e -αx +R(x)

[0105] Where I(x) is the function of the change of optical signal intensity with transmission distance, where x is the transmission distance of the optical signal in the optical cable, I0 is the initial intensity of the optical signal, that is, the intensity value of the optical signal when it enters the optical cable and starts to transmit, α is the attenuation coefficient of the optical signal in the optical cable, and R(x) is the reflected signal intensity function of the optical signal during the transmission process;

[0106] By fitting the collected optical signal data, the specific form of the attenuation coefficient α and the reflected signal intensity function R(x) is determined, thereby analyzing the transmission characteristics of the optical signal in the optical cable;

[0107] When an optical cable fails, the transmission characteristics of the optical signal will change significantly. At this time, the optical signal strength will drop sharply and the reflected signal strength will suddenly increase. By monitoring the changes in the optical signal strength and the reflected signal strength, combined with the above optical signal transmission characteristics analysis, when the fault judgment condition x is met, f , the fault point location can be determined:

[0108] x f = argmax{x|I(x)≤I0-ΔI th and R(x)≥R0+ΔR th}

[0109] where x f is the distance from the starting point to the fault point, I0 is the initial intensity of the optical signal, ΔI th is the threshold value of the optical signal intensity drop, R0 is the reference value of the reflected signal intensity under normal circumstances, ΔR th Threshold value added for reflected signals;

[0110] Based on the fault point location and the analysis results of optical signal transmission characteristics, decision support is provided for fault detection and repair.

[0111] A fiber core online rapid testing system, characterized in that it comprises an online OTDR testing unit, a cloud platform and a resource visualization system, characterized in that the online OTDR testing unit implements the steps of the method described in any one of claims 1 to 9 when executed with the cloud platform.

[0112] The resource visualization system provides visualization of optical fiber status and optical cable resources, including the optical fiber direction, connection points, occupancy of each port, optical fiber connection status, fault detection information, and distinguishes the quality status of optical cables based on GIS maps and optical cable quality coloring functions.

[0113] The specific steps are as follows:

[0114] S1: optical signal data acquisition; specifically: using an online OTDR test unit to collect an optical signal data sequence {I n}, whose sequence is represented as:

[0115]

[0116] Where t represents time, Δt represents the acquisition time interval, T is the total acquisition time, I n is the nth data point in the optical signal data sequence collected within the time interval [0, T] according to the collection time interval Δt.

[0117] S2: Edge computing performs preliminary processing, including data filtering and feature extraction;

[0118] The data filtering is specifically: using a finite impulse response FIR filter, whose filter coefficient sequence is {h k}, the length is M, the filtered optical signal data sequence {I′ n Through convolution operation, we get:

[0119]

[0120] Where I′ n is the nth data point in the optical signal data sequence after filtering, and its filter coefficient sequence is {h k}, M is the filter coefficient sequence {h k} length, I n-k is the filtered data point I′ in the original collected optical signal data sequence that is currently to be calculated n The relevant nk-th data point.

[0121] The feature extraction is specifically: extracting features from the filtered optical signal data and calculating the average intensity of the optical signal and the intensity variance σ 2 , the formula is:

[0122]

[0123]

[0124] in is the average intensity of the optical signal, and the filtered optical signal data is obtained from the M-1th data point to the last data point, that is, The data points are summed and averaged, which can roughly reflect the overall intensity level of the light signal during this period of time; σ 2 The optical signal intensity variance is used to measure the dispersion of the optical signal intensity relative to the average intensity within a certain time range. n With average intensity The square sum of the differences is obtained by corresponding averaging, which can be used to further analyze the stability and other characteristics of the optical signal; N is the number of data points involved in calculating the average intensity and variance, and its value is T is the total acquisition duration, Δt is the acquisition time interval, and M is the length of the filter coefficient sequence.

[0125] S3: Data reception and integration; specifically: using the cloud platform to receive the characteristic value data sequence from the online OTDR test unit, including the average intensity and the intensity variance σ 2 , these data are arranged in chronological order to form time series data, which can be expressed in vector form:

[0126]

[0127] in The average intensity of the optical signal received by the cloud platform from the online OTDR test unit at time t i The specific value of σ, which forms a time series data about the average intensity over time; 2 (t i ) is the optical signal strength variance received by the cloud platform from the online OTDR test unit at time t i The specific value of forms a time series data about the intensity variance as time goes by; X is a two-dimensional data matrix formed by arranging the eigenvalue data such as the average intensity and intensity variance of the optical signal received by the cloud platform at different times in time sequence; L is the length of the received data time series, that is, the number of groups of eigenvalue data received by the cloud platform at different times, which determines the number of rows of the data matrix X.

[0128] S4: Principal component analysis and data dimensionality reduction; Principal component analysis is used for data dimensionality reduction:

[0129]

[0130] Among them, C is the covariance matrix of the data matrix X, which is an important basis for data processing operations such as principal component analysis. is the vector mean of X, L is the length of the received data time series, that is, the number of rows of the data matrix X, and T is the total duration of the acquisition;

[0131] The principal component analysis is performed on the two dimensions of the received average intensity and intensity variance to calculate the covariance matrix C of the data matrix X; is the mean vector of X, solving for the eigenvalue λ of the covariance matrix C i and the corresponding eigenvector v i (i=1,2), by selecting the eigenvectors corresponding to the main eigenvalues ​​to form the projection matrix P, the original data is reduced in dimension to obtain the reduced-dimensional data Y:

[0132] Y=P T X

[0133] Where Y is the data after dimensionality reduction by principal component analysis, which is obtained by multiplying the original data matrix X with the transpose of the projection matrix P consisting of the eigenvectors corresponding to the main features. P is the projection matrix, which consists of the eigenvectors corresponding to the selected main eigenvalues. These eigenvectors are obtained by solving the eigenvalues ​​and eigenvectors of the covariance matrix C.

[0134] S5: Establishing an optical signal attenuation prediction model; specifically: establishing a prediction model by using a nonlinear regression model based on optical signal attenuation:

[0135] A(t)=β0+β1t+β2t 2 +…+β m t m +γ1z1(t)+γ2z2(t)+…+γ n z n (t

[0136] Where A(t) is the function of the optical signal attenuation over time, β i (i=0,1,…,m) are regression coefficients in the optical signal attenuation prediction model. The values ​​of these coefficients are determined by fitting the data using the least squares method, thereby determining the specific optical signal attenuation prediction model. j (j = 1, 2, ..., n is the regression coefficient associated with the vector z(t) composed of other factors affecting the optical signal attenuation in the optical signal attenuation prediction model, z j (t) is the function of the jth factor in the vector z(t) composed of other factors that affect the attenuation of the optical signal changing with time;

[0137] The actual observed optical signal attenuation data series {A k (t k )} and the corresponding predicted value is {A′ k (t k )}, set the error square sum function to:

[0138]

[0139] where E(β0,β1,…,β m ,γ1,…,γ n ) is the error sum of squares function, A k (t k ) is the kth data point in the actually observed optical signal attenuation data sequence, A′ k (t k ) is the corresponding predicted value, K is the length of the optical signal attenuation data sequence actually observed, that is, the number of attenuation data points measured at different times;

[0140] By solving the system of partial derivative equations that minimizes E:

[0141]

[0142] The optimal regression coefficient is obtained to determine the optical signal attenuation prediction model.

[0143] S6: Fault location and decision support. Specifically: Based on the established prediction model and the real-time received optical signal data, the transmission characteristics of the optical signal in the optical cable are analyzed, and the following relationship is established based on the light propagation theory:

[0144] I(x)=I0e -αx +R(x)

[0145] Where I(x) is the function of the change of optical signal intensity with transmission distance, where x is the transmission distance of the optical signal in the optical cable, I0 is the initial intensity of the optical signal, that is, the intensity value of the optical signal when it enters the optical cable and starts to transmit, α is the attenuation coefficient of the optical signal in the optical cable, and R(x) is the reflected signal intensity function of the optical signal during the transmission process;

[0146] By fitting the collected optical signal data, the specific form of the attenuation coefficient α and the reflected signal intensity function R(x) is determined, thereby analyzing the transmission characteristics of the optical signal in the optical cable;

[0147] When an optical cable fails, the transmission characteristics of the optical signal will change significantly. At this time, the optical signal strength will drop sharply and the reflected signal strength will suddenly increase. By monitoring the changes in the optical signal strength and the reflected signal strength, combined with the above optical signal transmission characteristics analysis, when the fault judgment condition x is met, f, the fault point location can be determined:

[0148] x f = argmax{x|I(x)≤I0-ΔI th and R(x)≥R0+ΔR th}

[0149] where x f is the distance from the starting point to the fault point, I0 is the initial intensity of the optical signal, ΔI th is the threshold value of the optical signal intensity drop, R0 is the reference value of the reflected signal intensity under normal circumstances, ΔR th Threshold value added for reflected signals;

[0150] Based on the fault point location and the analysis results of optical signal transmission characteristics, decision support is provided for fault detection and repair.

[0151] A fiber core online rapid testing system, characterized in that it comprises an online OTDR testing unit, a cloud platform and a resource visualization system, characterized in that the online OTDR testing unit implements the steps of the method described in any one of claims 1 to 9 when executed with the cloud platform.

[0152] The resource visualization system provides visualization of optical fiber status and optical cable resources, including the optical fiber direction, connection points, occupancy of each port, optical fiber connection status, fault detection information, and distinguishes the quality status of optical cables based on GIS maps and optical cable quality coloring functions.

Claims

1. A fiber core online rapid testing method, characterized in that: The following steps are involved: S1: optical signal data acquisition; S2: Edge computing performs preliminary processing; S3: data reception and integration; S4: principal component analysis and data dimensionality reduction; S5: Establish an optical signal attenuation prediction model; S6: Fault point location and decision support.

2. A fiber core online rapid testing method according to claim 1, characterized in that: Step S1 of optical signal data collection specifically includes: using an online OTDR test unit to collect an optical signal data sequence {I n }, whose sequence is represented as: Where t represents time, Δt represents the acquisition time interval, T is the total acquisition time, I n is the nth data point in the optical signal data sequence collected within the time interval [0, T] according to the collection time interval Δt.

3. A fiber core online rapid testing method according to claim 1, characterized in that: Step S2 edge computing performs preliminary processing, including data filtering and feature extraction; The data filtering is specifically: using a finite impulse response FIR filter, whose filter coefficient sequence is {h k }, the length is M, the filtered optical signal data sequence {I′ n Through convolution operation, we get: Where I′ n is the nth data point in the optical signal data sequence after filtering, and its filter coefficient sequence is {h k }, M is the filter coefficient sequence {h k } length, I n-k is the filtered data point I′ in the original collected optical signal data sequence that is currently to be calculated n The relevant nk-th data point.

4. A fiber core online rapid testing method according to claim 3, characterized in that: Step S2 edge computing performs preliminary processing, including data filtering and feature extraction; The feature extraction is specifically: extracting features from the filtered optical signal data and calculating the average intensity of the optical signal and intensity variance σ 2 , the formula is: in is the average intensity of the optical signal, σ 2 is the optical signal intensity variance, N is the number of data points involved in calculating the average intensity and variance, and its value is T is the total acquisition duration, Δt is the acquisition time interval, and M is the length of the filter coefficient sequence.

5. A fiber core online rapid testing method according to claim 1, characterized in that: Step S3 data reception and integration specifically includes: using the cloud platform to receive the characteristic value data sequence from the online OTDR test unit, including the average intensity and the intensity variance σ 2 , these data are arranged in chronological order to form time series data, which can be expressed in vector form: in The average intensity of the optical signal received by the cloud platform from the online OTDR test unit at time t i The specific value of σ 2 (t i ) is the optical signal strength variance received by the cloud platform from the online OTDR test unit at time t i , X is the two-dimensional data matrix formed by arranging the characteristic value data such as the average intensity and intensity variance of the optical signal received by the cloud platform at different times in chronological order, and L is the length of the received data time series.

6. A fiber core online rapid testing method according to claim 1, characterized in that: Step S4: principal component analysis and data dimensionality reduction are as follows: principal component analysis is used for data dimensionality reduction: Among them, C is the covariance matrix of the data matrix X, which is an important basis for data processing operations such as principal component analysis. is the vector mean of X, L is the length of the received data time series, that is, the number of rows of the data matrix X, and T is the total duration of the acquisition; The principal component analysis is performed on the two dimensions of the received average intensity and intensity variance to calculate the covariance matrix C of the data matrix X; is the mean vector of X, solving for the eigenvalue λ of the covariance matrix C i and the corresponding eigenvector v i (i=1, 2), by selecting the eigenvectors corresponding to the main eigenvalues ​​to form the projection matrix P, the original data is reduced in dimension to obtain the reduced-dimensional data Y: Y=P T X Where Y is the data after dimensionality reduction by principal component analysis, which is obtained by multiplying the original data matrix X with the transpose of the projection matrix P consisting of the eigenvectors corresponding to the main features. P is the projection matrix, which consists of the eigenvectors corresponding to the selected main eigenvalues. These eigenvectors are obtained by solving the eigenvalues ​​and eigenvectors of the covariance matrix C.

7. A fiber core online rapid testing method according to claim 1, characterized in that: Step S5 establishes the optical signal attenuation prediction model specifically by using a nonlinear regression model to establish a prediction model based on the optical signal attenuation: A(t)=β0+β1t+β2t 2 +…+b m t m +γ1z1(t)+γ2z2(t)+…+γ n z n (t) Where A(t) is the function of the optical signal attenuation over time, β i (i=0, 1, ..., m) are regression coefficients in the optical signal attenuation prediction model. The values ​​of these coefficients are determined by fitting the data using the least squares method, thereby determining the specific optical signal attenuation prediction model. j (j=1, 2, ..., n) is the regression coefficient associated with the vector z(t) composed of other factors affecting the optical signal attenuation in the optical signal attenuation prediction model. j (t) is the function of the jth factor in the vector z(t) composed of other factors that affect the attenuation of the optical signal changing with time; The actual observed optical signal attenuation data series {A k (t k )} and the corresponding predicted value is {A′ k (t k )}, set the error sum of squares function to: where E(β0, β1, …, β m ,γ1,…,γ n ) is the error sum of squares function, A k (t k ) is the kth data point in the actually observed optical signal attenuation data sequence, A′ k (t k ) is the corresponding predicted value, K is the length of the optical signal attenuation data sequence actually observed, that is, the number of attenuation data points measured at different times; By solving the system of partial derivative equations that minimizes E: The optimal regression coefficient is obtained to determine the optical signal attenuation prediction model.

8. A fiber core online rapid testing method according to claim 1, characterized in that: Step S6 of fault location and decision support is as follows: using the established prediction model and the optical signal data received in real time, analyzing the transmission characteristics of the optical signal in the optical cable, and establishing the following relationship based on the light propagation theory: I(x)=I0e -αx +R(x) Where I(x) is the function of the change of optical signal intensity with transmission distance, where x is the transmission distance of the optical signal in the optical cable, I0 is the initial intensity of the optical signal, that is, the intensity value of the optical signal when it enters the optical cable and starts to transmit, α is the attenuation coefficient of the optical signal in the optical cable, and R(x) is the reflected signal intensity function of the optical signal during the transmission process; By fitting the collected optical signal data, the specific form of the attenuation coefficient α and the reflected signal intensity function R(x) is determined, thereby analyzing the transmission characteristics of the optical signal in the optical cable; When an optical cable fails, the transmission characteristics of the optical signal will change significantly. At this time, the optical signal strength will drop sharply and the reflected signal strength will suddenly increase. By monitoring the changes in the optical signal strength and the reflected signal strength, combined with the above optical signal transmission characteristics analysis, when the fault judgment condition x is met, f , the fault point location can be determined: x f =argmax{x|I(x)≤I0-ΔI th and R(x)≥R0+ΔR th } where x f is the distance from the starting point to the fault point, I0 is the initial intensity of the optical signal, ΔI th is the threshold value of the optical signal intensity drop, R0 is the reference value of the reflected signal intensity under normal circumstances, ΔR th Threshold value added for reflected signals; Based on the fault point location and the analysis results of optical signal transmission characteristics, decision support is provided for fault detection and repair.

9. A fiber core online rapid testing system, characterized in that: The method comprises an online OTDR test unit, a cloud platform and a resource visualization system, and is characterized in that the online OTDR test unit implements the steps of the method described in any one of claims 1 to 9 when executed with the cloud platform.

10. A fiber core online rapid testing system according to claim 9, characterized in that: The resource visualization system provides the client with visualization of optical fiber status and optical cable resources, including the direction of optical fiber, connection points, occupancy of each port, connection status of optical fiber, and fault information. It also distinguishes the quality status of optical cables based on GIS maps and optical cable quality coloring functions.

Images