Unmanned aerial vehicle intelligent acceptance method and system for power transmission project

Abstract

The invention discloses an unmanned aerial vehicle intelligent acceptance method and system for a power transmission project, and relates to the technical field of power transmission project acceptance, and the system comprises a control center which is connected with a project collection module, a power transmission processing module, an intelligent analysis module and a comprehensive acceptance module. Performing frame image conversion on the power transmission video data to obtain an engineering pre-processing image, and setting a multilateral constraint criterion to perform constraint interception and classification extraction on the engineering pre-processing image to obtain an image weight feature sequence; performing mode conversion and truncation on the image weight feature sequence to obtain image weight slices; carrying out homomorphic replacement and dynamic correction on the acceptance standard data to obtain a corrected acceptance set, and carrying out project matching and judgment on the project to-be-inspected slices according to the corrected acceptance set to obtain qualified acceptance nodes and abnormal acceptance nodes; the acceptance speed is greatly improved, the acceptance period is shortened, and the quality management level of the power transmission project is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission project acceptance, and in particular to an unmanned aerial vehicle intelligent acceptance method and system for power transmission projects. Background Art

[0002] Transmission projects refer to projects that transmit the electricity produced by power plants to various substations through transmission lines and then distribute it to users. Therefore, it is necessary to inspect and accept transmission projects to ensure that the projects are successfully completed in accordance with design specifications and standards and can be put into operation safely and stably. Traditional transmission project acceptance mainly relies on manual inspection and ground instrument measurement. These methods are easily affected by factors such as terrain, weather, and measurement angle during the acceptance process, have visual blind spots, and are inefficient.

[0003] In recent years, the development of UAV technology has provided a new solution for the acceptance of power transmission projects. UAVs are used to replace manual work in high-altitude operations, reduce safety operation risks, quickly cover large areas of transmission lines, and reduce acceptance time. To this end, a UAV intelligent acceptance method and system for power transmission projects are provided. Summary of the invention

[0004] The object of the present invention is to provide a method and system for intelligent acceptance of unmanned aerial vehicles for power transmission projects.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] An unmanned aerial vehicle intelligent acceptance system for power transmission projects, comprising a control center, the control center being connected to an engineering acquisition module, a power transmission processing module, an intelligent analysis module and a comprehensive acceptance module;

[0007] The engineering acquisition module is used to collect pre-inspection engineering data, pre-process the pre-inspection engineering data, and obtain three-dimensional video data and thermal red geographic data;

[0008] The process of preprocessing the pre-test engineering data includes:

[0009] Set up a front-end collection terminal, collect criteria through the front-end collection terminal, and obtain acceptance standard data;

[0010] Set deployment criteria according to the power transmission project, intelligently deploy drones according to the deployment criteria, obtain deployment unmanned terminals, conduct safe collection of power transmission projects through the deployment unmanned terminals, obtain pre-verified project data, and mark the location where the deployment unmanned terminals conduct safe collection of power transmission projects as the collection point;

[0011] Perform digital image summarization on the pre-test engineering data to obtain picture video data and thermal red geographic data, perform image conversion on the picture video data to obtain three-dimensional video data, and the three-dimensional video data includes three-dimensional image data and power transmission video data.

[0012] The process of obtaining a constraint engineering image includes:

[0013] Perform frame-to-image conversion on the power transmission video data to obtain key video frames;

[0014] Continuously combining the obtained key video frames based on the order of frame-image conversion to obtain a key frame set, and marking the key frame set and the three-dimensional image data as an engineering preprocessing image;

[0015] A multilateral constraint criterion is set, and the engineering preprocessing image is constrained and intercepted according to the multilateral constraint criterion to obtain a constrained engineering image.

[0016] The process of obtaining the classified extracted block core includes:

[0017] A limited selection kernel is set according to the multilateral constraint criteria, an extraction starting point is set for the constrained engineering image, the obtained limited selection kernel is uploaded to the constrained engineering image based on the extraction starting point, and the corresponding range of the limited selection kernel in the constrained engineering image is marked as a constrained area;

[0018] According to the limited selection core constraint engineering image is classified and extracted to obtain the image weight feature sequence.

[0019] The process of classifying and extracting engineering images based on limited selection and verification constraints includes:

[0020] The obtained limited selection kernel is matched and convolved with the constraint area at the extraction starting point to obtain the classification extraction block kernel;

[0021] The shift spacing is set according to the multilateral constraint criteria, and the limited selection kernel is translated and slid according to the obtained shift spacing based on the constrained engineering image to reach the next constrained area. The limited selection kernel is matched and convolved with the constrained area at the extraction starting point until all areas of the constrained engineering image are covered, and the obtained classified extraction block kernel is weight-checked to obtain the block kernel weight;

[0022] The block core weights are repeatedly screened based on the order of translation and sliding to obtain the block replacement weight values, and the block replacement weight values ​​are sequentially combined based on the order of translation and sliding to obtain the image weight feature sequence.

[0023] The process of obtaining image weight slices includes:

[0024] Performing mode conversion on the obtained image weight feature sequence to obtain image weight variables;

[0025] The constrained weight coefficient is set according to the limited selected kernel, and the cardinality separation of the image weight variable is performed through the constrained weight coefficient to obtain the separated image weight coefficient;

[0026] Switching the form of the separated image weight coefficient to obtain an image weight data stream, and performing data truncation on the obtained image weight data stream to obtain an image weight slice;

[0027] Perform homomorphic replacement on the obtained hot red geographic data to obtain hot red geographic slices;

[0028] The obtained hot red geographic slices and image weight slices are marked as engineering test slices.

[0029] The process of isomorphic replacement of the obtained hot red geographic data includes:

[0030] Performing mode conversion on the hot red geographic data to obtain hot red geographic variables, performing cardinality separation on the hot red geographic variables by constraining weighted coefficients to obtain separated hot red geographic coefficients;

[0031] The separated hot red geographic coefficients are switched in form to obtain a hot red geographic data stream, and the hot red geographic data stream is truncated to obtain hot red geographic slices.

[0032] The process of obtaining qualified acceptance nodes and abnormal acceptance nodes includes:

[0033] Perform homomorphic permutation on the obtained acceptance standard data to obtain acceptance data slices;

[0034] The obtained acceptance data slices are combined into fragments to obtain an acceptance standard set, and the acceptance standard set is dynamically revised to obtain a revised acceptance set;

[0035] Obtain the pre-inspected engineering data corresponding to the engineering slice to be inspected, record it as the to-be-inspected matching data, perform project matching on the acceptance standard data according to the to-be-inspected matching data, and obtain the matching correction inspection collection;

[0036] Perform acceptance verification on the to-be-verified matching data according to the matching correction verification collection to obtain a verification matching set;

[0037] The result of the verification matching set is judged to obtain the judgment slice, and the node judgment of the pre-inspection engineering data is performed based on the judgment slice based on the centralized procurement point to obtain the qualified acceptance node and the abnormal acceptance node.

[0038] The process of determining the nodes of the centralized procurement points by determining the slices includes:

[0039] Obtaining the engineering slices to be tested corresponding to the pre-tested engineering data, and performing statistics on the results of the obtained engineering slices to be tested to obtain slice statistics;

[0040] According to the slice statistics, the centralized procurement points are assessed for safety and single-item acceptance is obtained. Based on the single-item acceptance, the centralized procurement points are security-verified to obtain qualified acceptance nodes and abnormal acceptance nodes.

[0041] Based on the above-mentioned UAV intelligent acceptance system for power transmission projects, the present invention also provides a UAV intelligent acceptance method for power transmission projects, comprising the following steps:

[0042] Step 1: Preprocess the collected pre-test engineering data to obtain three-dimensional video data and thermal red geographic data;

[0043] Step 2: Perform frame-to-image conversion and continuous combination on the power transmission video data to obtain a key frame set, set a multilateral constraint criterion to constrain the key frame set and three-dimensional image data to obtain a constrained engineering image, set a limited selection check based on the multilateral constraint criterion to classify and extract the constrained engineering image, and obtain an image weight feature sequence;

[0044] Step 3: Perform mode conversion on the image weight feature sequence to obtain image weight variables, set constraint weighting coefficients to perform cardinality separation on the image weight variables to obtain separated image weight coefficients, perform form switching and data truncation on the separated image weight coefficients to obtain image weight slices, perform homomorphic permutation on the hot red geographic data to obtain hot red geographic slices, and record the image weight slices and hot red geographic slices as engineering test slices;

[0045] Step 4: Perform homomorphic replacement and dynamic correction on the acceptance standard data to obtain the corrected verification collection, perform project matching on the engineering slices to be tested to obtain the matched corrected verification collection, accept and judge the engineering slices to be tested through the matched corrected verification collection to obtain the judged slices, perform node judgment on the centralized procurement points through the judged slices, and obtain qualified acceptance nodes and abnormal acceptance nodes.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The UAV is used to collect data on the entire transmission project, and the image features are extracted from the obtained video data to obtain high-definition feature maps. The high-definition feature maps are extracted and converted to obtain image weight slices, which is conducive to the accurate identification of the image information collected by the UAV, the discovery of minor defects, and the reliable operation of the transmission project.

[0048] The hot red geographic data and three-dimensional data collected by drones are converted homomorphically to obtain the same data form as image feature extraction, which is conducive to the same data features, facilitates acceptance under the same standards, improves acceptance speed, realizes data-based management, and improves the quality management level of power transmission projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION

[0051] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] like Figure 1 As shown, a UAV intelligent acceptance system for power transmission projects includes a control center, and the control center is connected to an engineering acquisition module, a power transmission processing module, an intelligent analysis module, and a comprehensive acceptance module;

[0053] The engineering acquisition module is used to collect pre-inspection engineering data, pre-process the pre-inspection engineering data, and obtain three-dimensional video data and thermal red geographic data. The specific process includes:

[0054] Set up a front-end collection terminal, collect criteria through the front-end collection terminal, and obtain acceptance standard data;

[0055] The acceptance standard data includes design parameters, national standards and industry standards, and project history data, wherein the design parameters represent the design data and design drawing information provided by the power transmission project at the design stage, and standard data for standardizing and unifying the construction of the power transmission project; the national standards and industry standards represent the construction and acceptance standards of the transmission line; and the project history data represent the historical acceptance data of similar projects, so as to obtain common quality problems in the construction of the same type of power transmission projects and improve the safety of acceptance;

[0056] Set deployment criteria according to the power transmission project, intelligently deploy drones according to the set deployment criteria, obtain deployment unmanned terminals, conduct safety collection of power transmission projects through deployment unmanned terminals, obtain pre-verified project data, and mark the location where the deployment unmanned terminals conduct safety collection in the power transmission project as the collection point;

[0057] The pre-inspection engineering data includes power transmission video data, thermal infrared data, three-dimensional data, point cloud data and geographic information data;

[0058] Furthermore, the deployment criteria indicate that the flight route of the drone for acceptance and the location of the collection points for the drone to collect the acceptance data of the power transmission project during the flight are determined according to the equipment and devices in the power transmission project;

[0059] The intelligent deployment means controlling the drone to fly according to the route of the deployment criteria, marking the drone that executes the deployment criteria as the deployment unmanned terminal, and collecting the acceptance data on the flight route in real time during the flight, which is recorded as the pre-inspection engineering data;

[0060] It should be further explained that, in the specific implementation process, the power transmission video data represents all video data that can be collected at the centralized collection point, the three-dimensional data represents the three-dimensional structure data that can be collected by all equipment in the power transmission project, and the geographic information data represents the geographic location-related data of the centralized collection point;

[0061] The obtained pre-test engineering data is summarized by digital graphics to obtain picture video data and thermal infrared geographic data, wherein the picture video data includes power transmission video data, three-dimensional data and point cloud data, and the thermal infrared geographic data includes thermal infrared data and geographic information data;

[0062] Performing image conversion on the obtained picture and video data to obtain three-dimensional video data;

[0063] Furthermore, the image conversion means converting the three-dimensional data and point cloud data in the picture video data into image form, and converting the point cloud data and the three-dimensional data into images through three-dimensional visualization technology, recorded as three-dimensional image data, and marking the obtained three-dimensional image data and transmission video data as three-dimensional video data.

[0064] The power transmission processing module is used to perform frame-to-image conversion on the power transmission video data to obtain the engineering preprocessing image, set a multilateral constraint criterion to perform constraint interception and classification extraction on the engineering preprocessing image, and obtain an image weight feature sequence. The specific process includes:

[0065] Performing frame-to-image conversion on the power transmission video data in the obtained three-dimensional video data to obtain a key video frame;

[0066] It should be further explained that, in a specific implementation process, the frame image conversion process includes:

[0067] An interval standard is set according to the obtained power transmission video data, wherein the interval standard indicates setting the time interval between extracting two adjacent frames of images. For example, if the interval standard is a, it indicates that the next frame of image is extracted after the interval time a.

[0068] The power transmission video data is captured at intervals according to the obtained interval standard to obtain key video frames, wherein the interval capture means extracting a frame of the power transmission video data at the next time point according to the time interval of the interval standard, and recording the image of the next frame as the key video frame; in particular, the interval capture is captured according to the video playback order of the power transmission video data, that is, the order of the obtained key video frames is the same as the video playback order of the power transmission video data;

[0069] Continuously combining the obtained key video frames based on the order of frame image conversion to obtain a key frame set, wherein the arrangement order of the key video frames in the obtained key frame set is arranged according to the order of interval capture;

[0070] Marking the obtained key image frame set and three-dimensional image data as engineering pre-processed images;

[0071] Setting a multilateral constraint criterion, wherein the multilateral constraint criterion includes a conveying size and a conveying channel;

[0072] The transmission size includes a horizontal size and a vertical size, wherein the horizontal size indicates the number of pixels of the image in the horizontal dimension, and the vertical size indicates the number of pixels of the image in the vertical dimension. For example, if the horizontal size is 224*224, it means that there are 224 pixels in the horizontal dimension and 224 pixels in the vertical direction.

[0073] The transmission channel indicates the number of colors allowed to pass through the image. For example, if the transmission channel is red, yellow and blue, it means that the three colors of red, yellow and blue are allowed to pass through. If there are other colors, they are not allowed to pass through, that is, after passing through the transmission channel, the color is filtered out;

[0074] According to the obtained multilateral constraint criteria, the engineering preprocessing image is constrained and intercepted to obtain a constrained engineering image, wherein, according to the key frame set and three-dimensional image data included in the engineering preprocessing image, the constrained engineering image includes a constrained key frame and a constrained three-dimensional image;

[0075] Furthermore, the constrained interception means that the engineering preprocessed image is cropped according to the multilateral constraint criterion, and the obtained constrained engineering image has the same number of pixels in the horizontal dimension and the vertical dimension, and the color range that the constrained engineering image can display is determined according to the transmission channel;

[0076] A limited selection kernel is set according to the obtained multilateral constraint criteria, wherein the size of the limited selection box is related to the conveying channel in the multilateral constraint size, that is, the limited selection box length = 2* conveying channel, the limited selection box width = 2* conveying channel, and the limited selection kernel is expressed in a matrix form composed of a number of elements, then the limited selection box length represents the number of matrix columns, and the limited selection box width represents the number of matrix rows, the obtained limited selection box length is marked as g, and the obtained limited selection box width is marked as k, then the obtained limited selection kernel is a matrix of k rows and g columns, wherein k = 1, 2, 3, ..., u1, u1 is a positive integer, and g = 1, 2, 3, ..., u2, u2 is a positive integer;

[0077] An extraction starting point is set for the obtained constrained engineering image, where the extraction starting point represents a movement starting point of a limited selection kernel on the constrained engineering image. In this embodiment, the extraction starting point is the upper left corner of the constrained engineering image.

[0078] Uploading the obtained limited selection core to the constrained engineering image based on the extraction starting point, and marking the corresponding range of the limited selection core in the constrained engineering image as a constrained area;

[0079] Classify and extract the engineering image based on the obtained limited selection core constraint to obtain the image weight feature sequence;

[0080] It should be further explained that, in the specific implementation process, the classification extraction process includes:

[0081] The obtained limited selection kernel is matched and convolved with the constraint area at the extraction starting point to obtain the classification extraction block kernel;

[0082] The matching convolution means convolving the elements in the obtained limited selection kernel with the pixels at the corresponding position of the constraint area to obtain the classification extraction block kernel, and the obtained classification extraction block kernel is expressed in the form of a matrix, and the elements in the matrix are the convolution results of the limited selection kernel at the corresponding position and the elements of the constraint area;

[0083] According to the obtained multilateral constraint criteria, a shift spacing is set, wherein the shift spacing represents a distance for limiting the selected core to move to the next position in the constraint engineering image, which is the same as the number of conveying channels;

[0084] Based on the constrained engineering image, the limited selection kernel is translated and slid according to the obtained shift interval to reach the next constrained area, and the classification extraction block kernel is obtained by matching the limited selection kernel with the constrained area at the extraction starting point, where the translation and sliding means moving the limited selection kernel from the current position to the distance of the shift interval in the constrained engineering image, and obtaining the corresponding range of the limited selection kernel in the constrained engineering image, which is the constrained area;

[0085] Repeat the process of obtaining the classification extraction block core based on the shift spacing until all areas of the constrained engineering image are covered, and count the obtained classification extraction block cores;

[0086] Perform weight check on the obtained classified extracted block core to obtain the block core weight;

[0087] Further, the weight check indicates that a core value calculation is performed on the classification extraction block core, and the core value calculation indicates that a value of a determinant of the classification extraction block core is calculated, that is, a block core weight, wherein, according to the expression form of the classification extraction block core being a matrix form, the core value calculation is to calculate the value of the determinant of the classification extraction block core;

[0088] Based on the order of translation and sliding, the obtained block core weights are repeatedly screened to obtain block replacement weight values, wherein the repeated screening means that the values ​​of the obtained block core weights are counted in the constrained engineering image, and the repeated block core weights are deleted according to the order of translation and sliding, that is, the first block core weight is retained according to the order of translation and sliding, and the repeated block core weights that appear for the second time and thereafter are deleted until the obtained block core weights are not repeated, and the obtained block core weights are marked as block replacement weight values;

[0089] The obtained block replacement weight values ​​are sequentially combined based on the order of translation and sliding to obtain an image weight feature sequence;

[0090] The sequence combination means combining the obtained block replacement weight values ​​in sequence according to the order of translation and sliding to obtain a sequence composed of weight values, which is the image weight feature sequence;

[0091] Further, according to the constraint keyframes and the constraint three-dimensional image included in the constraint engineering image, the image weight feature sequence includes a key weight feature sequence corresponding to the constraint keyframes and a three-dimensional weight feature sequence corresponding to the constraint three-dimensional image.

[0092] The intelligent analysis module is used to analyze the image weight feature sequence and the thermal red geographic data to obtain the engineering test slice. The specific process includes:

[0093] Performing mode conversion on the obtained image weight feature sequence to obtain image weight variables;

[0094] The mode conversion means converting the representation form of the image weight feature sequence into a signal form, that is, an image weight variable;

[0095] Perform mode conversion on the limited selected kernel to obtain the limited kernel variable, and set the constraint weight coefficient according to the limited selected kernel. Then the constraint weight coefficient = limited kernel variable Limit the kernel variable *β, β represents the balance factor, where " " represents the convolution symbol, that is, two identical restricted kernel variables are convolved and then multiplied by the balance factor to obtain the constrained weight coefficient;

[0096] The image weight variables are subjected to cardinality separation by constraining the weight coefficients to obtain the separated image weight coefficients;

[0097] Furthermore, the cardinality separation means uploading the obtained constraint weighting coefficient to the image weight variable, performing convolution on the corresponding position of the constraint weighting coefficient and the image weight variable, and then summing the convolution results of each corresponding position to obtain the separated image weight coefficient;

[0098] Switching the obtained separation image weight coefficients in a form to obtain an image weight data stream;

[0099] Furthermore, the form switching means replacing the obtained separated image weight coefficient with a data stream composed of random binary code elements;

[0100] Performing data truncation on the obtained image weight data stream to obtain image weight slices;

[0101] Furthermore, the data truncation means grouping the image weight data stream into groups of e binary symbols to obtain image weight slices. If the last group is less than e binary symbols, 0 is added after the insufficient position until e binary symbols are met. In this embodiment, e=7;

[0102] Perform homomorphic replacement on the obtained hot red geographic data to obtain hot red geographic slices;

[0103] It should be further explained that, in the specific implementation process, the isomorphic replacement process includes:

[0104] Performing mode conversion on the obtained hot red geographic data to obtain hot red geographic variables;

[0105] The hot red geographic variables are separated by constraining the weighted coefficients to obtain the separation hot red geographic coefficients;

[0106] Switching the obtained separated hot red geographic coefficients in a form to obtain a hot red geographic data stream;

[0107] Performing data truncation on the obtained hot red geographic data stream to obtain hot red geographic slices;

[0108] The obtained hot red geographic slices and image weight slices are marked as engineering test slices.

[0109] The comprehensive acceptance module is used to perform homomorphic replacement and dynamic correction on the acceptance standard data, obtain the corrected acceptance collection, match and judge the engineering slices to be inspected according to the corrected acceptance collection, and obtain qualified acceptance nodes and abnormal acceptance nodes. The specific process includes:

[0110] Perform homomorphic permutation on the obtained acceptance standard data to obtain acceptance data slices;

[0111] It should be further explained that, in the specific implementation process, the process of isomorphic replacement of acceptance criteria data includes:

[0112] Performing mode conversion on the obtained acceptance criterion data to obtain acceptance criterion variables;

[0113] Cardinality separation is performed on acceptance criteria variables by constraining weighted coefficients to obtain separation acceptance coefficients;

[0114] Switching the form of the obtained separation acceptance coefficient to obtain an acceptance standard data stream;

[0115] Performing data truncation on the obtained acceptance standard data stream to obtain acceptance data slices;

[0116] Slice and partition the obtained acceptance data slices to obtain normal codes and system codes;

[0117] Further, the slicing partitioning means marking the first 4 binary code elements as normal codes and marking the last 3 binary code elements as systematic codes in the acceptance criterion set;

[0118] According to the obtained acceptance data slices, the fragments are combined to obtain the acceptance standard set, and the obtained acceptance standard set is marked as , n represents the number of the acceptance criteria set after the fragments are combined, n=1, 2, 3, ..., u3, u3 is a positive integer;

[0119] Further, the fragment combination means that according to the order of data truncation, three consecutive acceptance data slices are grouped into a matrix form of 3 rows and 7 columns, that is, the acceptance standard set; in particular, if the last group is less than three acceptance data slices, it is supplemented forward until three acceptance data slices are satisfied as a group;

[0120] Dynamically modify the obtained acceptance criteria set to obtain a modified acceptance set;

[0121] Furthermore, the dynamic correction process includes:

[0122] The acceptance standard set is partitioned according to the normal code and the systematic code to obtain the normal code set and the systematic code set, wherein the normal code set represents a 3-row 4-column matrix formed by the first 4 binary code elements in the acceptance standard set, and the systematic code set represents a 3-row 3-column matrix formed by the last 3 binary code elements in the acceptance standard set;

[0123] Based on the system code set, the acceptance criteria set is restricted and transformed to obtain the revised acceptance set;

[0124] Furthermore, the restriction transformation means performing modulo-2 addition operations on the rows of the matrix in the acceptance standard set until the system code set becomes the unit matrix, then the normal code set is recorded and constituted with the unit matrix to form a revised test set, and the obtained revised test set is marked as ;

[0125] Obtain the pre-inspected engineering data corresponding to the engineering slice to be inspected, record it as the to-be-inspected matching data, perform project matching on the acceptance standard data according to the obtained to-be-inspected matching data, and obtain the matching correction inspection collection;

[0126] Further, the project matching means obtaining acceptance standard information matching the pre-tested engineering data in the acceptance standard data according to the pre-tested engineering data corresponding to the engineering slice to be tested, and obtaining the revised test collection corresponding to the matching acceptance standard information, which is recorded as the matched revised test collection;

[0127] Perform acceptance verification on the matched data to be verified based on the obtained matched correction verification set to obtain a verification matching set;

[0128] Further, the acceptance verification means that the engineering slices to be verified corresponding to the obtained matching data to be verified and the matching correction verification collection are subjected to matrix extraction to obtain a verification matching set, and the obtained verification matching set is marked as ,in, , according to each item of to-be-verified matching data is extracted and truncated to obtain a number of engineering to-be-verified slices, then m represents the number of the engineering to-be-verified slice included in each item of to-be-verified matching data, Indicates the engineering slice to be tested, m=1, 2, 3, ..., u4, u4 is a positive integer, express The matrix transpose of ;

[0129] Performing result determination on the obtained verification matching set to obtain determination slices, wherein the determination slices include qualified verification slices and abnormal verification slices;

[0130] Further, the result determination indicates that the calculation results of the verification matching set are classified, and when the verification matching set is a zero matrix, the engineering to-be-tested slice corresponding to the verification matching set is marked as a qualified verification slice, indicating that the part corresponding to the engineering to-be-tested slice in the pre-test engineering data is qualified;

[0131] When the verification matching set is a non-zero matrix, the engineering slice to be tested corresponding to the verification matching set is marked as an abnormal verification slice, indicating that the part corresponding to the engineering slice to be tested in the pre-verified engineering data is unqualified, that is, abnormal;

[0132] Based on the centralized procurement point, node judgment is performed on the pre-inspection engineering data according to the obtained judgment slices to obtain qualified acceptance nodes and abnormal acceptance nodes;

[0133] It should be further explained that, in the specific implementation process, the node determination process includes:

[0134] The engineering slices to be tested corresponding to the pre-verified engineering data are obtained, and the result judgment of the obtained engineering slices to be tested is counted to obtain the slice statistics, wherein the slice statistics include the qualified slice quantity and the abnormal slice quantity, wherein the qualified slice quantity represents the number of qualified verification slices in the same pre-verified engineering data, and the abnormal slice quantity represents the number of abnormal verification slices in the same pre-verified engineering data; for example, "the same pre-verified engineering data" can be represented as thermal infrared data in the pre-verified engineering data;

[0135] Make safety assessments on the centralized procurement points based on the slice statistics and obtain single-item acceptance;

[0136] Furthermore, the single acceptance qualification means that at the centralized procurement point, the abnormal slice quantity of the same pre-inspected engineering data is determined. If the abnormal slice quantity is equal to 0, it means that the pre-inspected engineering data is qualified, which is recorded as the single acceptance qualification. If the abnormal slice quantity is not equal to 0, it means that the pre-inspected engineering data is unqualified, which is recorded as the single acceptance failure.

[0137] Conduct safety verification on centralized procurement points based on the individual acceptances obtained, and obtain qualified acceptance nodes and abnormal acceptance nodes;

[0138] For the same centralized procurement point, if each item of pre-inspection engineering data passes the single acceptance, then the centralized procurement point is a qualified acceptance node; if not all items of pre-inspection engineering data pass the single acceptance, then the centralized procurement point is an abnormal acceptance node.

[0139] Based on the above-mentioned UAV intelligent acceptance system for power transmission projects, the present invention also provides a UAV intelligent acceptance method for power transmission projects, comprising the following steps:

[0140] Step 1: pre-process the collected pre-test engineering data to obtain three-dimensional video data and thermal red geographic data;

[0141] Step 2: Perform frame-to-image conversion and continuous combination on the power transmission video data to obtain a key frame set, set a multilateral constraint criterion to constrain the key frame set and three-dimensional image data to obtain a constrained engineering image, set a limited selection check based on the multilateral constraint criterion to classify and extract the constrained engineering image, and obtain an image weight feature sequence;

[0142] Step 3: Perform mode conversion on the image weight feature sequence to obtain image weight variables, set constraint weighting coefficients to perform cardinality separation on the image weight variables to obtain separated image weight coefficients, perform form switching and data truncation on the separated image weight coefficients to obtain image weight slices, perform homomorphic permutation on the hot red geographic data to obtain hot red geographic slices, and record the image weight slices and hot red geographic slices as engineering test slices;

[0143] Step 4: Perform homomorphic replacement and dynamic correction on the acceptance standard data to obtain the corrected verification collection, perform project matching on the engineering slices to be tested to obtain the matched corrected verification collection, accept and judge the engineering slices to be tested through the matched corrected verification collection to obtain the judged slices, perform node judgment on the centralized procurement points through the judged slices, and obtain qualified acceptance nodes and abnormal acceptance nodes.

[0144] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An unmanned aerial vehicle intelligent acceptance system for power transmission projects, comprising a control center, characterized in that: The control center is connected to an engineering acquisition module, a power transmission processing module, an intelligent analysis module and a comprehensive acceptance module; The engineering acquisition module is used to collect pre-inspection engineering data, pre-process the pre-inspection engineering data, and obtain three-dimensional video data and thermal red geographic data; The power transmission processing module is used to perform frame-to-image conversion and continuous combination on the power transmission video data to obtain a key frame set, set a multilateral constraint criterion to constrain the key frame set and the three-dimensional image data to obtain a constrained engineering image, set a limited selection check according to the multilateral constraint criterion to classify and extract the constrained engineering image, and obtain an image weight feature sequence; The intelligent analysis module is used to perform mode conversion on the image weight feature sequence to obtain the image weight variable, set the constraint weight coefficient to perform cardinality separation on the image weight variable to obtain the separated image weight coefficient, perform form switching and data truncation on the separated image weight coefficient to obtain the image weight slice, perform homomorphic permutation on the hot red geographic data to obtain the hot red geographic slice, and record the image weight slice and the hot red geographic slice as the engineering test slice; The comprehensive acceptance module is used to perform homomorphic replacement and dynamic correction on the acceptance standard data, obtain the corrected verification collection, perform project matching on the engineering slices to be tested, obtain the matched corrected verification collection, accept and judge the engineering slices to be tested through the matched corrected verification collection, obtain the judged slices, perform node judgment on the centralized procurement points through the judged slices, and obtain qualified acceptance nodes and abnormal acceptance nodes.

2. The UAV intelligent acceptance system for power transmission engineering according to claim 1 is characterized in that: The process of preprocessing the pre-test engineering data includes: Set up a front-end collection terminal, collect criteria through the front-end collection terminal, and obtain acceptance standard data; Set deployment criteria according to the power transmission project, intelligently deploy drones according to the deployment criteria, obtain deployment unmanned terminals, conduct safe collection of power transmission projects through the deployment unmanned terminals, obtain pre-verified project data, and mark the location where the deployment unmanned terminals conduct safe collection of power transmission projects as the collection point; Perform digital image summarization on the pre-test engineering data to obtain picture video data and thermal red geographic data, perform image conversion on the picture video data to obtain three-dimensional video data, and the three-dimensional video data includes three-dimensional image data and power transmission video data.

3. The UAV intelligent acceptance system for power transmission projects according to claim 2 is characterized in that: The process of obtaining a constraint engineering image includes: Perform frame-to-image conversion on the power transmission video data to obtain key video frames; Continuously combining the obtained key video frames based on the order of frame-image conversion to obtain a key frame set, and marking the key frame set and the three-dimensional image data as an engineering preprocessing image; A multilateral constraint criterion is set, and the engineering preprocessing image is constrained and intercepted according to the multilateral constraint criterion to obtain a constrained engineering image.

4. The UAV intelligent acceptance system for power transmission engineering according to claim 3 is characterized in that: The process of obtaining the classified extracted block core includes: A limited selection kernel is set according to the multilateral constraint criteria, an extraction starting point is set for the constrained engineering image, the obtained limited selection kernel is uploaded to the constrained engineering image based on the extraction starting point, and the corresponding range of the limited selection kernel in the constrained engineering image is marked as a constrained area; According to the limited selection core constraint engineering image is classified and extracted to obtain the image weight feature sequence.

5. The UAV intelligent acceptance system for power transmission projects according to claim 4 is characterized in that: The process of classifying and extracting engineering images based on limited selection and verification constraints includes: The obtained limited selection kernel is matched and convolved with the constraint area at the extraction starting point to obtain the classification extraction block kernel; The shift spacing is set according to the multilateral constraint criteria, and the limited selection kernel is translated and slid according to the obtained shift spacing based on the constrained engineering image to reach the next constrained area. The limited selection kernel is matched and convolved with the constrained area at the extraction starting point until all areas of the constrained engineering image are covered, and the obtained classified extraction block kernel is weight-checked to obtain the block kernel weight; The block core weights are repeatedly screened based on the order of translation and sliding to obtain the block replacement weight values, and the block replacement weight values ​​are sequentially combined based on the order of translation and sliding to obtain the image weight feature sequence.

6. The UAV intelligent acceptance system for power transmission engineering according to claim 5 is characterized in that: The process of obtaining image weight slices includes: Performing mode conversion on the obtained image weight feature sequence to obtain image weight variables; The constrained weight coefficient is set according to the limited selected kernel, and the cardinality separation of the image weight variable is performed through the constrained weight coefficient to obtain the separated image weight coefficient; Switching the form of the separated image weight coefficient to obtain an image weight data stream, and performing data truncation on the obtained image weight data stream to obtain an image weight slice; Perform homomorphic replacement on the obtained hot red geographic data to obtain hot red geographic slices; The obtained hot red geographic slices and image weight slices are marked as engineering test slices.

7. The UAV intelligent acceptance system for power transmission engineering according to claim 6 is characterized in that: The process of isomorphic replacement of the obtained hot red geographic data includes: Performing mode conversion on the hot red geographic data to obtain hot red geographic variables, performing cardinality separation on the hot red geographic variables by constraining weighted coefficients to obtain separated hot red geographic coefficients; The separated hot red geographic coefficients are switched in form to obtain a hot red geographic data stream, and the hot red geographic data stream is truncated to obtain hot red geographic slices.

8. The UAV intelligent acceptance system for power transmission engineering according to claim 7 is characterized in that: The process of obtaining qualified acceptance nodes and abnormal acceptance nodes includes: Perform homomorphic permutation on the obtained acceptance standard data to obtain acceptance data slices; The obtained acceptance data slices are combined into fragments to obtain an acceptance standard set, and the acceptance standard set is dynamically revised to obtain a revised acceptance set; Obtain the pre-inspected engineering data corresponding to the engineering slice to be inspected, record it as the to-be-inspected matching data, perform project matching on the acceptance standard data according to the to-be-inspected matching data, and obtain the matching correction inspection collection; Perform acceptance verification on the to-be-verified matching data according to the matching correction verification collection to obtain a verification matching set; The result of the verification matching set is judged to obtain the judgment slice, and the node judgment of the pre-inspection engineering data is performed based on the judgment slice based on the centralized procurement point to obtain the qualified acceptance node and the abnormal acceptance node.

9. The UAV intelligent acceptance system for power transmission engineering according to claim 8, characterized in that: The process of determining the nodes of the centralized procurement points by determining the slices includes: Obtaining the engineering slices to be tested corresponding to the pre-tested engineering data, and performing statistics on the results of the obtained engineering slices to be tested to obtain slice statistics; According to the slice statistics, the centralized procurement points are assessed for safety and single-item acceptance is obtained. Based on the single-item acceptance, the centralized procurement points are security-verified to obtain qualified acceptance nodes and abnormal acceptance nodes.

10. The unmanned aerial vehicle intelligent acceptance method of the unmanned aerial vehicle intelligent acceptance system for power transmission engineering according to any one of claims 1 to 9, characterized in that: The following steps are involved: Step 1: Preprocess the collected pre-test engineering data to obtain three-dimensional video data and thermal red geographic data; Step 2: Perform frame-to-image conversion and continuous combination on the power transmission video data to obtain a key frame set, set a multilateral constraint criterion to constrain the key frame set and three-dimensional image data to obtain a constrained engineering image, set a limited selection check based on the multilateral constraint criterion to classify and extract the constrained engineering image, and obtain an image weight feature sequence; Step 3: Perform mode conversion on the image weight feature sequence to obtain image weight variables, set constraint weighting coefficients to perform cardinality separation on the image weight variables to obtain separated image weight coefficients, perform form switching and data truncation on the separated image weight coefficients to obtain image weight slices, perform homomorphic permutation on the hot red geographic data to obtain hot red geographic slices, and record the image weight slices and hot red geographic slices as engineering test slices; Step 4: Perform homomorphic replacement and dynamic correction on the acceptance standard data to obtain the corrected verification collection, perform project matching on the engineering slices to be tested to obtain the matched corrected verification collection, accept and judge the engineering slices to be tested through the matched corrected verification collection to obtain the judged slices, perform node judgment on the centralized procurement points through the judged slices, and obtain qualified acceptance nodes and abnormal acceptance nodes.

Images