Overhead transmission line multi-source data fusion method and device, computer equipment, storage medium and computer program product

Abstract

The invention relates to an overhead transmission line multi-source data fusion method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring multi-source data of an overhead transmission line; performing feature extraction on the multi-source data according to data types to obtain multi-type feature data; performing fusion processing on the multiple types of feature data through a pixel-level fusion algorithm, a feature-level fusion algorithm and a decision-level fusion algorithm to obtain fusion data; wherein the fusion data is used for any one of operation monitoring, scheduling control, planning design and intelligent inspection of the overhead transmission line. By adopting the method, the fusion precision can be improved.

Description

Technical Field

[0001] This application relates to the technical field of power system automation, and particularly to a multi-source data fusion method, device, computer device, storage medium, and computer program product for overhead transmission lines. Background Art

[0002] In traditional technologies, the multi-source data acquisition of overhead transmission lines relies on various devices such as satellite maps, unmanned aerial vehicle (UAV) aerial survey images, lidar, Global Positioning System (GPS), and total stations. When traditional technologies utilize the fusion of these multi-source data, it is difficult to ensure a high-precision fusion effect. Summary of the Invention

[0003] Based on this, in view of the above technical problems, it is necessary to provide a multi-source data fusion method, device, computer device, computer-readable storage medium, and computer program product for overhead transmission lines that can improve the fusion accuracy.

[0004] In a first aspect, this application provides a multi-source data fusion method for overhead transmission lines, including:

[0005] Obtaining multi-source data of the overhead transmission line;

[0006] Performing feature extraction on the multi-source data according to the data type to obtain multi-type feature data;

[0007] Performing fusion processing on the multi-type feature data through a pixel-level fusion algorithm, a feature-level fusion algorithm, and a decision-level fusion algorithm to obtain fusion data;

[0008] Wherein, the fusion data is used for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

[0009] In one embodiment, the performing feature extraction on the multi-source data according to the data type to obtain multi-type feature data includes:

[0010] Performing data processing on the multi-source data to obtain processed data; the data processing includes at least one of format conversion, noise removal, coordinate system unification, data cropping, and registration;

[0011] Performing feature extraction on the processed data according to the data type to obtain multi-type feature data.

[0012] In one embodiment, the data processing includes the coordinate system unification, and the performing data processing on the multi-source data to obtain processed data includes:

[0013] Convert the multi-source data to the WGS84 coordinate system and establish the spatial correspondence between different data through spatial registration to obtain data with unified coordinates and spatial correlation.

[0014] In one embodiment, the multi-type feature data includes at least two of the following feature data:

[0015] The texture features, color features, and shape features of the image data;

[0016] The spatial distribution features, density features, and elevation change features of the lidar point cloud data;

[0017] The GPS coordinate features and time features corresponding to the coordinates of the GPS data;

[0018] The terrain undulation features, slope features, and aspect features of the elevation data;

[0019] The object contour features, type features, and location features of the topographic map;

[0020] The measurement point coordinate features, distance features, and angle features of the total station data;

[0021] The three-dimensional building structure features and texture features of the oblique data;

[0022] The planning area features, line features, and control point features of the planning data;

[0023] The line trend features, substation location features, and voltage level features of the power grid thematic data.

[0024] In one embodiment, the multi-type feature data is fused through a pixel-level fusion algorithm, a feature-level fusion algorithm, and a decision-level fusion algorithm to obtain fused data, including:

[0025] For the texture features, color features, and shape features of the image data, as well as the three-dimensional building structure features and texture features of the oblique data, a pixel-level fusion algorithm is used for fusion processing to obtain fused image data;

[0026] The fused image data and the first data are fused through a feature-level fusion algorithm to obtain a fused feature vector;

[0027] The fused vector feature and the second data are fused through a decision-level fusion algorithm to obtain the final fused data.

[0028] Among them, the first data includes at least one of the following:

[0029] The spatial distribution features, density features, and elevation change features of the lidar point cloud data;

[0030] The coordinate characteristics, distance characteristics, and angle characteristics of the measurement points of the total station data;

[0031] The second data includes at least one of the following:

[0032] The GPS coordinate characteristics of the GPS data and the time characteristics corresponding to the coordinates;

[0033] The terrain undulation characteristics, slope characteristics, and aspect characteristics of the elevation data;

[0034] The ground object contour characteristics, type characteristics, and location characteristics of the topographic map;

[0035] The coordinate characteristics, distance characteristics, and angle characteristics of the measurement points of the total station data;

[0036] The planned area characteristics, line characteristics, and control point characteristics of the planning data;

[0037] The line trend characteristics, substation location characteristics, and voltage level characteristics of the power grid special data.

[0038] In one embodiment, after the multi-type feature data is fused by pixel-level fusion algorithm, feature-level fusion algorithm, and decision-level fusion algorithm to obtain fused data, the method further includes:

[0039] Evaluating the quality of the fused data to obtain evaluation indicators;

[0040] Based on the evaluation indicators, performing error correction and data optimization on the fused data to obtain optimized data;

[0041] Using the optimized data for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

[0042] In a second aspect, the present application also provides an overhead transmission line multi-source data fusion device, including:

[0043] An acquisition module, configured to acquire multi-source data of the overhead transmission line;

[0044] An extraction module, configured to extract features from the multi-source data according to the data type to obtain multi-type feature data;

[0045] A fusion module, configured to fuse the multi-type feature data by pixel-level fusion algorithm, feature-level fusion algorithm, and decision-level fusion algorithm to obtain fused data;

[0046] Wherein, the fused data is used for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

[0047] In a third aspect, the present application further provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0048] In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in the first aspect is implemented.

[0049] In a fifth aspect, the present application further provides a computer program product, including a computer program. When the computer program is executed by a processor, the method described in the first aspect is implemented.

[0050] The above multi-source data fusion method, device, computer device, storage medium and computer program product for overhead transmission lines obtain multi-source data of overhead transmission lines; extract features from the multi-source data according to data types to obtain multi-type feature data; perform fusion processing on the multi-type feature data through pixel-level fusion algorithm, feature-level fusion algorithm and decision-level fusion algorithm to obtain fusion data; wherein, the fusion data is used for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line. Through this solution, first, multi-source data is obtained, widely collecting various relevant information, providing a rich data basis for subsequent fusion. Then, feature extraction is performed according to data types, accurately mining key features from different data to obtain multi-type feature data, ensuring that the effective information of the original data is fully extracted. The most crucial thing is to use pixel-level, feature-level and decision-level fusion algorithms to perform fusion processing on multi-type feature data. Pixel-level fusion directly processes the original feature data, retaining the most detailed information; feature-level fusion performs fusion after feature extraction, further integrating effective features and enhancing the representativeness of the data; decision-level fusion fuses from the independent decision results of each sensor, comprehensively making judgments from multiple aspects, reducing the error and uncertainty of a single decision. Through the progressive and complementary relationship of these three fusion algorithms, multi-type feature data is comprehensively and deeply fused. The finally obtained fusion data is more accurate and comprehensive, and can effectively improve the application effects of the fusion data of overhead transmission lines in various aspects such as operation monitoring, dispatching control, planning and design, and intelligent inspection. Description of the Drawings

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0052] Figure 1 It is a schematic flow chart of a multi-source data fusion method for an overhead transmission line in an embodiment;

[0053] Figure 2 It is a schematic flow chart of a fusion processing process in an embodiment;

[0054] Figure 3 It is a schematic flow chart of a post-fusion processing and optimization process in an embodiment;

[0055] Figure 4 It is a structural block diagram of a multi-source data fusion device for an overhead transmission line in an embodiment;

[0056] Figure 5 It is an internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0057] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0058] With the advancement of the intelligent and digital transformation of the power grid, the demand for refined management and efficient operation and maintenance of overhead transmission lines is becoming increasingly urgent. However, in the prior art, the heterogeneity, dispersion of multi-source data and the limitations of fusion methods have led to low efficiency in the comprehensive utilization of data, making it difficult to support the full life cycle management of transmission lines. For example, image data, laser point clouds, GPS data, etc. are stored independently, lacking a unified fusion framework and unable to provide high-precision and multi-dimensional decision-making support for services such as operation monitoring and intelligent inspection. Based on this, the present application proposes a multi-source data fusion method for overhead transmission lines, which effectively improves the data value through multi-level fusion strategies and optimization processing, and provides a reliable data basis for power grid services.

[0059] The multi-source data fusion method for overhead transmission lines proposed in this application is executed by a computer device with data processing capabilities. In actual application scenarios, the computer device with data processing capabilities that executes the multi-source data fusion method for overhead transmission lines proposed in this application can be of various types. For example, it may be a high-performance server set up in a power dispatching center. Such a server has powerful computing capabilities and large-capacity storage, can efficiently process a large amount of real-time collected overhead transmission line data, and achieve rapid fusion of multi-source data, providing timely and accurate data support for power dispatching decisions. It may also be industrial computers distributed at monitoring stations along the transmission line. They can perform preliminary processing and fusion of data such as line status and environmental parameters collected by various sensors at locations close to the data source, reducing data transmission pressure and ensuring the timeliness and stability of data processing. In addition, in some small-scale power operation and maintenance scenarios with a high degree of intelligence, the computer device may also be a portable intelligent terminal with certain data processing capabilities, such as a ruggedized tablet computer, which facilitates on-site operation and maintenance personnel to quickly perform fusion analysis on the collected multi-source data and promptly discover and handle line problems.

[0060] As Figure 1 shown in the figure, this embodiment provides a schematic flowchart of a multi-source data fusion method for overhead transmission lines, and the method includes the following steps:

[0061] 101. Obtain multi-source data of the overhead transmission line.

[0062] Among them, the above multi-source data may refer to data obtained through various different methods.

[0063] Among them, the multi-source data includes but is not limited to:

[0064] 1) Image data: High-resolution visible light, infrared, or radar images obtained by drones or satellites;

[0065] 2) LiDAR point cloud data: Three-dimensional point cloud data collected by Light Detection and Ranging (LiDAR);

[0066] 3) GPS data: Real-time coordinates and timestamp information of transmission line towers and key points of conductors;

[0067] 4) Elevation data: Digital Elevation Model (DEM) or Digital Terrain Model (DTM);

[0068] 5) Topographic map: Vectorized topographic map data, including ground object contours, types, and location information;

[0069] 6) Total station data: parameters such as the sag of the conductor and the inclination angle of the tower measured by the total station;

[0070] 7) Inclination data: 3D structure and texture information of buildings obtained by oblique photogrammetry;

[0071] 8) Planning data: information on the planned path of the transmission line, control points, and avoidance areas;

[0072] Power grid special data: data related to power grid operations such as the line direction, substation location, and voltage level.

[0073] 102. Extract features from the multi-source data according to the data type to obtain multi-type feature data.

[0074] In some embodiments, extracting features from the multi-source data according to the data type to obtain multi-type feature data includes: performing data processing on the multi-source data to obtain processed data; the data processing includes at least one of format conversion, noise removal, coordinate system unification, data clipping, and registration; extracting features from the processed data according to the data type to obtain multi-type feature data.

[0075] In some embodiments, the multi-type feature data includes at least two of the following feature data:

[0076] Texture features, color features, and shape features of image data;

[0077] Spatial distribution features, density features, and elevation change features of laser point cloud data;

[0078] GPS coordinate features and time features corresponding to the coordinates of GPS data;

[0079] Terrain undulation features, slope features, and aspect features of elevation data;

[0080] Feature of ground object contour, type feature, and location feature of topographic map;

[0081] Measuring point coordinate features, distance features, and angle features of total station data;

[0082] 3D structure features and texture features of inclination data;

[0083] Planning area features, line features, and control point features of planning data;

[0084] Line direction features, substation location features, and voltage level features of power grid special data.

[0085] The process of extracting features from the multi-source data according to the data type specifically includes the following steps 1021 and 1022:

[0086] 1021. Data preprocessing.

[0087] During the acquisition process of multi-source data, due to factors such as equipment and environment, there are often problems such as inconsistent formats, noise interference, coordinate system differences, and data redundancy. In order to improve data quality and facilitate subsequent feature extraction and fusion processing, a series of data processing operations need to be performed on the multi-source data.

[0088] In some embodiments, the above data processing includes coordinate system unification. The data processing of the multi-source data can be to convert the multi-source data to the World Geodetic System 1984 (WGS84) coordinate system and establish a spatial correspondence between different data through spatial registration to obtain data with unified coordinates and spatial association.

[0089] The above standardization processing of the multi-source data includes, but is not limited to:

[0090] (1) Format conversion: Convert vector data to a general format (such as Shapefile), and raster data to the Georeferenced Tagged Image File Format (GeoTIFF) format. The data formats generated by different data acquisition devices and systems are various. For example, image data may have formats such as Joint Photographic Experts Group (JPEG) and Tagged Image File Format (TIFF), and vector data may have formats such as Shapefile (SHP) and Geography Markup Language (GML). In order to unify the data format and facilitate subsequent processing, special data format conversion tools are used to convert various vector data to a general vector format, such as Shapefile format; convert raster data to a unified raster format, such as GeoTIFF format. During the conversion process, ensure the integrity and accuracy of the data and avoid data loss or damage.

[0091] Among them, Shapefile is a file format developed by the Environmental Systems Research Institute in the United States for storing geographic vector data (such as points, lines, polygons). GeoTIFF adds spatial reference information such as a geographic coordinate system on the basis of an image file and is a file format for storing raster data with geospatial information.

[0092] (2)Noise removal: Median filtering is used to reduce the noise of the image data, and outlier removal is performed on the laser point cloud.

[0093] Since noise inevitably exists in the data, its presence will affect the quality of the data and the accuracy of the analysis results. Methods such as filtering are used to remove the noise in the data. For example, for image data, algorithms such as Gaussian filtering are used for noise reduction. By performing weighted averaging on each pixel point in the image and its neighboring pixel points, the image is effectively smoothed, the interference of noise is reduced, and the clarity and readability of the image data are improved.

[0094] (3)Coordinate system unification: All data is converted to the WGS84 coordinate system, and a spatial correspondence is established through spatial registration.

[0095] Multi-source data may come from different measurement systems or geographic information systems, and there are differences in their coordinate systems. All data is converted to the same geographic coordinate system and projection coordinate system, such as being uniformly converted to the WGS84 coordinate system. Specifically, through professional geographic information processing software, multi-source data is converted to the WGS84 coordinate system, and a spatial registration algorithm is used to establish the spatial correspondence between different data, ensuring the consistency of the data in terms of spatial position, and obtaining data with unified coordinates and spatial association.

[0096] (4)Data cropping and registration: Crop the data based on the transmission line corridor range and register different data sources using control points.

[0097] According to the research area range of the overhead transmission line, the data is cropped to remove the irrelevant data parts and reduce the data processing volume. At the same time, through spatial registration technology, such as the registration method based on feature point matching, the spatial correspondence between different data is established, enabling the data from different data sources to be analyzed and fused in the same spatial framework.

[0098] Through the above data preprocessing steps, the quality and usability of multi-source data have been significantly improved. Format conversion unifies data from different sources into a common format, eliminating data processing obstacles caused by format differences, enabling subsequent data feature extraction and fusion operations to be carried out in a unified environment, and greatly improving the processing efficiency. The noise removal operation effectively reduces the interference information in the data, improves the clarity of image data and the accuracy of lidar point cloud data, providing a more reliable data basis for subsequent feature extraction. Coordinate system unification and spatial registration ensure the consistency of data from different data sources in terms of spatial position, establishing an accurate spatial correspondence relationship, enabling multi-source data to be analyzed and fused within the same spatial framework, avoiding fusion errors caused by inconsistent coordinates, and enhancing the accuracy and reliability of data fusion. Data cropping and registration remove irrelevant data parts, reduce the amount of data processing, improve the processing speed, and at the same time enable data from different data sources to be better matched and fused, further enhancing the effect of data fusion.

[0099] 1022. Feature extraction.

[0100] For the data after data processing, according to the characteristics of its data type, corresponding algorithms and technologies are used for feature extraction to obtain multi-type feature data.

[0101] Exemplarily, extracting key features for different data types may include:

[0102] Image data: Extract texture features, color features, and shape features. Among them, image data contains rich ground object information. By using edge detection algorithms, the edge features of ground objects in the image are extracted to obtain the texture features of the image; the color histogram is used to count the distribution of different colors in the image to extract color features; shape features are extracted through shape descriptor algorithms.

[0103] Lidar point cloud data: Extract spatial distribution features, density features, and elevation change features. Among them, lidar point cloud data can accurately reflect the three-dimensional spatial information of ground objects. Calculate the normal vector of the point cloud to describe the local surface direction of the point cloud and extract spatial distribution features; count the density of the point cloud to obtain density features; extract elevation change features by analyzing the elevation value changes of the point cloud. GPS data: Extract coordinate features and their corresponding time series features.

[0104] Elevation data: Use terrain analysis tools to extract slope, aspect, and terrain relief features. Elevation data is used to describe the undulation of the terrain. Through digital elevation model analysis technology, the slope and aspect of the terrain can be calculated to extract terrain relief features, slope features, and aspect features.

[0105] Topographic map: Extract the type, contour, and location features of ground objects based on vector data. The topographic map records various information of ground objects in detail. Through graphic recognition algorithms, the contour features of ground objects are extracted; according to the symbols and annotations of ground objects, the types of ground objects are identified and the type features are extracted; using coordinate information, the positions of ground objects are determined and the location features are extracted.

[0106] Total station data: Extract the coordinates, distances, and angle features of traverse survey points. The total station data are accurate data obtained through on-site measurement. The coordinate features of the measurement points are directly extracted; according to the measured distance and angle information, the distance features and angle features are extracted.

[0107] Oblique data: Extract the building structure features and surface textures through 3D reconstruction. Through 3D reconstruction technology, a 3D model of the building is constructed to extract the 3D structure features of the building; using image processing technology, the texture features of the building are extracted.

[0108] Planning data: Extract the boundaries of the planning area, the route paths, and the coordinates of control points. Extract the planning area features, route features, and control point features from planning drawings or electronic documents to provide a basis for the planning and design of routes.

[0109] Power grid special data: Extract the vector features of the line routes and the topological relationships of substations. The power grid special data mainly involve the relevant information of the power grid. Extract the line route features, substation location features, and voltage level features of the power grid for the operation monitoring and dispatching control of the power grid.

[0110] The feature extraction operations performed on different types of data effectively mine key information from multi-source data. The extraction of texture, color, and shape features of image data enables more accurate identification of the appearance and features of ground objects, providing an important basis for subsequent ground object classification and identification. The extraction of spatial distribution, density, and elevation change features of laser point cloud data accurately reflects the 3D spatial information of ground objects, contributing to the 3D modeling and analysis of the terrain and objects around transmission lines. The feature extraction of GPS data, elevation data, topographic map data, total station data, oblique data, planning data, and power grid special data, etc., obtains important information related to transmission lines from different angles, providing rich feature data for subsequent data fusion and applications. Through feature extraction, the original multi-source data are transformed into representative and distinguishable feature data, greatly improving the value and usability of the data, and providing a solid foundation for subsequent data fusion and applications such as operation monitoring, dispatching control, planning design, and intelligent inspection of transmission lines.

[0111] 103. Perform fusion processing on multi-type feature data through pixel-level fusion algorithms, feature-level fusion algorithms, and decision-level fusion algorithms to obtain fusion data.

[0112] The feature data is fused using pixel-level fusion algorithms, feature-level fusion algorithms, and decision-level fusion algorithms.

[0113] The above pixel-level fusion algorithms may include at least one of the following:

[0114] Weighted average method: Weighted fusion can be performed on the elevation information of image data and laser point clouds, and the weights are dynamically adjusted based on data confidence.

[0115] Intensity - Hue - Saturation (IHS) transformation method: The RGB space of the multispectral image is converted to the IHS space, the intensity component is replaced with the elevation information of the laser point cloud, and a fused image is generated after inverse transformation.

[0116] Principal component analysis (PCA): Principal component analysis is performed on multi-source images, and the secondary components are replaced to enhance feature expression.

[0117] The above feature-level fusion algorithms may include at least one of the following:

[0118] Scale-invariant feature transform (SIFT) algorithm: Scale-invariant feature points of images and point clouds can be extracted, and spatial alignment is achieved through descriptor matching.

[0119] Harris-Affine method: A matching method that combines affine invariant features.

[0120] Spatial structure matching: Utilize the linear features of transmission lines (such as conductors and towers) to match the geometric features of total station data and images.

[0121] The above decision-level fusion algorithms may include at least one of the following:

[0122] Voting method: Vote on the land cover classification results of multi-source data to determine the final category (such as vegetation, buildings).

[0123] Bayesian method: Combine the temporal features of GPS data and the topographic features of elevation data to calculate the optimal estimated value of the conductor sag.

[0124] Dempster - Shafer evidence theory (D - S evidence theory): Fuse the uncertainty information of planning data and power grid thematic data to optimize the line path decision.

[0125] In some embodiments, by way of example, as Figure 2 shown, is a schematic flowchart of a fusion processing procedure, which may include the following steps:

[0126] 201. For the texture features, color features, and shape features of the image data, as well as the three-dimensional building structure features and texture features of the oblique data, a pixel-level fusion algorithm is used for fusion processing to obtain fused image data.

[0127] Exemplarily, the weighted average method is adopted. The gray values or other values of the corresponding pixels of the image data and the oblique data participating in the fusion are weighted and summed according to certain weights to obtain the value of the fused pixel. In practical applications, the weights are determined according to factors such as the reliability and resolution of the image data and the oblique data. For example, if the resolution of the image data is higher and the expression of the texture features is more accurate, a higher weight is given to the image data; if the oblique data describes the three-dimensional structure of the building in more detail, the corresponding weight is given to the oblique data. By reasonably setting the weights and weighted-averaging the pixel values, the fused image retains both the rich texture and color information of the image data and incorporates the three-dimensional building structure features of the oblique data.

[0128] Exemplarily, the IHS transformation method is adopted. The color image is transformed from the RGB space to the IHS space, and the intensity I, hue H, and saturation S are separated. The relevant features of the oblique data are used to replace the I component, and then it is inverse-transformed back to the RGB space to achieve data fusion. This method can make the fused image retain both the color information of the original image and incorporate the features of the oblique data, enhancing the expression ability of the image for targets such as buildings.

[0129] Exemplarily, the principal component transformation method (PCA) is adopted. Principal component analysis is performed on the matrix composed of the image data and the oblique data, and the data is transformed to a new principal component space. The information of the oblique data is used to replace some principal components, and then it is inverse-transformed back to the original space to achieve the fusion purpose. The PCA method can effectively extract the main features of the data, remove the correlation between the data, make the fused image data more concise and effective, and highlight the texture and three-dimensional structure features of the building.

[0130] 202. The fused image data and the first data are fused using a feature-level fusion algorithm to obtain a fused feature vector.

[0131] Among them, the first data includes at least one of the following:

[0132] The spatial distribution features, density features, and elevation change features of the laser point cloud data;

[0133] The measurement point coordinate features, distance features, and angle features of the total station data.

[0134] Exemplarily, the SIFT (Scale-Invariant Feature Transform) algorithm is adopted. In the fused image data and laser point cloud data, etc., the SIFT algorithm is used to detect and extract feature points with scale invariance and rotation invariance, and the descriptors of the feature points are calculated. Feature-level fusion is achieved by matching the descriptors. This method is often used in image matching and the registration of point cloud and image, etc. For example, when monitoring the poles and towers of overhead transmission lines, the SIFT algorithm is used to match the pole and tower feature points in the fused image data and laser point cloud data to achieve accurate registration of the two, so as to obtain more comprehensive information about the poles and towers.

[0135] Exemplarily, the Harris - Affine affine invariant feature matching method is adopted. First, the Harris corner detection algorithm is used to extract corners, then the corner neighborhood is normalized through affine transformation to make it have affine invariance, and then feature matching is carried out. This method can achieve high-precision and high-stability wide-baseline image matching, and has a good effect on the fusion of image data at different perspectives and scales. When processing the image data of overhead transmission lines taken at different times, the Harris - Affine method can accurately match the feature points and provide a reliable basis for the change monitoring of the line.

[0136] Exemplarily, a line matching method based on spatial structure information is adopted. The line features in the fused image or point cloud are extracted, and the spatial position, direction, length and other structure information of the lines are used for matching and fusion. This method is applicable to ground objects with obvious line features, such as the poles and towers and conductors of transmission lines. By matching the line features, the position and orientation of the transmission line can be determined more accurately, and the accuracy of data fusion can be improved.

[0137] 203. The fused vector feature and the second data are fused and processed by a decision-level fusion algorithm to obtain the final fused data.

[0138] Among them, the second data includes at least one of the following:

[0139] The GPS coordinate feature of GPS data and the time feature corresponding to the coordinate;

[0140] The terrain undulation feature, slope feature, and aspect feature of elevation data;

[0141] The ground object contour feature, type feature, and position feature of topographic maps;

[0142] The measurement point coordinate feature, distance feature, and angle feature of total station data;

[0143] The planned area feature, line feature, and control point feature of planning data;

[0144] The line routing characteristics, substation location characteristics, and voltage level characteristics of the power grid special topic data.

[0145] Exemplarily, the voting method is adopted. Each data source independently conducts classification or decision-making, and the decision results of each data source are statistically voted. The final fusion decision is determined according to the voting results. For example, in the judgment of the ground object types around the overhead transmission line, the image data, topographic map data, and laser point cloud data respectively give their own judgment results, and the final ground object type is determined through majority voting. This method is simple and intuitive and is suitable for scenarios with high requirements for decision-making speed.

[0146] Exemplarily, the Bayesian method is adopted. Using Bayes' theorem, combined with the prior probabilities and conditional probabilities of each data source, the posterior probability of the fused result is calculated, and the decision is made according to the posterior probability. In practical applications, the probability distribution models of each data source and the statistical relationships between the data need to be known. For example, when predicting the fault probability of an overhead transmission line, combining multiple data sources such as historical fault data, real-time monitoring data, and environmental data, the Bayesian method is used to calculate the posterior probability of the fault occurrence, providing a scientific basis for the operation and maintenance of the line.

[0147] Exemplarily, the D - S evidence theory method is adopted. The information of each data source is regarded as evidence, and by calculating the basic probability assignment function, belief function, and likelihood function of the evidence, etc., the evidence is combined and fused, and the decision is made according to the fused evidence. This method can handle uncertain and inaccurate information and has strong robustness in data fusion. When dealing with complex power system data, the D - S evidence theory method can effectively fuse the information of multiple data sources, improving the accuracy and reliability of decision-making.

[0148] The SIFT algorithm, Harris - Affine method, and spatial structure matching method in the above feature-level fusion algorithms can accurately match the feature points and structure information of different data sources, realizing the spatial alignment and feature integration of the data. The SIFT algorithm extracts the scale-invariant feature points of the image and point cloud, and uses descriptor matching to achieve spatial alignment, which is commonly used in image matching and point...

[0149] The above multi-source data fusion method for overhead transmission lines first obtains multi-source data by widely collecting various relevant information, providing a rich data basis for subsequent fusion. Then, feature extraction is carried out according to the data type, accurately mining key features from different data to obtain multi-type feature data, ensuring that the effective information of the original data is fully extracted. The most crucial is to use pixel-level, feature-level, and decision-level fusion algorithms to fuse multi-type feature data. Pixel-level fusion directly processes the original feature data, retaining the most detailed information; feature-level fusion fuses after feature extraction to further integrate effective features and enhance the representativeness of the data; decision-level fusion fuses from the independent decision results of each sensor, comprehensively judging from multiple aspects to reduce the error and uncertainty of a single decision. Through the progressive and complementary relationship of these three fusion algorithms, multi-type feature data is comprehensively and deeply fused. The finally obtained fusion data is more accurate and comprehensive, and can effectively improve the application effects of the fusion data of overhead transmission lines in aspects such as operation monitoring, dispatching control, planning and design, and intelligent inspection.

[0150] After the multi-type feature data is fused by the pixel-level fusion algorithm, feature-level fusion algorithm, and decision-level fusion algorithm to obtain the fusion data, post-fusion processing and optimization can be carried out.

[0151] Exemplarily, as Figure 3 shown, it is a schematic flowchart of a post-fusion processing and optimization process, and this process can include the following steps:

[0152] 301. Evaluate the quality of the fusion data to obtain evaluation indicators.

[0153] Among them, calculate the quantitative indicators of the fusion data (such as information entropy, structural similarity index), and evaluate the fusion effect through visual inspection.

[0154] Exemplarily, quantitative and qualitative methods can be used to evaluate the quality of the fused data. Quantitative evaluation calculates indicators such as the information entropy and average gradient of the fused image to evaluate the information content and clarity of the fusion data. The larger the information entropy, the richer the information content of the fusion data; the larger the average gradient, the higher the clarity of the fusion data. Qualitative evaluation visually inspects the fusion effect to observe whether the fusion data retains the main features of the original data and whether there are obvious distortion or noise problems.

[0155] 302. Based on the evaluation indicators, correct the errors of the fusion data.

[0156] Correct registration deviation or classification error, for example, use total station data to correct the elevation error of the laser point cloud. Exemplarily, correct and repair the errors existing in the fused data according to the quality evaluation result. For example, if a registration error is found in the fused data, use an image registration algorithm to adjust the error; for the missing or incorrect parts of the data, use other relevant data to supplement and correct. Through error correction, improve the accuracy and reliability of the fused data.

[0157] Among them, according to the indicators obtained from the quality evaluation, for problems such as registration deviation and classification error existing in the fused data, use methods such as using total station data to correct the elevation error of the laser point cloud, use an image registration algorithm to adjust the error, and use relevant data to supplement and correct the missing or incorrect data, so as to improve the accuracy and reliability of the fused data.

[0158] 303. Optimize the data based on the evaluation indicators.

[0159] Based on the evaluation indicators, use methods such as smoothing, sharpening, contrast enhancement, and color adjustment to optimize the fused data, improve the visualization effect and usability of the data, and provide better data support for subsequent analysis and applications.

[0160] Exemplarily, methods such as smoothing and sharpening can be used to optimize the fused data, improve the visualization effect and usability of the data. Enhance the contrast of the fused image to make the ground objects in the image more clearly distinguishable; perform color adjustment to make the colors of the image more realistic and natural. Through data optimization, provide better data support for subsequent analysis and applications.

[0161] Different evaluation indicators can reflect different characteristics of the fused data. According to these characteristics, appropriate optimization methods can be selected targeted.

[0162] Exemplarily, based on the clarity index, if clarity evaluation indicators such as the average gradient are low, it indicates that the clarity of the fused data is insufficient. At this time, a sharpening method can be selected to enhance the edges and details of the image, improve the clarity of the data, and make the contour of the ground objects in the image more obvious.

[0163] Exemplarily, based on the information amount index, when information amount evaluation indicators such as the information entropy are not ideal, it means that the data contains insufficient information. In this case, a method of contrast enhancement can be tried, such as histogram equalization, which can stretch the gray value distribution of the image, make the bright part of the image brighter and the dark part darker, so as to highlight the detailed information in the image and increase the information amount of the data.

[0164] Exemplarily, based on the color performance metrics, if the evaluation metrics related to color indicate that the color of the fused data is not realistic and natural enough and there are problems such as color cast, color adjustment is required. According to the color evaluation metrics, a color correction algorithm can be adopted to adjust the color of the image to a state that conforms to the visual habit of the human eye, making the color of the image more vivid.

[0165] In some embodiments, the optimization parameters can be determined based on the evaluation metrics.

[0166] Exemplarily, when performing smoothing processing, common methods include mean filtering, Gaussian filtering, etc. These methods have corresponding parameters, such as the size of the filtering window. By continuously adjusting the parameters and combining the evaluation metrics to judge the processing effect, when the evaluation metrics reach the optimal value, the corresponding parameters are the more appropriate ones. For example, when using Gaussian filtering to perform smoothing processing on an image, if sharpness metrics such as average gradient reach a relatively good effect when the filtering window is 3x3 and the noise is effectively suppressed at the same time, then 3x3 can be used as the appropriate parameter for this smoothing processing.

[0167] Exemplarily, sharpening processing also needs to determine appropriate parameters. Taking Laplacian sharpening as an example, its sharpening degree can be controlled by adjusting the coefficient. By observing the change of the evaluation metrics, select the coefficient value that can achieve the best balance of metrics such as sharpness and information content. If the coefficient is too large, it may cause the image to be oversharpened and generate noise; if the coefficient is too small, the sharpening effect is not obvious. Therefore, determining the parameters based on the evaluation metrics can avoid these problems.

[0168] Exemplarily, during the optimization process, the evaluation metrics can also be continuously used to evaluate the effect of each operation and judge whether the expected optimization goal has been achieved.

[0169] Exemplarily, during the optimization process, when performing optimization processing on the fused data, the evaluation metrics can be calculated in real time. For example, during the process of contrast enhancement, each time the contrast parameter is adjusted, metrics such as information entropy and average gradient are calculated. If the metrics change in the expected direction, it indicates that the current optimization operation is effective; if the metrics do not improve or even deteriorate, the optimization method or parameters need to be adjusted.

[0170] Exemplarily, after the optimization processing is completed, the evaluation metrics before and after the processing are compared and analyzed. If the information entropy increases, the average gradient improves, and the color-related metrics are also improved, it indicates that the optimization processing has achieved good results; on the contrary, if some metrics do not reach the expected improvement, the reasons need to be further analyzed, and it may be necessary to reselect the optimization method or adjust the parameters for re-optimization.

[0171] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless there is a clear indication in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0172] Based on the same inventive concept, an embodiment of the present application further provides an overhead transmission line multi-source data fusion device for implementing the above-mentioned overhead transmission line multi-source data fusion method. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the overhead transmission line multi-source data fusion device provided below can refer to the limitations on the overhead transmission line multi-source data fusion method in the above text, and will not be repeated here.

[0173] In an exemplary embodiment, as Figure 4 shown, an overhead transmission line multi-source data fusion device is provided, including:

[0174] An acquisition module 401, configured to acquire multi-source data of an overhead transmission line;

[0175] An extraction module 402, configured to perform feature extraction on the multi-source data according to the data type to obtain multi-type feature data;

[0176] A fusion module 403, configured to perform fusion processing on the multi-type feature data through a pixel-level fusion algorithm, a feature-level fusion algorithm, and a decision-level fusion algorithm to obtain fusion data;

[0177] Wherein, the fusion data is used for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

[0178] In one of the embodiments, the extraction module 402 is specifically configured to:

[0179] Perform data processing on the multi-source data to obtain processed data; the data processing includes at least one of format conversion, noise removal, coordinate system unification, data cropping, and registration;

[0180] Perform feature extraction on the processed data according to the data type to obtain multi-type feature data.

[0181] In one embodiment, the extraction module 402 is specifically configured to:

[0182] Convert the multi-source data to the WGS84 coordinate system, and establish a spatial correspondence between different data through spatial registration to obtain data with unified coordinates and spatial association.

[0183] In one embodiment, the multi-type feature data includes at least two of the following feature data:

[0184] Texture features, color features, and shape features of image data;

[0185] Spatial distribution features, density features, and elevation change features of lidar point cloud data;

[0186] GPS coordinate features and time features corresponding to the coordinates of GPS data;

[0187] Terrain undulation features, slope features, and aspect features of elevation data;

[0188] Feature outline features, type features, and location features of topographic maps;

[0189] Measuring point coordinate features, distance features, and angle features of total station data;

[0190] Three-dimensional building structure features and texture features of tilt data;

[0191] Planning area features, line features, and control point features of planning data;

[0192] Line direction features, substation location features, and voltage level features of power grid thematic data.

[0193] In one embodiment, the fusion module 403 is specifically configured to:

[0194] For the texture features, color features, and shape features of the image data, and the three-dimensional building structure features and texture features of the tilt data, perform fusion processing using a pixel-level fusion algorithm to obtain fused image data;

[0195] Perform fusion processing on the fused image data and the first data using a feature-level fusion algorithm to obtain a fused feature vector;

[0196] Perform fusion processing on the fused vector feature and the second data using a decision-level fusion algorithm to obtain the final fused data.

[0197] Wherein, the first data includes at least one of the following:

[0198] The spatial distribution characteristics, density characteristics, and elevation change characteristics of the laser point cloud data;

[0199] The measurement point coordinate characteristics, distance characteristics, and angle characteristics of the total station data;

[0200] The second data includes at least one of the following:

[0201] The GPS coordinate characteristics of the GPS data and the time characteristics corresponding to the coordinates;

[0202] The terrain undulation characteristics, slope characteristics, and aspect characteristics of the elevation data;

[0203] The feature of ground object contour, type feature, and location feature of the topographic map;

[0204] The measurement point coordinate characteristics, distance characteristics, and angle characteristics of the total station data;

[0205] The planned area characteristics, line characteristics, and control point characteristics of the planning data;

[0206] The line direction characteristics, substation location characteristics, and voltage level characteristics of the power grid special data.

[0207] In one embodiment, the device further includes: a post-processing and optimization module 404, configured to:

[0208] Perform quality evaluation on the fused data to obtain evaluation indicators;

[0209] Based on the evaluation indicators, perform error correction and data optimization on the fused data to obtain optimized data;

[0210] Use the optimized data for any one of the operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

[0211] Each module in the above-mentioned overhead transmission line multi-source data fusion device can be implemented in whole or in part by software, hardware, and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

[0212] In an exemplary embodiment, a computer device is provided, and its internal structure diagram can be as Figure 5As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a multi-source data fusion method for overhead transmission lines.

[0213] Those skilled in the art can understand that Figure 5 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0214] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, it implements the multi-source data fusion method for overhead transmission lines in the above method embodiment.

[0215] In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, it implements the multi-source data fusion method for overhead transmission lines in the above method embodiment.

[0216] In an embodiment, a computer program product is provided, including a computer program. When the computer program is executed by the processor, it implements the multi-source data fusion method for overhead transmission lines in the above method embodiment.

[0217] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memory (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logics, data processing logics based on quantum computing, etc., without limitation.

[0218] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0219] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for fusion of multi-source data of overhead transmission lines, characterized in that: The method comprises: Acquire multi-source data on overhead transmission lines; Extracting features from the multi-source data according to data types to obtain multiple types of feature data; The multiple types of feature data are fused by a pixel-level fusion algorithm, a feature-level fusion algorithm, and a decision-level fusion algorithm to obtain fused data; The fused data is used for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

2. The method according to claim 1, characterized in that The extracting features of the multi-source data according to data types to obtain multiple types of feature data includes: Processing the multi-source data to obtain processed data; the data processing includes: at least one of format conversion, noise removal, coordinate system unification, data clipping and registration; Feature extraction is performed on the processed data according to the data type to obtain multiple types of feature data.

3. The method according to claim 2, characterized in that The data processing includes unifying the coordinate system, and the data processing of the multi-source data to obtain processed data includes: The multi-source data are converted to the WGS84 coordinate system, and the spatial correspondence between different data is established through spatial registration to obtain data with unified coordinates and spatial association.

4. The method according to claim 1, characterized in that The multiple types of feature data include at least two of the following feature data: Texture features, color features, and shape features of image data; The spatial distribution characteristics, density characteristics and elevation change characteristics of laser point cloud data; The GPS coordinate characteristics of GPS data and the time characteristics corresponding to the coordinates; Topographic relief characteristics, slope characteristics, and aspect characteristics of elevation data; The features of the contour, type and location of the features on the topographic map; The coordinate characteristics, distance characteristics, and angle characteristics of the measurement points of the total station data; Three-dimensional structural and texture features of buildings from tilted data; Planning area characteristics, line characteristics, and control point characteristics of planning data; The line direction characteristics, substation location characteristics, and voltage level characteristics of power grid thematic data.

5. The method according to claim 4, characterized in that The fusing the multiple types of feature data by using a pixel-level fusion algorithm, a feature-level fusion algorithm, and a decision-level fusion algorithm to obtain fused data includes: A pixel-level fusion algorithm is used to perform fusion processing on the texture features, color features and shape features of the image data, and the three-dimensional structure features and texture features of the building of the tilt data, so as to obtain fused image data; The fused image data and the first data are fused by using a feature-level fusion algorithm to obtain a fused feature vector; The fusion vector feature and the second data are fused by using a decision-level fusion algorithm to obtain the final fused data; Wherein, the first data is at least one of the following: The spatial distribution characteristics, density characteristics and elevation change characteristics of the laser point cloud data; The measurement point coordinate characteristics, distance characteristics, and angle characteristics of the total station data; The second data includes at least one of the following: The GPS coordinate characteristics of the GPS data and the time characteristics corresponding to the coordinates; The terrain undulation characteristics, slope characteristics, and aspect characteristics of the elevation data; The features of the contour, type and location of the features in the topographic map; The measurement point coordinate characteristics, distance characteristics, and angle characteristics of the total station data; The planning area characteristics, line characteristics, and control point characteristics of the planning data; The power grid thematic data includes line direction characteristics, substation location characteristics, and voltage level characteristics.

6. The method according to claim 1, characterized in that After fusing the multiple types of feature data using the pixel-level fusion algorithm, the feature-level fusion algorithm, and the decision-level fusion algorithm to obtain fused data, the method further includes: Performing quality evaluation on the fused data to obtain evaluation indicators; Based on the evaluation index, error correction and data optimization are performed on the fused data to obtain optimized data; The optimized data is used for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

7. An overhead power transmission line multi-source data fusion device, characterized in that: The device comprises: An acquisition module, used for acquiring multi-source data of overhead transmission lines; An extraction module, used for performing feature extraction on the multi-source data according to data types to obtain multiple types of feature data; A fusion module, used for fusing the multiple types of feature data through a pixel-level fusion algorithm, a feature-level fusion algorithm and a decision-level fusion algorithm to obtain fused data; The fused data is used for any one of operation monitoring, dispatching control, planning and design, and intelligent inspection of the overhead transmission line.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images