Transmission and transformation project progress identification method and device based on space-air-ground integration technology

Abstract

The invention provides a power transmission and transformation project progress identification method and device based on a space-air-ground integration technology, and relates to the field of project progress management. The method comprises the following steps: acquiring target image data in an engineering construction area through a plurality of acquisition devices, and preprocessing the target image data; performing feature extraction on the preprocessed image data to obtain project progress feature information corresponding to the project construction area; classifying the project progress feature information; based on the classified project progress feature information, matching is carried out in combination with a project progress feature library to output the current project progress corresponding to the project construction area; and if it is confirmed that the progress deviation value between the current project progress and the planned project progress is larger than the preset deviation value, project progress early warning is triggered. According to the method, the problem that the power transmission and transformation project progress is identified through a single monitoring technology, and the identification accuracy of the power transmission and transformation project progress is generally low is solved.

Description

Technical Field

[0001] The present application relates to the field of project progress management, and in particular to a method and device for identifying the progress of power transmission and transformation projects based on air-ground-space integrated technology. Background Art

[0002] Traditional progress monitoring of power transmission and transformation projects mainly relies on manual on-site inspection records. This method consumes a lot of manpower and material resources, has a long inspection cycle, cannot achieve real-time monitoring, and is prone to missing key progress nodes.

[0003] Traditional methods for identifying the progress of power transmission and transformation projects typically rely on a single technology, such as ground-based camera monitoring, which has a limited range. This makes it difficult to fully cover complex terrain and large construction areas, and limits information collection. Some methods also utilize drone aerial photography, a single monitoring technology, but this method performs poorly when monitoring progress in shadowed areas and has limitations in information collection. Therefore, identifying the progress of power transmission and transformation projects using a single monitoring technology generally suffers from low accuracy.

[0004] Therefore, there is an urgent need for a method and device for identifying the progress of power transmission and transformation projects based on integrated air-ground-space technology. Summary of the Invention

[0005] The present application provides a method and device for identifying the progress of power transmission and transformation projects based on integrated air-ground-space technology, which solves the common problem of low accuracy in identifying the progress of power transmission and transformation projects when using a single monitoring technology.

[0006] In a first aspect of the present application, a method for identifying the progress of a power transmission and transformation project based on integrated air-ground-space technology is provided, the method comprising: acquiring target image data in a construction area through multiple acquisition devices, and performing preprocessing operations on the target image data; performing feature extraction on the preprocessed image data to obtain construction progress feature information corresponding to the construction area; classifying the construction progress feature information; based on the classified construction progress feature information, matching is performed in combination with a construction progress feature library to output the current construction progress corresponding to the construction area; if it is confirmed that the progress deviation value between the current construction progress and the planned construction progress is greater than a preset deviation value, triggering a construction progress warning.

[0007] Optionally, target image data in the construction area is obtained through multiple acquisition devices, specifically including: periodically photographing the construction area through satellite remote sensing equipment to obtain satellite image data; taking multi-angle aerial photos of the construction area through unmanned aerial vehicle equipment to obtain aerial image data; taking close-up photos of the construction area through intelligent robot equipment to obtain complex area image data; and using satellite image data, aerial image data and complex area image data as target image data.

[0008] Optionally, the target image data is preprocessed, specifically including: performing radiometric correction on the satellite image data to eliminate radiometric errors, and performing geometric correction on the satellite image data to correct the satellite image data to the correct geographic coordinate system; performing stitching and image fusion on the aerial image data, and performing denoising on the fused image data; performing frame extraction on the complex area image data to decompose the video into single-frame images, and performing size normalization on the complex area image data and each frame image.

[0009] Optionally, feature extraction is performed on the preprocessed image data to obtain engineering progress feature information corresponding to the engineering construction area, specifically including: inputting the preprocessed target image data into a convolutional neural network model, and outputting engineering progress feature information based on the convolutional neural network model; performing high-level semantic feature extraction on the preprocessed target image data based on the engineering progress feature information, the high-level semantic features including substation building features and transmission line features.

[0010] Optionally, the project progress feature information is classified, specifically including: inputting the project progress feature information into a pre-trained convolutional neural network model constructed with a ResNet50 network as a backbone network; connecting a classification head including two fully connected layers to the back end of the ResNet50 network, wherein the first fully connected layer contains multiple neurons and uses a ReLU activation function, and the number of neurons in the second fully connected layer is consistent with the number of project progress stages; performing linear transformation processing on the project progress feature information through the first fully connected layer; and classifying the linearly transformed project progress feature information through the second fully connected layer.

[0011] Optionally, based on the classified project progress feature information, matching is performed in combination with a project progress feature library to output the current project progress corresponding to the project construction area, specifically including: establishing a project progress feature library based on the construction process and construction characteristics of the project construction area; using the data in the project progress feature library as training data to construct a project progress recognition model, and outputting the current project progress based on the project progress recognition model.

[0012] Optionally, the current project progress is output based on the project progress recognition model, specifically including: based on the project progress recognition model, using the Softmax activation function to generate a normalized probability distribution of each project progress stage; performing structured processing on the normalized probability distribution, and taking the project progress corresponding to the maximum probability value as the first candidate project progress; obtaining a historical project stage sequence based on historical image data, and constructing a stage confidence evaluation function based on the dynamic evolution trend between the first candidate project progress and the historical project stage sequence; outputting a first confidence score corresponding to the first candidate project progress based on the stage confidence evaluation function; if it is confirmed that the first confidence score is less than a preset threshold, returning to the image acquisition step and reacquiring the target image data; based on the target image data, outputting a second confidence score corresponding to the second candidate project progress through the project progress recognition model; if it is confirmed that the second confidence score is greater than or equal to the preset threshold, outputting the second candidate project progress as the current project progress.

[0013] In a second aspect of the present application, a device for identifying the progress of a power transmission and transformation project based on air-ground-space integration technology is provided. The device includes an acquisition module and a processing module, wherein: The acquisition module is used to acquire target image data in the construction area through multiple acquisition devices and perform preprocessing operations on the target image data.

[0014] The processing module is used to extract features from the pre-processed image data to obtain engineering progress feature information corresponding to the engineering construction area; classify the engineering progress feature information; match the classified engineering progress feature information with the engineering progress feature library to output the current engineering progress corresponding to the engineering construction area; if it is confirmed that the progress deviation value between the current engineering progress and the planned engineering progress is greater than the preset deviation value, a project progress warning is triggered.

[0015] In the third aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device performs any of the methods described above.

[0016] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to perform any of the above methods.

[0017] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. Acquire target image data in the construction area through multiple acquisition devices and pre-process the target image data; extract features from the pre-processed image data to obtain engineering progress feature information corresponding to the construction area; classify the engineering progress feature information; based on the classified engineering progress feature information, match it with the engineering progress feature library to output the current engineering progress corresponding to the construction area; if it is confirmed that the progress deviation value between the current engineering progress and the planned engineering progress is greater than the preset deviation value, trigger a project progress warning, thereby achieving full-process, full-space, and automated accurate identification and dynamic monitoring of the progress status of the power transmission and transformation project construction area. Without relying on manual inspections, it can efficiently grasp the changes in the project stages, improve the real-time and intelligent level of project management, promptly discover construction delays or abnormal progress, and assist the project scheduling system in decision-making adjustments and resource optimization allocation to ensure consistency and controllability between the construction progress and the planned goals, effectively reducing the risk of delays and construction costs.

[0018] 2. Perform radiometric correction on satellite image data to eliminate radiometric errors, and perform geometric correction on satellite image data to calibrate the satellite image data to the correct geographic coordinate system; perform stitching and image fusion on aerial image data, and perform denoising on the fused image data; perform frame extraction on complex area image data to decompose the video into single-frame images, and perform size normalization on complex area image data and each frame image, thereby constructing a multi-source target image data set with a unified spatial reference, clear visual quality, and standard input format, providing a high-quality and highly consistent input basis for feature extraction, project progress classification, and stage identification of subsequent convolutional neural network models, ensuring the model's stability and generalization capabilities when processing images across scales, devices, and time periods, improving the accuracy and robustness of progress identification, and providing reliable image expression support for full-cycle project monitoring and data-driven scheduling.

[0019] 3. Based on the construction process and construction characteristics of the construction area, a construction progress feature library is established; the data in the construction progress feature library is used as training data to construct a construction progress recognition model, and the current construction progress is output based on the construction progress recognition model, thereby realizing deep modeling and automatic recognition of the semantic features and progress stage information implicit in the construction images, so that the construction progress recognition model has end-to-end stage judgment capabilities. By learning the typical structural features of different construction stages in the image dimension, it can efficiently and accurately infer the construction stage status of the current construction area in actual operation, reduce the dependence on manual labeling and rule design, improve the recognition system's adaptability and judgment accuracy in complex scenes, multi-stage overlap and fuzzy boundary situations, enhance the system's generalization ability for multi-source image input, and provide a stable and intelligent recognition foundation for dynamic perception and scheduling control of construction progress. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flow chart of a method for identifying the progress of a power transmission and transformation project based on air-ground-ground integrated technology provided in an embodiment of the present application; Figure 2 This is a module diagram of a power transmission and transformation project progress identification device based on air-ground-space integration technology provided in an embodiment of the present application; Figure 3 This is a structural diagram of an electronic device provided in an embodiment of the present application.

[0021] Explanation of the reference numerals: 21, acquisition module; 22, processing module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0023] The terms used in the following examples of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application, the singular expressions "a", "an", "said", "above", "the", and "this" are intended to include plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more of the listed items.

[0024] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more.

[0025] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0026] Please refer to Figure 1 , which shows a flow chart of a method for identifying the progress of a power transmission and transformation project based on air-ground-space integrated technology provided in an embodiment of the present application. The flow chart mainly includes the following steps: S101 to S105.

[0027] Step S101: acquiring target image data in a construction area through a plurality of acquisition devices, and performing a pre-processing operation on the target image data.

[0028] Specifically, target image data within the construction area is acquired through multiple acquisition devices, including satellite remote sensing equipment, drones, and intelligent robotics. Different preprocessing operations are performed on the target image data based on the acquisition device to construct input data with analytical value.

[0029] In a possible implementation, step S101 also includes: periodically photographing the construction area through satellite remote sensing equipment to obtain satellite image data; performing multi-angle aerial photography of the construction area through drone equipment to obtain aerial image data; performing close-up photography of the construction area through intelligent robot equipment to obtain complex area image data; and using satellite image data, aerial image data, and complex area image data as target image data.

[0030] Specifically, satellite remote sensing equipment is used to periodically photograph the construction area. Satellite remote sensing equipment refers to an imaging payload system carried on an orbiting satellite platform, which usually includes visible light, multispectral or panchromatic sensors. This type of equipment obtains wide-area, panoramic images of the project through periodic orbital coverage without the need for on-site deployment. Periodic photography refers to repeated observations of the same construction area at set time intervals, such as daily, weekly, or every ten days, to obtain time-series image data that reflects the macro-progress of the project. Satellite image data has a wide coverage area and strong geographic registration capabilities, and can provide information such as the overall progress of the construction area, site layout, and changes in material stacking.

[0031] Multi-angle aerial photography of the construction area is carried out using drone equipment. The drone equipment includes fixed-wing drones and multi-rotor drones, equipped with high-resolution optical cameras, and has the ability to flexibly perform low-altitude imaging tasks at different heights, angles, and paths. The so-called multi-angle aerial photography refers to the drone collecting images of the target area from multiple perspectives such as vertical, oblique, and surround shooting directions according to the preset mission path, so as to enhance the image's ability to express three-dimensional structures and the visibility of target details. The drone's task scheduling is centrally configured by the management platform. It can choose between immediate execution mode for emergency photography, scheduled execution mode for planned sampling, and periodic execution mode for continuous monitoring. For example, immediate execution, scheduled execution, and periodic execution (daily, weekly, and monthly) can ensure continuous coverage of key stages and key structures during the construction process.

[0032] Close-up photography of construction areas is performed using intelligent robotic equipment. Intelligent robotic equipment refers to working platforms deployed on the ground at the construction site, capable of autonomous walking, positioning, navigation, and image acquisition. These include wheeled robots, tracked robots, or robots with biomimetic gaits. This type of equipment can enter complex areas difficult for humans to reach, such as the interior of pipe corridors, around foundation pits, and between tower bases, enabling high-precision, fixed-point photography of detailed locations. Close-range photography involves the robot approaching the target structural unit within tens of centimeters to several meters, using its onboard high-magnification imaging module to capture high-texture, high-definition images of the area. This is used to identify construction details and construct feature markers.

[0033] Satellite imagery, aerial photography, and complex regional imagery acquired by the three aforementioned types of equipment are combined as target image data to form a collection of raw images with spatially hierarchical distribution characteristics. This collection encompasses multiple granularity perspectives, from macroscopic (satellite) to mesoscale (drone) to microscopic (robot) in the spatial dimension. In the temporal dimension, it integrates multiple timescales, from periodic monitoring (satellite) to medium-frequency patrols (drone) to real-time capture (robot). This provides a comprehensive, detailed, and timely data foundation for subsequent project progress analysis.

[0034] In one possible implementation, step S101 further includes: performing radiometric correction on the satellite image data to eliminate radiometric errors, and performing geometric correction on the satellite image data to correct the satellite image data to a correct geographic coordinate system; performing stitching and image fusion on the aerial image data, and performing denoising on the fused image data; performing frame extraction on the complex area image data to decompose the video into single-frame images, and performing size normalization on the complex area image data and each frame image.

[0035] Specifically, satellite image data has significant radiometric errors and geometric distortions due to its acquisition process being affected by atmospheric conditions, solar radiation angles, and sensor response differences. To improve image radiometric consistency, radiometric correction is required. Radiometric correction includes atmospheric correction, sensor gain normalization, and surface reflectivity inversion, and its mathematical expression can be expressed as:

[0036]

[0037] in, is the top-of-atmosphere radiance, is the pixel gray value, and are the calibration offset and gain coefficient of the sensor respectively, is the surface reflectivity, is the distance from the Earth to the Sun (astronomical units), is the solar irradiance in this band, The above correction steps can significantly reduce the image brightness deviation caused by imaging time or sensor differences.

[0038] Then, geometric correction is performed to eliminate the spatial offset caused by factors such as imaging angle changes and orbital disturbances. By selecting ground control points (GCPs) with clear spatial coordinates, a polynomial mapping relationship between the image and the geographic reference map is constructed. The commonly used geometric transformation model is a quadratic or cubic polynomial function:

[0039] in, Represents the original image pixel coordinates, represents the coordinates of the image after correction, and is the fitting coefficient, Indicates the order of the polynomial geometric transformation. This process aligns the imagery with the GIS geographic coordinate system, ensuring spatial consistency across time series data.

[0040] Secondly, for aerial image data, due to the overlapping areas and perspective differences between images caused by multi-angle and multi-track imaging by drones, stitching and fusion processing are required. The stitching process first uses feature matching algorithms such as SIFT (Scale-Invariant Feature Transform) or SURF (Speeded Robust Features) to extract image key points and calculate feature descriptors. Feature points are paired using nearest neighbor search and ratio test methods, and the affine transformation matrix is ​​estimated using the RANSAC algorithm:

[0041] Among them, the affine matrix parameters 、 、 Obtained by fitting matching points. After geometric alignment, image fusion is performed to eliminate brightness discontinuities in overlapping areas. Fusion methods can include pyramid fusion, multi-resolution weighted averaging, or Poisson fusion. The final output is a complete, continuous, and uniformly viewed image of the construction area.

[0042] After the fused image is stitched, there is local noise accumulation. To improve visual clarity, it is necessary to perform denoising on the image. Common methods include median filtering and Gaussian filtering. The Gaussian filter expression is:

[0043] in, Represents a two-dimensional Gaussian function at position The weight value when The standard deviation of the Gaussian kernel determines the degree of blur; the kernel is convolved with the image to achieve noise smoothing.

[0044] Finally, for complex area image data, since its acquisition method is often video stream data collected by robotic equipment, frame extraction is required to achieve static image processing requirements. Frame extraction is based on the set frame rate Extract a sequence of image frames at equal intervals from a video stream , such that:

[0045] in, Indicates the time interval between two consecutive frame images. The subsequent frame image and the original image need to be normalized to the specified input size. , commonly used methods are bilinear interpolation or areal interpolation:

[0046] in The current interpolation calculation Rank Column interpolation weights, For the original picture, This is the scaled image. By unifying the image resolution and aspect ratio, batch consistency and feature scale stability are guaranteed during subsequent model processing.

[0047] Step S102 : performing feature extraction on the pre-processed image data to obtain engineering progress feature information corresponding to the engineering construction area.

[0048] In a possible implementation, step S102 further includes: inputting the preprocessed target image data into a convolutional neural network model, and outputting engineering progress feature information based on the convolutional neural network model; and extracting high-level semantic features from the preprocessed target image data based on the engineering progress feature information, where the high-level semantic features include substation building features and transmission line features.

[0049] Specifically, the convolutional neural network model utilizes a deep network architecture with ResNet50 as its backbone. Through multi-level convolutional layers, residual connections, and a global receptive field construction mechanism, it extracts characteristic features such as edges, textures, structures, and spatial distribution from the input image layer by layer. The model accepts image input in a uniform format and sequentially undergoes operations such as convolution kernel scanning, activation function mapping, pooling dimensionality reduction, and residual skipping. Ultimately, it forms a compact representation of the project progress in a high-level feature map, capturing the key semantic features of the target image in its current state.

[0050] Subsequently, based on the project progress information, targeted high-level semantic feature extraction is performed to further identify typical structural differences and visual morphological changes exhibited by substation buildings and transmission lines at different construction stages. High-level semantic features refer to high-order spatial structural information and image content semantic labels that can be used to distinguish different construction stages. They provide engineering semantic clarity and construction stage identification. Specifically, substation building features at different construction stages include: ground texture, color distribution, and obstacle outlines during the site leveling stage; foundation pit shape, soil color and texture, and construction machinery appearance during the foundation construction stage; building structure outline, building material color and texture, and construction personnel and equipment distribution during the civil construction stage; and electrical equipment appearance during the equipment installation stage. Transmission line features at different construction stages include: foundation shape and size, foundation surface texture, and surrounding soil disturbance during the foundation construction stage; tower appearance, color and material characteristics, and tower construction scene characteristics during the tower assembly stage; and visual characteristics of conductors and lightning conductors, equipment characteristics, and line sag during the line stringing stage.

[0051] Step S103: classify the engineering progress characteristic information.

[0052] In a possible implementation, step S103 further includes: inputting the project progress feature information into a pre-trained convolutional neural network model constructed with the ResNet50 network as the backbone network; connecting a classification head including two fully connected layers to the back end of the ResNet50 network, wherein the first fully connected layer contains multiple neurons and uses a ReLU activation function, and the number of neurons in the second fully connected layer is consistent with the number of project progress stages; performing linear transformation processing on the project progress feature information through the first fully connected layer; and classifying the linearly transformed project progress feature information through the second fully connected layer.

[0053] Specifically, the engineering progress feature information extracted in the previous stage is used as input and sent to a pre-trained convolutional neural network model built with the ResNet50 network as the backbone network to perform engineering progress classification operations. The ResNet50 network is a deep residual structure network that contains 50 network layers and has residual connection units. Its core advantage is that it can effectively alleviate the gradient vanishing and degradation problems in deep network training, thereby ensuring the effective transmission and expression of deep semantic features. The model has been pre-trained on large-scale image recognition datasets (such as ImageNet) and has excellent image understanding and feature generalization capabilities. Among them, the model training parameters are: Optimizer: The Adam optimizer is selected. The Adam optimizer combines the advantages of the adaptive gradient algorithm (AdaGrad) and the root mean square propagation algorithm (RMSProp). It can converge quickly in the early stage of training and stably optimize the model parameters in the later stage; Loss function: The cross entropy loss function is used, and the formula is:

[0054] in, represents the total cross entropy loss, is the total number of samples, is the total number of categories, Indicates the Does the sample belong to class (true label); The model predicts The samples belong to The probability of the class, is the logarithm of the predicted probability.

[0055] After the project progress feature information completes convolution, normalization, and pooling in the ResNet50 backbone network, a customized classification head is connected to the network tail. The classification head consists of two fully connected layers. The first fully connected layer performs a linear transformation on the high-dimensional feature vector output by the convolutional network. The number of neurons in the first layer is set to an intermediate dimension (for example, 512 or 1024), and the ReLU activation function is used to achieve nonlinear enhancement of the feature space. Its mathematical expression is:

[0056] in Represents the engineering progress feature vector output by ResNet50, and are the weight matrix and bias vector of the first fully connected layer, Output features for the first layer.

[0057] then, The feature vector is input to the second fully connected layer for classification. The number of neurons in this layer is equal to the total number of project progress classification categories (e.g., 8 stages). Nonlinear activation is no longer used, but the original classification vector is directly output. The probability distribution corresponding to each category is then calculated through the Softmax function:

[0058] in and For the The weights and biases corresponding to the categories, is the total number of categories, Indicates that the input feature belongs to The probability of the project progress stage.

[0059] Finally, the model output is a length of The probability vector represents the matching probability of the current input image at each construction progress stage. The index with the highest probability value is used as the construction progress stage to which the current image belongs, thus completing the classification and recognition of the preprocessed target image. This step accurately maps complex image semantic features into a predefined construction stage space through a combination of linear transformation and nonlinear mapping, achieving end-to-end inference from image to construction semantic state. It should be noted that during training, before each round of training, the model is set to training mode (model.train()), with the Dropout and BN layers enabled. After each round of training, the model is switched to evaluation mode (model.eval()), with the Dropout and BN layers disabled. Model performance is evaluated on the validation set, monitoring changes in loss and accuracy. If the validation set accuracy does not improve significantly after 10 consecutive rounds of training, a learning rate decay strategy (multiplying the learning rate by 0.5) is implemented to prevent the model from falling into a local optimum until it reaches a stable and high accuracy on the validation set.

[0060] Step S104 : Based on the classified engineering progress feature information, matching is performed in combination with the engineering progress feature library to output the current engineering progress corresponding to the engineering construction area.

[0061] Specifically, based on the classified construction progress feature information, a feature matching operation is performed using a pre-built construction progress feature library. This identifies the standard progress template that most closely matches the current feature information and, based on this, outputs the current construction progress stage of the construction area. The construction progress feature library contains typical image feature representation structures for different stages, providing semantic comparison and judgment support for the classification results. The matching results verify the stage accuracy of the classification output and improve the stability and interpretability of the recognition.

[0062] In a possible implementation, step S104 further includes: establishing a project progress feature library based on the construction process and construction characteristics of the project construction area; using the data in the project progress feature library as training data to construct a project progress recognition model, and outputting the current project progress based on the project progress recognition model.

[0063] Specifically, in this step, different construction stages of a substation have different characteristics. For example, the foundation construction stage can be determined by the pouring of the tower base, and the main structure construction stage can be determined by the degree of completion of the building structure. These characteristics are quantified and classified to establish a project progress feature library. See Table 1 for the feature information of the project progress feature library.

[0064] [Table 1]

[0065] Collect a large amount of image data from the sky and ground at different stages of the power transmission and transformation project area, including satellite remote sensing images (resolution up to 0.5 meters), drone aerial images (resolution 1-5 cm) and high-definition photos collected by ground equipment.

[0066] Organize professional power engineering personnel and image annotation teams to annotate images according to the project progress stages (foundation construction, tower assembly, line construction, equipment installation, commissioning and acceptance, etc., a total of 8 stages).

[0067] The LabelMe tool is used to accurately label key areas and features related to project progress in the image, such as the tower foundation pouring area and the transmission tower erection structure, forming a dataset with accurate progress stage labels to form a project progress feature library.

[0068] Afterwards, the data in the project progress feature library was used as training data to construct a project progress recognition model. The model was constructed as follows: First, based on the multi-stage construction sample images contained in the project progress feature library, the images were classified and sorted according to the annotated project progress labels to form a standard training sample set with stage identification; secondly, a convolutional neural network model with ResNet50 as the backbone network was adopted, which received the above sample images as input, extracted multi-scale image semantic features through multi-layer convolutional layers, and connected to a fully connected classification layer at the end of the model to output a classification vector corresponding to the number of project stage categories; during the training stage, the model used the cross-entropy loss function as the objective function to minimize the error between the predicted classification and the true label.

[0069] To enhance the generalization ability of the model, data augmentation strategies are introduced during the training process, including image rotation, horizontal flipping, brightness perturbation, and Gaussian blur, to construct a richer input distribution. After each round of training, the accuracy and loss indicators are evaluated in the validation set. If the accuracy does not improve after several consecutive rounds of verification, the learning rate is automatically adjusted and the optimal model weights are saved.

[0070] After training, the model is able to input any pre-processed image and output the corresponding engineering stage prediction results, providing a stable prediction foundation for subsequent stage classification, confidence inference, and dynamic judgment. The stage probability distribution output by the model serves as the source of the Softmax normalized probability.

[0071] In a possible implementation, step S104 further includes: based on the project progress recognition model, using the Softmax activation function to generate a normalized probability distribution for each project progress stage; performing structured processing on the normalized probability distribution, and taking the project progress corresponding to the maximum probability value as the first candidate project progress; obtaining a historical project stage sequence based on historical image data, and constructing a stage confidence evaluation function based on the dynamic evolution trend between the first candidate project progress and the historical project stage sequence; outputting a first confidence score corresponding to the first candidate project progress based on the stage confidence evaluation function; if it is confirmed that the first confidence score is less than a preset threshold, returning to the image acquisition step and reacquiring the target image data; based on the target image data, outputting a second confidence score corresponding to the second candidate project progress through the project progress recognition model; if it is confirmed that the second confidence score is greater than or equal to the preset threshold, outputting the second candidate project progress as the current project progress.

[0072] Specifically, first, the target image is input into the project progress recognition model, and the normalized probability distribution of the stage prediction is generated based on the Softmax function of the classification layer at the end of the model. Let the unnormalized classification vector output by the model be ,in Indicates the total number of predefined project progress stages, that is, the number of stage categories set in the model classification task, such as site leveling, foundation construction, civil construction, equipment installation, etc. Indicates the engineering progress identification model for the The unnormalized scores output by each engineering progress stage, the Softmax output probability vector is:

[0073] The normalized vector Represents the predicted confidence probability at each stage, Indicates that the normalized image belongs to The predicted confidence probability of a project progress stage, that is, the relative probability that the image is judged to be in this stage, Represents the model for all stages ∈{1,…, The output logit score is then structured and the phase index with the largest probability value is defined as the first candidate project progress:

[0074] Then, the current target area is extracted from the engineering monitoring system in the past The historical stage sequence corresponding to the moment , For the The historical stages corresponding to the moments are constructed, and a stage consistency model based on time evolution constraints is constructed. The consistency between the historical stage sequence and the stage confidence evaluation function is defined as follows:

[0075] in, Represents the confidence probability corresponding to the first candidate stage in the Softmax output; is the stage consistency indicator function, when When it is set to 1, otherwise it is set to 0; is the deep feature vector of the current image, For the The feature vector of the image corresponding to the moment; represents the cosine similarity function, which is used to measure the high-dimensional semantic consistency between the current image and the historical images; and satisfy , is an adjustable confidence weighting coefficient that converts the confidence scores of the above stages into With the set threshold For comparison, if , then the current model output is considered to be sufficiently stable and the output is directly As the actual project progress output of the former target area; if , the current model output result is considered to be insufficiently stable, and the image re-collection mechanism is triggered, returning to the image acquisition step to collect target image data at a new perspective or time node; the model reasoning is performed again based on the re-collected image to obtain the second candidate project progress and its corresponding stage confidence score , the score calculation is still carried out according to the above formula. , then The actual engineering progress of the current target area is output as the final recognition result; if . , it means that after the model re-collects the target image and performs secondary inference, the second candidate project progress output by the model still does not reach the preset confidence threshold, that is, the current prediction result is uncertain, unstable or inconsistent with the historical evolution trend in terms of probability credibility, historical consistency, feature stability or inter-class boundary clarity. At this time, at least one of the following processing logics should be executed: block the current round of project progress judgment process and do not output any valid stage judgment results to prevent misjudgment; generate abnormal labeling information, and package and record the input data, Softmax probability distribution, stage confidence and historical stage sequence corresponding to the image; enter the manual review or intelligent scheduling intervention process, and send a "stage recognition failure" signal to the project monitoring platform through the prompt system interface, and optionally request a larger viewing angle, higher resolution or cross-device supplementary image data; count the frequency of stage recognition failure in the current area, if multiple times occur in adjacent continuous time periods or adjacent spatial units If a situation is identified, it is considered a decline in model robustness or data collection anomalies, and the system can initiate model retraining recommendations or data quality warning mechanisms. This strategy combines the confidence of the current classification output, the continuity of historical stage evolution, and feature stability to significantly improve recognition robustness and engineering semantic interpretation capabilities in scenarios with blurred stage boundaries or ambiguous images.

[0076] Step S105: If it is determined that the progress deviation value between the current project progress and the planned project progress is greater than the preset deviation value, a project progress warning is triggered.

[0077] Specifically, the identified current project progress is compared with the planned progress, and the progress deviation between the two is calculated. Based on whether this deviation exceeds a preset deviation threshold, a decision is made as to whether to trigger a project progress warning. Specifically, the system extracts the planned project progress stage that should be reached at the current moment based on the time-stage mapping relationship set in the project scheduling plan. The system then quantifies the stage-level difference between this stage and the actual project progress stage identified by the model. The deviation between the actual and planned progress is then calculated, taking into account the stage number sequence, duration span, or the extension of key nodes.

[0078] If it is determined that the deviation value is greater than the preset threshold, that is, the current project is significantly lagging behind or abnormally ahead of schedule, the system will trigger the project progress warning mechanism, output abnormal prompt information on the monitoring interface, and simultaneously notify the project management personnel and on-site construction managers through the set notification path (such as scheduling platform notification, SMS reminder, email push), and at the same time mark the current deviation time, location unit and judgment basis, so as to timely analyze the cause and adjust the construction rhythm.

[0079] If the deviation does not exceed the preset threshold, meaning the current project progress is within the planned tolerance, no warning is issued. The system then deems the current construction status to be consistent with the expected pace. While slight time differences, overlapping phase boundaries, or minor misjudgments due to multi-source observation errors may exist, these misjudgments are insufficient to disrupt construction scheduling. The system records the current status as "normal" or "warning-free" and maintains continuous tracking, without triggering any manual intervention or scheduling response. This strategy helps reduce false alarm rates and improves the practicality and credibility of warnings.

[0080] Please refer to Figure 2 , which shows a module schematic diagram of a power transmission and transformation project progress identification device based on air-ground integrated technology provided by an embodiment of the present application, the device includes an acquisition module 21 and a processing module 22, wherein, The acquisition module 21 is used to acquire target image data in the construction area through multiple acquisition devices and perform preprocessing operations on the target image data.

[0081] The processing module 22 is used to extract features from the pre-processed image data to obtain engineering progress feature information corresponding to the engineering construction area; classify the engineering progress feature information; match the classified engineering progress feature information with the engineering progress feature library to output the current engineering progress corresponding to the engineering construction area; if it is confirmed that the progress deviation value between the current engineering progress and the planned engineering progress is greater than a preset deviation value, a project progress warning is triggered.

[0082] In one possible embodiment, the acquisition module 21 is used to acquire target image data in the construction area through multiple acquisition devices, specifically including: periodically photographing the construction area through satellite remote sensing equipment to acquire satellite image data; performing multi-angle aerial photography of the construction area through unmanned aerial vehicle equipment to acquire aerial image data; performing close-up photography of the construction area through intelligent robot equipment to acquire complex area image data; and using satellite image data, aerial image data, and complex area image data as target image data.

[0083] In one possible embodiment, the acquisition module 21 is used to perform radiation correction on the satellite image data to eliminate radiation errors, and perform geometric correction on the satellite image data to correct the satellite image data to the correct geographic coordinate system; perform splicing and image fusion on the aerial image data, and perform denoising on the fused image data; perform frame extraction on the complex area image data to decompose the video into single frame images, and perform size normalization on the complex area image data and each frame image.

[0084] In one possible implementation, the processing module 22 is used to perform feature extraction on the preprocessed image data to obtain engineering progress feature information corresponding to the engineering construction area, specifically including: inputting the preprocessed target image data into a convolutional neural network model, and outputting engineering progress feature information based on the convolutional neural network model; performing high-level semantic feature extraction on the preprocessed target image data based on the engineering progress feature information, the high-level semantic features including substation building features and transmission line features.

[0085] In one possible implementation, the processing module 22 is used to classify the project progress feature information, specifically including: inputting the project progress feature information into a pre-trained convolutional neural network model constructed with the ResNet50 network as the backbone network; connecting a classification head including two fully connected layers to the back end of the ResNet50 network, wherein the first fully connected layer contains multiple neurons and uses a ReLU activation function, and the number of neurons in the second fully connected layer is consistent with the number of project progress stages; performing linear transformation processing on the project progress feature information through the first fully connected layer; and classifying the linearly transformed project progress feature information through the second fully connected layer.

[0086] In one possible implementation, the processing module 22 is configured to perform matching based on the classified project progress feature information in combination with a project progress feature library to output the current project progress corresponding to the project construction area. Specifically, the processing module 22 includes: establishing a project progress feature library based on the construction process and construction characteristics of the project construction area; constructing a project progress recognition model using the data in the project progress feature library as training data, and outputting the current project progress based on the project progress recognition model.

[0087] In one possible implementation, the processing module 22 is configured to output the current project progress based on the project progress recognition model, specifically including: generating a normalized probability distribution for each project progress stage based on the project progress recognition model using a Softmax activation function; performing structured processing on the normalized probability distribution, and taking the project progress corresponding to the maximum probability value as the first candidate project progress; acquiring a historical project stage sequence based on historical image data, and constructing a stage confidence evaluation function based on the dynamic evolution trend between the first candidate project progress and the historical project stage sequence; outputting a first confidence score corresponding to the first candidate project progress based on the stage confidence evaluation function; if it is confirmed that the first confidence score is less than a preset threshold, returning to the image acquisition step and reacquiring the target image data; based on the target image data, outputting a second confidence score corresponding to the second candidate project progress through the project progress recognition model; if it is confirmed that the second confidence score is greater than or equal to the preset threshold, outputting the second candidate project progress as the current project progress.

[0088] It should be noted that the above embodiments provide devices that implement their functions using only the division of the above functional modules as examples. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiments are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0089] This application also provides an electronic device. Figure 3 , Figure 3 3 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. The electronic device may include: at least one processor 301, at least one communication bus 302, a user interface 303, at least one network interface 304, and a memory 305.

[0090] The communication bus 302 is used to implement the connection and communication between these components.

[0091] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0092] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0093] The processor 301 may include one or more processing cores. Using various interfaces and circuits, the processor 301 connects to various components within the server. It executes instructions, programs, code sets, or instruction sets stored in the memory 305, as well as accesses data stored in the memory 305, to perform various server functions and process data. Optionally, the processor 301 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 301 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing content displayed on the display screen; and the modem handles wireless communications. It is understood that the modem may not be integrated into the processor 301 but implemented as a separate chip.

[0094] Among them, the memory 305 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may also optionally be at least one storage device located away from the aforementioned processor 301. Refer to Figure 3 The memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a power transmission and transformation project progress identification application based on air-ground-space integrated technology.

[0095] exist Figure 3In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call the power transmission and transformation project progress identification application based on the air-ground-space integrated technology stored in the memory 305. When executed by one or more processors 301, the electronic device executes one or more of the methods described in the above embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should know that this application is not limited to the described order of actions, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for this application.

[0096] The present application also provides a computer-readable storage medium storing instructions, which, when executed by one or more processors, enable an electronic device to execute one or more of the methods described in the above embodiments.

[0097] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0098] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely schematic, such as the division of units, which is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interface, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0099] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0100] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0101] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of this application, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of this application. The aforementioned memory includes various media that can store program code, such as USB flash drives, mobile hard drives, magnetic disks, or optical disks.

[0102] The above descriptions are merely exemplary embodiments disclosed in this application and are not intended to limit the scope of this application. That is, any equivalent changes and modifications made based on the teachings disclosed in this application are still within the scope of this application.

[0103] This application is intended to cover any modifications, uses or adaptations disclosed in this application, which follow the general principles disclosed in this application and include common knowledge or customary technical means in the technical field not disclosed in this application.

Claims

1. A method for identifying the progress of power transmission and transformation projects based on air-ground-space integrated technology, characterized in that: The method comprises: Acquire target image data in the construction area through multiple acquisition devices, and perform preprocessing operations on the target image data; Performing feature extraction on the pre-processed image data to obtain engineering progress feature information corresponding to the engineering construction area; classifying the engineering progress characteristic information; Based on the classified engineering progress feature information, matching is performed in combination with an engineering progress feature library to output the current engineering progress corresponding to the engineering construction area; If it is confirmed that the progress deviation value between the current project progress and the planned project progress is greater than the preset deviation value, a project progress warning is triggered.

2. The method according to claim 1, characterized in that The method of acquiring target image data in the construction area by using multiple acquisition devices specifically includes: Periodically photograph the construction area using satellite remote sensing equipment to obtain satellite image data; Use drone equipment to take multi-angle aerial photos of the construction area to obtain aerial image data; Use intelligent robotic equipment to take close-up photos of the construction area to obtain image data of complex areas; The satellite image data, the aerial image data and the complex area image data are used as the target image data.

3. The method according to claim 2, characterized in that The pre-processing operation on the target image data specifically includes: Performing radiometric correction on the satellite image data to eliminate radiometric errors, and performing geometric correction on the satellite image data to correct the satellite image data to a correct geographic coordinate system; Performing stitching and image fusion on the aerial image data, and performing denoising on the fused image data; Frame extraction is performed on the complex area image data to decompose the video into single-frame images, and size normalization processing is performed on the complex area image data and each frame image.

4. The method according to claim 1, wherein The performing feature extraction on the pre-processed image data to obtain the engineering progress feature information corresponding to the engineering construction area specifically includes: Inputting the preprocessed target image data into a convolutional neural network model, and outputting the engineering progress feature information based on the convolutional neural network model; High-level semantic features are extracted from the pre-processed target image data based on the engineering progress feature information, where the high-level semantic features include substation building features and transmission line features.

5. The method according to claim 1, characterized in that The classifying of the engineering progress characteristic information specifically includes: Inputting the engineering progress feature information into a pre-trained convolutional neural network model built with a ResNet50 network as the backbone network; A classification head including two fully connected layers is connected to the back end of the ResNet50 network, wherein the first fully connected layer contains multiple neurons and uses a ReLU activation function, and the number of neurons in the second fully connected layer is consistent with the number of project progress stages; Performing linear transformation processing on the engineering progress feature information through the first fully connected layer; The second fully connected layer classifies the engineering progress feature information after the linear transformation.

6. The method according to claim 1, characterized in that The classified engineering progress feature information is matched with an engineering progress feature library to output the current engineering progress corresponding to the engineering construction area, specifically including: Establishing the project progress feature database based on the construction process and construction characteristics of the project construction area; The data in the engineering progress feature library is used as training data to construct an engineering progress recognition model, and the current engineering progress is output based on the engineering progress recognition model.

7. The method according to claim 1, characterized in that Outputting the current project progress based on the project progress identification model specifically includes: Based on the project progress recognition model, a Softmax activation function is used to generate a normalized probability distribution of each project progress stage; Performing structural processing on the normalized probability distribution, and taking the project progress corresponding to the maximum probability value as the first candidate project progress; Acquire a historical engineering phase sequence based on historical image data, and construct a phase confidence evaluation function based on a dynamic evolution trend between the first candidate project progress and the historical engineering phase sequence; Outputting a first confidence score corresponding to the first candidate project progress based on the stage confidence evaluation function; If it is confirmed that the first confidence score is less than a preset threshold, returning to the image acquisition step and reacquiring the target image data; Based on the target image data, outputting a second confidence score corresponding to a second candidate project progress through the project progress recognition model; If it is confirmed that the second confidence score is greater than or equal to the preset threshold, the second candidate project progress is output as the current project progress.

8. The power transmission and transformation project progress identification device based on air-ground integrated technology is characterized by: The device includes an acquisition module and a processing module, wherein: The acquisition module is used to acquire target image data in the construction area through multiple acquisition devices and perform preprocessing operations on the target image data; The processing module is used to extract features from the pre-processed image data to obtain engineering progress feature information corresponding to the engineering construction area; classify the engineering progress feature information; match the classified engineering progress feature information with an engineering progress feature library to output the current engineering progress corresponding to the engineering construction area; and trigger an engineering progress warning if it is confirmed that the progress deviation value between the current engineering progress and the planned engineering progress is greater than a preset deviation value.

9. An electronic device, characterized in that: The electronic device comprises a processor, a communication bus, a user interface, a network interface and a memory, wherein the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is performed.

Images