Digital supervision system and method for electric power engineering field quality

Abstract

The embodiment of the invention provides a digital supervision system and method for electric power engineering field quality, and belongs to the technical field of digital quality supervision. The digital supervision method comprises the following steps: acquiring standard parameter data and historical image data of an electric power tower in a power grid system; constructing a structure digital extraction model of the electric power tower; training the structure digital extraction model according to the historical image data of the electric power tower; according to the method, the standard parameter data and the historical image data of the electric power tower are acquired, and the structure digital extraction model is trained according to the historical image data, so that the real-time construction image of the current electric power tower can be converted into the structure quality information and the parameter quality information of the current electric power tower; and finally, according to the structure quality information and the parameter quality information, carrying out digital construction quality evaluation on the tower body quality of the current power tower.

Description

Technical Field

[0001] The present invention relates to the technical field of digital quality supervision, and in particular to a digital supervision system and method for on-site quality of electric power engineering. Background Art

[0002] With the continuous development of domestic industry and the improvement of living standards, domestic electricity consumption is also increasing year by year. In the construction of the power system, the main focus is on high-voltage transmission lines, including the construction of various power towers, such as tension towers, right-angle towers, corner towers, etc.

[0003] In the process of erecting power towers, double horizontal arm brackets are generally used to assemble the tower body layer by layer from bottom to top. At the power engineering site, the assembly quality of each layer of the tower body affects the assembly effect of the multiple layers of the tower body above. Therefore, the assembly quality of each layer of the tower body needs to be strictly inspected before the next layer is assembled. At present, at the power engineering site, the quality supervision of each layer of the tower body is done manually. However, this method has subjective errors on the one hand, resulting in poor inspection results. On the other hand, it is impossible to conduct an overall inspection of the tower body at a high altitude, which makes the inspection target of each layer of the tower body poor, and thus leads to low inspection efficiency.

[0004] The inventors of the present application discovered during the process of implementing the present invention that the above-mentioned solutions in the prior art have the defects of poor manual inspection effect and low efficiency. Summary of the invention

[0005] The purpose of the embodiments of the present invention is to provide a digital supervision system and method for the on-site quality of an electric power project, which has the function of good and efficient digital quality detection of each tower body.

[0006] In order to achieve the above-mentioned purpose, an embodiment of the present invention provides a digital supervision method for the on-site quality of a power project, comprising:

[0007] Obtain standard parameter data and historical image data of power towers in the power grid system;

[0008] Constructing a digital structural extraction model of the power tower;

[0009] Training the structure digital extraction model based on historical image data of the power tower;

[0010] Obtaining real-time construction images of the power towers at the current project site;

[0011] Inputting the real-time construction image into the trained structure digital extraction model to obtain the structural quality information of the power tower at the current project site;

[0012] Acquiring parameter quality information of the power tower at the current engineering site according to the real-time construction image and standard parameter data of the power tower at the current engineering site;

[0013] The digital construction quality assessment of the power tower is performed based on the structural quality information and parameter quality information of the power tower in the current project site.

[0014] Optionally, the standard parameter data includes the length of the main pipe and the inclination angle of the main pipe.

[0015] Optionally, the structure digitization extraction model includes a U-Net network model.

[0016] Optionally, training the structure digital extraction model according to the historical image data of the power tower includes:

[0017] Preprocessing the historical image data of the power pole tower and dividing it into a training set, a validation set and a test set, wherein the training set, the validation set and the test set include power pole tower pixel labels and background pixel labels;

[0018] Using the training set to train the structure digital extraction model;

[0019] Using the validation set to adjust model hyperparameters of the structure digitization extraction model;

[0020] The test set is used to evaluate the performance of the structure digital extraction model.

[0021] Optionally, obtaining the structural quality information of the power tower in the current project site includes:

[0022] Obtain characteristic images of four sides of the construction layer of the current power tower;

[0023] Acquire characteristic images of four sides corresponding to the current construction layer of the power tower in the historical image data;

[0024] Obtaining the similarity between the feature map of each side surface of the current construction layer of the power pole tower and the feature map of the corresponding side surface in the historical image data;

[0025] According to formula (1), the structural quality information of the power tower at the current project site is obtained.

[0026]

[0027] Among them, Q1 is the structural quality information of the power tower in the current project site, σ i is the similarity coefficient of the ith side of the construction layer of the power tower, SIM iis the similarity between the feature map of the i-th side of the construction layer of the power pole tower and the feature map of the corresponding side in the historical image data, i is an integer number, and i≤N, N is the total number of sides of the construction layer of the power pole tower, N=4.

[0028] Optionally, obtaining parameter quality information of the power tower in the current engineering site according to the real-time construction image and standard parameter data of the power tower in the current engineering site includes:

[0029] Obtaining an image facing the main tube in the historical image, labeling the main tube in the image facing the main tube, and constructing a training set of main tube data;

[0030] Build a supervisory digital network model;

[0031] Using the supervisor data training set to train the supervisor digital network model;

[0032] Inputting the real-time construction image into the trained main digitized network model to obtain the main feature map of the real-time construction image;

[0033] The parameter quality information of the current power tower is obtained according to the main characteristic map of the real-time construction image.

[0034] Optionally, obtaining the parameter quality information of the current power tower according to the main characteristic graph of the real-time construction image includes:

[0035] Obtain pixel coordinates of feature points at both ends of the main tube in the main tube feature map of the real-time construction image;

[0036] Convert the pixel coordinates of the feature points at both ends of the main tube into world coordinates;

[0037] Acquire the real-time parameter data of the main pipe of the construction layer of the current power tower according to the world coordinates of both ends of the main pipe;

[0038] According to formula (2), the parameter quality information of the construction layer of the current power tower is obtained.

[0039]

[0040] Among them, Q2 is the parameter quality information of the construction layer main pipe of the current power pole tower, γ1 is the length coefficient of the construction layer main pipe of the current power pole tower, l0 is the length of the construction layer main pipe of the current power pole tower, l B is the standard length of the main pipe corresponding to the construction layer of the current power pole tower, γ2 is the inclination coefficient of the main pipe of the construction layer of the current power pole tower, θ0 is the inclination angle of the main pipe of the construction layer of the current power pole tower, θ BIt is the standard inclination angle of the main pipe corresponding to the construction layer of the current power pole tower.

[0041] Optionally, performing a digital construction quality assessment of the power tower according to the structural quality information and parameter quality information of the power tower at the current project site includes:

[0042] According to formula (3), the overall quality information of the current power tower is obtained.

[0043] Q z =μ1·Q1+μ2·Q2, (3)

[0044] Among them, Q z is the overall quality information of the current power pole tower, μ1 is the structural coefficient of the current power pole tower, and μ2 is the parameter coefficient of the current power pole tower.

[0045] Optionally, performing the digital construction quality assessment of the power tower according to the structural quality information and parameter quality information of the power tower in the current project site further includes:

[0046] Determine whether the overall quality information of the power tower is within a preset overall quality threshold;

[0047] In the case where it is determined that the overall quality information of the power pole tower is within the preset overall quality threshold range, it is determined whether the structural quality information of the power pole tower is within the preset structural quality threshold range;

[0048] In the case where it is determined that the current structural quality information of the power pole tower is within a preset structural quality threshold range, determining whether the current parameter quality information of the power pole tower is within a preset parameter quality threshold range;

[0049] In the case where it is determined that the parameter quality information of the current power pole tower is within a preset parameter quality threshold range, determining that the construction quality of the current power pole tower is qualified;

[0050] In the case where it is determined that the parameter quality information of the current power pole tower is not within the preset parameter quality threshold range, inspecting the supervisor of the construction layer of the current power pole tower;

[0051] In the case where it is determined that the structural quality information of the current power pole tower is not within a preset structural quality threshold range, inspecting the tower material structure of the construction layer of the current power pole tower;

[0052] When it is determined that the overall quality information of the current power pole tower is not within the preset overall quality threshold range, it is determined that the construction quality of the current power pole tower is unqualified, and an early warning is issued.

[0053] On the other hand, the present invention also provides a digital supervision system for power engineering site quality, comprising:

[0054] A drone, wherein the drone includes a high-definition camera, and the high-definition camera is used to shoot real-time construction images of power poles and towers at the power engineering site at a fixed point;

[0055] The image processing module is communicatively connected to the drone and is used to execute any of the digital supervision methods described above.

[0056] Through the above technical scheme, the present invention provides a digital supervision system and method for the on-site quality of power engineering projects. By acquiring standard parameter data and historical image data of power towers and training a structural digital extraction model based on the historical image data, the real-time construction image of the current power tower can be converted into the structural quality information and parameter quality information of the current power tower. Finally, the tower body quality of the current power tower is digitally evaluated based on the structural quality information and parameter quality information. The use of this digital supervision method can comprehensively and intelligently evaluate the construction quality of each layer of the tower body. On the one hand, it can effectively reduce the error of manual inspection and improve the inspection accuracy; on the other hand, it can quickly evaluate the current construction quality and diagnose the type of construction quality defects, that is, it provides a quality defect detection target, thereby improving the inspection efficiency.

[0057] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0059] Figure 1 is a flow chart of a digital supervision method for power engineering site quality according to an embodiment of the present invention;

[0060] Figure 2 It is a flowchart of training a structure digital extraction model in a digital supervision method for power engineering site quality according to an embodiment of the present invention;

[0061] Figure 3 It is a flow chart of obtaining structural quality information in a digital supervision method for on-site quality of power engineering according to an embodiment of the present invention;

[0062] Figure 4 It is a flow chart of obtaining a main characteristic diagram in a digital supervision method for power engineering site quality according to an embodiment of the present invention;

[0063] Figure 5 It is a flow chart of obtaining parameter quality information in a digital supervision method for power engineering site quality according to an embodiment of the present invention;

[0064] Figure 6 The present invention is a flowchart of the construction quality assessment of power towers in a digital supervision method for power engineering site quality according to an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0066] Figure 1 1 is a flow chart of a digital supervision method for power engineering site quality according to an embodiment of the present invention. Figure 1 In the present invention, the digital supervision method may include:

[0067] In step S1, standard parameter data and historical image data of power towers in the power grid system are obtained. Among them, the types of power towers may include tension towers, corner towers, straight towers, etc. At the same time, tension towers may also include tension towers of various heights. Therefore, the relevant data of power towers mentioned here include relevant data of power towers of various types and various heights. Considering the simplicity of description, this application will introduce intelligent supervision for power towers of one type and one height in the following, and the rest of the types / heights can be analogized. Specifically, the composition of the overall structure of the power tower may include main pipes, cross materials, auxiliary materials, ground wire brackets and cross arms, etc., and the specific models / sizes of the main pipes and cross materials between layers may be different. Considering that the overall structure of the power tower is mainly supported by the main pipe, the standard parameter data mainly monitors the type, length and inclination angle of the main pipe, but is not limited to monitoring the parameters of other components such as cross materials and auxiliary materials. For the historical image data, it may include multi-angle image data and overall image data of each tower body in the power tower.

[0068] In step S2, a structural digital extraction model of the power pole is constructed. Considering the influence of various background noises when shooting the image of the power pole, in order to reduce the interference of the background on the image of the power pole, the structural digital extraction model may include a U-Net network model. Specifically, the U-Net network model can accurately separate the power pole / tower body and the background, and can also maintain the integrity of the image details.

[0069] In step S3, a structure digital extraction model is trained based on the historical image data of the power tower. The structure digital extraction model is trained using the historical image data of the power tower, that is, the U-Net network model is trained to improve the segmentation effect of the model on the power tower / tower body and the background in the image.

[0070] In step S4, a real-time construction image of the power tower in the current engineering site is obtained. In the process of building the power tower, a double-arm holding pole is generally used to build the tower body layer by layer, and the effect of each layer of the tower body is the top priority of the power engineering. At the same time, the construction stability of each layer of the tower body will affect the tower body above it. Therefore, in order to avoid the subsequent dismantling and reconstruction of the tower body, and to ensure the efficiency and effect of the power engineering, it is necessary to identify and diagnose the current tower body after each layer of the tower body is built. The real-time construction image is a multi-angle real-time image after a certain layer of the tower is built at the current power engineering site. For the convenience of expression, a certain layer of the tower body that has been built is called a construction layer. Specifically, the real-time construction image may include multi-angle and multi-(standard) point images of the middle layer of the construction layer.

[0071] In step S5, the real-time construction image is input into the trained structure digital extraction model to obtain the structural quality information of the power tower in the current project site. The real-time construction image is input into the trained structure digital extraction model to obtain the characteristic image of the tower body of the current construction layer. By comparing the characteristic image with the characteristic image of the layer in the historical image, the structural quality information of the current construction layer can be determined. That is, the occurrence of tower material leakage, wrong installation, etc. can be effectively avoided.

[0072] In step S6, the parameter quality information of the power tower in the current engineering site is obtained according to the real-time construction image and the standard parameter data of the power tower in the current engineering site. The parameter data of the main pipe of the current construction layer can be obtained according to the real-time construction image, and the parameter quality information of the current construction layer can be determined by comparing it with the standard parameter data of the main pipe of the layer. That is, it can avoid the occurrence of misinstallation and offset of the main pipe of the current construction layer.

[0073] In step S7, the digital construction quality of the power tower is evaluated based on the structural quality information and parameter quality information of the power tower in the current project site. After obtaining the structural quality information and parameter quality information of the current power tower, the overall quality evaluation of the construction layer of the current power tower can be performed to ensure that the current construction layer meets the construction requirements.

[0074] In step S1 to step S7, the standard parameter data and historical image data of the main pipe in the power tower are first obtained, and then the structural digital extraction model of the power tower is constructed. The structural digital extraction model is trained using the historical image data so that the structural digital extraction model can effectively segment the power tower / tower body and the background. At the same time, the real-time construction image of the power tower in the current project site is obtained, specifically, the real-time image of the tower body of the construction layer, and the real-time construction image is input into the trained structural digital extraction model to obtain the feature map of the tower body of the layer, and the feature map of the layer is compared with the feature map of the layer in the historical image data to obtain the structural quality information of the power tower. In addition, the parameter data of the main pipe of the construction layer is obtained according to the real-time construction image, and the parameter quality information of the power tower is obtained by comparing it with the standard parameter data of the main pipe of the layer. Finally, the two are comprehensively evaluated to evaluate the construction quality of the current power tower and realize the digital supervision of the construction of the power tower.

[0075] The quality supervision of each layer of the traditional power tower is done by manual inspection. However, this method has subjective errors on the one hand, resulting in poor inspection results. On the other hand, it is impossible to conduct an overall inspection of the tower body at a high position, which makes the inspection of each layer of the tower body poorly targeted, and thus leads to low inspection efficiency. In this embodiment of the present invention, a method of combining structural quality information and parameter quality information to evaluate the construction quality of the tower body can be used to conduct a digital construction quality assessment of each layer of the tower body in an overall and intelligent manner. On the one hand, it can effectively reduce the errors of manual inspection and improve the inspection effect; on the other hand, it can quickly evaluate the current construction quality and diagnose the type of construction quality defects, that is, it provides a quality defect detection target, thereby improving the inspection efficiency.

[0076] In this embodiment of the present invention, in order to accurately segment and extract the tower body in the real-time construction image, it is also necessary to train the structure digital extraction model. The specific training steps can be as follows: Figure 2 Specifically, Figure 2 In the embodiment, the digital supervision method may further include:

[0077] In step S30, the historical image data of the power tower is preprocessed and divided into a training set, a validation set and a test set, wherein the training set, the validation set and the test set include power tower pixel labels and background pixel labels. The preprocessing of the historical image may include image cropping to obtain multi-angle images of each layer of the tower body, and then normalizing each image to adjust the size to be consistent. The preprocessed data division may include a training set, a validation set and a test set known to those skilled in the art. Specifically, the training set is used to train and update model parameters, the validation set is used for model selection and parameter adjustment, and the test set is used to evaluate the performance of the model. The training set, the validation set and the test set include images and labels of the power tower / tower body and the background area, which are usually provided in the form of a mask, and the power tower / tower body and the background are represented by different labels (such as 1 and 0), respectively.

[0078] In step S31, the training set is used to train the structure digital extraction model. The image and the corresponding mask are input into the U-Net model, the U-Net model is trained, and the model parameters are updated. Specifically, the U-Net model first performs multi-layer convolution and pooling operations on the input image to extract the features of the image. The feature map is then gradually upsampled to obtain the final feature map, which is compared with the true value, and the model parameters are continuously iterated to optimize the model parameters to achieve model training, that is, training to distinguish between background and power poles / tower bodies. In addition, in the task of background removal, the U-Net network model usually undergoes a threshold processing step, in which all pixels predicted to be background are set to transparent or white, while the foreground objects, that is, the power poles, remain unchanged.

[0079] In step S32, the validation set is used to adjust the model hyperparameters of the structure digitization extraction model.

[0080] In step S33, the test set is used to evaluate the performance of the structure digital extraction model.

[0081] In step S30 to step S33, the historical image data of the power tower is preprocessed and divided into a training set, a validation set and a test set. The training set, the validation set and the test set include power tower pixel labels and background pixel labels, that is, the structural digital extraction model can perform background removal operations on the power tower image to obtain a feature map of the power tower / tower body. Specifically, the feature map of the power tower / tower body output by the structural digital extraction model can facilitate the subsequent calculation and acquisition of the structural quality information of the tower body of the current construction layer.

[0082] In this embodiment of the present invention, after obtaining the real-time construction image, it is input into the trained structure digital extraction model to obtain the tower body feature map of the current construction layer, and the structural quality information of the construction layer is obtained according to the feature map. The specific steps can be as follows: Figure 3 Specifically, Figure 3 In the embodiment, the digital supervision method may further include:

[0083] In step S50, feature maps of the four sides of the construction layer of the current power tower are obtained. Each layer of the power tower has four sides, so a drone can be used to shoot real-time construction images of the four sides of the current construction layer, and the real-time construction images are input into the structure digital extraction model to obtain feature maps of the four sides. Specifically, there is no requirement for the shooting points of the real-time construction images of the four sides, as long as there are shooting point images corresponding to the layer in the historical image data. However, considering the balance of the subsequent feature map similarity fusion, the effectiveness of the similarity calculation, and the standardization of the real-time construction image shooting, the shooting points of the four sides of the tower body in this application may include positions on the median vertical planes of the four sides at a preset distance from the center of the tower body. In this way, on the one hand, the consistency of multiple images can be guaranteed, and on the other hand, limiting the shooting points of the images can also improve the accuracy of the subsequent similarity calculation.

[0084] In step S51, the feature maps of the four sides corresponding to the construction layer of the current power pole tower in the historical image data are obtained. The feature maps of the four sides of the corresponding layer in the historical image data are also the historical images taken at the same point as above. The historical images are input into the structure digital extraction model to obtain the feature maps corresponding to the feature maps of the four sides of the construction layer of the current power pole tower, which can also be called standard feature maps.

[0085] In step S52, the similarity between the feature map of each side of the current power tower construction layer and the feature map of the corresponding side in the historical image data is obtained. The similarity between the two feature maps can be calculated by cosine similarity, gray value comparison, etc. known to those skilled in the art.

[0086] In step S53, the structural quality information of the power tower in the current project site is obtained according to formula (1).

[0087]

[0088] Among them, Q1 is the structural quality information of the power tower in the current project site, σ i is the similarity coefficient of the ith side of the construction layer of the power tower, SIM iis the similarity between the feature map of the i-th side of the construction layer of the power pole and the feature map of the corresponding side in the historical image data, i is an integer number, and i≤N, N is the total number of sides of the construction layer of the power pole, N=4. Specifically, for the similarity coefficients of the four sides, it is generally assumed that the weights are equal, that is, σ i =1 / 4. However, considering that different obstructions are captured from different sides during the construction of power towers, the similarity coefficient for the side with more obstructions, that is, for the side with relatively greater difficulty in extracting the tower features at the construction layer, can be set lower, and vice versa, it can be set higher, and the sum of the four similarity coefficients is guaranteed to be 1, so as to reduce the impact of feature extraction accuracy on the structural quality information of the construction layer.

[0089] In step S50 to step S53, the real-time construction images of the construction layer at the preset points on the four sides are first obtained, and the feature map of each side is obtained according to the structure digital extraction model. At the same time, the historical images of the same points on the four sides of the construction layer of the current power pole are obtained from the historical image data, and the corresponding feature maps are obtained according to the structure digital extraction model. Then, the similarity between the feature maps of the four sides of the construction layer and the corresponding historical feature maps is calculated respectively, and the structural quality information of the current power pole construction layer can be obtained.

[0090] In this embodiment of the present invention, after obtaining the real-time construction image of the construction layer of the current power pole tower, the parameters of the main controller in the construction image can also be obtained to calculate the parameter quality information of the current power pole tower. The specific steps can be as follows: Figure 4 Specifically, Figure 4 In the embodiment, the digital supervision method may further include:

[0091] In step S60, the image facing the main tube in the historical image is obtained, and the main tube in the image facing the main tube is labeled to construct a training set of main tube data. Among them, the image facing the main tube can be achieved by shooting at preset points. Specifically, the shooting point of the image facing the main tube can also be located on the median vertical plane of the side, and the distance between the drone and the center of the tower body is adjusted so that the line connecting the drone and the axis of the main tube can be perpendicular to the two diagonal main tube axis planes. In addition, in addition to the training set of main tube data, a verification set and a test set similar to the above-mentioned model training can also be included.

[0092] In step S61, a main tube digital network model is constructed, wherein the main tube digital network model may include but is not limited to a neural network, a U-Net network model, etc., and is mainly used to perform feature recognition on the main tube in the real-time construction image.

[0093] In step S62, the main digital network model is trained using the main data training set.

[0094] In step S63, the real-time construction image is input into the trained digital network model of the main body to obtain the main body feature map of the real-time construction image. The real-time construction image is the real-time construction image obtained by the above-mentioned shooting point. For the four main bodies of the construction layer, four construction images can be included. The implementation construction image is input into the digital network model of the main body to obtain the main body feature map in the current construction layer, which includes the feature information of the main body. Specifically, the real-time construction image of each main body can also include two, that is, the shooting point is symmetrical about the axis of the main body. When the parameter quality information of the main body is calculated later, the values ​​corresponding to the two images can also be calculated respectively and averaged to further improve the accuracy of parameter quality information acquisition.

[0095] In step S64, the parameter quality information of the current power pole is obtained according to the main characteristic map of the real-time construction image. The step of obtaining the parameter quality information of the current power pole can be as follows: Figure 5 Specifically, Figure 5 In the embodiment, the artificial supervision intelligent method may further include:

[0096] In step S640, the pixel coordinates of the feature points at both ends of the main tube in the main tube feature map of the real-time construction image are obtained. The feature points at both ends of the main tube and the feature points can be obtained through the network model. Specifically, the feature points may include the feature points at the center of the ends.

[0097] In step S641, the pixel coordinates of the feature points at both ends of the main tube are converted to world coordinates. The conversion of pixel coordinates to world coordinates may include the sequential transformation of pixel coordinates, image coordinates, camera coordinates, and world coordinates known to those skilled in the art, which will not be described in detail here. Specifically, when converting coordinates, it is necessary to obtain the intrinsic parameters and extrinsic parameters of the drone camera.

[0098] In step S642, the real-time parameter data of the main pipe of the construction layer of the current power pole tower is obtained according to the world coordinates of the two ends of the main pipe, wherein the main pipe length and the inclination angle of the main pipe of the construction layer of the current power pole tower can be obtained according to the world coordinates of the two ends of the main pipe.

[0099] In step S643, the parameter quality information of the construction layer of the current power tower is obtained according to formula (2).

[0100]

[0101] Among them, Q2 is the parameter quality information of the construction layer main pipe of the current power pole tower, that is, the parameter quality information of the current power pole tower, γ1 is the length coefficient of the construction layer main pipe of the current power pole tower, l0 is the length of the construction layer main pipe of the current power pole tower, l Bis the standard length of the main pipe corresponding to the construction layer of the current power pole tower, γ2 is the inclination coefficient of the main pipe of the construction layer of the current power pole tower, θ0 is the inclination angle of the main pipe of the construction layer of the current power pole tower, θ B is the standard inclination angle of the main pipe corresponding to the construction layer of the current power tower. Specifically, for the length coefficient γ1 of the main pipe and the inclination coefficient γ2 of the main pipe, it is generally believed that the weights of the two are equivalent, that is, γ1=γ2=1 / 2.

[0102] In step S640 to step S643, the pixel coordinates of the feature points at both ends of the main tube are first obtained, and the pixel coordinates are converted into world coordinates. Then, the main tube length and the inclination angle of the main tube of the current main tube are obtained according to the world coordinates of the two ends of the main tube. The length and inclination angle of the current main tube are compared with the length and inclination angle of the standard main tube of the construction layer, and the parameter quality information of the main tube of the construction layer can be obtained. In addition, the world coordinates of the two ends of the main tube of the current construction layer can also be obtained by binocular vision detection known to those skilled in the art, which can effectively improve the accuracy of obtaining the world coordinates of the two ends of the main tube.

[0103] In step S60 to step S64, a main body data training set is first constructed from the main body's facing image in the historical image, and the main body digital network model is trained. The real-time construction image of the current construction layer is input into the model to obtain a main body feature map, and the parameter quality information of the power tower can be obtained according to the main body feature map, that is, the parameter quality information of the main body of the construction layer.

[0104] In this embodiment of the present invention, after obtaining the structural quality information and parameter quality information of the construction layer of the current power pole tower, the construction quality of the construction layer of the current power pole tower can be evaluated / assessed. The specific evaluation steps can be as follows: Figure 6 Specifically, Figure 6 In the embodiment, the digital supervision method may further include:

[0105] In step S70, the overall quality information of the current power tower is obtained according to formula (3):

[0106] Q z =μ1·Q1+μ2·Q2, (3)

[0107] Among them, Q zis the overall quality information of the current power tower, μ1 is the structural coefficient of the current power tower, and μ2 is the parameter coefficient of the current power tower. Specifically, in the process of assembling the power tower, the tower structure / volume of the construction layer becomes smaller and smaller, that is, the features of the construction layer become denser and denser, resulting in a decrease in the accuracy of feature extraction. Therefore, the structural coefficient and parameter coefficient can be adjusted according to the number of layers of the current construction layer. Specifically, the initial structural coefficient and parameter coefficient are equal, both of which are 0.5. However, as the number of layers continues to increase, the structural coefficient gradually decreases, and the parameter coefficient gradually increases. Specifically, the changes in the structural coefficient and parameter coefficient can be obtained according to formula (4),

[0108]

[0109] Among them, a is greater than the total number of tower layers of the power tower and can be set larger, and x is the number of tower layers corresponding to the current construction layer.

[0110] In step S71, it is determined whether the overall quality information of the current power tower is within a preset overall quality threshold range.

[0111] In step S72, when it is determined that the overall quality information of the current power pole is within the preset overall quality threshold range, it is determined whether the structural quality information of the current power pole is within the preset structural quality threshold range.

[0112] In step S73, when it is determined that the structural quality information of the current power pole is within the preset structural quality threshold range, it is determined whether the parameter quality information of the current power pole is within the preset parameter quality threshold range.

[0113] In step S74, when it is determined that the parameter quality information of the current power tower is within the preset parameter quality threshold range, it is determined that the construction quality of the current power tower is qualified. If the overall quality information, structural quality information and parameter quality information of the power tower are all within the corresponding threshold range, it means that the construction quality of the current power tower is qualified.

[0114] In step S75, when it is determined that the parameter quality information of the current power pole tower is not within the preset parameter quality threshold range, the main pipe of the construction layer of the current power pole tower is checked. If the parameter quality information is not within the preset parameter quality threshold range, it means that there may be an assembly problem with the main pipe, so the main pipe can be further checked.

[0115] In step S76, when it is determined that the structural quality information of the current power tower is not within the preset structural quality threshold range, the construction layer tower material structure of the current power tower is inspected. If the structural quality information is not within the preset structural quality threshold range, it means that there may be problems with the structure, so the structure can be further inspected.

[0116] In step S77, when it is determined that the overall quality information of the current power pole tower is not within the preset overall quality threshold range, the construction quality of the current power pole tower is determined to be unqualified, and an early warning is issued. If the overall quality information is not within the preset overall quality threshold range, it means that there is a greater risk in the current construction layer, and an early warning is issued and a more careful inspection is performed.

[0117] In step S70 to step S77, the overall quality information of the power tower is first obtained according to the structural quality information and the parameter instruction information, and then the overall quality information, the structural quality information and the parameter quality information are judged in turn to evaluate the construction quality of the current construction layer.

[0118] On the other hand, the present invention also provides a digital supervision system for the quality of power engineering sites. Specifically, the digital supervision system may include a drone and an image processing module. Specifically, in order to obtain detailed features of the construction layer, the drone may include a high-resolution high-definition camera.

[0119] The high-definition camera is used to capture the real-time construction images of the power towers at the power engineering site at a fixed point, and the image processing module is connected to the UAV communication to execute any of the above digital supervision methods.

[0120] Through the above technical scheme, a digital supervision system and method for the on-site quality of power engineering projects provided by the present invention can convert the real-time construction image of the current power tower into the structural quality information and parameter quality information of the current power tower by acquiring the standard parameter data and historical image data of the power tower and training the structural digital extraction model according to the historical image data, and finally perform a digital construction quality assessment on the tower body quality of the current power tower according to the structural quality information and the parameter quality information; the use of this digital supervision method can comprehensively and intelligently assess the construction quality of each layer of the tower body, which can effectively reduce the error of manual inspection and improve the inspection effect on the one hand; on the other hand, it can quickly assess the current construction quality and diagnose the type of construction quality defects, that is, provide a quality defect detection target, thereby improving the inspection efficiency.

[0121] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0122] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0125] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0126] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0127] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0128] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0129] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A digital supervision method for the on-site quality of power engineering projects, characterized in that: include: Obtain standard parameter data and historical image data of power towers in the power grid system; Constructing a digital structural extraction model of the power tower; Training the structure digital extraction model based on historical image data of the power tower; Obtaining real-time construction images of the power towers at the current project site; Inputting the real-time construction image into the trained structure digital extraction model to obtain the structural quality information of the power tower at the current project site; Acquiring parameter quality information of the power tower at the current engineering site according to the real-time construction image and standard parameter data of the power tower at the current engineering site; The digital construction quality assessment of the power tower is performed based on the structural quality information and parameter quality information of the power tower in the current project site.

2. The digital supervision method according to claim 1 is characterized in that: The standard parameter data includes the length of the main tube and the inclination angle of the main tube.

3. The digital supervision method according to claim 2 is characterized in that: The structure digitization extraction model includes a U-Net network model.

4. The digital supervision method according to claim 3 is characterized in that: Training the structure digital extraction model according to the historical image data of the power tower includes: Preprocessing the historical image data of the power pole tower and dividing it into a training set, a validation set and a test set, wherein the training set, the validation set and the test set include power pole tower pixel labels and background pixel labels; Using the training set to train the structure digital extraction model; Using the validation set to adjust model hyperparameters of the structure digitization extraction model; The test set is used to evaluate the performance of the structure digital extraction model.

5. The digital supervision method according to claim 4 is characterized in that: The structural quality information of the power towers in the current project site is obtained including: Obtain characteristic images of four sides of the construction layer of the current power tower; Acquire characteristic images of four sides corresponding to the current construction layer of the power tower in the historical image data; Obtaining the similarity between the feature map of each side surface of the current construction layer of the power pole tower and the feature map of the corresponding side surface in the historical image data; According to formula (1), the structural quality information of the power tower at the current project site is obtained. Among them, Q1 is the structural quality information of the power tower in the current project site, σ i is the similarity coefficient of the ith side of the construction layer of the power tower, SIM i is the similarity between the feature map of the i-th side of the construction layer of the power pole tower and the feature map of the corresponding side in the historical image data, i is an integer number, and i≤N, N is the total number of sides of the construction layer of the power pole tower, N=4.

6. The digital supervision method according to claim 5 is characterized in that: Acquiring parameter quality information of the power tower at the current engineering site according to the real-time construction image and the standard parameter data of the power tower at the current engineering site includes: Obtaining an image facing the main tube in the historical image, labeling the main tube in the image facing the main tube, and constructing a training set of main tube data; Build a supervisory digital network model; Using the supervisor data training set to train the supervisor digital network model; Inputting the real-time construction image into the trained main digitized network model to obtain the main feature map of the real-time construction image; The parameter quality information of the current power tower is obtained according to the main characteristic map of the real-time construction image.

7. The digital supervision method according to claim 6, characterized in that: Acquiring the parameter quality information of the current power tower according to the main characteristic graph of the real-time construction image includes: Obtain pixel coordinates of feature points at both ends of the main tube in the main tube feature map of the real-time construction image; Convert the pixel coordinates of the feature points at both ends of the main tube into world coordinates; Acquire the real-time parameter data of the main pipe of the construction layer of the current power tower according to the world coordinates of both ends of the main pipe; According to formula (2), the parameter quality information of the construction layer of the current power tower is obtained. Among them, Q2 is the parameter quality information of the construction layer main pipe of the current power pole tower, γ1 is the length coefficient of the construction layer main pipe of the current power pole tower, l0 is the length of the construction layer main pipe of the current power pole tower, l B is the standard length of the main pipe corresponding to the construction layer of the current power pole tower, γ2 is the inclination coefficient of the main pipe of the construction layer of the current power pole tower, θ0 is the inclination angle of the main pipe of the construction layer of the current power pole tower, θ B It is the standard inclination angle of the main pipe corresponding to the construction layer of the current power pole tower.

8. The digital supervision method according to claim 7, characterized in that: The digital construction quality assessment of the power tower based on the structural quality information and parameter quality information of the power tower at the current project site includes: According to formula (3), the overall quality information of the current power tower is obtained. Q z =μ1·Q1+μ2·Q2, (3) Among them, Q z is the overall quality information of the current power pole tower, μ1 is the structural coefficient of the current power pole tower, and μ2 is the parameter coefficient of the current power pole tower.

9. The digital supervision method according to claim 8, characterized in that: The digital construction quality assessment of the power tower according to the structural quality information and parameter quality information of the power tower at the current project site also includes: Determine whether the overall quality information of the power tower is within a preset overall quality threshold; In the case where it is determined that the overall quality information of the power pole tower is within the preset overall quality threshold range, it is determined whether the structural quality information of the power pole tower is within the preset structural quality threshold range; In the case where it is determined that the current structural quality information of the power pole tower is within a preset structural quality threshold range, determining whether the current parameter quality information of the power pole tower is within a preset parameter quality threshold range; In the case where it is determined that the parameter quality information of the current power pole tower is within a preset parameter quality threshold range, determining that the construction quality of the current power pole tower is qualified; In the case where it is determined that the parameter quality information of the current power pole tower is not within the preset parameter quality threshold range, inspecting the supervisor of the construction layer of the current power pole tower; In the case where it is determined that the structural quality information of the current power pole tower is not within a preset structural quality threshold range, inspecting the tower material structure of the construction layer of the current power pole tower; When it is determined that the overall quality information of the current power pole tower is not within the preset overall quality threshold range, it is determined that the construction quality of the current power pole tower is unqualified, and an early warning is issued.

10. A digital supervision system for the quality of power engineering site, characterized in that: include: A drone, wherein the drone includes a high-definition camera, and the high-definition camera is used to shoot real-time construction images of power poles and towers at the power engineering site at a fixed point; An image processing module is communicatively connected to the drone and is used to execute the digital supervision method as described in any one of claims 1-9.

Images