Transmission line defect identification and early warning method and system based on remote sensing monitoring

Abstract

The present invention relates to a method and system for identifying and warning defects of transmission lines based on remote sensing monitoring, and belongs to the technical field. The present invention constructs a training set according to the historical defect image information of the transmission line, and constructs a transmission line defect recognition model according to the training set, thereby identifying the remote sensing image information of the transmission line through the transmission line defect recognition model, obtaining the defect feature data of the transmission line in each area, and issuing an early warning according to the defect feature data of the transmission line in each area, generating early warning information, and finally setting a priority maintenance plan according to the early warning information, and allocating work tasks based on the priority maintenance plan. The present invention processes the target defect features in the training set through a cyclic spatial attention mechanism, so that the attention is focused on the target defect features in the training set, can suppress the interference of multi-scale features on the transmission line defect recognition model, and can improve the recognition accuracy of transmission line defect recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission line monitoring, and in particular to a power transmission line defect identification and early warning method and system based on remote sensing monitoring. Background Art

[0002] In recent years, drones, as a high-tech inspection tool, have been rapidly promoted and applied in the power grid industry. By operating drones to fly along the line, taking photos or laser scanning at the same time, combined with intelligent AI data processing software, hidden dangers and defects in the line can be discovered quickly and accurately. With the continuous improvement of the number and skill level of drones in the national power grid industry, the "drone patrol era" of power grid inspection has arrived. Domestic drone technology is already at the international leading level. Some manufacturers apply image deep learning to the front end of drone automatic inspection, using the high-performance image automatic recognition chip carried by the drone to lock the target in real time and take pictures; at the same time, some drones are put into flight on automatic routes, saving a lot of manpower, so that inspection personnel can be liberated from a lot of physical labor and put into higher-level data analysis and processing. The operation mode and personnel structure of the operation and maintenance team are also optimized, and the reliability of the power grid is further improved. However, in the existing technology, due to the interference of multi-scale features in the training process of the deep learning model, the recognition accuracy of the defective transmission line is too low, which is not conducive to the rapid and accurate recognition of transmission line defects. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a method and system for identifying and warning transmission line defects based on remote sensing monitoring.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] The first aspect of the present invention provides a method for identifying and warning transmission line defects based on remote sensing monitoring, comprising the following steps:

[0006] Construct an integrated control network for unmanned aerial vehicles, optimize the integrated control network for unmanned aerial vehicles, obtain an optimized integrated control network for unmanned aerial vehicles, and control unmanned aerial vehicles through the integrated control network for unmanned aerial vehicles to obtain remote sensing image information of power transmission lines;

[0007] Obtain historical defect image information of the transmission line through big data, build a training set based on the historical defect image information of the transmission line, and build a transmission line defect recognition model based on the training set;

[0008] The remote sensing image information of the transmission line is recognized through the transmission line defect recognition model to obtain the defect feature data of the transmission line in each area, and an early warning is issued based on the defect feature data of the transmission line in each area to generate early warning information;

[0009] Set a priority maintenance plan based on the early warning information, and assign work tasks based on the priority maintenance plan.

[0010] Furthermore, in this method, a UAV integrated control network is constructed, and the UAV integrated control network is optimized to obtain an optimized UAV integrated control network, which specifically includes:

[0011] Construct a UAV integrated control network, test the UAV integrated control network, obtain the historical UAV control quantity upper limit change characteristics of the UAV integrated control network, and build a UAV control quantity upper limit characteristic prediction model based on convolutional neural network;

[0012] Inputting the historical UAV control quantity upper limit change characteristics of the UAV integrated control network into the UAV control quantity upper limit feature prediction model for training, and obtaining the trained UAV control quantity upper limit feature prediction model;

[0013] Obtain the upper limit change characteristics of the number of drones controlled by the drone integrated control network within a preset time and input them into the trained upper limit characteristic prediction model of the number of drones controlled to make predictions, and obtain the upper limit characteristics of the number of drones controlled by the drone integrated control network at the current timestamp;

[0014] Get the real-time number of drone controls of the drone integrated control network. When the real-time number of drone controls is greater than the upper limit of the number of drone controls of the drone integrated control network at the current timestamp, adjust the real-time number of drone controls according to the upper limit of the number of drone controls of the drone integrated control network at the current timestamp to get the optimized drone integrated control network.

[0015] Furthermore, in this method, historical defect image information of the transmission line is obtained through big data, and a training set is constructed based on the historical defect image information of the transmission line, and a transmission line defect recognition model is constructed based on the training set, specifically including:

[0016] Obtain historical defect image information of the transmission line through big data, and classify the defect types of the historical defect image information of the transmission line to obtain image information of each defect type, and construct a training set of each defect type based on the image information of each defect type;

[0017] The image information in the training set of each defect type is input into the feature pyramid network to obtain the defect features in each image information, and determine whether there are at least two defect features in the image information, and introduce a cyclic spatial attention mechanism;

[0018] When there are at least two defect features in the image information, the defect feature corresponding to the defect type label information of the current image information is obtained as the relevant defect feature; when there is only one defect feature in the image information, the defect feature corresponding to the defect type label information of the current image information is used as the relevant defect feature;

[0019] The relevant defect features are input into the cyclic spatial attention mechanism to focus on the relevant defect features, obtain the attention feature map, build a transmission line defect recognition model based on the deep neural network, and input the attention feature map in the training set into the transmission line defect recognition model for encoding learning.

[0020] Furthermore, in this method, the remote sensing image information of the transmission line is recognized by the transmission line defect recognition model, the defect feature data of the transmission line in each area is obtained, and an early warning is issued according to the defect feature data of the transmission line in each area, and early warning information is generated, which specifically includes:

[0021] Inputting the remote sensing image information of the transmission line into the transmission line defect recognition model for recognition, obtaining the defect feature data of the transmission line in each area, and presetting the warning level standard;

[0022] According to the warning level standard, the defect characteristic data of the transmission line in each area are divided into warning levels, the defect characteristic warning level of the transmission line in each area is obtained, and the geographical location information of the target area where the transmission line has defects is obtained;

[0023] An early warning is generated based on the early warning level of the defect characteristics of the transmission line in each area, the geographical location information of the target area where the transmission line has defects, and the defect characteristic data of the transmission line in each area to generate early warning information.

[0024] Furthermore, in this method, a priority maintenance plan is set according to the early warning information, specifically including:

[0025] According to the early warning information, the geographical location information of the target area where the transmission line has defects is obtained, and a search tag is constructed based on the geographical location information of the target area where the transmission line has defects, and a search is performed according to the search tag to obtain the power supply area of ​​the transmission line;

[0026] Acquire quantity information of supply users according to the power supply area of ​​the transmission line, construct a supply user quantity ranking table, input the quantity information of supply users into the supply user quantity ranking table for ranking, and obtain a supply quantity ranking result from large to small;

[0027] A priority maintenance plan in the target area is set based on the supply quantity sorting result from large to small, and the priority maintenance plan is output.

[0028] Furthermore, in this method, work tasks are allocated based on the priority maintenance plan, specifically including:

[0029] Obtain the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects, and calculate the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the power line has defects based on the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects;

[0030] A genetic algorithm is introduced, and the genetic algebra is set according to the genetic algorithm. The work position allocation of the maintenance personnel is initialized according to the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the transmission line has defects, and the work task location of each maintenance personnel is obtained;

[0031] Determine whether the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is greater than a preset Euclidean distance value; when the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is not greater than the preset Euclidean distance value, output the work task location of each maintenance personnel;

[0032] When the Euclidean distance value between the location of the maintenance personnel and the geographic location information of the target area where the transmission line has defects is greater than the preset Euclidean distance value, the work task location of each maintenance personnel is reallocated until the Euclidean distance value between the location of the maintenance personnel and the geographic location information of the target area where the transmission line has defects is no greater than the preset Euclidean distance value.

[0033] A second aspect of the present invention provides a transmission line defect recognition and early warning system based on remote sensing monitoring. The transmission line defect recognition and early warning system based on remote sensing monitoring includes a memory and a processor. The memory includes a transmission line defect recognition and early warning method program based on remote sensing monitoring. When the transmission line defect recognition and early warning method program based on remote sensing monitoring is executed by the processor, the following steps are implemented:

[0034] Construct an integrated control network for unmanned aerial vehicles, optimize the integrated control network for unmanned aerial vehicles, obtain an optimized integrated control network for unmanned aerial vehicles, and control unmanned aerial vehicles through the integrated control network for unmanned aerial vehicles to obtain remote sensing image information of power transmission lines;

[0035] Obtain historical defect image information of the transmission line through big data, build a training set based on the historical defect image information of the transmission line, and build a transmission line defect recognition model based on the training set;

[0036] The remote sensing image information of the transmission line is recognized through the transmission line defect recognition model to obtain the defect feature data of the transmission line in each area, and an early warning is issued based on the defect feature data of the transmission line in each area to generate early warning information;

[0037] Set a priority maintenance plan based on the early warning information, and assign work tasks based on the priority maintenance plan.

[0038] The third aspect of the present invention provides a computer-readable storage medium, which includes a program for a transmission line defect identification and early warning method based on remote sensing monitoring. When the program for a transmission line defect identification and early warning method based on remote sensing monitoring is executed by a processor, any step of the transmission line defect identification and early warning method based on remote sensing monitoring is implemented.

[0039] The present invention solves the defects existing in the background technology and has the following beneficial effects:

[0040] The present invention constructs an unmanned aerial vehicle integrated control network, optimizes the unmanned aerial vehicle integrated control network, obtains the optimized unmanned aerial vehicle integrated control network, controls the unmanned aerial vehicle through the unmanned aerial vehicle integrated control network to obtain remote sensing image information of the transmission line, and then obtains the historical defect image information of the transmission line through big data, and constructs a training set according to the historical defect image information of the transmission line, and constructs a transmission line defect recognition model according to the training set, so that the remote sensing image information of the transmission line is recognized by the transmission line defect recognition model, the defect feature data of the transmission line in each area is obtained, and an early warning is performed according to the defect feature data of the transmission line in each area, and early warning information is generated, and finally a priority maintenance plan is set according to the early warning information, and work tasks are assigned based on the priority maintenance plan. The present invention processes the target defect features in the training set through a cyclic spatial attention mechanism, so that the attention is focused on the target defect features in the training set, which can suppress the interference of multi-scale features on the transmission line defect recognition model, and can improve the recognition accuracy of transmission line defect recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0042] Figure 1 The overall method flow chart of the transmission line defect identification and early warning method based on remote sensing monitoring is shown;

[0043] Figure 2A first method flow chart of a transmission line defect identification and early warning method based on remote sensing monitoring is shown;

[0044] Figure 3 A second method flow chart of the transmission line defect identification and early warning method based on remote sensing monitoring is shown;

[0045] Figure 4 The system block diagram of the transmission line defect identification and early warning system based on remote sensing monitoring is shown. DETAILED DESCRIPTION

[0046] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0047] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0048] like Figure 1 As shown, the first aspect of the present invention provides a method for identifying and warning transmission line defects based on remote sensing monitoring, comprising the following steps:

[0049] S102: constructing a drone integrated control network, optimizing the drone integrated control network, obtaining an optimized drone integrated control network, and controlling the drone through the drone integrated control network to obtain remote sensing image information of the power transmission line;

[0050] S104: Acquire historical defect image information of the transmission line through big data, construct a training set based on the historical defect image information of the transmission line, and construct a transmission line defect recognition model based on the training set;

[0051] S106: recognizing remote sensing image information of the transmission line through the transmission line defect recognition model, obtaining defect feature data of the transmission line in each area, and issuing an early warning based on the defect feature data of the transmission line in each area to generate early warning information;

[0052] S108: Setting a priority maintenance plan according to the warning information, and allocating work tasks based on the priority maintenance plan.

[0053] It should be noted that the present invention processes the target defect features in the training set through a cyclic spatial attention mechanism, so that the attention is focused on the target defect features in the training set, which can suppress the interference of multi-scale features on the transmission line defect recognition model and improve the recognition accuracy of transmission line defect recognition.

[0054] like Figure 2 As shown, further, in this method, a UAV integrated control network is constructed, and the UAV integrated control network is optimized to obtain an optimized UAV integrated control network, which specifically includes:

[0055] S202: constructing a drone integrated control network, and obtaining the historical drone control quantity upper limit change characteristics of the drone integrated control network by testing the drone integrated control network, and constructing a drone control quantity upper limit characteristic prediction model based on a convolutional neural network;

[0056] S204: inputting the historical UAV control quantity upper limit change characteristics of the UAV integrated control network into the UAV control quantity upper limit feature prediction model for training, and obtaining the trained UAV control quantity upper limit feature prediction model;

[0057] S206: Obtain the upper limit change feature of the number of drones controlled by the drone integrated control network within a preset time and input it into the trained upper limit feature prediction model of the number of drones controlled to perform prediction, and obtain the upper limit feature of the number of drones controlled by the drone integrated control network at the current timestamp;

[0058] S208: Obtain the real-time number of drone controls of the drone integrated control network. When the real-time number of drone controls is greater than the upper limit characteristic of the number of drone controls of the drone integrated control network at the current timestamp, adjust the real-time number of drone controls according to the upper limit characteristic of the number of drone controls of the drone integrated control network at the current timestamp to obtain the optimized drone integrated control network.

[0059] It should be noted that, in fact, since there is an upper limit to information transmission, there is an upper limit to the number of drone controls in the drone integrated control network; since the drone integrated control network is controlled by multiple devices, the performance of the terminal equipment will degrade to a certain extent after a certain number of years of use (such as a decrease in the amount of information transmission per unit time), which leads to an upper limit to the number of drone controls in the drone integrated control network. This method can adjust the real-time number of drone controls according to the upper limit characteristics of the number of drone controls in the drone integrated control network at the current timestamp, so that the number of drone controls controlled by the human-machine integrated control network meets the predetermined requirements, thereby ensuring the stability of the drone integrated control network.

[0060] like Figure 3 As shown, further, in this method, historical defect image information of the transmission line is obtained through big data, and a training set is constructed based on the historical defect image information of the transmission line, and a transmission line defect recognition model is constructed based on the training set, specifically including:

[0061] S302: Obtain historical defect image information of the transmission line through big data, and classify the defect types of the historical defect image information of the transmission line to obtain image information of each defect type, and construct a training set of each defect type according to the image information of each defect type;

[0062] S304: inputting the image information in the training set of each defect type into the feature pyramid network, obtaining the defect features existing in each image information, and determining whether there are at least two defect features existing in the image information, and introducing a cyclic spatial attention mechanism;

[0063] S306: When there are at least two defect features in the image information, the defect feature corresponding to the defect type label information of the current image information is obtained as the relevant defect feature; when there is only one defect feature in the image information, the defect feature corresponding to the defect type label information of the current image information is used as the relevant defect feature;

[0064] S308: Input the relevant defect features into the cyclic spatial attention mechanism to focus the attention on the relevant defect features, obtain the attention feature map, build a transmission line defect recognition model based on the deep neural network, and input the attention feature map in the training set into the transmission line defect recognition model for encoding learning.

[0065] It should be noted that the defect image information includes crack defect images, corrosion defect images, defect images after burning and other data. Since the image information in the training set of each defect type may contain at least two defect types in one image, such as crack defects, corrosion defects, etc., during training, the training sets of multiple defect types should be input into the deep neural network for training. When label information of at least two defect types appears in an image during training, it will be disturbed. By integrating the recurrent spatial attention mechanism, this method can focus on the relevant defect features, suppress the interference of multi-scale defect features on the transmission line defect recognition model, and thus improve the prediction accuracy of the model.

[0066] Furthermore, in this method, the remote sensing image information of the transmission line is recognized by the transmission line defect recognition model, the defect feature data of the transmission line in each area is obtained, and an early warning is issued according to the defect feature data of the transmission line in each area, and early warning information is generated, which specifically includes:

[0067] Inputting the remote sensing image information of the transmission line into the transmission line defect recognition model for recognition, obtaining the defect feature data of the transmission line in each area, and presetting the warning level standard;

[0068] According to the warning level standard, the defect characteristic data of the transmission line in each area are divided into warning levels, the defect characteristic warning level of the transmission line in each area is obtained, and the geographical location information of the target area where the transmission line has defects is obtained;

[0069] An early warning is generated based on the early warning level of the defect characteristics of the transmission line in each area, the geographical location information of the target area where the transmission line has defects, and the defect characteristic data of the transmission line in each area to generate early warning information.

[0070] It should be noted that the standards for warning levels can be set according to the type of defects and the size of the defects, where the warning levels include low warning levels, medium warning levels, high warning levels, etc.

[0071] Furthermore, in this method, a priority maintenance plan is set according to the early warning information, specifically including:

[0072] According to the early warning information, the geographical location information of the target area where the transmission line has defects is obtained, and a search tag is constructed based on the geographical location information of the target area where the transmission line has defects, and a search is performed according to the search tag to obtain the power supply area of ​​the transmission line;

[0073] Acquire quantity information of supply users according to the power supply area of ​​the transmission line, construct a supply user quantity ranking table, input the quantity information of supply users into the supply user quantity ranking table for ranking, and obtain a supply quantity ranking result from large to small;

[0074] A priority maintenance plan in the target area is set based on the supply quantity sorting result from large to small, and the priority maintenance plan is output.

[0075] It should be noted that, since the maintenance personnel may be limited, a more reasonable priority maintenance plan can be formulated through this method.

[0076] Furthermore, in this method, work tasks are allocated based on the priority maintenance plan, specifically including:

[0077] Obtain the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects, and calculate the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the power line has defects based on the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects;

[0078] A genetic algorithm is introduced, and the genetic algebra is set according to the genetic algorithm. The work position allocation of the maintenance personnel is initialized according to the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the transmission line has defects, and the work task location of each maintenance personnel is obtained;

[0079] Determine whether the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is greater than a preset Euclidean distance value; when the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is not greater than the preset Euclidean distance value, output the work task location of each maintenance personnel;

[0080] When the Euclidean distance value between the location of the maintenance personnel and the geographic location information of the target area where the transmission line has defects is greater than the preset Euclidean distance value, the work task location of each maintenance personnel is reallocated until the Euclidean distance value between the location of the maintenance personnel and the geographic location information of the target area where the transmission line has defects is no greater than the preset Euclidean distance value.

[0081] It should be noted that, through this method, the transmission lines to be repaired in the target area can be configured so that each maintenance area can respond quickly and the work tasks of the maintenance personnel are more reasonably allocated.

[0082] In addition, controlling the UAV to obtain remote sensing image information of the transmission line through the UAV integrated control network may also include the following steps:

[0083] The flight route of the current UAV is obtained through the UAV integrated control network, and the remote sensing image feature data information under each influencing factor is obtained through big data, and the remote sensing image feature data information under each influencing factor is input into the graph neural network; the influencing factor is used as the first graph node of the graph neural network, and the remote sensing image feature data information is used as the second graph node of the graph neural network, an adjacency matrix is ​​constructed according to the first graph node and the second graph node, and the adjacency matrix is ​​input into the knowledge graph for storage; the influencing factor information of the UAV's flight route within a preset range within a preset time is obtained, and the influencing factor information of the UAV's flight route within a preset range within the preset time is input into the knowledge graph to obtain the remote sensing image prediction feature data information of each timestamp; when the remote sensing image prediction feature data information is greater than the preset remote sensing image feature threshold, the corresponding time period is used as the working task period of the UAV, and the corresponding time period is used as the working task period of the UAV, and data collection is performed during the working task period of the UAV to obtain remote sensing image information of the transmission line.

[0084] It should be noted that the influencing factors include visibility, haze and other data. This method can further improve the rationality of drone data collection. The remote sensing image feature data information includes clarity, brightness and other data.

[0085] In addition, the method may also include the following steps: obtaining remote sensing image feature information of each camera parameter set under each influencing factor and the influencing factor of the area where the current drone is located through big data, and obtaining the remote sensing image feature information of the drone for each camera parameter according to the remote sensing image feature information of each camera parameter set under the influencing factor and the influencing factor of the area where the current drone is located; obtaining the adjustable range of the camera parameters of the camera of the current drone, and judging whether there is at least one camera parameter among the camera parameters that makes the remote sensing image feature information of the drone greater than the preset remote sensing image feature threshold; when there is at least one camera parameter among the camera parameters When the parameters make the remote sensing image feature information of the UAV greater than the data of the preset remote sensing image feature threshold, a camera parameter will be randomly output to make the remote sensing image feature information of the UAV greater than the data of the preset remote sensing image feature threshold and used as the camera parameter of the UAV; when there is no camera parameter among the various camera parameters that makes the remote sensing image feature information of the UAV greater than the data of the preset remote sensing image feature threshold, the data collection period of the UAV, the flight route of the UAV and the data collection point of the UAV are adjusted until at least one camera parameter among the various camera parameters makes the remote sensing image feature information of the UAV greater than the data of the preset remote sensing image feature threshold.

[0086] It should be noted that due to the influence of the environment, no matter how the camera parameters are adjusted, there is no parameter that makes the remote sensing image feature information of the drone greater than the preset remote sensing image feature threshold, so that the image of the predetermined standard cannot be obtained. This method can further improve the rationality of data collection.

[0087] like Figure 4 As shown, the second aspect of the present invention provides a transmission line defect recognition and early warning system 4 based on remote sensing monitoring. The transmission line defect recognition and early warning system 4 based on remote sensing monitoring includes a memory 41 and a processor 42. The memory 41 includes a transmission line defect recognition and early warning method program based on remote sensing monitoring. When the transmission line defect recognition and early warning method program based on remote sensing monitoring is executed by the processor 42, the following steps are implemented:

[0088] Construct an integrated control network for unmanned aerial vehicles, optimize the integrated control network for unmanned aerial vehicles, obtain an optimized integrated control network for unmanned aerial vehicles, and control unmanned aerial vehicles through the integrated control network for unmanned aerial vehicles to obtain remote sensing image information of power transmission lines;

[0089] Obtain historical defect image information of the transmission line through big data, build a training set based on the historical defect image information of the transmission line, and build a transmission line defect recognition model based on the training set;

[0090] The remote sensing image information of the transmission line is recognized through the transmission line defect recognition model to obtain the defect feature data of the transmission line in each area, and an early warning is issued based on the defect feature data of the transmission line in each area to generate early warning information;

[0091] Set a priority maintenance plan based on the early warning information, and assign work tasks based on the priority maintenance plan.

[0092] Furthermore, in this system, a drone integrated control network is constructed, and the optimized drone integrated control network is obtained by optimizing the drone integrated control network, which specifically includes:

[0093] Construct a UAV integrated control network, test the UAV integrated control network, obtain the historical UAV control quantity upper limit change characteristics of the UAV integrated control network, and build a UAV control quantity upper limit characteristic prediction model based on convolutional neural network;

[0094] Inputting the historical UAV control quantity upper limit change characteristics of the UAV integrated control network into the UAV control quantity upper limit feature prediction model for training, and obtaining the trained UAV control quantity upper limit feature prediction model;

[0095] Obtain the upper limit change characteristics of the number of drones controlled by the drone integrated control network within a preset time and input them into the trained upper limit characteristic prediction model of the number of drones controlled to make predictions, and obtain the upper limit characteristics of the number of drones controlled by the drone integrated control network at the current timestamp;

[0096] Get the real-time number of drone controls of the drone integrated control network. When the real-time number of drone controls is greater than the upper limit of the number of drone controls of the drone integrated control network at the current timestamp, adjust the real-time number of drone controls according to the upper limit of the number of drone controls of the drone integrated control network at the current timestamp to get the optimized drone integrated control network.

[0097] Furthermore, in this system, historical defect image information of the transmission line is obtained through big data, and a training set is constructed based on the historical defect image information of the transmission line, and a transmission line defect recognition model is constructed based on the training set, specifically including:

[0098] Obtain historical defect image information of the transmission line through big data, and classify the defect types of the historical defect image information of the transmission line to obtain image information of each defect type, and construct a training set of each defect type based on the image information of each defect type;

[0099] The image information in the training set of each defect type is input into the feature pyramid network to obtain the defect features in each image information, and determine whether there are at least two defect features in the image information, and introduce a cyclic spatial attention mechanism;

[0100] When there are at least two defect features in the image information, the defect feature corresponding to the defect type label information of the current image information is obtained as the relevant defect feature; when there is only one defect feature in the image information, the defect feature corresponding to the defect type label information of the current image information is used as the relevant defect feature;

[0101] The relevant defect features are input into the cyclic spatial attention mechanism to focus on the relevant defect features, obtain the attention feature map, build a transmission line defect recognition model based on the deep neural network, and input the attention feature map in the training set into the transmission line defect recognition model for encoding learning.

[0102] Furthermore, in this system, the remote sensing image information of the transmission line is recognized through the transmission line defect recognition model, the defect feature data of the transmission line in each area is obtained, and an early warning is issued according to the defect feature data of the transmission line in each area, and early warning information is generated, specifically including:

[0103] Inputting the remote sensing image information of the transmission line into the transmission line defect recognition model for recognition, obtaining the defect feature data of the transmission line in each area, and presetting the warning level standard;

[0104] According to the warning level standard, the defect characteristic data of the transmission line in each area are divided into warning levels, the defect characteristic warning level of the transmission line in each area is obtained, and the geographical location information of the target area where the transmission line has defects is obtained;

[0105] An early warning is generated based on the early warning level of the defect characteristics of the transmission line in each area, the geographical location information of the target area where the transmission line has defects, and the defect characteristic data of the transmission line in each area to generate early warning information.

[0106] Furthermore, in this system, a priority maintenance plan is set according to the early warning information, including:

[0107] According to the early warning information, the geographical location information of the target area where the transmission line has defects is obtained, and a search tag is constructed based on the geographical location information of the target area where the transmission line has defects, and a search is performed according to the search tag to obtain the power supply area of ​​the transmission line;

[0108] Acquire quantity information of supply users according to the power supply area of ​​the transmission line, construct a supply user quantity ranking table, input the quantity information of supply users into the supply user quantity ranking table for ranking, and obtain a supply quantity ranking result from large to small;

[0109] A priority maintenance plan in the target area is set based on the supply quantity sorting result from large to small, and the priority maintenance plan is output.

[0110] Furthermore, in this system, work tasks are allocated based on the priority maintenance plan, including:

[0111] Obtain the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects, and calculate the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the power line has defects based on the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects;

[0112] A genetic algorithm is introduced, and the genetic algebra is set according to the genetic algorithm. The work position allocation of the maintenance personnel is initialized according to the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the transmission line has defects, and the work task location of each maintenance personnel is obtained;

[0113] Determine whether the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is greater than a preset Euclidean distance value; when the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is not greater than the preset Euclidean distance value, output the work task location of each maintenance personnel;

[0114] When the Euclidean distance value between the location of the maintenance personnel and the geographic location information of the target area where the transmission line has defects is greater than the preset Euclidean distance value, the work task location of each maintenance personnel is reallocated until the Euclidean distance value between the location of the maintenance personnel and the geographic location information of the target area where the transmission line has defects is no greater than the preset Euclidean distance value.

[0115] The third aspect of the present invention provides a computer-readable storage medium, which includes a program for a transmission line defect identification and early warning method based on remote sensing monitoring. When the program for a transmission line defect identification and early warning method based on remote sensing monitoring is executed by a processor, any step of the transmission line defect identification and early warning method based on remote sensing monitoring is implemented.

[0116] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0117] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0118] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0119] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0120] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0121] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for identifying and warning transmission line defects based on remote sensing monitoring, characterized in that: The following steps are involved: Constructing an unmanned aerial vehicle integrated control network, optimizing the unmanned aerial vehicle integrated control network to obtain an optimized unmanned aerial vehicle integrated control network, and controlling the unmanned aerial vehicle through the unmanned aerial vehicle integrated control network to obtain remote sensing image information of the power transmission line; Acquire historical defect image information of the transmission line through big data, construct a training set based on the historical defect image information of the transmission line, and construct a transmission line defect recognition model based on the training set; Recognize the remote sensing image information of the transmission line through the transmission line defect recognition model, obtain the defect feature data of the transmission line in each area, and issue an early warning based on the defect feature data of the transmission line in each area to generate early warning information; Setting a priority maintenance plan according to the warning information, and allocating work tasks based on the priority maintenance plan; Obtaining historical defect image information of the transmission line through big data, building a training set based on the historical defect image information of the transmission line, and building a transmission line defect recognition model based on the training set, specifically including: Acquire historical defect image information of the transmission line through big data, classify the defect types of the historical defect image information of the transmission line, acquire image information of each defect type, and construct a training set of each defect type according to the image information of each defect type; Input the image information in the training set of each defect type into the feature pyramid network, obtain the defect features in each image information, and determine whether there are at least two defect features in the image information, and introduce a cyclic spatial attention mechanism; When there are at least two defect features in the image information, the defect feature corresponding to the defect type label information of the current image information is obtained as the relevant defect feature; when there is only one defect feature in the image information, the defect feature corresponding to the defect type label information of the current image information is used as the relevant defect feature; The relevant defect features are input into the cyclic spatial attention mechanism so that the attention is focused on the relevant defect features, an attention feature map is obtained, a transmission line defect recognition model is constructed based on a deep neural network, and the attention feature map is input into the transmission line defect recognition model for encoding learning.

2. A method for identifying and warning transmission line defects based on remote sensing monitoring according to claim 1, characterized in that: Constructing a UAV integrated control network, and optimizing the UAV integrated control network to obtain an optimized UAV integrated control network, specifically including: Constructing a drone integrated control network, and obtaining the historical drone control quantity upper limit change characteristics of the drone integrated control network by testing the drone integrated control network, and constructing a drone control quantity upper limit characteristic prediction model based on a convolutional neural network; Inputting the historical UAV control quantity upper limit change characteristics of the UAV integrated control network into the UAV control quantity upper limit characteristic prediction model for training, and obtaining the trained UAV control quantity upper limit characteristic prediction model; Obtaining the upper limit change characteristics of the number of drones controlled by the drone integrated control network within a preset time and inputting them into the trained upper limit characteristic prediction model for the number of drones controlled to perform prediction, and obtaining the upper limit characteristic of the number of drones controlled by the drone integrated control network at the current timestamp; The real-time number of drone controls of the drone integrated control network is obtained. When the real-time number of drone controls is greater than the upper limit characteristic of the number of drone controls of the drone integrated control network at the current timestamp, the real-time number of drone controls is adjusted according to the upper limit characteristic of the number of drone controls of the drone integrated control network at the current timestamp to obtain an optimized drone integrated control network.

3. The method for identifying and warning transmission line defects based on remote sensing monitoring according to claim 1 is characterized in that: The remote sensing image information of the transmission line is recognized by the transmission line defect recognition model, the defect feature data of the transmission line in each area is obtained, and an early warning is issued according to the defect feature data of the transmission line in each area to generate early warning information, which specifically includes: Inputting the remote sensing image information of the transmission line into the transmission line defect recognition model for recognition, obtaining the defect feature data of the transmission line in each area, and presetting the warning level standard; According to the warning level standard, the defect feature data of the transmission line in each area is divided into warning levels, the defect feature warning level of the transmission line in each area is obtained, and the geographical location information of the target area where the transmission line has defects is obtained; An early warning is generated based on the defect characteristic early warning level of the transmission line in each area, the geographical location information of the target area where the transmission line has defects, and the defect characteristic data of the transmission line in each area, and early warning information is generated.

4. The method for identifying and warning transmission line defects based on remote sensing monitoring according to claim 1 is characterized in that: A priority maintenance plan is set according to the warning information, specifically including: Acquire geographical location information of a target area where a transmission line defect exists according to the warning information, construct a search tag based on the geographical location information of the target area where a transmission line defect exists, and perform a search according to the search tag to acquire a power supply area of ​​the transmission line; Acquire quantity information of supply users according to the power supply area of ​​the transmission line, and construct a supply user quantity ranking table, input the quantity information of the supply users into the supply user quantity ranking table for ranking, and obtain a supply quantity ranking result from large to small; A priority maintenance plan in the target area is set based on the supply quantity sorting result from large to small, and the priority maintenance plan is output.

5. The method for identifying and warning transmission line defects based on remote sensing monitoring according to claim 1 is characterized in that: Based on the priority maintenance plan, work tasks are allocated, specifically including: Obtaining the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects, and calculating the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the power line has defects according to the distribution characteristic information of the current maintenance personnel and the geographical location information of the target area where the power line has defects; Introducing a genetic algorithm, setting a genetic algebra according to the genetic algorithm, initializing the maintenance personnel work position allocation according to the Euclidean distance value between the location of each maintenance personnel and the geographical location information of the target area where the transmission line has defects, and obtaining the work task location of each maintenance personnel; Determine whether the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is greater than a preset Euclidean distance value; when the Euclidean distance between the location of the maintenance personnel and the geographical location information of the target area where the transmission line has defects is not greater than the preset Euclidean distance value, output the work task location of each maintenance personnel; When the Euclidean distance value between the location of the maintenance personnel and the geographical location information of the target area with defects in the transmission line is greater than the preset Euclidean distance value, the work task location of each maintenance personnel is reallocated until the Euclidean distance value between the location of the maintenance personnel and the geographical location information of the target area with defects in the transmission line is no greater than the preset Euclidean distance value.

6. A transmission line defect identification and early warning system based on remote sensing monitoring, characterized in that: The transmission line defect recognition and early warning system based on remote sensing monitoring includes a memory and a processor, wherein the memory includes a transmission line defect recognition and early warning method program based on remote sensing monitoring, and when the transmission line defect recognition and early warning method program based on remote sensing monitoring is executed by the processor, the following steps are implemented: Constructing an unmanned aerial vehicle integrated control network, optimizing the unmanned aerial vehicle integrated control network to obtain an optimized unmanned aerial vehicle integrated control network, and controlling the unmanned aerial vehicle through the unmanned aerial vehicle integrated control network to obtain remote sensing image information of the power transmission line; Acquire historical defect image information of the transmission line through big data, construct a training set based on the historical defect image information of the transmission line, and construct a transmission line defect recognition model based on the training set; Recognize the remote sensing image information of the transmission line through the transmission line defect recognition model, obtain the defect feature data of the transmission line in each area, and issue an early warning based on the defect feature data of the transmission line in each area to generate early warning information; Setting a priority maintenance plan according to the warning information, and allocating work tasks based on the priority maintenance plan; Obtaining historical defect image information of the transmission line through big data, building a training set based on the historical defect image information of the transmission line, and building a transmission line defect recognition model based on the training set, specifically including: Acquire historical defect image information of the transmission line through big data, classify the defect types of the historical defect image information of the transmission line, acquire image information of each defect type, and construct a training set of each defect type according to the image information of each defect type; Input the image information in the training set of each defect type into the feature pyramid network, obtain the defect features in each image information, and determine whether there are at least two defect features in the image information, and introduce a cyclic spatial attention mechanism; When there are at least two defect features in the image information, the defect feature corresponding to the defect type label information of the current image information is obtained as the relevant defect feature; when there is only one defect feature in the image information, the defect feature corresponding to the defect type label information of the current image information is used as the relevant defect feature; The relevant defect features are input into the cyclic spatial attention mechanism so that the attention is focused on the relevant defect features, an attention feature map is obtained, a transmission line defect recognition model is constructed based on a deep neural network, and the attention feature map is input into the transmission line defect recognition model for encoding learning.

7. A computer-readable storage medium, characterized in that: The computer-readable memory medium includes a transmission line defect identification and early warning method program based on remote sensing monitoring. When the transmission line defect identification and early warning method program based on remote sensing monitoring is executed by a processor, the steps of the transmission line defect identification and early warning method based on remote sensing monitoring as described in any one of claims 1 to 5 are implemented.

Images