1. A digital management method for optical cable resources is characterized by comprising the following steps:

step S1, obtaining a paper data picture of the fiber core wiring information from the operation and maintenance data, and identifying the fiber core wiring information in the paper data picture by using an OCR technology;

step S2, converting the identified fiber core wiring information into structured data, associating the obtained structured data with fiber core service data and transmission segment data in a resource system, updating the resource system according to the associated fiber core service data and transmission segment data, completing the association of operation and maintenance data with the resource system, and forming fiber core digital service information;

step S3, obtaining optical cable fiber core attenuation curve data from the fiber core test data, and analyzing the fiber core attenuation curve data of a plurality of optical fibers in the optical cable fiber core attenuation curve data to obtain fiber core basic data and splice closure data;

step S4, collecting data to be correlated of each system, wherein the data to be correlated comprises power transmission line data, splice closure spatial data and tower spatial data acquired from a GIS system, optical fiber resource data acquired from a resource system, and fiber core basic data and splice closure data acquired from fiber core test data;

step S5, according to the data to be correlated of each system, correlating the GIS system, the fiber core test data and the resource system by using a machine learning method to form the correlation relation of the complete physical route of the optical cable, and according to the correlation relation of the complete physical route of the optical cable, extracting the characteristic value by using a theme model to form fiber core digital spatial information;

and step S6, merging the fiber core digital service information and the fiber core digital spatial information to obtain complete digital information of the fiber core, completing data association among operation and maintenance data, a GIS system, fiber core test data and a resource system, and digitally managing the optical cable resources according to the complete digital fiber core information.

2. The digital management method for optical cable resources as claimed in claim 1, wherein the specific process of identifying the core distribution information in the paper quality data picture by using the OCR technology in step S1 is as follows:

step S11, cutting the obtained paper data pictures, and dividing the forms in the paper data pictures according to the optical cable source to enable the wiring information in each paper data picture to be the fiber core information of the same optical cable;

s12, preprocessing the cut paper data picture to obtain a picture to be detected;

s13, performing target detection, target positioning and feature point detection on the picture to be detected by using a YOLO neural network algorithm, and segmenting all table areas in the picture to be detected to obtain a table picture;

s14, performing semantic segmentation on the content in the table picture by using a U-net algorithm, identifying and reading cells in the table picture, sharpening the read table picture to obtain coordinate data of the cells, and finally obtaining the coordinate data of all the cells in the table picture;

step S15, extracting region-of-interest data in the table picture according to the ROI technology, wherein the region-of-interest data comprises region content, coordinate information of the region, cell content in the region and coordinate data of corresponding cells;

step S16, marking the table picture according to the ROI data to obtain a plurality of marked ROI areas, and acquiring cells in each marked ROI area;

and step S17, inputting the cells in the marked ROI area into a CNN image recognition model for character recognition to obtain the fiber core wiring information recorded in the table in the paper data picture.

3. The digital management method for optical cable resources as claimed in claim 2, wherein the specific process of step S12 is as follows:

step S121, performing uniform formatting, graying and binarization processing on the cut paper data picture in sequence to obtain a picture to be processed;

s122, obtaining the outline edge of the table in the picture to be processed, and obtaining an enclosing matrix and an inclination angle of the table outline according to the outline edge of the table;

step S123, inputting the inclination angle of the table outline into the picture azimuth checking model, and judging whether the inclination angle of the table outline is normal or not;

step S124, if the determination result of the step S123 is abnormal, performing rotation correction on the to-be-processed picture according to a normal angle to obtain a table profile of a normal inclination angle, and executing the step S125, and if the determination result of the step S123 is normal, executing the step S125;

step S125, extracting an image area in the table outline with the normal inclination angle to obtain an independent identification image;

and S126, carrying out image enhancement processing on the independent identification image, wherein the image enhancement processing comprises image sharpening, smoothing and denoising processing, and obtaining the image to be detected.

4. The digital management method for optical cable resources as claimed in claim 2, wherein the specific process of step S16 is as follows:

step S161, acquiring a corresponding form template according to the region-of-interest data in the form picture, and acquiring a corresponding point-of-interest template according to the acquired form template;

step S162, coordinate data of four corners of a cell in a table picture are obtained, and whether the scaling of the table picture and a corresponding interest point template are matched or not is judged according to the coordinate data of the four corners of the cell;

step S163, if the determination result of step S162 is not matching, scaling the table picture so that the scaling of the interest point template matches the scaling of the table picture, and then executing step S164; if the determination result of step S162 is matching, step S164 is executed;

and S164, when the table picture is matched with the scaling of the interest point template, calculating an ROI (region of interest) matched with the cells in the table picture according to the interest point template and marking the ROI to obtain the cells in the marked ROI.

5. The method for digitally managing optical cable resources according to claim 4, wherein the specific process of associating the configuration data with the core service data in the resource system in step S2 includes:

step S21, performing text error correction on the fiber core wiring information in the identified paper data picture according to a communication service dictionary and an n-gram algorithm to obtain fiber core wiring information to be converted;

step S22, constructing a service fiber core pair by using fiber core distribution information to be converted according to the obtained interest point template, storing the service fiber core pair according to the format of the optical cable section name, the fiber core pair, the fiber core service data and the opposite end fiber core pair and the format of a table, and finally generating the structured data of the fiber core distribution information;

step S23, acquiring the name of the optical cable section to which the fiber core wiring information belongs according to the generated structured data, acquiring fiber core service data corresponding to the name of the optical cable section from a resource system according to the name of the optical cable section, calculating the structured data and the fiber core service data by using a WMD algorithm to obtain a first feature corpus, and calculating the structured data and the fiber core service data by using a Smooth Inverse Frequency algorithm to obtain a second feature corpus;

step S24, inputting the first characteristic linguistic data and the second characteristic linguistic data into an SVM (support vector machine) to obtain the distribution probability of the incidence relation between the structured data and the fiber core service data in the resource system;

and S25, sequencing the distribution probability of the incidence relation, and taking the incidence relation of the maximum distribution probability to obtain the incidence of the structured data and the fiber core service data in the resource system, and finally finishing the incidence of the resource system and the fiber core service data in the operation and maintenance data.

6. The method for digital management of optical cable resources as claimed in claim 5, wherein the specific process of associating the structured data with the transmission segment data in the resource system in step S2 is as follows:

obtaining fiber core service data according to the structured data of the fiber core distribution information to form a fiber core service data set A, and obtaining the name of the optical cable section in the resource system and the relation between the corresponding optical cable section and the transmission section to form a data set B;

and calculating the support degrees of the data sets A and B according to the optical cable section names in the data sets A and B to obtain the incidence relation between the fiber core wiring information and the transmission section.

7. The digital management method for optical cable resources as claimed in claim 1, wherein the specific process of step S3 is as follows:

step S31, respectively preprocessing the attenuation curve data of the optical fibers to obtain a plurality of preprocessed data;

step S32, obtaining a mixed curve S (l) by superimposing a plurality of preprocessed data:

S(l)＝d₁(l)+d₂(l)+…+d_n(l)

Y＝[y₁(l)，y₂(l)，…，y_i(l)]

D＝[d₁(l)，d₂(l)，…，d_i(l)]

wherein l represents the length of the cable, y_i(l) Representing the attenuation value of the ith fiber at length l, d_i(l) Representing attenuation value y of corresponding i-th optical fiber_i(l) Vector after preconditioning;

step S33, extracting features of the mixed curve S (l), where the extracted features include a signal peak, a height of the peak, a signal average, a signal peak width, a variance, and a root mean square, and the extracted feature set is denoted by T ═ T [ T ]₁,t₂,...,t_n]；

Step S34, carrying out data annotation on each point in the mixed curve S (l) to obtain an annotation result, wherein the annotation result is 1, which indicates that the junction box is present; the labeling result is to indicate whether it is a closure;

step S35, inputting the mixed curve S (l), the labeling result and the feature set T into an offline training classification model together for offline training, and obtaining a trained classification model after training, wherein the offline training classification model is a classifier II;

s36, carrying out on-line prediction by using the trained classification model, outputting all points with the labeling result of 1 according to the classification model, and outputting the distance l corresponding to the points_j，(j＝1,2,...,m)，l_jThe j-th splice closure position of the optical cable, m being the number of all pointsCounting;

step S37 according to l_jAnd calculating attenuation value a of point j from mixing curve S (l)_jAccording to l_jAnd a_jClosure data L and D are obtained, wherein,

a_j＝f(l_j,S(l))

L＝[l₁,l₂,...,l_m]

D＝[a₁,a₂,...,a_m]

and f is an algorithm function, the algorithm function comprises attenuation data difference and filtering processes, L represents a position point, and D represents attenuation values of all splicing boxes of the optical cable.

8. The optical cable resource digital management method according to claim 1, wherein the data to be associated in step S4 further includes the following data:

spatial data of the tower: the method comprises the steps of determining GIS coordinates, names, maintenance units, transmission line names, construction time and whether connection exists or not, wherein the GIS coordinates, the names, the maintenance units, the transmission line names and the construction time correspond to towers;

splice closure spatial data: the method comprises the following steps of (1) including the names of a tower to which a splicing box belongs and the tower;

fiber core test data: including fibre core basic data and closure data, the fibre core basic data includes the start and end station name of fibre core, fibre core end-to-end total length, wavelength, affiliated optical cable model, decay and incident point, and the incident point is by the data pair: (core distance, attenuation); splice closure data includes splice closure location and its attenuation;

fiber resource data: the method comprises the optical cable, the optical cable section name, the optical cable model, the start-stop station name, the associated transmission line, service information and the use.

9. The optical cable resource digital management method according to claim 8, wherein the specific process of step S5 is as follows:

step S51, formatting the data to be associated, processing the data to a format conforming to a feature extraction algorithm, and obtaining feature extraction data;

step S52, extracting the following feature values from the feature extraction data according to different data sources:

pole tower characteristic value: the method comprises the steps of including the name of a power transmission line and GIS coordinates corresponding to a tower;

fiber core test characteristic value: the method comprises the name of a start-stop station of a fiber core and the type of the optical cable to which the start-stop station belongs;

resource characteristic value: the method comprises the following steps of (1) including the optical cable model, the starting and stopping station name and the associated transmission line name;

step S53, constructing an association algorithm model, wherein the algorithm model comprises a WMD algorithm, a Smooth Inverse Frequency algorithm and an SVM support vector machine, inputting the extracted characteristic values into the constructed association algorithm model, and finally obtaining the association relationship between the tower and the power transmission line, the association relationship between the power transmission line and the optical cable section, the association relationship between the optical cable section and the fiber core test data and the association relationship between the splice box and the optical cable;

s54, forming a relation data vector containing the event, the tower, the optical cable and the station according to the association relation obtained in the step S53, and formatting the obtained relation data vector;

step S55, extracting the relation features of the relation data vectors after the formatting processing to obtain the associated data features;

s56, inputting the correlation data characteristics into the correlation algorithm model constructed in the S53 to obtain the correlation relationship among the tower, the optical cable and the junction box, obtaining the fiber core distance and attenuation of the event point in the fiber core test data, and GIS coordinates corresponding to the optical cable, the tower and the tower, and calculating and judging the GIS coordinates of the tower and the junction box by taking the GIS coordinates, the distance and the attenuation as calculation and judgment references according to the correlation information of the tower, the optical cable and the junction box after correlation;

and S57, associating the tower, the optical cable, the power transmission line and the splice closure according to the GIS coordinates of the tower and the splice closure and all data of the data to be associated, thereby splicing the association relation of the complete physical route of the optical cable.

Background

The power communication optical fiber information is managed in different systems, but various information exists in an isolated island in a chimney mode, and an information complete optical fiber portrait system such as associated resources, assets, power transmission lines, optical fiber tests and the like does not exist, so that manual inquiry is required in each subsystem, and then the information is connected in series in a manual mode. The concrete expression is as follows: the maintenance of the optical fiber distribution service information needs to be manually entered into the system one by one. The island phenomenon of information exists in each system related to power communication, the related information of the current communication resources is stored in the system in a specific field according to the characteristics of a service system, few intersections exist among the systems, and particularly for optical fiber communication, no complete service information system can be completely stringed. The following problems exist in the management of optical fiber communication resources:

(1) the ODF information is not maintained completely, the actual wiring information is stored and managed by a maintenance unit in a paper drawing mode, unified input and management in a resource system are not performed, or data input and managed in the resource system and the paper drawing are independently managed without linkage, and the actual wiring information and the resource system have deviation;

(2) the fiber core test data are discrete and are stored in the test instrument, and uniform digital management is not performed;

(3) the asset and resource data of the optical fiber are managed in different systems without sharing and association;

(4) the physical fiber is not digitally managed.

(5) Complete physical space routing of optical cables, transmission lines and splice closures is not formed.

Therefore, data of each link needs to be digitized, so that a complete digitized optical fiber is formed, a data basis is provided for applications such as fault analysis, simulation analysis and digital twinning, and optical fiber resource information is effectively managed.

Disclosure of Invention

The invention aims to solve the technical problem of how to carry out association and digital management on optical cable resource information in each system, and aims to provide an optical cable resource digital management method which realizes association of power communication network resource data, distribution data, optical fiber test data and transmission line data by digitalizing optical cable physical resources, fiber core service information and fiber core space information so as to carry out digital management on optical cable resources.

The invention is realized by the following technical scheme:

a digital management method for optical cable resources comprises the following steps:

step S1, obtaining a paper data picture of the fiber core wiring information from the operation and maintenance data, and identifying the fiber core wiring information in the paper data picture by using an OCR technology;

step S2, converting the identified fiber core wiring information into structured data, associating the obtained structured data with fiber core service data and transmission segment data in a resource system, updating the resource system according to the associated fiber core service data and transmission segment data, completing the association of operation and maintenance data with the resource system, and forming fiber core digital service information; wherein, the fiber core service data: the method comprises the name of the optical cable section to which the fiber core belongs, the service use of the fiber core, the number of the fiber cores and the name of a start-stop station of the fiber core;

step S3, obtaining optical cable fiber core attenuation curve data from the fiber core test data, and analyzing the fiber core attenuation curve data of a plurality of optical fibers in the optical cable fiber core attenuation curve data to obtain fiber core basic data and splice closure data;

step S4, collecting data to be correlated of each system, wherein the data to be correlated comprises power transmission line data, splice closure spatial data and tower spatial data acquired from a GIS system, optical fiber resource data acquired from a resource system, and fiber core basic data and splice closure data acquired from fiber core test data;

step S5, according to the data to be correlated of each system, correlating the GIS system, the fiber core test data and the resource system by using a machine learning method to form the correlation relation of the complete physical route of the optical cable, and according to the correlation relation of the complete physical route of the optical cable, extracting the characteristic value by using a theme model to form fiber core digital spatial information;

and step S6, merging the fiber core digital service information and the fiber core digital spatial information to obtain complete digital information of the fiber core, completing data association among operation and maintenance data, a GIS system, fiber core test data and a resource system, and digitally managing the optical cable resources according to the complete digital fiber core information.

The information about optical fibers in the existing power communication resources is quite discrete and exists in different systems, if the optical fiber information needs to be inquired, manual inquiry is often carried out in each subsystem, then the information is connected in series in a manual mode, time and labor are consumed, each system related to the existing power communication has an information island phenomenon, the information related to the optical fiber communication resources is stored in the system in a specific field according to the characteristics of service systems, few intersections exist among the systems, and no complete management system capable of completely stringing services exists, so that the fiber core service information is digitalized, and the digitalized fiber core service information (wiring data) is bound with a resource system; the optical cable physical resource digitization is realized by analyzing optical cable fiber core attenuation curve data, relevant data of each system are correlated by a machine learning method, so that the correlation relation of the complete physical route of the optical cable is formed to realize fiber core spatial information digitization, and finally fiber core digitized service information and fiber core digitized spatial information are combined to realize the digital management of the optical cable resources. The method is characterized in that the power communication network optical fiber resource data, the optical fiber distribution data, the optical fiber test data and the transmission line (optical cable path routing) data are respectively bound with the optical cable optical fiber data in the asset data to realize association through the process, the asset system and the resource system are directly bound because the resource data in the asset system are static fiber core data comprising fixed information such as fiber core production model, producer and the like, and when the data in the resource system are associated with other data, the association of the asset system and other systems is completed, so that complete digital information of the fiber core including the fiber core asset data is realized, the machine replaces manual work, the working efficiency is greatly improved, and the problems of mechanicalness, repeatability and complexity existing in the past depending on manual association are solved.

Further, the concrete process of identifying the core wiring information in the paper quality data picture by using the OCR technology in step S1 is as follows:

step S11, cutting the obtained paper data pictures, and dividing the forms in the paper data pictures according to the optical cable source to enable the wiring information in each paper data picture to be the fiber core information of the same optical cable;

s12, preprocessing the cut paper data picture to obtain a picture to be detected;

s13, performing target detection, target positioning and feature point detection on the picture to be detected by using a YOLO neural network algorithm, and segmenting all table areas in the picture to be detected to obtain a table picture;

s14, performing semantic segmentation on the content in the table picture by using a U-net algorithm, identifying and reading cells in the table picture, sharpening the read table picture to obtain coordinate data of the cells, and finally obtaining the coordinate data of all the cells in the table picture;

step S15, extracting region-of-interest data in the table picture according to the ROI technology, wherein the region-of-interest data comprises region content, coordinate information of the region, cell content in the region and coordinate data of corresponding cells;

step S16, marking the table picture according to the ROI data to obtain a plurality of marked ROI areas, and acquiring cells in each marked ROI area;

and step S17, inputting the cells in the marked ROI area into a CNN image recognition model for character recognition to obtain the fiber core wiring information recorded in the table in the paper data picture.

Further, the specific process of step S12 is:

step S121, performing uniform formatting, graying and binarization processing on the cut paper data picture in sequence to obtain a picture to be processed;

s122, obtaining the outline edge of the table in the picture to be processed, and obtaining an enclosing matrix and an inclination angle of the table outline according to the outline edge of the table;

step S123, inputting the inclination angle of the table outline into the picture azimuth checking model, and judging whether the inclination angle of the table outline is normal or not;

step S124, if the determination result of the step S123 is abnormal, performing rotation correction on the to-be-processed picture according to a normal angle to obtain a table profile of a normal inclination angle, and executing the step S125, and if the determination result of the step S123 is normal, executing the step S125;

step S125, extracting an image area in the table outline with the normal inclination angle to obtain an independent identification image;

and S126, carrying out image enhancement processing on the independent identification image, wherein the image enhancement processing comprises image sharpening, smoothing and denoising processing, and obtaining the image to be detected.

Further, the specific process of step S16 is:

step S161, acquiring a corresponding form template according to the region-of-interest data in the form picture, and acquiring a corresponding point-of-interest template according to the acquired form template;

step S162, coordinate data of four corners of a cell in a table picture are obtained, and whether the scaling of the table picture and a corresponding interest point template are matched or not is judged according to the coordinate data of the four corners of the cell;

step S163, if the determination result of step S162 is not matching, scaling the table picture so that the scaling of the interest point template matches the scaling of the table picture, and then executing step S164; if the determination result of step S162 is matching, step S164 is executed;

and S164, when the table picture is matched with the scaling of the interest point template, calculating an ROI (region of interest) matched with the cells in the table picture according to the interest point template and marking the ROI to obtain the cells in the marked ROI.

Further, the specific process of associating the configuration data with the core service data in the resource system in step S2 includes:

step S21, performing text error correction on the fiber core wiring information in the identified paper data picture according to a communication service dictionary and an n-gram algorithm to obtain fiber core wiring information to be converted;

step S22, constructing a service fiber core pair by using fiber core distribution information to be converted according to the obtained interest point template, storing the service fiber core pair according to the format of the optical cable segment name, the fiber core pair, the service distribution information and the opposite end fiber core pair and the format of a table, and finally generating the structured data of the fiber core distribution information;

step S23, acquiring the name of the optical cable section to which the fiber core wiring information belongs according to the generated structured data, acquiring fiber core service data corresponding to the name of the optical cable section from a resource system according to the name of the optical cable section, calculating the structured data and the fiber core service data by using a WMD algorithm to obtain a first feature corpus, and calculating the structured data and the fiber core service data by using a Smooth Inverse Frequency algorithm to obtain a second feature corpus;

step S24, inputting the first characteristic linguistic data and the second characteristic linguistic data into an SVM (support vector machine) to obtain the distribution probability of the incidence relation between the structured data and the fiber core service data in the resource system;

and S25, sequencing the distribution probability of the incidence relation, and taking the incidence relation of the maximum distribution probability to obtain the incidence of the structured data and the fiber core service data in the resource system, and finally finishing the incidence of the resource system and the fiber core service data in the operation and maintenance data.

Further, the specific process of associating the structured data with the transmission segment data in the resource system in step S2 is as follows:

obtaining fiber core service data according to the structured data of the fiber core distribution information to form a fiber core service data set A, and obtaining the name of the optical cable section in the resource system and the relation between the corresponding optical cable section and the transmission section to form a data set B;

and calculating the support degrees of the data sets A and B according to the optical cable section names in the data sets A and B to obtain the incidence relation between the fiber core wiring information and the transmission section.

Further, the specific process of step S3 is:

step S31, respectively preprocessing the attenuation curve data of the optical fibers to obtain a plurality of preprocessed data;

step S32, obtaining a mixed curve S (l) by superimposing a plurality of preprocessed data:

S(l)＝d₁(l)+d₂(l)+…+d_n(l)

Y＝[y₁(l)，y₂(l)，…，y_i(l)]

D＝[d₁(l)，d₂(l)，…，d_i(l)]

wherein l represents the length of the cable, y_i(l) Representing the attenuation value of the ith fiber at length l, d_i(l) Representing attenuation value y of corresponding i-th optical fiber_i(l) Vector after preconditioning;

step S33, extracting features of the mixed curve S (l), where the extracted features include a signal peak, a height of the peak, a signal average, a signal peak width, a variance, and a root mean square, and the extracted feature set is denoted by T ═ T [ T ]₁,t₂,...,t_n]；

Step S34, carrying out data annotation on each point in the mixed curve S (l) to obtain an annotation result, wherein the annotation result is 1, which indicates that the junction box is present; the labeling result is to indicate whether it is a closure;

step S35, inputting the mixed curve S (l), the labeling result and the feature set T into an offline training classification model together for offline training, and obtaining a trained classification model after training, wherein the offline training classification model is a classifier II;

s36, carrying out on-line prediction by using the trained classification model, outputting all points with the labeling result of 1 according to the classification model, and outputting the distance l corresponding to the points_j，(j＝1,2,...,m)，l_jThe position of the j-th splice closure of the optical cable, wherein m is the number of all points;

step S37 according to l_jAnd calculating attenuation value a of point j from mixing curve S (l)_jAccording to l_jAnd a_jClosure data L and D are obtained, wherein,

a_j＝f(l_j,S(l))

L＝[l₁,l₂,...,l_m]

D＝[a₁,a₂,...,a_m]

and f is an algorithm function, the algorithm function comprises attenuation data difference and filtering processes, L represents a position point, and D represents attenuation values of all the splicing boxes of the optical cable.

Further, the data to be associated in step S4 specifically includes the following data:

spatial data of the tower: the method comprises the steps of determining GIS coordinates, names, maintenance units, transmission line names, construction time and whether connection exists or not, wherein the GIS coordinates, the names, the maintenance units, the transmission line names and the construction time correspond to towers;

splice closure spatial data: the method comprises the following steps of (1) including the names of a tower to which a splicing box belongs and the tower;

fiber core test data: including fibre core basic data and closure data, the fibre core basic data includes the start and end station name of fibre core, fibre core end-to-end total length, wavelength, affiliated optical cable model, decay and incident point, and the incident point is by the data pair: (core distance, attenuation); splice closure data includes splice closure location and its attenuation;

fiber resource data: the method comprises the optical cable, the name of the optical cable section, the type of the optical cable, the name of a start-stop station, an associated power transmission line, service information, application and the like.

Further, the specific process of step S5 is:

step S51, formatting the data to be associated, processing the data to a format conforming to a feature extraction algorithm, and obtaining feature extraction data;

step S52, extracting the following feature values from the feature extraction data according to different data sources:

pole tower characteristic value: the method comprises the steps of including the name of a power transmission line and GIS coordinates corresponding to a tower;

fiber core test characteristic value: the method comprises the name of a start-stop station of a fiber core and the type of the optical cable to which the start-stop station belongs;

resource characteristic value: the method comprises the following steps of (1) including the optical cable model, the starting and stopping station name and the associated transmission line name;

step S53, constructing an association algorithm model, wherein the algorithm model comprises a WMD algorithm, a Smooth Inverse Frequency algorithm and an SVM support vector machine, inputting the extracted characteristic values into the constructed association algorithm model, and finally obtaining the association relationship between the tower and the power transmission line, the association relationship between the power transmission line and the optical cable section, the association relationship between the optical cable section and the fiber core test data and the association relationship between the splice box and the optical cable section;

s54, forming a relation data vector containing the event, the tower, the optical cable and the station according to the association relation obtained in the step S53, and formatting the obtained relation data vector;

step S55, extracting the relation features of the relation data vectors after the formatting processing to obtain the associated data features;

s56, inputting the correlation data characteristics into the correlation algorithm model constructed in the S53 to obtain the correlation relationship among the tower, the optical cable and the junction box, obtaining the fiber core distance and attenuation of the event point in the fiber core test data, and GIS coordinates corresponding to the optical cable, the tower and the tower, and calculating and judging the GIS coordinates of the tower and the junction box by taking the GIS coordinates, the distance and the attenuation as calculation and judgment references according to the correlation information of the tower, the optical cable and the junction box after correlation;

and S57, associating the tower, the optical cable, the power transmission line and the splice closure according to the GIS coordinates of the tower and the splice closure and all data of the data to be associated, thereby splicing the association relation of the complete physical route of the optical cable.

Compared with the prior art, the invention has the following advantages and beneficial effects:

1. according to the digital management method for the optical cable resources, disclosed by the invention, the optical cable physical resources are digitized, the fiber core service information is digitized and the fiber core space information is digitized, so that the association of the resource data, the asset data, the distribution data, the optical fiber test data and the power transmission line data of a power communication network is realized, the association of the optical fiber data in each communication system is completed, the work of replacing manpower by a machine is realized, the working efficiency is greatly improved, and the problems of mechanicalness, repeatability and complexity existing in the past depending on manual association are solved;

2. the digital management method for the optical cable resources comprehensively utilizes the advantages of various text similarity models, overcomes the problem of low correlation accuracy in the traditional technology based on the deterministic correlation rule, improves the model prediction accuracy, greatly improves the working efficiency, provides more multidimensional and larger information quantity correlation data for the large data analysis of mass data, and provides an intelligent technical support for the operation management of the power industry basic communication network;

3. the invention relates to a digital management method for optical cable resources, which associates fiber core distribution information in operation and maintenance data with fiber core service data in a resource system, perfects optical fiber distribution service information and realizes the comprehensive maintenance of the optical fiber distribution service information.

Drawings

The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:

FIG. 1 is a schematic view of the overall flow chart of the present invention;

FIG. 2 is a flow chart of core wiring information identification in a paper document picture;

fig. 3 is different paper material pictures obtained in the embodiment, wherein example (1) and example (2) represent two different table types obtained; example (3) represents an ROI cell region divided from a table;

FIG. 4 is core attenuation curve data based on OTDR testing.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one of ordinary skill in the art that: it is not necessary to employ these specific details to practice the present invention. In other instances, well-known structures, circuits, materials, or methods have not been described in detail so as not to obscure the present invention.

Throughout the specification, reference to "one embodiment," "an embodiment," "one example," or "an example" means: the particular features, structures, or characteristics described in connection with the embodiment or example are included in at least one embodiment of the invention. Thus, the appearances of the phrases "one embodiment," "an embodiment," "one example" or "an example" in various places throughout this specification are not necessarily all referring to the same embodiment or example. Furthermore, the particular features, structures, or characteristics may be combined in any suitable combination and/or sub-combination in one or more embodiments or examples. Further, those of ordinary skill in the art will appreciate that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

In the description of the present invention, it is to be understood that the terms "front", "rear", "left", "right", "upper", "lower", "vertical", "horizontal", "high", "low", "inner", "outer", etc. indicate orientations or positional relationships based on those shown in the drawings, and are only for convenience of description and simplicity of description, and do not indicate or imply that the referenced devices or elements must have a particular orientation, be constructed and operated in a particular orientation, and therefore, are not to be construed as limiting the scope of the present invention.

Example 1

As shown in fig. 1, the method for digitally managing optical cable resources according to the present invention implements digital management of optical cable resources by digitizing optical cable physical resources, digitizing fiber core service information, and digitizing fiber core spatial information, and specifically includes the following steps:

(1) fiber core service information digitization:

step S1, obtaining a paper data picture of the fiber core wiring information from the operation and maintenance data, and identifying the fiber core wiring information in the paper data picture by using an OCR technology;

specifically, as shown in fig. 2, the specific process of identifying the core wiring information in the paper quality data picture by using the OCR technology in step S1 is as follows:

step S11, cutting the obtained paper data pictures, and dividing the forms in the paper data pictures according to the optical cable source to enable the wiring information in each paper data picture to be the fiber core information of the same optical cable;

s12, preprocessing the cut paper data picture to obtain a picture to be detected;

s13, performing target detection, target positioning and feature point detection on the picture to be detected by using a YOLO neural network algorithm, and segmenting all table areas in the picture to be detected to obtain a table picture;

process of detection using YOLO neural network algorithm: the input picture is scaled, the YOLO network divides the input picture into SxS grids, if the central point of an object in the picture falls into a certain grid, the corresponding grid is responsible for predicting the size and the category of the object; sending the picture to be detected into a convolutional neural network for prediction; the confidence degree threshold processing is performed according to the prediction result to obtain a final detection result, which is output as a table picture in this embodiment.

S14, performing semantic segmentation on the content in the table picture by using a U-net algorithm, identifying and reading cells in the table picture, sharpening the read table picture to obtain coordinate data of the cells, and finally obtaining the coordinate data of all the cells in the table picture;

specifically, the digital processing of convolution and pooling is carried out on the table picture, then all horizontal and vertical lines of the table in the table picture are detected, and finally the boundary coordinates of the table are obtained; and positioning each cell according to the detected horizontal and vertical lines of the table, and recording the coordinates of each cell.

Step S15, extracting region-of-interest data in the table picture according to the ROI technology, wherein the region-of-interest data comprises region content, coordinate information of the region, cell content in the region and coordinate data of corresponding cells; the region of interest (ROI) is a region to be processed, which is defined by a frame, a circle, an ellipse, an irregular polygon, etc. from a processed image in machine vision and image processing, and is called as a region of interest; as shown in example (3) in fig. 3, the divided regions of interest serve as basic information units of the data conversion structure data.

Step S16, marking the table picture according to the ROI data to obtain a plurality of marked ROI areas, and acquiring cells in each marked ROI area;

and step S17, inputting the cells in the marked ROI area into a CNN image recognition model for character recognition to obtain the fiber core wiring information recorded in the table in the paper data picture.

Step S2, converting the identified fiber core wiring information into structured data, associating the obtained structured data with fiber core service data and transmission segment data in a resource system, updating the resource system according to the associated fiber core service data and transmission segment data, completing the association of operation and maintenance data with the resource system, and forming fiber core digital service information; wherein, the fiber core service data: the method comprises the name of the optical cable section to which the fiber core belongs, the service use of the fiber core, the number of the fiber cores and the name of a start-stop station of the fiber core;

specifically, the process of associating the structured data with the core service data in the resource system in step S2 includes:

step S21, performing text error correction on the fiber core wiring information in the identified paper data picture according to a communication service dictionary and an n-gram algorithm to obtain fiber core wiring information to be converted; the method using the n-gram algorithm comprises the following steps: if the sentence in the text is S ═ { w1, w2, …, wn }, it can be converted into:

p(s) ═ P (w1, w 2.., wn) ═ P (w1) · P (w2| w1) · P (w3| w2, w1) · P (wn | wn-1, wn-2.., w2, w1), where P(s) is a language model, i.e., a model used to calculate the legal probability of a sentence, and wn represents each word that makes up the sentence.

Step S22, constructing a service fiber core pair by using fiber core distribution information to be converted according to the obtained interest point template, storing the service fiber core pair according to the format of the optical cable segment name, the fiber core pair, the service distribution information and the opposite end fiber core pair and the format of a table, and finally generating the structured data of the fiber core distribution information, wherein the storage format of the structured data is shown in Table 1;

TABLE 1

Optical cable segment name	Fiber core pair	Use (service)	Fiber core pair at opposite ends
				Xxx optical cable section	(B01，B02)	Xxx link	(A1，A2)

Step S23, constructing a correlation algorithm model, acquiring the name of an optical cable section to which fiber core wiring information belongs according to the generated structured data, acquiring fiber core service data corresponding to the name of the optical cable section from a resource system according to the name of the optical cable section, and inputting the structured data and the fiber core service data into the algorithm model, wherein the algorithm model comprises a WMD algorithm, a Smooth Inverse Frequency algorithm and an SVM support vector machine, specifically, the WMD algorithm is used for calculating the structured data and the fiber core service data to obtain a first feature corpus, and the Smooth Inverse Frequency algorithm is used for calculating the structured data and the fiber core service data to obtain a second feature corpus;

step S24, inputting the first characteristic linguistic data and the second characteristic linguistic data into an SVM (support vector machine) to obtain the distribution probability of the incidence relation between the structured data and the fiber core service data in the resource system;

and S25, sequencing the distribution probability of the incidence relation, and taking the incidence relation of the maximum distribution probability to obtain the incidence of the structured data and the fiber core service data in the resource system, and finally finishing the incidence of the resource system and the fiber core service data in the operation and maintenance data.

Specifically, the specific process of associating the structured data with the transmission segment data in the resource system in step S2 is as follows:

obtaining fiber core service data according to the structured data of the fiber core distribution information, wherein the fiber core service data comprises a fiber core service purpose name and a start-stop station name to form a fiber core service data set A, and the fiber cable section name in the resource system and the corresponding relationship between the fiber cable section and the transmission section to form a data set B;

and calculating the support degrees of the data sets A and B according to the optical cable section names in the data sets A and B to obtain the incidence relation between the fiber core wiring information and the transmission section.

The image technology is used for preprocessing the picture in the step S12, and the preprocessing includes processing problems such as picture inclination, blur, underexposure, table inclination and the like, and the specific process is as follows:

step S121, carrying out uniform formatting on the cut paper data pictures in sequence to obtain pictures with the sizes being 512 x 512, carrying out graying processing on the pictures and carrying out binarization processing on the pictures, and converting the pictures into a form more suitable for human or machine analysis processing to obtain a picture to be processed;

s122, obtaining the outline edge of the table in the picture to be processed, and obtaining an enclosing matrix and an inclination angle of the table outline according to the outline edge of the table;

step S123, inputting the inclination angle of the table outline into the picture azimuth checking model, and judging whether the inclination angle of the table outline is normal or not;

step S124, if the determination result of the step S123 is abnormal, performing rotation correction on the to-be-processed picture according to a normal angle to obtain a table profile of a normal inclination angle, and executing the step S125, and if the determination result of the step S123 is normal, executing the step S125;

step S125, extracting an image area in the table outline with the normal inclination angle to obtain an independent identification image;

s126, carrying out image enhancement processing on the independent identification image, wherein the image enhancement processing comprises image sharpening, smoothing and denoising processing, and obtaining a picture to be detected; the image enhancement is to suppress useless information and improve the use value of an image; in this embodiment, the picture enhancement method utilizes spatial domain processing and transformation thereof, removes noise using low-frequency filtering, and enhances a signal using high-frequency filtering, so that a picture is clear:

where x, y are the position/coordinates of the pixel in the picture; k, l are the positions/coordinates in the convolution kernel, the coordinates of the center point are (0, 0); f (k, l) is the weight parameter on (k, l) in the convolution kernel; i (x + k, y + l) is the picture pixel value corresponding to f (k, l); h (x, y) is the result of the filtering/convolution of the (x, y) pixel in the picture.

Specifically, the specific process of marking the ROI area in step S16 is as follows:

step S161, acquiring a corresponding form template according to the region-of-interest data in the form picture, and acquiring a corresponding point-of-interest template according to the acquired form template; because the wiring information tables of the regions are not consistent, as different paper data pictures of the example (1) and the example (2) provided in fig. 3, the used Excel templates are different, so that an interest point template needs to be established according to the characteristics of the regions when the ROI region is extracted, wherein the interest point template defines the format to be converted and defines the specific meaning of the target region;

step S162, coordinate data of four corners of a cell in a table picture are obtained, and whether the scaling of the table picture and a corresponding interest point template are matched or not is judged according to the coordinate data of the four corners of the cell;

step S163, if the determination result of step S162 is not matching, scaling the table picture so that the scaling of the interest point template matches the scaling of the table picture, and then executing step S164; if the determination result of step S162 is matching, step S164 is executed;

and S164, when the table picture is matched with the scaling of the interest point template, calculating an ROI (region of interest) matched with the cells in the table picture according to the interest point template and marking the ROI to obtain the cells in the marked ROI.

(2) Optical cable physical resource digitalization: the optical fiber end-to-end of the optical cable mainly comprises key points such as a starting point station, a splicing point, a terminal station and the like, and the digitization of physical resources mainly focuses on the position of the corresponding point and data such as attenuation information, total attenuation, total length and the like of the corresponding point; the optical cable physical resource digitalization mainly analyzes the optical cable fiber core attenuation curve data to obtain the basic data of the whole-process length, the total attenuation, the average attenuation and the like of the optical cable, and then further analyzes the attenuation curve based on artificial intelligence to obtain the position and the attenuation of a splice closure, wherein the specific process comprises the following steps:

step S3, obtaining optical cable fiber core attenuation curve data from the fiber core test data, and analyzing the fiber core attenuation curve data of a plurality of optical fibers in the optical cable fiber core attenuation curve data to obtain fiber core basic data and splice closure data;

specifically, the analysis process is as follows:

step S31, respectively preprocessing the attenuation curve data of the optical fibers to obtain a plurality of preprocessed data;

step S32, obtaining a mixed curve S (l) by superimposing a plurality of preprocessed data:

S(l)＝d₁(l)+d₂(l)+…+d_n(l)

Y＝[y₁(l)，y₂(l)，…，y_i(l)]

D＝[d₁(l)，d₂(l)，…，d_i(l)]

wherein l represents the length of the cable, y_i(l) Representing the attenuation value of the ith fiber at length l, d_i(l) Representing attenuation value y of corresponding i-th optical fiber_i(l) Vector after preconditioning;

step S33, feature extraction is performed on the mixed curve S (l), the extracted features include a signal peak, a height of the peak, a signal average, a signal peak width, a variance, a root mean square, and the like, the set of extracted features is represented by T, where T is [ T ═ T [ [ T ]₁,t₂,...,t_n]；

Step S34, carrying out data annotation on each point in the mixed curve S (l) to obtain an annotation result, wherein the annotation result is 1, which indicates that the junction box is present; the labeling result is to indicate whether it is a closure;

step S35, inputting the mixed curve S (l), the labeling result and the feature set T into an offline training classification model together for offline training, and obtaining a trained classification model after training, wherein the offline training classification model is a binary classifier which can adopt different binary classification methods such as a random forest classification method, a decision tree classification method, gradient pressurization and the like;

s36, carrying out on-line prediction by using the trained classification model, outputting all points with the labeling result of 1 according to the classification model, and outputting the distance l corresponding to the points_j，(j＝1,2,...,m)，l_jThe position of the j-th splice closure of the optical cable, wherein m is the number of all points;

step S37 according to l_jAnd calculating attenuation value a of point j from mixing curve S (l)_jAccording to l_jAnd a_jClosure data L and D are obtained, wherein,

a_j＝f(l_j,S(l))

L＝[l₁,l₂,...,l_m]

D＝[a₁,a₂,...,a_m]

and f is an algorithm function, the algorithm function comprises attenuation data difference and filtering processes, L represents a position point, and D represents attenuation values of all the splicing boxes of the optical cable.

(3) Digitizing the spatial information of the fiber core:

step S4, collecting data to be correlated of each system, wherein the data to be correlated comprises power transmission line data, splice closure spatial data and tower spatial data acquired from a GIS system, optical fiber resource data acquired from a resource system, and fiber core basic data and splice closure data acquired from fiber core test data; the data to be associated specifically includes the following data:

spatial data of the tower: the method comprises the steps of determining GIS coordinates, names, maintenance units, transmission line names, construction time and whether connection exists or not, wherein the GIS coordinates, the names, the maintenance units, the transmission line names and the construction time correspond to towers;

splice closure spatial data: the method comprises the following steps of (1) including the names of a tower to which a splicing box belongs and the tower;

fiber core test data: including fibre core basic data and closure data, the fibre core basic data includes the start and end station name of fibre core, fibre core end-to-end total length, wavelength, affiliated optical cable model, decay and incident point, and the incident point is by the data pair: (core distance, attenuation); splice closure data includes splice closure location and its attenuation;

fiber resource data: the method comprises the optical cable, the name of the optical cable section, the type of the optical cable, the name of a start-stop station, an associated power transmission line, service information, application and the like.

Step S5, according to the data to be correlated of each system, correlating the GIS system, the fiber core test data and the resource system by using a machine learning method to form the correlation relation of the complete physical route of the optical cable, and according to the correlation relation of the complete physical route of the optical cable, extracting the characteristic value by using a theme model to form fiber core digital spatial information;

specifically, the specific process of step S5 is:

step S51, formatting the data to be associated, and processing the data to a format conforming to the feature extraction algorithm, for example: xxx towers/names/power transmission lines, and obtaining feature extraction data;

step S52, extracting the following feature values from the data obtained in step S51 according to different data sources:

pole tower characteristic value: the method comprises the steps of including the name of a power transmission line and GIS coordinates corresponding to a tower;

fiber core test characteristic value: the method comprises the name of a start-stop station of a fiber core and the type of the optical cable to which the start-stop station belongs;

resource characteristic value: the method comprises the following steps of (1) including the optical cable model, the starting and stopping station name and the associated transmission line name;

step S53, constructing a correlation algorithm model, wherein the algorithm model comprises a WMD algorithm, a Smooth Inverse Frequency algorithm and an SVM support vector machine, inputting the extracted characteristic values into the constructed correlation algorithm model, specifically, calculating each characteristic value pairwise by using the WMD algorithm and the Smooth Inverse Frequency algorithm respectively, inputting the calculation result into the SVM support vector machine, and finally obtaining the correlation relationship between a tower and a transmission line, the correlation relationship between the transmission line and an optical cable section, the correlation relationship between an optical cable section and fiber core test data and the correlation relationship between a splice box of the optical fiber test data and the optical cable section; the incidence relation formed by the tower and the transmission line corresponds to the name of the incidence transmission line in the resource system, and the incidence of the resource system and the GIS system, the incidence of the resource system and the fiber core test data and the incidence of the GIS system and the fiber core test data are completed through the incidence relation among the data;

for example, the association process of the tower data and the transmission line data of the GIS system is as follows:

a1, inputting the name of the power transmission line and the name of the power transmission line corresponding to the tower, and obtaining a first characteristic corpus through a WMD algorithm;

a2, inputting the name of the transmission line and the name of the transmission line corresponding to the tower, and obtaining a second feature corpus through a Smooth Inverse Frequency algorithm; the specific process is as follows:

a21: extracting required characteristic text information, wherein the characteristic text information is the name of the power transmission line;

a22: merging the characteristic text information into one text information through aggregation;

a23: then splitting words and sentences in the text information to obtain split sentences;

a24: calculating the similarity of the sentences through the split sentences to obtain a similarity matrix;

a25: storing the similar matrix as a graph form, wherein the sentence is a point, and the similarity is an edge;

a26: and obtaining sentences with high similarity as second feature linguistic data through matrix operation.

A3, inputting the obtained first characteristic corpus and the obtained second characteristic corpus into an SVM (support vector machine) to obtain the association relation between the tower and the transmission line.

Similarly, according to the steps A1-A3, the probability distribution of the power transmission line and the optical cable section is obtained; obtaining the relation between the optical cable section and the optical fiber test data; the relationship of the closure to the cable segment.

S54, forming a relation data vector containing the event, the tower, the optical cable and the station according to the association relation obtained in the step S53, and formatting the obtained relation data vector;

step S55, extracting the relation features of the relation data vectors after the formatting processing to obtain the associated data features;

step S56, inputting the associated data characteristics into the associated algorithm model constructed in the step S53 to obtain the associated relation among the tower, the optical cable and the junction box, specifically, calculating the associated data characteristics by using a WMD algorithm and a Smooth Inverse Frequency algorithm respectively, inputting the calculation result into an SVM support vector machine to obtain the associated relation among the tower, the optical cable and the junction box, obtaining the fiber core distance and attenuation of an event point in fiber core test data, and GIS coordinates corresponding to the optical cable, the tower and the tower, and calculating the GIS coordinates of the tower and the junction box by taking the GIS coordinates, the distance and the attenuation as calculation judgment bases according to the associated information of the tower, the optical cable and the junction box after association;

and S57, associating the tower, the optical cable, the power transmission line and the splice closure according to the GIS coordinates of the tower and the splice closure and all data of the data to be associated, thereby splicing the association relation of the complete physical route of the optical cable.

And step S6, merging the fiber core digital service information and the fiber core digital spatial information to obtain complete digital information of the fiber core, completing data association among operation and maintenance data, a GIS system, fiber core test data and a resource system, and digitally managing the optical cable resources according to the complete digital fiber core information.

In addition, because the resource data in the asset system is static fiber core data which comprises fixed information of fiber core production model, manufacturer and the like, the resource system and the asset related data are directly bound, and a complete optical fiber resource full life cycle management system is formed by combining the generated complete digital fiber core information and work orders, defects, overhaul and the like in the operation and maintenance system. The full life cycle management system based on the optical fiber resource can provide basic data basis for advanced service application of a service system, and can realize the following advanced application:

1. n-x analysis: the analysis capability of the N-x analysis is improved, the analysis of finer dimensions can be provided, and the method is not limited to optical cable sections and equipment, and can also bring splice boxes, towers and the like into the analysis result;

2. optical fiber transmission simulation: the data based on the full life cycle comprises the introduction of parameters such as optical fiber attenuation, optical fiber nonlinearity and the like, so that the simulation capability can be provided for newly-built optical fibers, the decision basis is provided for investment construction, and meanwhile, the risk point simulation test can be provided for operation and maintenance;

3. asset investment analysis: resources related to the optical fiber are managed on the basis of management, and whether decision basis is provided for investment construction of the optical fiber is determined by analyzing the service life, damage and distribution condition of the optical fiber;

4. network risk analysis: based on the data of the full life cycle, the risk assessment of the optical fiber communication network is realized, the possible risk points are estimated, and the corresponding plan is provided, so that the reliability of the power communication network is improved.

To better illustrate the beneficial effects of the present invention, as shown in fig. 4, fig. 4(a) shows typical optical fiber test data (curve), and fig. 4(b) is a relationship curve between the accuracy of digitizing the optical fiber physical information by applying artificial intelligence and different iteration times, which corresponds to three different classification methods (random forest classification method, decision tree classification method, gradient boosting classification method). The accuracy rate refers to the probability that the identified optical fiber splice closure information conforms to the actual situation through different algorithms; when a random forest classification method, a decision tree classification method or a gradient pressurization classification method is applied, good accuracy is obtained after 200 iterations, wherein the accuracy of single fiber core data after 200 iterations of the random forest classification method and the gradient pressurization classification method reaches 95%. After the superposition of 10 fiber core data is applied, the output result reaches 100% after 200 iterations. The result obtained after the physical resources of the optical cable are digitized is used as the input information of the digitization of the space resources of the optical cable, the coordinate information of the power transmission line of the power system is superposed, the geographic position of each splicing box of the optical cable can be accurately positioned, and accurate space-time data is provided for the first-aid repair and fault positioning of the optical cable.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.