Electric power inspection image intelligent identification system based on unmanned aerial vehicle

Abstract

The invention discloses an electric power inspection image intelligent identification system based on an unmanned aerial vehicle, and relates to the technical field of electric power inspection, the unmanned aerial vehicle is used for carrying a high-definition camera to carry out image acquisition, the unmanned aerial vehicle can complete data acquisition in a short time, data accuracy is ensured, and acquired data is quantized into a site layout evaluation value; the standardized and quantitative evaluation mode eliminates man-made subjective differences, guarantees comparable and reliable evaluation results of different sites, provides exact basis for whether the electrical equipment can enter the site for assembly, reduces the risk of difficulty in subsequent assembly caused by potential problems of the site, automatically adjusts the hovering shooting angle and height of the unmanned aerial vehicle, and improves the efficiency of the unmanned aerial vehicle. The most valuable picture can be captured in the whole assembling process, and clear and high-quality image data are provided for monitoring operation and problem analysis. Along with the change of key areas in the assembly process, the unmanned aerial vehicle flexibly adapts, overcomes the monitoring dead angle limitation of a fixed camera, comprehensively considers the clean full view of a site, and avoids the problems of one-sided and omission of manual inspection.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power inspection, and in particular to an intelligent recognition system for electric power inspection images based on unmanned aerial vehicles. Background Art

[0002] With the continuous development of the power industry, the scale of power equipment is getting larger and larger, and the technical content is also increasing. For example, transformers, switch cabinets and other equipment in large substations are large in size and complex in structure, and the requirements for the site during the assembly process are very strict. In addition, the connection and coordination between equipment are more precise, and traditional inspection methods are difficult to meet the comprehensive monitoring needs of the assembly process of these complex equipment. Therefore, a power inspection image intelligent recognition system based on drones is needed.

[0003] Prior art, such as the invention application patent with announcement number: CN117217739A, discloses an intelligent power inspection system, which belongs to the field of power inspection and solves the problem of how to plan inspection routes and integrate multiple data to evaluate the operating status of drones to improve the efficiency of power inspection; the map acquisition module acquires the electronic map of the power equipment in the target section, and sends the acquired electronic map of the power equipment to the route acquisition module; the route acquisition module acquires the inspection route of the target section in combination with the electronic map of the power equipment in the target section, and sends the electronic map of the power equipment marked with the inspection route to the cloud platform for storage; the equipment inspection module inspects each inspection route of the target section according to the inspection route on the power map, and collects the drone operating status data during the inspection and sends it to the flight evaluation module; the flight evaluation module analyzes the drone operating status data, calculates the operation evaluation coefficient, and determines whether the corresponding drone is in normal operating state according to the operation evaluation coefficient.

[0004] In view of the above scheme, the inventors of the present application have found that the above technology has at least the following technical problems: 1. The prior art lacks the ability to obtain site layout data through drones, and has long relied on manual measurement. Manual measurement is easily affected by the accuracy of measuring tools, the operating level of surveyors and subjective factors, resulting in inaccurate data. For example, when measuring the slope of the ground, manual measurement may not be able to accurately capture slight slope changes, thereby affecting the judgment of the flatness of the site. For multiple equipment assembly sites of large power companies, this process will be very time-consuming and slow down the progress of preliminary preparations for equipment assembly. Moreover, manual evaluation is difficult to complete data collection and analysis of multiple sites in a short period of time, and it is impossible to promptly discover unqualified site layouts. Without a drone system, it is difficult to monitor and update the site layout, and it is impossible to promptly discover and re-evaluate whether the site layout still meets the equipment assembly requirements.

[0005] 2. The existing technology cannot conveniently obtain the working space contact data during the equipment assembly process, including space occupancy, personnel operating radius and the distance between personnel and assembly equipment. In the absence of real-time drone shooting and image analysis, obtaining these data requires manual observation and estimation, which is not only prone to errors, but also difficult to accurately count without interfering with normal operations, and difficult to adjust the observation angle in time according to the actual situation of the assembly process. During the equipment assembly process, as the assembly work progresses, the key areas and spatial relationships that need to be paid attention to at different stages will change. Without the intelligent adjustment function of the drone, there may be blind spots for observation, and potential problems in the working space cannot be fully and timely discovered, such as the risk of collision between personnel and equipment, unreasonable space utilization, etc., and the safety distance and operating space between personnel and equipment cannot be effectively monitored, which will increase the risk of safety accidents. For example, when personnel are operating large equipment or working in a narrow space, if there is no timely monitoring and early warning, it is easy to cause danger due to improper operation or insufficient space, endangering the life safety of the staff and the integrity of the equipment.

[0006] 3. It is difficult for existing technologies to fully obtain information such as the debris coverage rate, remaining ratio of channel width, and volume ratio of obstructions after the equipment assembly construction site is completed. Manual inspections may miss some areas, especially those that are difficult to reach or have poor visibility, resulting in inaccurate and incomplete assessments of the site cleanliness, and failure to promptly detect unqualified cleaning work. During the manual inspection process, it may be necessary to wait until all cleaning work is completed before inspection. At this time, if it is found that the cleaning is not up to standard, re-cleaning will waste more time and resources. Moreover, for large sites, the manual inspection cycle is long, and it is impossible to monitor the progress and quality of cleaning in real time. For unqualified cleaning work, it may not be possible to promptly and effectively notify relevant personnel and take appropriate measures. For example, minor cleaning problems may not be dealt with in a timely manner, gradually accumulating into serious problems, or serious cleaning problems may not attract the attention of sufficiently high management, resulting in delays in the problem. Summary of the invention

[0007] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide an intelligent recognition system for power inspection images based on drones.

[0008] In order to solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides an intelligent recognition system for power inspection images based on drones, including: a power pre-assembly inspection module: used to perform initial drone inspections before the assembly of each equipment of the target power company, and to determine whether the site layout corresponding to each equipment assembly site is qualified. If the site layout corresponding to a certain equipment assembly site is unqualified, an early warning prompt will be issued.

[0009] Inspection module during power assembly: It is used to conduct drone process inspections when the target power company's equipment is being assembled, and set several collection time points to obtain the work space contact data corresponding to each equipment assembly site at each collection time point, and then adjust the hovering shooting angle and height of the drone corresponding to each equipment assembly site at each collection time point.

[0010] Post-assembly inspection module for electric power: It is used to conduct drone inspection after the target power company completes the assembly of various equipment, so as to analyze the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site, and evaluate whether the cleaning work completed at each equipment assembly construction site is qualified.

[0011] Preferably, the power pre-assembly inspection module further includes a site layout data acquisition unit and a site layout evaluation value analysis unit.

[0012] The site layout data acquisition unit is used to use a drone to collect images of each equipment assembly site, and then collect site layout data corresponding to each equipment assembly site. The site layout data includes ground slope, actual assembly area error rate, and corresponding distances and heights of surrounding buildings.

[0013] The site layout evaluation value analysis unit is used to analyze the site layout data corresponding to each equipment assembly site to obtain the site layout evaluation value corresponding to each equipment assembly site.

[0014] Preferably, the specific judgment process for judging whether the site layout corresponding to each equipment assembly site is qualified is as follows: comparing the site layout evaluation value corresponding to each equipment assembly site with the site layout evaluation value corresponding to a set standard equipment assembly site; if the site layout evaluation value corresponding to a certain equipment assembly site is greater than or equal to the site layout evaluation value corresponding to the set standard equipment assembly site, then the site layout corresponding to the equipment assembly site is judged to be qualified; if the site layout evaluation value corresponding to a certain equipment assembly site is less than the site layout evaluation value corresponding to the set standard equipment assembly site, then the site layout corresponding to the equipment assembly site is judged to be unqualified.

[0015] Preferably, the work space connection data includes common space occupancy rate of assembly, operation radius of each person and the distance between each person and assembly equipment.

[0016] Preferably, the work space contact data corresponding to each equipment assembly is obtained at each collection time point, and the specific collection process is as follows: A1. The three-dimensional map model of each equipment assembly site of the target power company is imported into the control software of the UAV. The UAV flies according to the planned path, and at each collection time point, each assembly area is photographed using the onboard high-definition camera.

[0017] A2. Using the images collected by the drone, count the number of pixels occupied by people and equipment in each equipment assembly site at each collection time point, denoted as The total number of pixels of each equipment assembly site at each acquisition time point is counted and recorded as Among them, g represents the number corresponding to each collection time point, g is a positive integer, h represents the number corresponding to each device, h is a positive integer, substitute into the calculation formula: The space occupancy rate ρ corresponding to each device assembly at each collection time point is obtained gy .

[0018] A3. Using the images collected by the drone, the target detection algorithm is used to identify the gender and arm length of each person, and the type of tools used by each person is further identified. For the types of tools used by each person, the tool types used by each person in the image are matched with the tool types in the tool library through a pre-established tool library, and then the length of the tool used by each person is obtained, and the arm length of each person and the length of the tool used are accumulated to obtain the accumulated value corresponding to the arm length of each person and the length of the tool used, and the accumulated value is recorded as the person's operating radius, and the person's operating radius corresponding to each equipment assembly at each acquisition time point is obtained.

[0019] A4. Use image recognition technology to identify assembly equipment through images captured by drones. Determine the position of the assembly equipment in the image based on its appearance, logo and color. Combine the drone's positioning system and image information to determine the three-dimensional coordinate position of the assembly equipment in the field and the coordinate position of each person. Use the distance formula between two points in space to calculate the distance between each person and the assembly equipment at each acquisition time point.

[0020] Preferably, the hovering shooting angle and height corresponding to the drone at each equipment assembly site are adjusted at each collection time point, and the specific analysis process is as follows: B1. Obtain the operating space connection evaluation value corresponding to each equipment assembly site at each collection time point, and compare the operating space connection evaluation value corresponding to each equipment assembly site at each collection time point with the operating space connection evaluation value corresponding to the hovering shooting angle of each drone in the database. If the operating space connection evaluation value corresponding to a certain equipment assembly site at a certain collection time point is the same as the operating space connection evaluation value corresponding to the hovering shooting angle of a certain drone in the database, then the hovering shooting angle of the drone in the database is used as the hovering shooting angle corresponding to the drone at the equipment assembly site at that collection time point.

[0021] B2. Compare the work space connection evaluation value corresponding to each equipment assembly site at each collection time point with the work space connection evaluation value corresponding to the hovering shooting height of each drone in the database. If the work space connection evaluation value corresponding to a certain equipment assembly site at a certain collection time point is the same as the work space connection evaluation value corresponding to the hovering shooting height of a drone in the database, then the hovering shooting height of the drone in the database will be used as the hovering shooting height corresponding to the drone of the equipment assembly site at that collection time point.

[0022] Preferably, the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site is analyzed. The specific analysis process is as follows: C1. Obtain the debris coverage rate, channel width remaining ratio and obstruction volume ratio corresponding to the completion of each equipment assembly construction site.

[0023] C2. Input the debris coverage rate, remaining ratio of channel width and volume ratio of obstruction corresponding to the completion of each equipment assembly construction site into the environmental cleanliness assessment analysis model, and output the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site.

[0024] Preferably, the evaluation of whether the cleaning work completed at each equipment assembly construction site is qualified is specifically carried out as follows: D1. Compare the environmental cleaning evaluation value corresponding to the completion of each equipment assembly construction site with the set environmental cleaning evaluation value corresponding to the completion of a standard equipment assembly construction site. If the environmental cleaning evaluation value corresponding to the completion of a certain equipment assembly construction site is greater than the set environmental cleaning evaluation value corresponding to the completion of a standard equipment assembly construction site, it means that the cleaning work completed at the equipment assembly construction site is unqualified.

[0025] D2. If the cleaning work completed at an equipment assembly construction site is unqualified, and the environmental cleaning assessment value corresponding to the completion of the equipment assembly construction site does not exceed 30% of the set environmental cleaning assessment value corresponding to the completion of the standard equipment assembly construction site, it will be recorded as a slight cleaning failure, and an early warning message will be sent to the mobile phone work APP of the team leader responsible for the site cleaning finishing work and the on-site supervisor, and the early warning voice will be broadcast in a regular loop through the drone's intelligent voice broadcast terminal.

[0026] D3. If the cleaning work completed at an equipment assembly construction site is unqualified, and the environmental cleaning assessment value corresponding to the completion of the equipment assembly construction site exceeds 30% of the set environmental cleaning assessment value corresponding to the completion of the standard equipment assembly construction site, it will be recorded as a serious cleaning failure. Not only will an early warning message be sent to the mobile phone work APP of the team leader and on-site supervisor responsible for the site cleaning finishing work, but an email warning will also be sent to higher-level personnel such as the person in charge of the construction project and the head of the safety and quality control department. In addition, an alarm signal with a red flashing light and a high-decibel continuous beep will be emitted through the drone's intelligent voice broadcast terminal.

[0027] The beneficial effects of the present invention are as follows: 1. In the embodiment of the present invention, a drone equipped with a high-definition camera is used for image acquisition, and in conjunction with a site layout data acquisition unit, site layout data can be quickly and accurately collected, including ground slope, actual assembly area error rate, surrounding building distance and height, and other information, to avoid the drawbacks of large manual measurement errors and low efficiency. For example, in the preparation of a large substation equipment assembly site, for a large area with undulating terrain, a drone can complete data acquisition in a short time, saving a lot of time and energy compared with manual measurement, ensuring data accuracy, quantifying the collected data into a site layout evaluation value, and comparing it with the standard value to determine whether the site is qualified. This standardized and quantitative evaluation method eliminates subjective differences, ensures that the evaluation results of different sites are comparable and reliable, provides a definite basis for whether the power equipment can be assembled on site, and reduces the risk of subsequent assembly difficulties caused by potential site problems. Once the site layout evaluation is unqualified, an early warning prompt is immediately triggered, and the relevant responsible personnel can know the problem at the first time and rectify the site in time to avoid discovering site defects after the equipment assembly starts, resulting in idle work, rework, delays in construction period and increased costs. For example, if it is found that the distance between the site and the high-voltage tower is not up to standard, the plan can be adjusted in advance to ensure the safety and compliance prerequisites of equipment assembly.

[0028] 2. In the embodiment of the present invention, at each stage of equipment assembly, drones are used to obtain working space connection data according to set collection time points, covering key information such as space occupancy rate, personnel operating radius, and distance between personnel and equipment. Based on these data, the construction party can clearly grasp the space utilization status of the work site and the rationality of personnel operating space, and timely adjust the material stacking and personnel standing positions, optimize the operation process, and improve assembly efficiency. For example, when assembling equipment in a small distribution room, tools can be reasonably arranged to avoid crowded space hindering construction. By comparing the working space connection evaluation value with the database information, the drone's hovering shooting angle and height are automatically adjusted to ensure that the most valuable images can be captured throughout the assembly process, providing clear and high-quality image data for monitoring operations and analyzing problems. As key areas of the assembly process change, drones adapt flexibly, overcome the limitations of blind spots in fixed camera monitoring, and monitor the safety and standardization of the working space throughout the entire process. This effectively prevents safety hazards such as personnel mistakenly operating close to running equipment and working in dangerous spaces. While improving operational safety, the quality of equipment assembly is more guaranteed due to the smooth process and reasonable space, avoiding assembly defects caused by collisions and restricted operation, reducing the cost of later debugging and maintenance, and ensuring that the assembly of power equipment is carried out in accordance with quality and on time.

[0029] 3. The embodiment of the present invention comprehensively considers the overall picture of site cleanliness, eliminates one-sided and missed problems in manual inspection, and can accurately locate and quantify the evaluation of situations such as waste accumulation in corners and passages blocked by large debris, objectively reflects the actual level of site cleanliness, and meets the high-standard cleaning requirements of power sites. According to the comparison results of the cleaning evaluation value and the standard, it divides the level of minor and serious cleaning failures and matches the corresponding early warning methods. Minor problems are specifically notified to grassroots executive personnel, who are urged to conduct a quick review and cleanup with the help of mobile APP and voice broadcasts; serious problems are notified to senior managers through multiple channels, and with the help of eye-catching sound and light alarms, sufficient resources are allocated for rectification. This hierarchical mechanism rationally allocates management energy, ensures that cleaning problems are handled in an orderly and efficient manner, allows the site to quickly meet the delivery and operation standards, and provides a safe and clean environment for subsequent operation and maintenance personnel for daily inspections and maintenance operations, avoids delays in operation and maintenance work and damage to equipment due to poor site cleaning, extends the service life of power equipment, and ensures the long-term stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] 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, other drawings can be obtained based on these drawings without paying creative work.

[0031] Figure 1 This is a schematic diagram of the connection of the system modules of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Embodiments of the present invention Figure 1 As shown, an intelligent recognition system for power inspection images based on drones includes: a power assembly pre-inspection module, a power assembly mid-inspection module, a power assembly post-inspection module and a database.

[0034] The power assembly mid-inspection module is connected to the power assembly pre-inspection module and the power assembly post-inspection module respectively, and the database is connected to the power assembly mid-inspection module.

[0035] Power pre-assembly inspection module: It is used to conduct initial drone inspections before the target power company assembles its equipment, and to determine whether the site layout corresponding to each equipment assembly site is qualified. If the site layout corresponding to a certain equipment assembly site is unqualified, an early warning prompt will be issued.

[0036] In a specific embodiment, the power pre-assembly inspection module further includes a site layout data acquisition unit and a site layout evaluation value analysis unit.

[0037] The site layout data acquisition unit is used to use a drone to collect images of each equipment assembly site, and then collect site layout data corresponding to each equipment assembly site. The site layout data includes ground slope, actual assembly area error rate, and corresponding distances and heights of surrounding buildings.

[0038] It should be noted that the drone is equipped with a laser radar device. The laser radar measures the distance between an object and the equipment by emitting a laser beam and receiving reflected light. When collecting site images, the laser radar can quickly obtain elevation data of each point on the ground. Importing this data into geographic information system (GIS) software or specialized analysis tools can generate a high-precision digital elevation model (DEM), and then obtain the ground slope. The images taken by the drone are spliced ​​to form a complete site plan. During the shooting process, reference objects of known size are placed in the site. The scale of the image is determined based on the pixel size and actual size of these reference objects in the image. Then, the image analysis software is used to accurately outline the boundaries of the area for equipment assembly in the site design plan, calculate the pixel area of ​​the area in the spliced ​​image, and then convert it into the actual area according to the scale. This actual area is compared with the assembly area of ​​the design plan, and the actual assembly area error rate is calculated according to the formula error rate = (actual area-planned area) ÷ planned area × 100%. The positioning system of the drone provides its own location information. It is also equipped with a laser rangefinder. During the shooting process, the laser rangefinder directly measures the distance from the drone to different locations of the surrounding buildings. By integrating positioning information and ranging data, combined with parameters such as shooting angle, the distance between the building and each location on the site can be accurately calculated. For the measurement of building height, after obtaining the distance between the drone and the bottom and top of the building, the height of the building can be calculated using simple geometric relationships, such as the Pythagorean theorem.

[0039] The site layout evaluation value analysis unit is used to analyze the site layout data corresponding to each equipment assembly site to obtain the site layout evaluation value corresponding to each equipment assembly site.

[0040] In another specific embodiment, the analysis obtains the site layout evaluation value corresponding to each equipment assembly site. The specific analysis process is as follows: the ground slope corresponding to each equipment assembly site, the actual assembly area error rate, and the distance and height corresponding to each surrounding building are respectively recorded as q y 、s y , and y represents the number corresponding to each device, y is a positive integer, v represents the number corresponding to each building, v is a positive integer, substitute into the calculation formula:

[0041]

[0042] The site layout evaluation value α corresponding to each equipment assembly site is obtained y , where q′, s′, t′, and m′ are the standard ground slope, standard actual assembly area error rate, standard distance and standard height of surrounding buildings corresponding to the set equipment assembly site, respectively, and η 1 , η 2 , η 3 , η 4 They are respectively the weight factor corresponding to the ground slope of the set equipment assembly site, the weight factor corresponding to the error rate of the actual assembly area, the weight factor corresponding to the distance from surrounding buildings, and the weight factor corresponding to the height.

[0043] It should be noted that η 1 , η 2 , η 3 , η 4 Both are greater than 0 and less than 1.

[0044] It should also be noted that through the summary of a large amount of research data and experimental data, the standard ground slope corresponding to the equipment assembly site, the standard actual assembly area error rate, the standard distance corresponding to the surrounding buildings, and the standard height are set by professional institutions and research institutions. At the same time, based on the professional knowledge and research basis of experts in the field, and discussed and confirmed with industry organizations or professional institutions, the experts set the weight factors corresponding to the ground slope of the equipment assembly site, the weight factors corresponding to the actual assembly area error rate, the weight factors corresponding to the distance to the surrounding buildings, and the weight factors corresponding to the height according to their own experience and knowledge.

[0045] In another specific embodiment, the determination of whether the site layout corresponding to each equipment assembly site is qualified is performed as follows: the site layout evaluation value corresponding to each equipment assembly site is compared with the site layout evaluation value corresponding to a set standard equipment assembly site; if the site layout evaluation value corresponding to a certain equipment assembly site is greater than or equal to the site layout evaluation value corresponding to the set standard equipment assembly site, then the site layout corresponding to the equipment assembly site is determined to be qualified; if the site layout evaluation value corresponding to a certain equipment assembly site is less than the site layout evaluation value corresponding to the set standard equipment assembly site, then the site layout corresponding to the equipment assembly site is determined to be unqualified.

[0046] In the embodiment of the present invention, a drone equipped with a high-definition camera is used for image acquisition, and in conjunction with a site layout data acquisition unit, site layout data can be quickly and accurately collected, including ground slope, actual assembly area error rate, surrounding building distance and height, and other information, to avoid the drawbacks of large manual measurement errors and low efficiency. For example, in the preparation of a large substation equipment assembly site, for a large area with undulating terrain, a drone can complete data acquisition in a short time, saving a lot of time and energy compared with manual measurement, ensuring data accuracy, quantifying the collected data into a site layout evaluation value, and comparing it with the standard value to determine whether the site is qualified. This standardized and quantitative evaluation method eliminates subjective differences, ensures that the evaluation results of different sites are comparable and reliable, provides a definite basis for whether power equipment can be assembled on site, and reduces the risk of subsequent assembly difficulties caused by potential site problems. Once the site layout evaluation is unqualified, an early warning prompt is immediately triggered, and the relevant responsible personnel can know the problem at the first time and rectify the site in time to avoid discovering site defects after the equipment assembly starts, resulting in idle work, rework, delays in construction period and increased costs. For example, if it is found that the distance between the site and the high-voltage tower is not up to standard, the plan can be adjusted in advance to ensure the safety and compliance prerequisites of equipment assembly.

[0047] Inspection module during power assembly: It is used to conduct drone process inspections when the target power company's equipment is being assembled, and set several collection time points to obtain the work space contact data corresponding to each equipment assembly site at each collection time point, and then adjust the hovering shooting angle and height of the drone corresponding to each equipment assembly site at each collection time point.

[0048] In a specific embodiment, the work space connection data includes common space occupancy rate of assembly, operation radius of each person, and the distance between each person and assembly equipment.

[0049] In a specific embodiment, the work space contact data corresponding to each equipment assembly is obtained at each collection time point, and the specific collection process is as follows: A1. The three-dimensional map model of each equipment assembly site of the target power company is imported into the control software of the UAV. The UAV flies according to the planned path, and at each collection time point, the high-definition camera on board is used to shoot each assembly area.

[0050] A2. Using the images collected by the drone, count the number of pixels occupied by people and equipment in each equipment assembly site at each collection time point, denoted as The total number of pixels of each equipment assembly site at each acquisition time point is counted and recorded as Among them, g represents the number corresponding to each collection time point, g is a positive integer, h represents the number corresponding to each device, h is a positive integer, substitute into the calculation formula: The space occupancy rate ρ corresponding to each device assembly at each collection time point is obtained gy .

[0051] A3. Using the images collected by the drone, the target detection algorithm is used to identify the gender and arm length of each person, and the type of tools used by each person is further identified. For the types of tools used by each person, the tool types used by each person in the image are matched with the tool types in the tool library through a pre-established tool library, and then the length of the tool used by each person is obtained, and the arm length of each person and the length of the tool used are accumulated to obtain the accumulated value corresponding to the arm length of each person and the length of the tool used, and the accumulated value is recorded as the person's operating radius, and the person's operating radius corresponding to each equipment assembly at each acquisition time point is obtained.

[0052] A4. Use image recognition technology to identify assembly equipment through images captured by drones. Determine the position of the assembly equipment in the image based on its appearance, logo and color. Combine the drone's positioning system and image information to determine the three-dimensional coordinate position of the assembly equipment in the field and the coordinate position of each person. Use the distance formula between two points in space to calculate the distance between each person and the assembly equipment at each acquisition time point.

[0053] In a specific embodiment, the hovering shooting angle and height corresponding to the drone of each equipment assembly site are adjusted at each collection time point, and the specific analysis process is as follows: B1. Obtain the operating space connection evaluation value corresponding to each equipment assembly site at each collection time point, and compare the operating space connection evaluation value corresponding to each equipment assembly site at each collection time point with the operating space connection evaluation value corresponding to the hovering shooting angle of each drone in the database. If the operating space connection evaluation value corresponding to a certain equipment assembly site at a certain collection time point is the same as the operating space connection evaluation value corresponding to the hovering shooting angle of a certain drone in the database, then the hovering shooting angle of the drone in the database is used as the hovering shooting angle corresponding to the drone of the equipment assembly site at the collection time point.

[0054] B2. Compare the work space connection evaluation value corresponding to each equipment assembly site at each collection time point with the work space connection evaluation value corresponding to the hovering shooting height of each drone in the database. If the work space connection evaluation value corresponding to a certain equipment assembly site at a certain collection time point is the same as the work space connection evaluation value corresponding to the hovering shooting height of a drone in the database, then the hovering shooting height of the drone in the database will be used as the hovering shooting height corresponding to the drone of the equipment assembly site at that collection time point.

[0055] In the embodiment of the present invention, at each stage of equipment assembly, the drone is used to obtain the working space connection data according to the set collection time point, covering key information such as space occupancy rate, personnel operation radius and the distance between personnel and equipment. Based on these data, the construction party can clearly grasp the space utilization status and the rationality of the personnel operation space at the work site, timely adjust the material stacking and personnel position, optimize the operation process, and improve the assembly efficiency. For example, when assembling equipment in a small distribution room, the tool placement is reasonably arranged to avoid space congestion hindering the construction. By comparing the working space connection evaluation value with the database information, the drone hovering shooting angle and height are automatically adjusted to ensure that the most valuable pictures can be captured throughout the assembly process, providing clear and high-quality image data for monitoring operations and analyzing problems. As the key areas of the assembly process change, the drone flexibly adapts, overcomes the limitations of the fixed camera monitoring blind spots, and monitors the safety and standardization of the working space without omission throughout the process, effectively preventing safety hazards such as personnel operating close to the running equipment and operating in dangerous spaces by mistake. While improving the safety of the operation, the quality of equipment assembly is more guaranteed due to the smooth process and reasonable space, avoiding assembly defects caused by collisions and limited operations, reducing the cost of later debugging and maintenance, and ensuring that the assembly of power equipment is carried out in accordance with quality and on time.

[0056] Post-assembly inspection module for electric power: It is used to conduct drone inspection after the target power company completes the assembly of various equipment, so as to analyze the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site, and evaluate whether the cleaning work completed at each equipment assembly construction site is qualified.

[0057] In a specific embodiment, the process of obtaining the operation space connection evaluation value corresponding to each equipment assembly site at each collection time point is as follows: Denote the space occupancy rate, the operation radius of each person, and the distance between each person and the assembly equipment corresponding to each equipment assembly site obtained at each collection time point as and Let y represent the number corresponding to each equipment, where y is a positive integer, g represent the number corresponding to each collection time point, where g is a positive integer, and h represent the number corresponding to each person, where h is a positive integer. Substitute them into the calculation formula: to obtain the operation space connection evaluation value corresponding to each equipment assembly site at each collection time point where w′, f′, and k′ are respectively the standard space occupancy rate, the standard operation radius of personnel, and the standard distance between personnel and the assembly equipment corresponding to the set equipment assembly site, and υ 1 , υ 2 , υ 3 are respectively the weight factors corresponding to the space occupancy rate of the equipment assembly site, the weight factor corresponding to the operation radius of personnel, and the weight factor corresponding to the distance between personnel and the assembly equipment.

[0058] It should be noted that υ 1 , υ 2 , υ 3 are all greater than 0 and less than 1.

[0059] It should also be noted that through the summary of a large amount of research data and experimental data. According to the settings of professional institutions and research institutions, the standard space occupancy rate, the standard operation radius of personnel, and the standard distance between personnel and the assembly equipment corresponding to the equipment assembly site are determined. At the same time, based on the professional knowledge and research basis of domain experts, discussions and confirmations are carried out with industry organizations or professional institutions. The weight factors corresponding to the space occupancy rate of the equipment assembly site, the weight factor corresponding to the operation radius of personnel, and the weight factor corresponding to the distance between personnel and the assembly equipment are set by experts according to their own experience and knowledge.

[0060] In another specific embodiment, the environmental cleanliness evaluation value corresponding to the completion of each equipment assembly construction site is analyzed. The specific analysis process is as follows: C1. Obtain the debris coverage rate, the remaining ratio of the channel width, and the proportion of the blocked object volume corresponding to the completion of each equipment assembly construction site.

[0061] C2. Input the debris coverage rate, the remaining ratio of the channel width, and the proportion of the blocked object volume corresponding to the completion of each equipment assembly construction site into the environmental cleanliness evaluation analysis model, and output the environmental cleanliness evaluation value corresponding to the completion of each equipment assembly construction site.

[0062] The analysis process of the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site is as follows: the debris coverage rate, channel width remaining ratio and obstruction volume ratio corresponding to the completion of each equipment assembly construction site are normalized, and the debris coverage rate, channel width remaining ratio and obstruction volume ratio corresponding to the completion of each equipment assembly construction site after the processing are recorded as Z y , H y and F y , substitute into the analysis formula Obtain the environmental cleanliness assessment value δ corresponding to the completion of each equipment assembly construction site y , They are respectively the weight coefficient corresponding to the coverage rate of completed debris at the set equipment assembly construction site, the weight coefficient corresponding to the remaining proportion of the channel width, and the weight coefficient corresponding to the volume proportion of the obstruction, where y represents the number corresponding to each equipment and y is a positive integer.

[0063] It should be noted that Both are greater than 0 and less than 1.

[0064] It should also be noted that, based on the expertise and research of experts in the field, and after discussion and confirmation with industry organizations or professional institutions, the experts set the weight coefficients corresponding to the coverage rate of completed debris at the equipment assembly construction site, the weight coefficient corresponding to the remaining proportion of the channel width, and the weight coefficient corresponding to the volume proportion of the obstruction based on their own experience and knowledge.

[0065] In another specific embodiment, the evaluation of whether the cleaning work completed at each equipment assembly construction site is qualified is performed as follows: D1. Compare the environmental cleaning evaluation value corresponding to the completion of each equipment assembly construction site with the set environmental cleaning evaluation value corresponding to the completion of a standard equipment assembly construction site. If the environmental cleaning evaluation value corresponding to the completion of a certain equipment assembly construction site is greater than the set environmental cleaning evaluation value corresponding to the completion of a standard equipment assembly construction site, it means that the cleaning work completed at the equipment assembly construction site is unqualified.

[0066] D2. If the cleaning work completed at an equipment assembly construction site is unqualified, and the environmental cleaning assessment value corresponding to the completion of the equipment assembly construction site does not exceed 30% of the set environmental cleaning assessment value corresponding to the completion of the standard equipment assembly construction site, it will be recorded as a slight cleaning failure, and an early warning message will be sent to the mobile phone work APP of the team leader responsible for the site cleaning finishing work and the on-site supervisor, and the early warning voice will be broadcast in a regular loop through the drone's intelligent voice broadcast terminal.

[0067] D3. If the cleaning work completed at an equipment assembly construction site is unqualified, and the environmental cleaning assessment value corresponding to the completion of the equipment assembly construction site exceeds 30% of the set environmental cleaning assessment value corresponding to the completion of the standard equipment assembly construction site, it will be recorded as a serious cleaning failure. Not only will an early warning message be sent to the mobile phone work APP of the team leader and on-site supervisor responsible for the site cleaning finishing work, but an email warning will also be sent to higher-level personnel such as the person in charge of the construction project and the head of the safety and quality control department. In addition, an alarm signal with a red flashing light and a high-decibel continuous beep will be emitted through the drone's intelligent voice broadcast terminal.

[0068] The embodiments of the present invention comprehensively consider the overall picture of site cleanliness, eliminate one-sided and missed problems in manual inspections, and accurately locate and quantify the evaluation of situations such as waste accumulation in corners and passages blocked by large debris, objectively reflect the true level of site cleanliness, and meet the high-standard cleaning requirements of power sites. According to the comparison results of the cleaning evaluation value and the standard, the levels of minor and serious cleaning failures are divided and matched with corresponding early warning methods. Minor problems are specifically notified to grassroots executive personnel, who are urged to conduct a quick review and cleanup with the help of mobile APP and voice broadcasts; serious problems are notified to senior managers through multiple channels, and with the help of eye-catching sound and light alarms, sufficient resources are allocated for rectification. This hierarchical mechanism rationally allocates management energy, ensures that cleaning problems are handled in an orderly and efficient manner, allows the site to quickly meet the delivery and operation standards, and provides a safe and clean environment for subsequent operation and maintenance personnel for daily inspections and maintenance operations, avoids delays in operation and maintenance work and damage to equipment due to poor site cleaning, extends the service life of power equipment, and ensures the long-term stable operation of the power system.

[0069] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they shall all fall within the protection scope of the present invention.

Claims

1. An intelligent recognition system for power inspection images based on drones, characterized in that: include: Pre-assembly inspection module for electric power: used to conduct initial inspection by drones before the target power company assembles its equipment, and to determine whether the site layout corresponding to each equipment assembly site is qualified. If the site layout corresponding to a certain equipment assembly site is unqualified, an early warning prompt will be issued; Inspection module for power assembly: It is used to conduct drone inspections when the target power company's equipment is being assembled, and set several collection time points to obtain the work space contact data corresponding to each equipment assembly site at each collection time point, and then adjust the hovering shooting angle and height of the drone corresponding to each equipment assembly site at each collection time point; Post-assembly inspection module for electric power: It is used to conduct drone inspection after the target power company completes the assembly of various equipment, so as to analyze the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site, and evaluate whether the cleaning work completed at each equipment assembly construction site is qualified.

2. The intelligent recognition system for power inspection images based on drones as claimed in claim 1, characterized in that: The power assembly pre-inspection module also includes a site layout data acquisition unit and a site layout evaluation value analysis unit; The site layout data acquisition unit is used to use the drone to collect images of each equipment assembly site, and then collect the site layout data corresponding to each equipment assembly site, the site layout data including the ground slope, the actual assembly area error rate, and the distance and height corresponding to each surrounding building; The site layout evaluation value analysis unit is used to analyze the site layout data corresponding to each equipment assembly site to obtain the site layout evaluation value corresponding to each equipment assembly site.

3. The intelligent recognition system for power inspection images based on drones as claimed in claim 2, characterized in that: The analysis obtains the site layout evaluation value corresponding to each equipment assembly site. The specific analysis process is as follows: The ground slope corresponding to each equipment assembly site, the actual assembly area error rate, and the distance and height of the surrounding buildings are recorded as q y 、s y , and y represents the number corresponding to each device, y is a positive integer, v represents the number corresponding to each building, v is a positive integer, substitute into the calculation formula: The site layout evaluation value α corresponding to each equipment assembly site is obtained y , where q′, s′, t′, and m′ are the standard ground slope, standard actual assembly area error rate, standard distance and standard height of the equipment assembly site, respectively; η1, η2, η3, and η4 are the weight factors corresponding to the ground slope, actual assembly area error rate, distance and height of the surrounding buildings, respectively.

4. The intelligent recognition system for power inspection images based on drones as claimed in claim 3 is characterized in that: The specific process of judging whether the site layout corresponding to each equipment assembly site is qualified is as follows: The site layout assessment value corresponding to each equipment assembly site is compared with the site layout assessment value corresponding to the set standard equipment assembly site. If the site layout assessment value corresponding to an equipment assembly site is greater than or equal to the site layout assessment value corresponding to the set standard equipment assembly site, the site layout corresponding to the equipment assembly site is judged to be qualified. If the site layout assessment value corresponding to an equipment assembly site is less than the site layout assessment value corresponding to the set standard equipment assembly site, the site layout corresponding to the equipment assembly site is judged to be unqualified.

5. The intelligent recognition system for power inspection images based on drones as claimed in claim 1, characterized in that: The work space connection data includes common space occupancy rate of assembly, operation radius of each person and the distance between each person and assembly equipment.

6. The intelligent recognition system for power inspection images based on drones as claimed in claim 5, characterized in that: The operation space connection data corresponding to each equipment assembly is obtained at each collection time point, and the specific collection process is as follows: A1. Import the 3D map model of each equipment assembly site of the target power company into the control software of the drone. The drone flies according to the planned path and uses the high-definition camera on board to take pictures of each assembly area at each collection time point. A2. Using the images collected by the drone, count the number of pixels occupied by people and equipment in each equipment assembly site at each collection time point, denoted as The total number of pixels of each equipment assembly site at each acquisition time point is counted and recorded as Among them, g represents the number corresponding to each collection time point, g is a positive integer, h represents the number corresponding to each device, h is a positive integer, substitute into the calculation formula: The space occupancy rate ρ corresponding to each device assembly at each collection time point is obtained gy ; A3. Using the images collected by the drone, the target detection algorithm is used to identify the gender and arm length of each person, and the type of tool used by each person is further identified. For the tool types used by each person, the tool types used by each person in the image are matched with the tool types in the tool library through a pre-established tool library, and the length of the tool used by each person is obtained. The arm length of each person and the length of the tool used are accumulated to obtain the accumulated value corresponding to the arm length of each person and the length of the tool used, and the accumulated value is recorded as the person's operating radius to obtain the person's operating radius corresponding to each equipment assembly at each acquisition time point; A4. Use image recognition technology to identify assembly equipment through images captured by drones. Determine the position of the assembly equipment in the image based on its appearance, logo and color. Combine the drone's positioning system and image information to determine the three-dimensional coordinate position of the assembly equipment in the field and the coordinate position of each person. Use the distance formula between two points in space to calculate the distance between each person and the assembly equipment at each acquisition time point.

7. The intelligent recognition system for power inspection images based on drones as claimed in claim 6, characterized in that: The hovering shooting angle and height of the drone corresponding to each equipment assembly site are adjusted at each collection time point. The specific analysis process is as follows: B1. Obtain the operation space connection evaluation value corresponding to each equipment assembly site at each collection time point, and compare the operation space connection evaluation value corresponding to each equipment assembly site at each collection time point with the operation space connection evaluation value corresponding to each drone hovering shooting angle in the database. If the operation space connection evaluation value corresponding to a certain equipment assembly site at a certain collection time point is the same as the operation space connection evaluation value corresponding to a certain drone hovering shooting angle in the database, then use the drone hovering shooting angle in the database as the drone hovering shooting angle corresponding to the equipment assembly site at the collection time point; B2. Compare the work space connection evaluation value corresponding to each equipment assembly site at each collection time point with the work space connection evaluation value corresponding to the hovering shooting height of each drone in the database. If the work space connection evaluation value corresponding to a certain equipment assembly site at a certain collection time point is the same as the work space connection evaluation value corresponding to the hovering shooting height of a drone in the database, then the hovering shooting height of the drone in the database will be used as the hovering shooting height corresponding to the drone of the equipment assembly site at that collection time point.

8. The intelligent recognition system for power inspection images based on drones as claimed in claim 1, characterized in that: The specific acquisition process of obtaining the work space connection evaluation value corresponding to each equipment assembly site at each collection time point is as follows: The space occupancy rate of each equipment assembly site, the operating radius of each person, and the distance between each person and the assembly equipment at each collection time point are recorded as and y represents the number corresponding to each device, y is a positive integer, g represents the number corresponding to each collection time point, g is a positive integer, h represents the number corresponding to each person, h is a positive integer, substitute into the calculation formula: The work space connection evaluation value corresponding to each equipment assembly site at each collection time point is obtained Among them, w′, f′, and k′ are the standard space occupancy rate, standard personnel operating radius, and standard distance between personnel and assembly equipment corresponding to the set equipment assembly site, respectively; υ1, υ2, and υ3 are the weight factors corresponding to the space occupancy rate, personnel operating radius, and distance between personnel and assembly equipment of the set equipment assembly site, respectively.

9. The intelligent recognition system for power inspection images based on drones as claimed in claim 8, characterized in that: The analysis is done on the environmental cleanliness assessment values ​​corresponding to the completion of each equipment assembly construction site. The specific analysis process is as follows: C1. Obtain the debris coverage rate, remaining ratio of channel width and volume ratio of obstructions corresponding to the completion of each equipment assembly construction site; C2. Input the debris coverage rate, remaining ratio of channel width and volume ratio of obstruction corresponding to the completion of each equipment assembly construction site into the environmental cleanliness assessment analysis model, and output the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site.

10. The intelligent recognition system for power inspection images based on drones as claimed in claim 9, characterized in that: The specific evaluation process of evaluating whether the cleaning work completed at each equipment assembly construction site is qualified is as follows: D1. Compare the environmental cleanliness assessment value corresponding to the completion of each equipment assembly construction site with the environmental cleanliness assessment value corresponding to the completion of the set standard equipment assembly construction site. If the environmental cleanliness assessment value corresponding to the completion of a certain equipment assembly construction site is greater than the environmental cleanliness assessment value corresponding to the completion of the set standard equipment assembly construction site, it means that the cleaning work of the equipment assembly construction site is unqualified; D2. If the cleaning work completed at a certain equipment assembly construction site is unqualified, and the environmental cleaning assessment value corresponding to the completion of the equipment assembly construction site does not exceed 30% of the set environmental cleaning assessment value corresponding to the completion of the standard equipment assembly construction site, it will be recorded as a slight cleaning failure, and an early warning message will be sent to the mobile phone work APP of the team leader responsible for the site cleaning and the on-site supervisor, and the early warning voice will be broadcast regularly and cyclically through the intelligent voice broadcast terminal of the drone; D3. If the cleaning work completed at an equipment assembly construction site is unqualified, and the environmental cleaning assessment value corresponding to the completion of the equipment assembly construction site exceeds 30% of the set environmental cleaning assessment value corresponding to the completion of the standard equipment assembly construction site, it will be recorded as a serious cleaning failure. Not only will an early warning message be sent to the mobile phone work APP of the team leader and on-site supervisor responsible for the site cleaning finishing work, but an email warning will also be sent to higher-level personnel such as the person in charge of the construction project and the head of the safety and quality control department. In addition, an alarm signal with a red flashing light and a high-decibel continuous beep will be emitted through the drone's intelligent voice broadcast terminal.

Images