Power grid engineering construction progress monitoring method and device and electronic equipment

Abstract

The present application relates to the technical field of power grid engineering construction, and discloses a power grid engineering construction progress monitoring method, device and electronic equipment, which is based on remote sensing time sequence images of a target monitoring area, constructs a plurality of grid cell data corresponding to the remote sensing time sequence images according to the time requirement of the power grid engineering construction progress, realizes automatic monitoring of the power grid engineering construction progress, does not need a large number of personnel investment, and achieves the purposes of significantly reducing the monitoring cost, improving the monitoring efficiency and monitoring comprehensiveness. The progress stage recognition model is used to identify the progress stage and the confidence degree corresponding to the grid cell data, so that the current progress time corresponding to the grid cell data and the total progress time of the power grid line of the target monitoring area can be determined, and the accuracy of the power grid engineering construction progress monitoring is improved.

Description

Technical Field

[0001] This invention relates to the field of power grid engineering construction technology, specifically to a method, device, and electronic equipment for monitoring the progress of power grid engineering construction. Background Technology

[0002] With the rapid advancement of the State Grid construction, the scale of transmission line projects is expanding daily, especially the construction of ultra-high voltage transmission lines, which involves thousands of kilometers of line corridors and hundreds of tower foundations. Ensuring the orderly progress of power grid construction projects is crucial for monitoring the project's progress.

[0003] In related technologies, project progress monitoring mainly relies on manual on-site inspections, requiring monitoring personnel to regularly conduct on-site surveys and record the project construction progress. This method has problems such as high labor intensity, long monitoring cycle, high cost, and limited coverage, and it is difficult to achieve all-weather, high-precision and automated monitoring of large-scale power grid projects. Summary of the Invention

[0004] This invention provides a method, device, and electronic equipment for monitoring the construction progress of power grid projects, in order to solve the problems of relying on manual on-site inspections, which have the disadvantages of high labor intensity, long monitoring cycle, high cost, and limited coverage, and make it difficult to achieve all-weather, high-precision, and automated monitoring of large-scale power grid projects.

[0005] In a first aspect, the present invention provides a method for monitoring the construction progress of power grid projects, the method comprising: Based on the remote sensing time-series images of the target monitoring area, and in accordance with the time requirements of the power grid project construction schedule, multiple grid cell data corresponding to the remote sensing time-series images are constructed. For each grid cell data, extract multiple temporal feature vectors corresponding to the grid cell data; For each grid cell data, the various time-series feature vectors corresponding to the grid cell data are input into a pre-trained progress stage recognition model to obtain the progress stage and confidence level corresponding to the grid cell data. For each grid cell data, the current progress time corresponding to the grid cell data is determined based on the confidence level of the grid cell data and the preset weight corresponding to the progress stage to which the grid cell data belongs. The total progress time of the power grid lines in the target monitoring area is determined based on the current progress time corresponding to the data of each grid cell.

[0006] This invention, through the implementation of the above-described embodiments, constructs multiple grid cell data corresponding to the remote sensing time-series images of the target monitoring area according to the time requirements of power grid project construction progress. This enables automated monitoring of power grid project construction progress without requiring a large number of personnel, significantly reducing monitoring costs and improving monitoring efficiency and comprehensiveness. By using a progress stage identification model to identify the progress stage and confidence level corresponding to the grid cell data, it is easier to determine the current progress time corresponding to the grid cell data and the total progress time of the power grid lines in the target monitoring area, thereby improving the accuracy of power grid project construction progress monitoring.

[0007] In some specific implementations, the power grid engineering construction progress monitoring method in this embodiment further includes: Obtain the historical progress time corresponding to each grid cell data; For each grid cell data, the current progress time and historical progress time corresponding to the grid cell data are input into the pre-trained progress prediction model to predict the expected progress time corresponding to the grid cell data. For each grid cell data, calculate the deviation between the current progress time and the expected progress time corresponding to the grid cell data; For each grid cell data, when the deviation between the current progress time corresponding to the grid cell data exceeds a preset threshold, an anomaly warning action is performed on the grid cell data to obtain an anomaly warning information.

[0008] By implementing the above-described embodiments, this invention performs anomaly warning actions on grid cell data, enabling maintenance personnel to promptly identify problems during construction and thus avoid power grid losses.

[0009] In some specific implementations, the power grid engineering construction progress monitoring method in this embodiment further includes: Obtain the cloud coverage rate for each grid cell; For each grid cell data, when the cloud coverage exceeds the preset coverage, the grid cell data is subjected to cloud occlusion processing to obtain the cloud data processing result.

[0010] By implementing the above-described embodiments, this invention performs anomaly warning actions on grid cell data, enabling maintenance personnel to promptly identify problems during construction and thus avoid power grid losses.

[0011] In some specific implementations, the power grid engineering construction progress monitoring method in this embodiment further includes: For each grid cell data, a progress monitoring report and visualization charts are generated based on the current progress time corresponding to the grid cell data, the total progress time of the power grid lines in the target monitoring area, abnormal early warning information, and cloud data processing results.

[0012] By implementing the above-described embodiments, this invention presents automated, high-precision, and all-weather monitoring of the progress of power grid engineering construction.

[0013] In some specific implementations, the remote sensing time-series images of the target monitoring area are obtained through the following preprocessing steps; Acquire the original time-series images of the remote sensing time-series images of the target monitoring area; The original time-series images are sequentially processed with radiometric correction, geometric fine correction, image registration, cloud detection, and image fusion to obtain remote sensing time-series images of the target monitoring area.

[0014] By performing the above-described embodiments, the present invention preprocesses the remote sensing time-series images of the target monitoring area to ensure the accuracy of the remote sensing time-series images of the target monitoring area.

[0015] In some specific implementations, based on remote sensing time-series images of the target monitoring area, and according to the time requirements of power grid engineering construction progress, multiple grid cell data corresponding to the remote sensing time-series images are constructed, including: Based on the remote sensing time-series images of the target monitoring area, determine the node data and edge data of the power grid topology; The target monitoring area is divided into multiple grid units according to the preset size; Node data and edge data are mapped into multiple grid cells to obtain multiple grid cell data corresponding to remote sensing time series images. Each grid cell data includes grid attribute information. Based on the construction schedule requirements of the power grid project, the schedule requirements are added to the grid attribute information of each grid cell data to obtain multiple grid cell data corresponding to the remote sensing time series image.

[0016] By implementing the above-described embodiments, the present invention facilitates high-precision monitoring of the target monitoring area.

[0017] In some specific implementations, the time-series feature vector includes: a spectral time-series feature vector and a spectral rate of change feature vector. For each grid cell data, multiple time-series feature vectors corresponding to the grid cell data are extracted, including: For each grid cell data, obtain the near-infrared band reflectance, red band reflectance, short-wave infrared band reflectance, and bare soil index corresponding to the grid cell data; For each grid cell data, the normalized vegetation index corresponding to the grid cell data is calculated based on the near-infrared band reflectance and red band reflectance of the grid cell data. For each grid cell data, the normalized building index corresponding to the grid cell data is calculated based on the shortwave infrared band reflectance and near-infrared band reflectance corresponding to the grid cell data. Based on the normalized vegetation index, normalized building index, and bare soil index corresponding to each grid cell data, they are arranged into a spectral temporal feature vector according to the time series. Calculate the spectral change rate between two adjacent time points based on the spectral time series feature vector; Based on the spectral change rate at two adjacent time points, arrange them into a spectral change rate feature vector according to the time series.

[0018] By implementing the above-described embodiments, this invention extracts various temporal feature vectors corresponding to grid cell data, which facilitates accurate identification of progress stages.

[0019] In some specific implementations, the temporal feature vector includes: a texture temporal feature vector and a texture change rate feature vector. For each grid cell data, multiple temporal feature vectors corresponding to the grid cell data are extracted, and the feature vector also includes: Obtain the grayscale parameter values, grayscale co-occurrence normalization parameters, and correlation parameters corresponding to the data of two adjacent grid cells; Calculate the contrast parameter of two adjacent grid cell data based on the gray level parameter value and gray level co-occurrence normalization parameter corresponding to the data of two adjacent grid cells. Based on the contrast and correlation parameters corresponding to the data of two adjacent grid cells, the texture temporal feature vector is arranged according to the time series. Calculate the texture change rate between two adjacent time points based on the texture temporal feature vector; Based on the texture change rate at two adjacent time points, arrange them into a texture change rate feature vector according to the time series.

[0020] By implementing the above-described embodiments, this invention extracts various temporal feature vectors corresponding to grid cell data, which facilitates accurate identification of progress stages.

[0021] In some specific implementations, the temporal feature vector includes: morphological temporal feature vector and change temporal feature vector. For each grid cell data, multiple temporal feature vectors corresponding to the grid cell data are extracted, and the implementation also includes: For each grid cell, obtain the area change parameters and perimeter change parameters corresponding to the grid cell data; For each grid cell data, the shape index change parameter corresponding to the grid cell data is calculated based on the area change parameter and perimeter change parameter of the grid cell. Based on the area change parameters, perimeter change parameters, and shape index change parameters corresponding to each grid cell data, they are arranged into a morphological temporal feature vector according to the time series.

[0022] By implementing the above-described embodiments, this invention extracts various temporal feature vectors corresponding to grid cell data, which facilitates accurate identification of progress stages.

[0023] In some specific implementations, the progress stage identification model is trained through the following steps: Acquire training samples of remote sensing images at various progress stages; Extract training features from remote sensing images for each progress stage; For each stage of the remote sensing image training features, the remote sensing image training features are input into the stage recognition model for recognition. The recognition results of the remote sensing image training features and the loss between the recognition results of the remote sensing image training features and the true labels of the remote sensing image training features are obtained. The loss is used to update the parameter values ​​of the stage recognition model. The stage recognition model consists of an input layer, a temporal convolutional layer, an attention layer, a fully connected layer and an output layer.

[0024] By implementing the above-described embodiments, the present invention facilitates the accurate identification of the progress stage of power grid construction projects.

[0025] Secondly, the present invention also provides a power grid engineering construction progress monitoring device, the device comprising: The grid data construction module is used to construct multiple grid cell data corresponding to the remote sensing time series images based on the target monitoring area and in accordance with the time requirements of the power grid project construction schedule. The temporal feature extraction module is used to extract multiple temporal feature vectors corresponding to each grid cell data. The progress stage identification module is used to input multiple time-series feature vectors corresponding to each grid cell data into a pre-trained progress stage identification model for identification, so as to obtain the progress stage and confidence level corresponding to the grid cell data. The current progress determination module is used to determine the current progress time of each grid cell data based on the confidence level of the grid cell data and the preset weight of the progress stage to which the grid cell data belongs. The overall progress determination module is used to determine the total progress time of the power grid lines in the target monitoring area based on the current progress time corresponding to the data of each grid unit.

[0026] Thirdly, the present invention also provides an electronic device, comprising: The memory and the processor are interconnected and communicate with each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the power grid construction progress monitoring method in the first aspect or any embodiment of the first aspect. Attached Figure Description

[0027] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0028] Figure 1 This is a flowchart illustrating a method for monitoring the construction progress of power grid projects according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating another method for monitoring the construction progress of a power grid project according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating a method for monitoring the construction progress of a power grid project according to an embodiment of the present invention. Figure 4 This is a schematic diagram of the computer equipment system architecture in the power grid engineering construction progress monitoring method according to an embodiment of the present invention; Figure 5 This is a structural block diagram of a power grid engineering construction progress monitoring device according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0031] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0032] This embodiment provides a method for monitoring the construction progress of power grid projects, which can be used in electronic devices such as mobile phones, tablets, desktop computers, laptops, and servers. Figure 1 This is a flowchart of a power grid engineering construction progress monitoring method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Based on the remote sensing time-series images of the target monitoring area, construct multiple grid cell data corresponding to the remote sensing time-series images according to the time requirements of the power grid project construction schedule.

[0033] Specifically, the target monitoring area can be determined based on the design path of the transmission line project, including the line corridor and buffer zones on both sides, which are typically 200 meters on each side of the line. The remote sensing time-series imagery of the target monitoring area includes tower images and images of the conductor segments between the towers.

[0034] The specific time series of remote sensing images can be set according to the requirements of project progress monitoring, with a recommended cycle of 5-10 days. Remote sensing image sources include Gaofen series satellites (GF-1, GF-2, GF-6), Sentinel-2, Landsat series, etc., with a spatial resolution requirement of better than 2 meters.

[0035] In some specific implementations, the remote sensing time-series images of the target monitoring area are obtained through the following preprocessing steps: Step a1: Obtain the original time-series image of the remote sensing time-series image of the target monitoring area.

[0036] Step a2 involves sequentially performing radiometric correction, geometric fine correction, image registration, cloud detection, and image fusion processing on the original time-series images to obtain remote sensing time-series images of the target monitoring area.

[0037] For the radiometric correction example, the digital quantization values ​​(DN values) of the original time-series imagery are converted to apparent reflectance to eliminate the effects of atmospheric and illumination conditions. The radiometric correction formula is as follows:

[0038] in, Apparent reflectance, Radiance, The Earth-Sun distance (astronomical units). ESUN Solar spectral irradiance, This is the solar zenith angle.

[0039] For the geometric fine correction example, ground control points (GCPs) and digital elevation models (DEMs) are used. The geometric fine correction of the image is achieved through a polynomial fitting model, and the correction accuracy is controlled within 0.5 pixels.

[0040] For the image registration example, using an image from a certain period as a reference, images from other periods are accurately registered to the reference image coordinate system, with the registration error controlled within 0.3 pixels.

[0041] For the cloud detection example, a cloud detection algorithm is used to identify cloud regions in the image and generate a cloud mask. Cloud detection employs a multi-feature fusion method, integrating spectral features, texture features, and temperature features. The cloud detection probability calculation formula is as follows:

[0042] in, fspectral This refers to spectral characteristics (such as reflectance in the blue light band). ftexture For texture features (such as local binary pattern variance). fthermal These are thermal infrared characteristics (such as brightness and temperature). w1, w2, w3 These are the weighting coefficients.

[0043] For the image fusion example, for multispectral images, full-color sharpening is used to improve spatial resolution, and contrast and edge enhancement are performed to improve target recognition.

[0044] In this embodiment of the application, the remote sensing time-series images of the target monitoring area are preprocessed through the above steps a1-a2 to ensure the accuracy of the remote sensing time-series images of the target monitoring area.

[0045] Furthermore, traditional methods typically identify targets using single remote sensing images of the monitored area. However, power grid engineering is a gradual construction process, with different surface features at each stage, from obstacle clearing, foundation excavation, foundation pouring, tower erection to conductor stringing. Therefore, single remote sensing images cannot accurately identify the temporal changes in the project's progress stages, ultimately leading to inaccurate progress monitoring.

[0046] Therefore, the embodiments of this application use remote sensing time-series images of the target monitoring area to ensure the continuity of the remote sensing images of the target monitoring area, thereby improving the accuracy of progress monitoring.

[0047] In some specific implementations, step S101 involves constructing multiple grid cell data corresponding to the remote sensing time-series images of the target monitoring area, according to the time requirements of the power grid project construction schedule, including: Step b1: Determine the node data and edge data of the power grid topology map based on the remote sensing time-series images of the target monitoring area.

[0048] Specifically, since the remote sensing time-series imagery of the target monitoring area includes tower images and conductor segments between towers, when determining the power grid topology, the line path can be viewed as a series of nodes and edges. Node data represents towers, and edge data represents conductor segments between towers. The path is stored using a vector data structure, and each node contains attributes such as latitude and longitude coordinates, tower type, and foundation type.

[0049] Step b2: Divide the target monitoring area into multiple grid units according to the preset size.

[0050] Specifically, the target monitoring area can be configured with a preset size of 50 meters × 50 meters. In this embodiment, data processing tools can be used to divide the target monitoring area into multiple grid units. Each grid unit serves as the basic unit for progress monitoring, and the preset size can be flexibly set in practical applications.

[0051] Step b3: Map the node data and edge data into multiple grid cells to obtain multiple grid cell data corresponding to the remote sensing time series image. Each grid cell data includes grid attribute information.

[0052] Specifically, embodiments of this application may also employ data processing tools to map node data and edge data into multiple grid cells to obtain multiple grid cell data corresponding to remote sensing time-series images.

[0053] Grid attribute information includes, but is not limited to, grid identity (ID), center coordinates, tower segment to which it belongs, current construction stage, and progress percentage.

[0054] Step b4: Based on the construction schedule requirements of the power grid project, add the schedule requirements to the grid attribute information of each grid cell data to obtain multiple grid cell data corresponding to the remote sensing time series image.

[0055] Specifically, the construction schedule requirements for power grid projects include time requirements for various different stages. These stages include: obstacle clearing, foundation excavation, foundation pouring, tower erection, conductor stringing, and acceptance. Time requirements for these different stages can be added to each grid cell data.

[0056] This application embodiment divides the remote sensing time-series images of the target monitoring area into grids to facilitate the analysis of the construction progress of power grid projects.

[0057] Step S102: For each grid cell data, extract multiple temporal feature vectors corresponding to the grid cell data.

[0058] In a specific example, the temporal feature vector includes: spectral temporal feature vector, spectral rate of change feature vector, texture temporal feature vector, texture rate of change feature vector, morphological temporal feature vector, and change temporal feature vector.

[0059] In some specific implementations, step S102 above involves extracting multiple temporal feature vectors corresponding to each grid cell data, including: Step c1: For each grid cell data, obtain the near-infrared band reflectance, red band reflectance, short-wave infrared band reflectance, and bare soil index corresponding to the grid cell data.

[0060] Specifically, for each grid cell data, the near-infrared reflectance corresponding to the grid cell data can be represented by NIR, the red reflectance by Red, the short-wave infrared reflectance by SWIR, and the bare soil index by BI.

[0061] Step c2: For each grid cell data, calculate the normalized vegetation index corresponding to the grid cell data based on the near-infrared band reflectance and red band reflectance of the grid cell data.

[0062] For each grid cell of data, the normalized vegetation index is calculated and expressed by the following formula:

[0063] in, The normalized vegetation index corresponds to the grid cell data. The near-infrared reflectance corresponding to the grid cell data. This represents the red band reflectance corresponding to the grid cell data.

[0064] Step c3: For each grid cell data, calculate the normalized building index corresponding to the grid cell data based on the shortwave infrared band reflectance and near-infrared band reflectance corresponding to the grid cell data.

[0065] For each grid cell of data, the normalized building index corresponding to the grid cell data is calculated, expressed by the following formula:

[0066] in, The normalized building index corresponding to the grid cell data. The shortwave infrared reflectance corresponding to the grid cell data. This represents the near-infrared reflectance corresponding to the grid cell data.

[0067] Step c4: Arrange the normalized vegetation index, normalized building index and bare soil index of each grid cell data into a spectral time series feature vector according to the time series.

[0068] Arranged into a spectral temporal feature vector according to the time series, it is expressed by the following formula:

[0069] Where t is the time sequence number.

[0070] Step c5: Calculate the spectral change rate between two adjacent time points based on the spectral time series feature vector.

[0071] Specifically, the rate of spectral change between two adjacent time points is expressed by the following formula:

[0072] in, The spectral change rate is the rate of change between two adjacent time points. Let be the normalized vegetation index corresponding to the t-th time point. is the normalized vegetation index corresponding to the (t-1)th time point.

[0073] Step c6: Arrange the spectral change rate feature vectors according to the time series based on the spectral change rate of two adjacent time points.

[0074] Specifically, the spectral change rate eigenvector is arranged according to the time series and expressed by the following formula:

[0075] in, This is the eigenvector of the spectral change rate. Let be the spectral change rate at the t-th time point.

[0076] In some specific implementations, step S102 above, which extracts multiple temporal feature vectors corresponding to each grid cell data, further includes: Step d1: Obtain the grayscale parameter values, grayscale co-occurrence normalization parameters, and correlation parameters corresponding to the data of two adjacent grid cells.

[0077] Specifically, the grayscale parameter values ​​corresponding to the data of two adjacent grid cells are respectively used as i and j This indicates that the gray-level co-occurrence normalization parameter is used... The correlation parameter is represented by... Cor express.

[0078] Step d2: Calculate the contrast parameter of the data of two adjacent grid cells based on the grayscale parameter values ​​and grayscale co-occurrence normalization parameters corresponding to the data of two adjacent grid cells.

[0079] Specifically, the contrast parameter between two adjacent grid cells is calculated using the following formula:

[0080] in, The contrast parameter for data between two adjacent grid cells. i The grayscale parameter is the first adjacent network unit. The grayscale parameter is the first adjacent network unit. This refers to the gray-level co-occurrence normalization parameter corresponding to the data of two adjacent grid cells.

[0081] Step d3: Based on the contrast parameters and correlation parameters corresponding to the data of two adjacent grid cells, arrange them into a texture temporal feature vector according to the time series.

[0082] Specifically, the texture temporal feature vector is arranged according to the time series and expressed by the following formula:

[0083] Step d4: Calculate the texture change rate between two adjacent time points based on the texture temporal feature vector.

[0084] Specifically, the texture change rate between two adjacent time points is expressed by the following formula:

[0085] in, The texture change rate at two adjacent time points. Let be the normalized building index corresponding to the t-th time point. is the normalized building index corresponding to the (t-1)th time point.

[0086] Step d5 arranges the texture change rate feature vectors according to the time series based on the texture change rate of two adjacent time points.

[0087] Specifically, the texture change rate feature vector is arranged according to the time series and expressed by the following formula:

[0088] in, This is the texture change rate feature vector. Let be the texture change rate at time t.

[0089] Step S102 above, which extracts multiple temporal feature vectors corresponding to each grid cell data, also includes: Step e1: For each grid cell data, obtain the area change parameters and perimeter change parameters of the grid cell data.

[0090] Specifically, for each grid cell data point, the area change parameter corresponding to the grid cell data is used... This indicates that the perimeter variation parameter corresponding to the grid cell data is used... express.

[0091] Step e2: For each grid cell data, calculate the shape index change parameter corresponding to the grid cell data based on the area change parameter and perimeter change parameter corresponding to the grid cell.

[0092] Specifically, the shape exponential change parameter corresponding to the grid cell data is calculated using the following formula:

[0093] in, The shape exponential variation parameter corresponding to the grid cell data. The perimeter variation parameter corresponding to the grid cell data. This refers to the area variation parameter corresponding to the grid cell data.

[0094] Step e3: Based on the area change parameters, perimeter change parameters, and shape index change parameters corresponding to each grid cell data, arrange them into a morphological temporal feature vector according to the time series.

[0095] Specifically, the morphological time-series feature vectors, arranged according to the time series, are expressed by the following formula:

[0096] in, This is a morphological temporal feature vector. Let be the area change parameter corresponding to the t-th time point. Let be the perimeter change parameter corresponding to the t-th time point. The shape exponential change parameter corresponds to the t-th time point.

[0097] Step S103: For each grid cell data, input the various time-series feature vectors corresponding to the grid cell data into the pre-trained progress stage recognition model to obtain the progress stage and confidence level corresponding to the grid cell data.

[0098] In some specific implementations, the progress stage identification model is trained through the following steps: Step f1: Obtain training samples of remote sensing images at various progress stages.

[0099] Specifically, the various progress phases include: obstacle clearing, foundation excavation, foundation pouring, pole erection, conductor stringing, and acceptance. The training samples of remote sensing images for each of these multiple progress phases contain no fewer than 500 samples of each category.

[0100] Step f2: Extract training features from remote sensing images for each progress stage.

[0101] Specifically, embodiments of this application can extract training features of remote sensing images for each progress stage using a temporal convolutional neural network.

[0102] Step f2: For the remote sensing image training features of each progress stage, input the remote sensing image training features into the progress stage recognition model for recognition, and obtain the recognition result of the remote sensing image training features and the loss between the recognition result of the remote sensing image training features and the true label of the remote sensing image training features. The loss is used to update the parameter values ​​of the progress stage recognition model. The progress stage recognition model consists of an input layer, a temporal convolutional layer, an attention layer, a fully connected layer and an output layer.

[0103] Specifically, the input layer receives training features of remote sensing images extracted for each progress stage using a temporal convolutional neural network, as described in the example above; the temporal convolutional layer uses dilated causal convolution to extract temporal features, with the dilation coefficient increasing exponentially, including coefficients such as 1, 2, 4, and 8.

[0104] The attention layer employs a self-attention mechanism to capture key temporal features. The attention weight calculation formula is as follows:

[0105] in, For query vector, For key vectors, Let be the dimension of the key vector. A fully connected layer maps temporal features to the classification space. The output layer uses... Softmax The function outputs the probability distribution for each category. For model training and optimization, the loss between the recognition results of the remote sensing image training features and the true labels of the remote sensing image training features is determined using the cross-entropy loss function. This application's embodiments use... Adam The optimizer trains the model. The loss function is calculated as follows:

[0106] Where C represents the number of categories. yi For real labels, i To predict probabilities, data augmentation techniques are used to expand the training samples, including temporal interpolation, noise addition, and temporal pruning.

[0107] This application's embodiments utilize temporal feature extraction and deep learning classification to accurately identify the stage of engineering construction. In a practical application of a certain ultra-high voltage project, the accuracy rate of progress stage identification can reach 92.3%.

[0108] Step S104: For each grid cell data, determine the current progress time corresponding to the grid cell data based on the confidence level corresponding to the grid cell data and the preset weight corresponding to the progress stage to which the grid cell data belongs.

[0109] Specifically, the progress stages of the grid cell data include: not started, clearing obstacles, foundation excavation, foundation pouring, tower erection in progress, conductor erection in progress, and completed. For each grid cell data point, a preset weight corresponding to each progress stage is assigned. For example, the weight allocation scheme is as follows: The pre-set weights for the stages not yet started are as follows: obstacle clearing stage (10%), foundation excavation stage (30%), foundation excavation stage (50%), tower erection stage (70%), conductor erection stage (90%), and acceptance stage (20%).

[0110] The current progress time of the grid cell data is determined using the following formula:

[0111] in, The current progress time corresponding to the grid cell data. The preset weights are assigned to the progress stages to which the grid cell data belongs. This represents the confidence level corresponding to the grid cell data.

[0112] Step S105: Determine the total progress time of the power grid lines in the target monitoring area based on the current progress time corresponding to the data of each grid cell.

[0113] Determine the total progress time of the power grid lines in the target monitoring area, including: Step f1: Determine all grid cell data corresponding to the tower and all grid cell data corresponding to the conductor segment from the target monitoring area.

[0114] Step f2: Calculate the current progress time for each tower based on the current progress time corresponding to each grid cell data and the grid cell data corresponding to the tower.

[0115] The current progress time for each tower is expressed by the following formula:

[0116] in, The current progress time for each tower. The current progress time corresponding to the grid cell data. The number of grid cells corresponding to each tower. Any grid cell corresponding to each tower i The distance to the center of all grid cells to which the tower belongs.

[0117] Step f3: Calculate the total progress time of the power grid lines in the target monitoring area based on the current progress time corresponding to each tower.

[0118] The total progress time of the power grid lines in the target monitoring area is expressed by the following formula:

[0119] in, The total progress time of the power grid lines in the target monitoring area. The current progress time for each tower, where M is the total number of towers. For towers j The length of the conductor segment it represents.

[0120] In some optional implementations, the power grid engineering construction progress monitoring method in this application embodiment further includes: calculating the completion rate of different progress stages.

[0121] Specifically, the different progress stages are: obstacle clearing stage, foundation excavation stage, foundation pouring stage, tower erection stage, conductor stringing stage, and acceptance stage.

[0122] The current progress time at different progress stages is expressed by the following formula:

[0123] in, For the first i Completion rate of each progress phase For the first i The number of completed phases in each progress stage. For the first i The number of grid cells in each progress stage.

[0124] Traditional manual on-site inspections require significant manpower and resources. For example, on a 500-kilometer transmission line project, a comprehensive inspection every 10 days would require 5-8 workers working continuously for 7-10 days, resulting in annual monitoring costs exceeding 2 million yuan. Furthermore, geographical limitations prevent manual inspections in dangerous areas such as mountains, forests, and uninhabited zones, leading to monitoring blind spots. In other words, manual monitoring is inefficient, costly, and yields incomplete results. Traditional manual on-site inspections suffer from high labor intensity, long monitoring cycles, high costs, and limited coverage, making it difficult to achieve 24 / 7, high-precision, and automated monitoring of large-scale power grid projects. Therefore, this application embodiment, based on remote sensing time-series imagery of the target monitoring area, constructs multiple grid cell data corresponding to the remote sensing time-series imagery according to the time requirements of power grid engineering construction progress. This enables automated monitoring of power grid engineering construction progress without requiring a large number of personnel, significantly reducing monitoring costs and improving monitoring efficiency and comprehensiveness. A progress stage identification model identifies the progress stage and confidence level corresponding to the grid cell data, facilitating the determination of the current progress time corresponding to the grid cell data and the total progress time of the power grid lines in the target monitoring area, thereby improving the accuracy of power grid engineering construction progress monitoring.

[0125] This embodiment provides a method for monitoring the construction progress of power grid projects, which can be used in electronic devices such as mobile phones, tablets, desktop computers, laptops, and servers. Figure 2 This is a flowchart of a power grid engineering construction progress monitoring method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S101: Based on the remote sensing time-series images of the target monitoring area, construct multiple grid cell data corresponding to the remote sensing time-series images according to the time requirements of the power grid project construction schedule.

[0126] Step S102: For each grid cell data, extract multiple temporal feature vectors corresponding to the grid cell data.

[0127] Step S103: For each grid cell data, input the various time-series feature vectors corresponding to the grid cell data into the pre-trained progress stage recognition model to obtain the progress stage and confidence level corresponding to the grid cell data.

[0128] Step S104: For each grid cell data, determine the current progress time corresponding to the grid cell data based on the confidence level corresponding to the grid cell data and the preset weight corresponding to the progress stage to which the grid cell data belongs.

[0129] Step S105: Determine the total progress time of the power grid lines in the target monitoring area based on the current progress time corresponding to the data of each grid cell.

[0130] For details of steps S101-S105 above, please refer to the above embodiment, and they will not be repeated here.

[0131] Step S106: Obtain the historical progress time corresponding to each grid cell data.

[0132] Step S107: For each grid cell data, input the current progress time and historical progress time corresponding to the grid cell data into the pre-trained progress prediction model to predict the expected progress time corresponding to the grid cell data.

[0133] Specifically, the schedule prediction model uses an LSTM model. The model takes the historical schedule time corresponding to each grid cell as input and outputs the schedule prediction values ​​for several future periods. The update formula for the LSTM model is as follows:

[0134]

[0135]

[0136]

[0137]

[0138]

[0139] in, ft For the Gate of Oblivion it For input gate, ot For output gate, Ct In cellular state, ht In hidden state, W b represents the weight and bias, respectively. Sigmoid function.

[0140] The estimated time for grid cell data is expressed by the following formula:

[0141] in, Tcurrent The current progress time corresponding to the grid cell data. Pcurrent This is the current progress stage. Rpredicted To predict the progress growth rate, T As a unit of time, This represents the estimated timeframe for the grid cell data.

[0142] Step S108: For each grid cell data, calculate the deviation between the current progress time and the expected progress time corresponding to the grid cell data.

[0143] Step S109: For each grid cell data, when the target completion time corresponding to the grid cell data exceeds the preset progress time, perform an abnormal warning action on the grid cell data and obtain abnormal warning information.

[0144] For example, for each grid cell data, when the deviation between the current progress time and the expected progress time corresponding to the grid cell data exceeds a preset threshold (such as 15%), a progress delay warning is triggered.

[0145] In other examples, a local anomaly warning is triggered when the current progress time of any grid cell data is less than the current progress time of its adjacent grid cell data.

[0146] This application embodiment performs anomaly warning actions on grid cell data, which helps maintenance personnel to promptly detect problems during construction and thus avoid power grid losses.

[0147] In some optional implementations, the power grid engineering construction progress monitoring method in the embodiments of this application, in Figure 2 The process also includes the following steps: Step S110: Obtain the cloud coverage rate of each grid cell.

[0148] Step S111: For each grid cell data, when the cloud coverage exceeds the preset coverage, perform cloud occlusion processing on the grid cell data to obtain the cloud data processing result.

[0149] Specifically, the cloud coverage rate of each grid cell data can be specifically used as... Cratio It indicates that the preset coverage rate can be 30%, and this preset coverage rate can be flexibly set according to the actual scenario.

[0150] Specifically, for low-quality phases, temporal interpolation can be used for compensation. Cubic spline interpolation is employed, using data from adjacent high-quality phases to estimate the characteristic values ​​of the occluded phase. Interpolation function. S(t) The calculation formula is as follows:

[0151] in, a,b,c,d These are the interpolation coefficients. ti For interpolation node time.

[0152] When the quality of the remote sensing image corresponding to the grid cell data is poor, multi-source data such as drone aerial images and ground monitoring equipment data can be integrated to supplement the missing information.

[0153] Satellite remote sensing imagery is easily affected by cloud cover, especially in rainy southern regions, making it difficult to acquire continuous high-quality imagery. Traditional intelligent cloud cover handling mechanisms cannot effectively compensate for poor image quality at a particular time by utilizing imagery from other time periods, leading to interruptions in monitoring continuity.

[0154] Therefore, this application embodiment performs cloud occlusion processing on the grid cell data to improve the data quality of remote sensing image data. Through mechanisms such as cloud detection, temporal interpolation, and multi-source data fusion, this invention effectively overcomes the impact of cloud occlusion on monitoring continuity. In tests conducted in rainy southern regions, the effective monitoring rate reached 96.5%, a 25% improvement compared to traditional methods.

[0155] In some optional implementations, the power grid engineering construction progress monitoring method in the embodiments of this application, in Figure 2 The process also includes the following steps: Step S112: For each grid cell data, generate a progress monitoring report and visualization chart based on the current progress time corresponding to the grid cell data, the total progress time of the power grid lines in the target monitoring area, the abnormal early warning information, and the cloud data processing results.

[0156] Specifically, the progress monitoring report and visualization charts include the current progress time for each grid cell, the total progress time for the power grid lines in the target monitoring area, anomaly warning information, and cloud data processing results. The target monitoring report can also use a progress map to represent the progress stages of each grid cell with different colors. The visualization charts include a progress curve table, which displays the trend of progress changes.

[0157] This application's embodiments present automated, high-precision, and 24 / 7 monitoring of power grid engineering construction progress through progress monitoring reports and visual charts.

[0158] like Figure 3 The diagram shown is a simplified overall flowchart of the power grid engineering construction progress monitoring method in this application embodiment: Figure 3 The process involves acquiring remote sensing time-series images, image preprocessing, constructing multiple grid cell data corresponding to the remote sensing time-series images, extracting time-series feature vectors, identifying progress stages using a progress stage identification model, determining the current progress time and the total progress time of the power grid corresponding to the grid cell data, predicting the expected progress time, issuing anomaly warnings, handling cloud obstruction, and generating progress monitoring reports and visualization charts.

[0159] like Figure 4 The diagram shown is a schematic representation of the computer equipment system architecture in the power grid engineering construction progress monitoring method according to an embodiment of this application. Figure 4In the process, remote sensing time-series images are first acquired through the system's input layer. Then, the preprocessing module performs radiometric correction, geometric correction, image registration, and cloud detection. Next, the time-series feature extraction module extracts change time-series feature vectors, morphological time-series feature vectors, spectral time-series feature vectors, and texture time-series feature vectors. The process then uses a deep learning model to identify the progress stage. This deep learning model is a progress stage identification model, which includes a time-series convolutional layer, an attention layer, a fully connected layer, and an output layer. Finally, the output module outputs anomaly warnings, progress warnings, progress quantification assessments, and progress stage classifications.

[0160] A specific example of the power grid engineering construction progress monitoring method in this application embodiment is as follows: A provincial power company is responsible for constructing a 500 kV ultra-high voltage transmission line, 280 kilometers long, traversing various terrains including mountains, plains, and rivers, with a total of 560 towers. The total investment for the project is approximately 1.5 billion yuan, and the planned construction period is 18 months.

[0161] The project faces the following challenges: 1) The terrain along the route is complex, with mountainous sections accounting for 60%, making manual inspection difficult and costly; 2) The project has a tight schedule and requires frequent progress monitoring to ensure timely completion; 3) The construction units are scattered, requiring unified management of the progress of each section; 4) Some areas are ecologically sensitive and require reduced human intervention.

[0162] To address the aforementioned issues, the power company adopted the power grid construction progress monitoring method described in this application.

[0163] 1. Data Acquisition Configuration Remote sensing impact data sources: GF-2 satellite (spatial resolution 0.8 meters) was used as the primary data source, supplemented by Sentinel-2 (spatial resolution 10 meters); Monitoring cycle: Images are acquired every 7 days, with a total of approximately 80 images acquired throughout the entire project period; Monitoring range: 200-meter buffer zones on both sides of the line, with a total area of ​​approximately 112 square kilometers; Grid unit division: A 50m × 50m grid is used, which is divided into approximately 44,800 grid units.

[0164] 2. System Deployment Hardware configuration: Deployed in a high-performance computing cluster, equipped with 4 GPU servers (each equipped with 8 NVIDIA V100 GPUs), 20 CPU compute nodes, and a 500TB storage array; Software architecture: A microservice architecture is adopted, including image preprocessing service, feature extraction service, model inference service, progress calculation service, and visualization service; the number of samples correctly identified and the accuracy of identification during the construction phase are also considered. Data processing workflow: After the images are received, they automatically enter the processing pipeline, and the time from the original images to the generation of the final report does not exceed 24 hours.

[0165] 3. Model Training Sample construction: Historical images of 10 power transmission line projects built in the province over the past 5 years were collected, and approximately 50,000 samples were manually annotated, covering 8 construction phases; Model configuration: TCN-Attention model, 4 temporal convolutional layers, dilation coefficient [1, 2, 4, 8], hidden layer dimension 256, 8 attention heads; Training parameters: learning rate 0.001, batch size 32, number of training epochs 100, early stopping strategy is used to prevent overfitting.

[0166] The effects achieved in this embodiment are as follows: 1. Monitoring efficiency has been greatly improved. Traditional manual inspections are conducted every 10 days, requiring 20 staff members to work continuously for 5 days each time, with an annual labor cost of approximately 1.8 million yuan. After adopting the implementation method described in this application, the monitoring cycle is shortened to 7 days, requiring only 2 technicians for system maintenance and result review, with an annual cost of approximately 350,000 yuan, representing a cost reduction of 80.6%.

[0167] 2. Significantly improved recognition accuracy During the project implementation, 200 grid cells were randomly selected for manual verification. The identification results during the construction phase are shown in Table 1 below: Table 1. Identification Results of the Construction Phase

[0168] The average recognition accuracy reached 93.0%, which is 43 percentage points higher than the traditional single-phase method (about 65%).

[0169] 3. Accurate quantitative assessment of progress Taking the progress in the 12th month as an example, the system calculates the progress at each level as follows: Overall progress at the line level: 67.8% Progress of sub-projects: Road clearing work: 100% (completed) Foundation engineering: 85.3% (foundations for 476 / 560 towers have been completed) Tower engineering: 45.2% (253 / 560 towers have been erected) Conductor Engineering: 12.5% ​​(35 / 280 km of conductor has been erected) Progress comparison at the section level (selecting 3 representative sections) is shown in Table 2 below: Table 2 Comparison of Progress at the Section Level

[0170] The system monitoring results were largely consistent with the manually reported results, with an average deviation of 4.1%. Section B showed a larger deviation; on-site verification revealed that this section had misreported progress, which the system promptly detected and issued an early warning.

[0171] 4. Accurate and reliable schedule forecasting Based on historical progress data from the first nine months, the system predicts the progress for months 10-12. The prediction results are compared with the actual progress in Table 3 below: Table 3 Comparison of Predicted Results and Actual Progress

[0172] The average forecast error over three months was 1.2%, indicating high forecast accuracy and providing a reliable basis for project scheduling.

[0173] 5. Timely and effective early warning of anomalies The system triggered 12 warnings during the project implementation, including: Progress delay warnings: 5 times, mainly due to weather conditions (continuous rainfall) causing construction to be suspended; Local anomaly warnings: 4 times, 3 instances of delayed tower foundation construction and 1 instance of abnormally slow conductor erection were detected; Construction anomaly warnings: 3 times, identifying 2 foundations that have not been poured for a long time after excavation and 1 tower that has not been erected with conductors for a long time after its erection.

[0174] All early warnings were communicated to the construction and supervision units immediately, enabling timely measures to avoid project delays. For example, the delay warning for Section B prompted the construction unit to increase resource investment, ultimately catching up with the planned schedule.

[0175] 6. Effective cloud cover handling During the project, severe cloud cover occurred in southern China during the rainy season (months 5-7), with approximately 40% of the imagery obscured by clouds. The system effectively compensated for the obscured data through temporal interpolation and multi-source data fusion. The relationship between cloud coverage and effective monitoring rate is shown in Table 4 below: Table 4 Relationship between cloud cover rate and effective monitoring rate

[0176] As shown in Table 4 above, the overall effective monitoring rate reached 96.8%, which is 26.8 percentage points higher than that of traditional methods (approximately 70%).

[0177] This embodiment also provides a power grid engineering construction progress monitoring device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0178] This embodiment provides a power grid engineering construction progress monitoring device, such as... Figure 5 As shown, the device includes: The grid data construction module 501 is used to construct multiple grid cell data corresponding to the remote sensing time series images based on the remote sensing time series images of the target monitoring area, according to the time requirements of the power grid project construction schedule. The temporal feature extraction module 502 is used to extract multiple temporal feature vectors corresponding to each grid cell data for each grid cell data. The progress stage identification module 503 is used to input multiple time-series feature vectors corresponding to each grid cell data into a pre-trained progress stage identification model for identification, so as to obtain the progress stage and confidence level corresponding to the grid cell data. The current progress determination module 504 is used to determine the current progress time of each grid cell data according to the confidence level of the grid cell data and the preset weight of the progress stage to which the grid cell data belongs. The overall progress determination module 505 is used to determine the total progress time of the power grid lines in the target monitoring area based on the current progress time corresponding to the data of each grid unit.

[0179] The power grid construction progress monitoring device provided in this embodiment of the invention can execute the power grid construction progress monitoring method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method. Further functional descriptions of the various modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0180] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0181] The following is a detailed reference. Figure 6 , Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0182] The following is a detailed reference. Figure 6This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 601, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 602 or a program loaded from memory 608 into random access random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the electronic device. The processor 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0183] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0184] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a memory 608, or installed from a ROM 602. When the computer program is executed by the processor 601, it performs the functions defined in the power grid engineering construction progress monitoring method of the embodiments of the present invention.

[0185] Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0186] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the power grid engineering construction progress monitoring method shown in the above embodiments is implemented.

[0187] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0188] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for monitoring the construction progress of a power grid project, characterized in that, The method includes: Based on the remote sensing time-series images of the target monitoring area, and in accordance with the time requirements of the power grid project construction schedule, multiple grid cell data corresponding to the remote sensing time-series images are constructed. For each grid cell data, extract multiple temporal feature vectors corresponding to the grid cell data; For each grid cell data, multiple time-series feature vectors corresponding to the grid cell data are input into a pre-trained progress stage recognition model to obtain the progress stage and confidence level corresponding to the grid cell data. For each grid cell data, the current progress time corresponding to the grid cell data is determined based on the confidence level corresponding to the grid cell data and the preset weight corresponding to the progress stage to which the grid cell data belongs. The total progress time of the power grid lines in the target monitoring area is determined based on the current progress time corresponding to the data of each grid cell.

2. The method according to claim 1, characterized in that, The method further includes: Obtain the historical progress time corresponding to each grid cell data; For each grid cell data, the current progress time and historical progress time corresponding to the grid cell data are input into a pre-trained progress prediction model to predict the expected progress time corresponding to the grid cell data. For each grid cell data, calculate the deviation between the current progress time and the expected progress time corresponding to the grid cell data; For each grid cell data, when the deviation between the current progress time corresponding to the grid cell data exceeds a preset threshold, an abnormal warning action is performed on the grid cell data to obtain abnormal warning information.

3. The method according to claim 2, characterized in that, The method further includes: Obtain the cloud coverage rate for each grid cell; For each grid cell data, when the cloud coverage exceeds a preset coverage, the grid cell data is subjected to cloud occlusion processing to obtain the cloud data processing result.

4. The method according to claim 3, characterized in that, The method further includes: For each grid cell data, a progress monitoring report and visualization chart are generated based on the current progress time corresponding to the grid cell data, the total progress time of the power grid lines in the target monitoring area, the abnormal early warning information, and the cloud data processing results.

5. The method according to claim 1, characterized in that, The remote sensing time-series images of the target monitoring area are obtained through the following preprocessing steps; Acquire the original time-series images of the remote sensing time-series images of the target monitoring area; The original time-series images are sequentially subjected to radiometric correction, geometric fine correction, image registration, cloud detection, and image fusion processing to obtain remote sensing time-series images of the target monitoring area.

6. The method according to any one of claims 1 to 5, characterized in that, Based on the remote sensing time-series imagery of the target monitoring area, and in accordance with the time requirements of the power grid project construction schedule, multiple grid cell data corresponding to the remote sensing time-series imagery are constructed, including: Based on the remote sensing time-series images of the target monitoring area, determine the node data and edge data of the power grid topology; The target monitoring area is divided into multiple grid units according to a preset size; The node data and the edge data are mapped into the multiple grid cells to obtain multiple grid cell data corresponding to the remote sensing time series image. Each grid cell data includes grid attribute information. Based on the construction schedule requirements of the power grid project, the schedule requirements are added to the grid attribute information of each grid cell data to obtain multiple grid cell data corresponding to the remote sensing time series image.

7. The method according to claim 1, characterized in that, The time-series feature vector includes: a spectral time-series feature vector and a spectral rate of change feature vector. For each grid cell data, multiple time-series feature vectors corresponding to the grid cell data are extracted, including: For each grid cell data, obtain the near-infrared band reflectance, red band reflectance, short-wave infrared band reflectance, and bare soil index corresponding to the grid cell data; For each grid cell data, the normalized vegetation index corresponding to the grid cell data is calculated based on the near-infrared band reflectance and the red band reflectance corresponding to the grid cell data. For each grid cell data, the normalized building index corresponding to the grid cell data is calculated based on the shortwave infrared band reflectance and near-infrared band reflectance corresponding to the grid cell data. Based on the normalized vegetation index, the normalized building index, and the bare soil index corresponding to each grid cell data, they are arranged into a spectral temporal feature vector according to the time series. Calculate the spectral change rate between two adjacent time points based on the spectral time series feature vector; Based on the spectral change rate at two adjacent time points, arrange them into a spectral change rate feature vector according to the time series.

8. The method according to claim 7, characterized in that, The temporal feature vector includes: a texture temporal feature vector and a texture change rate feature vector. For each grid cell data, multiple temporal feature vectors corresponding to the grid cell data are extracted, and the feature vector also includes: Obtain the grayscale parameter values, grayscale co-occurrence normalization parameters, and correlation parameters corresponding to the data of two adjacent grid cells; Calculate the contrast parameter of the two adjacent grid cell data based on the gray level parameter value and gray level co-occurrence normalization parameter corresponding to the data of the two adjacent grid cells. Based on the contrast parameters and correlation parameters corresponding to the data of two adjacent grid cells, they are arranged into a texture temporal feature vector according to the time sequence. Based on the texture temporal feature vector, calculate the texture change rate between two adjacent time points; Based on the texture change rate of two adjacent time points, a texture change rate feature vector is arranged according to the time sequence.

9. The method according to claim 7, characterized in that, The temporal feature vector includes: morphological temporal feature vector and change temporal feature vector. For each grid cell data, multiple temporal feature vectors corresponding to the grid cell data are extracted, and the feature vector also includes: For each grid cell data, obtain the area change parameters and perimeter change parameters corresponding to the grid cell data; For each grid cell data, the shape index change parameter corresponding to the grid cell data is calculated based on the area change parameter and perimeter change parameter of the grid cell. Based on the area change parameters, perimeter change parameters, and shape index change parameters corresponding to each grid cell data, they are arranged into a morphological temporal feature vector according to the time series.

10. The method according to claim 1, characterized in that, The progress stage identification model is trained through the following steps: Acquire training samples of remote sensing images at various progress stages; Extract training features from remote sensing images for each progress stage; For each progress stage, the remote sensing image training features are input into the progress stage recognition model for recognition. The recognition result of the remote sensing image training features and the loss between the recognition result of the remote sensing image training features and the true label of the remote sensing image training features are obtained. The loss is used to update the parameter values ​​of the progress stage recognition model. The progress stage recognition model is composed of an input layer, a temporal convolutional layer, an attention layer, a fully connected layer and an output layer.

11. A power grid construction progress monitoring device, characterized in that, The device includes: The grid data construction module is used to construct multiple grid cell data corresponding to the remote sensing time series images of the target monitoring area according to the time requirements of the power grid project construction schedule. The temporal feature extraction module is used to extract multiple temporal feature vectors corresponding to each grid cell data. The progress stage identification module is used to input multiple time-series feature vectors corresponding to each grid cell data into a pre-trained progress stage identification model for identification, so as to obtain the progress stage and confidence level corresponding to the grid cell data. The current progress determination module is used to determine the current progress time corresponding to each grid cell data based on the confidence level of the grid cell data and the preset weight corresponding to the progress stage to which the grid cell data belongs. The overall progress determination module is used to determine the total progress time of the power grid lines in the target monitoring area based on the current progress time corresponding to the data of each grid unit.

12. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the power grid engineering construction progress monitoring method according to any one of claims 1 to 10.

Images