An overhead line three-dimensional unstructured data extraction classification method

Abstract

The present application relates to a kind of overhead line three-dimensional unstructured data extraction classification method, comprising the following steps: step S1: the data that can be automatically extracted is imported into data analysis library, then the data type analysis of each data sample is carried out by naive Bayes algorithm to obtain underlying data;Step S2: the underlying data in step S1 is abstracted, and according to different data types and data sources, automatically select the method of data analysis and extraction, and the extracted data is saved to database;Step S3: according to the data extracted, target data calculation is carried out, and the corresponding storage mode is set according to the data type.The present application realizes the automatic extraction classification of space distance, safety distance, tower weight, concrete quantity, avoidance area and other data, reduces the amount of manual work, and provides data extraction technical service for subsequent related project research.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to an overhead line three-dimensional unstructured data extraction and classification method and system. BACKGROUND

[0002] The overhead line mainly refers to an overhead open line, which is erected above the ground and is a power transmission line for transmitting electric energy by fixing power transmission conductors on the pole tower erected on the ground through insulators. The overhead line is convenient to erect and maintain and has low cost, but is prone to be affected by weather and environment (such as strong wind, lightning, contamination, ice and snow) to cause faults, and occupies a large land area of the entire power transmission corridor, which is easy to cause electromagnetic interference to the surrounding environment.

[0003] In the prior art, when the structural data of the overhead line machine is processed and classified, manual operation is required, which is not only time-consuming and laborious, but also prone to errors, so an overhead line three-dimensional unstructured data extraction and classification method is needed to solve the above problems. SUMMARY

[0004] Therefore, the purpose of the present application is to provide an overhead line three-dimensional unstructured data extraction and classification method and system, which classifies unstructured data and manages the classified data in a database, realizes automatic extraction and classification of data such as spatial distance, safety distance, tower weight, concrete amount and avoidance area, reduces the amount of manual work, and provides data extraction technical services for subsequent related project research.

[0005] To achieve the above purpose, the present application adopts the following technical solutions:

[0006] An overhead line three-dimensional unstructured data extraction and classification method comprises the following steps:

[0007] Step S1: importing data that can be automatically extracted into a data analysis library, and then performing data type analysis on each data sample through a naive Bayes algorithm to obtain underlying data;

[0008] Step S2: abstracting the underlying data in step S1, and automatically selecting a data analysis and extraction method according to different data types and data sources, and saving the extracted data to a database;

[0009] Step S3: performing target data calculation according to the extracted data, and setting a corresponding storage mode according to the data type.

[0010] Further, the step S1 is specifically:

[0011] Each data sample is described by an n-dimensional feature vector to represent the values of n attributes, that is:

[0012] X = {x1, x2, ..., x} n}

[0013] Suppose there are m classes, labeled C1, C2, ..., Cn. m This means that, given an unknown data sample X, if the Naive Bayes classification method assigns the unknown data sample X to class C... i Then it must be

[0014] P(C i |X)>P(C j |X)1≤j≤m,j≠i,

[0015] According to Bayes' theorem, since P(X) is constant for all classes, maximizing the posterior probability P(C) is the optimal approach. i |X) can be transformed into maximizing the prior probability P(X|C) i )P(C i For a sample X of unknown category, first calculate the probability P(X|Ci)P(Ci) of X belonging to each category Ci, and then select the category with the highest probability as its category.

[0016] Furthermore, the automatically extractable data includes spatial distance, safety distance, avoidance zone, tower weight, concrete volume data samples, and conductor sag distance to the ground; the spatial distance and safety distance are numerical data; the avoidance zone is spatial topology data; the tower weight is automatically summarized based on the weight of the tower body and connecting legs; and the extraction of concrete volume data is calculated based on the foundation type and foundation geometry.

[0017] Furthermore, in step S1, if the training dataset has many attributes and tuples, calculate P(X|C i The cost will be greater than the preset value. Assume that the values ​​of each attribute are independent, and the prior probability P(x1|C) is greater than the preset value. i ), P(x2|C i ), …, P(x n |C i (Obtained from the training dataset)

[0018] Furthermore, the distance between the conductor sag and the ground is specifically as follows: based on the elevation of the suspension points at both ends of the conductor, the span, and the ground elevation, then applying the formula... L: span, f: sag, h: height difference, theta: observation angle, a: vertical distance from suspension point to instrument, alpha: height difference angle, +: when the instrument is closer to the near suspension point and farther to the far suspension point, take "+", otherwise take "-", the calculation of crossing span distance, first, the sag of the tension section where crossing exists is calculated, then the distance value of the sag to the crossing object is calculated and compared with the required value in the specification to form the corresponding result; the automatic extraction of the avoidance area, the series of spatial coordinates of the line path are automatically extracted according to the corresponding attributes of the vector / raster data of the thematic layer.

[0019] Further, in step S2, for GIS data extraction, the GIS data content is returned back through the query port provided in the service, and all intervals are accessed by stack through bisection, and recursive calls are made.

[0020] Further, in step S2, for numerical data, a trigger is used for data extraction, a trigger is built at the extraction end, when the source data changes, the corresponding trigger writes the changed data into a temporary database, and the extraction thread extracts data from the temporary table, and for the data samples of the avoidance area, tower weight and concrete quantity, a log-based data extraction method is used, in the process of collecting the database log, the required data is extracted by collecting the log.

[0021] Further, in step S2, for the extraction of crossing span distance, the spatial calculation function of PostGIS can be used to first extract the sag of the tension section where crossing exists, and then calculate the distance value of the sag to the crossing object.

[0022] Further, in step S3, for numerical data, a regression calculation method is used, and a sequential storage structure is used for data storage, and for the data samples of the avoidance area, tower weight and concrete quantity, a parallel algorithm is used for calculation, and a linked storage structure is used for data storage.

[0023] An overhead line three-dimensional unstructured data extraction and classification system, comprising a processor, a memory and a computer program stored on the memory, when the processor executes the computer program, specifically executes the steps in the overhead line three-dimensional unstructured data extraction and classification method as described above.

[0024] Compared with the prior art, the present application has the following beneficial effects:

[0025] The present application adopts the form of layer or data layering, classifies and manages the unstructured data by using the database, realizes the automatic extraction and classification of the data such as spatial distance, safety distance, tower weight, concrete quantity and avoidance area, reduces the labor operation amount, and provides data extraction technical services for subsequent related project research. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a flow chart of the method of the present application. DETAILED DESCRIPTION

[0027] The present application is further described below in conjunction with the accompanying drawings and examples.

[0028] Reference should be made to Figure 1 The present application provides a three-dimensional unstructured data extraction and classification method for overhead lines, comprising the following steps:

[0029] S1, analysis of data types that can be automatically extracted

[0030] The data that can be automatically extracted is imported into a data analysis library, and then data type analysis is performed on each data sample by a naive Bayes algorithm to obtain underlying data, wherein

[0031] Each data sample is described by an n-dimensional feature vector to describe the values of n attributes, i.e., X={x1, x2,…, xn}, assuming that there are m classes, respectively represented by C1, C2,…, Cm, and given an unknown data sample X (i.e., without class label), if the naive Bayes classification method allocates the unknown data sample X to the class C n , then it must be m i

[0032] P(C i |X)>P(C j |X)1≤j≤m,j≠i,

[0033] According to Bayes' theorem,

[0034] Since P(X) is constant for all classes, maximizing the posterior probability P(C i |X) can be converted to maximizing the prior probability P(X|C i )P(C i ), for an unknown class sample X, the probability P(X|Ci)P(Ci) of X belonging to each class Ci can be calculated first, and then the class with the largest probability is selected as its class. If the training data set has many attributes and tuples, the overhead of calculating P(X|C i ) can be very large. Assuming that the values of the attributes are independent of each other, the prior probabilities P(x1|C i ), P(x2|C i ),…, P(x n |C i ) can be obtained from the training data set.

[0035] ​​In the embodiment, the data that can be automatically extracted includes data samples such as spatial distance, safety distance, avoidance area, tower weight, and concrete amount, and the spatial distance and safety distance are numerical data, the avoidance area is spatial topology data, the tower weight needs to be automatically summarized based on the tower body and leg weight, and the extraction of the concrete amount and other data needs to be calculated based on the foundation type and foundation geometric size. The formula for calculating the ground distance of the conductor sag is based on the suspension point elevation of the two ends of the conductor, the span, and the ground elevation, and then the formula is used for calculation. The calculation of the crossing distance needs to first calculate the sag of the tension section that exists in the crossing, then calculate the distance value of the sag to the crossing object and compare it with the required value in the specification to form the corresponding result. The avoidance area is automatically extracted based on the series of spatial coordinates of the line path and the corresponding properties of the vector / raster data of the thematic layer.

[0036] S2, bottom layer engineering data and GIS data extraction

[0037] The bottom layer data in step S1 is analyzed, and according to different data types and data sources, the method of data analysis and extraction is automatically selected, and the extracted data is saved to the database. For GIS data extraction, the GIS data content is returned back through the query port provided in the service, and all intervals are accessed by bisection and stack, and recursive call. For numerical data, the data extraction is performed using the trigger method. The trigger is built at the extraction end. When the source data changes, the corresponding trigger writes the changed data to the temporary database, and the extraction thread extracts the data from the temporary table. For data samples such as avoidance area, tower weight, and concrete amount, the log-based data extraction method is used. In the process of collecting database logs, the required data is extracted by collecting logs. For the extraction of the crossing distance, the spatial calculation function of PostGIS can be used to first extract the sag of the tension section that exists in the crossing, and then calculate the distance value of the sag to the crossing object.

[0038] S3, automatic calculation and saving of target data

[0039] After the bottom layer related data is extracted in step S2, the target data is calculated according to the data to be extracted, and the corresponding storage method is set according to the data type.

[0040] For numerical data, the regression calculation method is used, and the sequential storage structure is used for data storage. For data samples such as avoidance area, tower weight, and concrete amount, parallel algorithm is used for calculation, and link storage structure is used for data storage.

[0041] In the embodiment, the three-dimensional unstructured data extraction classification method of the overhead line adopts the form of layers or data layers, classifies the unstructured data, and classifies and manages the data by using a database, so as to realize automatic extraction and classification of data such as spatial distance, safety distance, tower weight, concrete quantity, and avoidance area, reduce the labor operation amount, and provide data extraction technical services for subsequent related project research.

[0042] The above merely describes preferred embodiments of the present application, and any equivalent changes and modifications made within the scope of the present application should be included in the scope of the present application.

Claims

1. A method for extracting and classifying three-dimensional unstructured data of overhead power lines, characterized in that, Includes the following steps: Step S1: Import the automatically extractable data into the data analysis library, and then use the Naive Bayes algorithm to perform data type analysis on each data sample to obtain the underlying data; Step S2: Extract the underlying data from Step S1, and automatically select data analysis and extraction methods according to different data types and data sources, and save the extracted data to the database; Step S3: Calculate the target data based on the extracted data, and set up the corresponding storage method according to the data type; Step S1 specifically involves: Suppose that each data sample is described by an n-dimensional feature vector representing the values ​​of n attributes, i.e.: X={x1,x2,…,x n } Suppose there are m classes, labeled C1, C2, ..., Cn. m This means that, given an unknown data sample X, if the Naive Bayes classification method assigns the unknown data sample X to class C... i Then it must be P(C i |X)>P(C j |X),1≤j≤m,j≠i, According to Bayes' theorem, since P(X) is constant for all classes, maximizing the posterior probability P(C) is the optimal approach. i |X) can be transformed into maximizing the prior probability P(X|C) i )P(C i For a sample X of unknown category, first calculate the probability P(X|Ci)P(Ci) that X belongs to each category Ci, and then select the category with the highest probability as its category. The automatically extractable data includes spatial distance, safety distance, clearance zone, tower weight, concrete volume data samples, and conductor sag distance to the ground; the spatial distance and safety distance are numerical data; the clearance zone is spatial topology data; the tower weight is automatically summarized based on the weight of the tower body and its legs; and the extraction of concrete volume data is calculated based on the foundation type and foundation geometry. In step S1, if the training dataset has many attributes and tuples, calculate P(X|C i The cost will be greater than the preset value. Assume that the values ​​of each attribute are independent, and the prior probability P(x1|C) is greater than the preset value. i ), P(x2|C i ), …, P(x n |C i (Obtained from the training dataset) 2. The method for extracting and classifying three-dimensional unstructured data of overhead power lines according to claim 1, characterized in that, The distance between the conductor sag and the ground is specifically determined as follows: based on the elevation of the suspension points at both ends of the conductor, the span, and the ground elevation, the formula is then applied. l: span, f: sag, h: elevation difference, θ: observation angle, a: vertical distance from suspension point to instrument, α: elevation difference angle, ±: when the instrument's near suspension point is lower than the far suspension point, take "+", otherwise take "-". Calculation of the crossing distance requires first calculating the sag of the tension section with crossing, then calculating the distance of the sag to the crossing object and comparing it with the standard requirement value to form the corresponding result; the automatic extraction of the avoidance zone is performed automatically based on the series of spatial coordinates of the line path according to the corresponding attributes of the vector / raster data of the thematic layer.

3. The method for extracting and classifying three-dimensional unstructured data of overhead power lines according to claim 1, characterized in that, In step S2, the GIS data is extracted through the query port provided in the service, the GIS data content is returned, and all intervals are stored in a stack using a binary search method, as well as recursive calls.

4. The method for extracting and classifying three-dimensional unstructured data of overhead power lines according to claim 1, characterized in that, In step S2, for numerical data, a trigger method is used for data extraction. A trigger is created at the extraction end. When the source data changes, the corresponding trigger writes the changed data into a temporary database. The extraction thread extracts data from the temporary table. For data samples of avoidance zone, tower weight, and concrete quantity, a log-based data extraction method is used. During the collection of database logs, the required data is extracted by collecting logs.

5. The method for extracting and classifying three-dimensional unstructured data of overhead power lines according to claim 1, characterized in that, In step S2, for the extraction of the crossing distance, the spatial calculation function of PostGIS can be used to first extract the sag of the tension section with crossing, and then calculate the distance value of the sag to the crossing object.

6. The method for extracting and classifying three-dimensional unstructured data of overhead power lines according to claim 1, characterized in that, In step S3, regression calculation method is used for numerical data and sequential storage structure is used for data storage. For data samples of avoidance zone, tower weight, and concrete quantity, parallel algorithm is used for calculation and linked storage structure is used for data storage.

7. A three-dimensional unstructured data extraction and classification system for overhead power lines, characterized in that, It includes a processor, a memory, and a computer program stored in the memory. When the processor executes the computer program, it specifically performs the steps in the three-dimensional unstructured data extraction and classification method for overhead lines as described in any one of claims 1-6.

Images