Distributed photovoltaic super-capacity intelligent detection method and system based on unmanned aerial vehicle

Abstract

The invention relates to a distributed photovoltaic super-capacity intelligent detection method and system. The method specifically comprises the steps that specific information of a photovoltaic panel is collected and obtained through an unmanned aerial vehicle; the ground receiving device extracts the video information collected by the unmanned aerial vehicle, extracts and analyzes the photovoltaic panorama, and obtains the specific number and information of the photovoltaic panel assembly; according to the specific photovoltaic panel information and the registration data information in the system, the ground receiving device calculates the virtual mark condition and obtains the total over-capacity; the ground receiving equipment calculates and obtains all dismantling paths and calculates an optimal dismantling path; and generating a specific dismantling scheme according to the optimal dismantling path, and sending the specific dismantling scheme to a related personnel port through the cloud management system. According to the invention, through the unmanned aerial vehicle, large-area distributed photovoltaic can be rapidly covered, efficient and continuous inspection is carried out, the inspection time is greatly shortened, the working efficiency is improved, and the standardization and intelligence level of photovoltaic inspection is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic overcapacity detection, and specifically relates to a distributed photovoltaic overcapacity intelligent detection method and system based on unmanned aerial vehicle. Background Art

[0002] Under the existing technology, the specific method for staff to detect whether the photovoltaic panel is over capacity is:

[0003] Auditors calculate the overcapacity rate of all photovoltaic customers in the system based on the audit rules for distributed photovoltaics exceeding theoretical values ​​after meter reading at the beginning of the month, screen out power generation customers suspected of overcapacity, and generate marketing audit anomalies in the system.

[0004] The marketing inspection abnormality details and auxiliary judgment system information are sent to the person in charge of electricity inspection at the grassroots unit to notify him to check and verify on-site whether the abnormality is true. After receiving the clue, the inspector needs to retrieve the registered capacity of the photovoltaic power generation user, which is calculated and planned based on the area of ​​the installation site, the specifications of the photovoltaic panels (such as the power of each photovoltaic panel), the rated power of the inverter, and the grid-connected agreement and other related factors to complete the preparation work before the on-site inspection.

[0005] The inspectors will go to the site to conduct an on-site inventory of the installed photovoltaic panels and take photos to record the actual number of photovoltaic panels installed. According to the actual number of photovoltaic panels installed and the single panel capacity marked on the nameplate, the actual grid-connected photovoltaic panel capacity is calculated and compared with the actual registered photovoltaic panel capacity. If the actual grid-connected photovoltaic panel capacity exceeds the actual registered photovoltaic panel capacity, then there is an overcapacity; otherwise, there is no overcapacity.

[0006] The traditional manual method of checking whether photovoltaic power generation households are over capacity has the following disadvantages:

[0007] Inefficiency: Manual inspection usually requires inspectors to measure and record each photovoltaic panel one by one. For large-scale photovoltaic power plants or distributed photovoltaic users, the inspection process is time-consuming and inefficient. Accuracy is affected by human factors: During manual inspection, factors such as the inspector's skill level, concentration, and the accuracy of the measuring tools may affect the accuracy of over-capacity judgment, and there is a risk of misjudgment or omission. High safety risk: Manual inspection may require climbing on the roof or working at high altitudes outdoors, especially in complex terrain or severe weather conditions, which poses a high safety risk. Inconvenient data recording and analysis: During manual inspection, data recording and analysis usually rely on paper records or simple electronic equipment, and the efficiency of data collation and analysis is low, which is not conducive to long-term tracking and comparison of power plant operation status.

[0008] In order to solve the above problems, an innovative distributed photovoltaic capacity detection method and system is urgently needed. Summary of the invention

[0009] To overcome the shortcomings of the prior art, the present invention provides a distributed photovoltaic overcapacity intelligent detection method and system based on drones, which can quickly and effectively detect overcapacity through drones, and generate a removal plan if overcapacity exists, including:

[0010] Step S1, obtaining specific information of photovoltaic panels through drone collection;

[0011] Step S2, the ground receiving equipment extracts information from the video information collected by the drone, and extracts and analyzes the photovoltaic panoramic image to obtain the specific photovoltaic panel component number and information;

[0012] Step S3, based on the specific photovoltaic panel information and the registered data information in the system, the ground receiving device calculates the virtual label situation and obtains the total excess capacity;

[0013] Step S4, the ground receiving device calculates and obtains all removal paths, and calculates the best removal path;

[0014] Step S5, generating a specific demolition plan based on the optimal demolition path and sending it to the relevant personnel port through the cloud management system.

[0015] Preferably, the method of obtaining the photovoltaic panel information comprises installing a camera device and a positioning device on the drone, collecting panoramic video information of the photovoltaic panels through the camera device of the drone, obtaining the location information of the photovoltaic panels through the positioning device of the drone, and querying the photovoltaic panel information registered in the cloud management system through the location information.

[0016] Preferably, the steps of extracting information from the video information and extracting and analyzing the photovoltaic panoramic image to obtain the specific photovoltaic panel number and information are as follows:

[0017] Step S21, constructing a FasterRCNN network model to identify photovoltaic panel information by extracting feature points, setting the photovoltaic module recognition image data set of each brand as a training set, and training the FasterRCNN network model through the training set;

[0018] Step S22, extracting information from the video information to obtain photovoltaic panel information, specifically:

[0019] Step S221, using a small target image enhancement technology to capture a panoramic area of ​​the photovoltaic panel in the video information;

[0020] Step S222, using the remap algorithm of OpenCV to perform dedistortion processing on the photovoltaic panel area image;

[0021] Step S223, matching the feature points of the panoramic image with the feature points of the photovoltaic panel plan view by using the surf algorithm, marking the matching area, performing image segmentation according to the matching situation, and obtaining image information of a single photovoltaic module;

[0022] Step S224, numbering according to the position;

[0023] Step S225, using the trained network model to identify a single photovoltaic module, and outputting the identification result and photovoltaic panel image label information via TXT text content;

[0024] Step S226, obtaining specific brand information of photovoltaic panels registered in the current area from the cloud management system for matching, and obtaining the capacity data information of each photovoltaic panel component according to the photovoltaic panel brand information identified by the model.

[0025] Preferably, the obtaining of the virtual overcapacity and the obtaining of the total overcapacity is specifically as follows:

[0026] The UAV is provided with a standard capacity photovoltaic panel. The tilt angle of the ground photovoltaic panel is obtained by processing and analyzing the video image information. The angle of the photovoltaic panel carried by the UAV is adjusted to make the tilt angle of the UAV photovoltaic panel the same as the tilt angle of the ground photovoltaic panel. The power generation data per unit time of the standard capacity photovoltaic panel is obtained. According to the photovoltaic panel information in the image information, the false marking situation is obtained. If there is a false marking situation, the false marking coefficient Z is configured. n ;

[0027] Calculate the total capacitance of the photovoltaic modules and the total capacitance of the capacitor plates registered in the database, and compare the two capacitances to obtain the total excess capacitance.

[0028] Preferably, the obtaining of all removal paths and calculating the best removal path is specifically as follows:

[0029] Step S41, obtaining the removal path of all photovoltaic panels;

[0030] Step S42, calculating the impact factor and the total score of each demolition path;

[0031] Step S43, calculating and obtaining the best demolition path;

[0032] According to a distributed photovoltaic super-capacity intelligent detection method based on drones according to claim 5, it is characterized in that the acquisition of the removal path of all photovoltaic panels is specifically:

[0033] Step S411, starting with number 1, sets the excess capacity of the photovoltaic panel numbered m to a m ;

[0034] Step S412, find the adjacent photovoltaic panel bn , and calculate the total capacity of the photovoltaic panels under all current paths and place the photovoltaic panel capacity value in the array U n middle;

[0035] Step S413, sequentially transform the array U n The sum of the current capacity values ​​of each path in Q n With the total excess capacity Q c Compare, capacity value Q n Less than the total excess capacity Q c Jump to the next step when the excess capacity value Q n Greater than or equal to the total excess capacity Q c Export the current path when

[0036] Step S414, search for all photovoltaic panels around the photovoltaic panel in the above step, and calculate the total capacity Q of the photovoltaic panels under all current paths n And place the array U n middle;

[0037] Step S415, repeat the above steps until the total capacity of all paths is Q n Are greater than or equal to the total excess capacity Q c , then all the paths starting with number 1 are obtained.

[0038] Preferably, the calculation of the impact factor is specifically performed by

[0039] E=Φ(β0+β1x1+…+β n x n )

[0040] The model is constructed, where E is the probability of the photovoltaic panel being removed, Φ is the probability density function of the standard normal distribution, β n is the Probit regression coefficient under the nth independent variable, x n is the nth independent variable, which is a parameter that affects the use of photovoltaic panels, including service life, photovoltaic brand, corresponding brand usage years, past maintenance status, and past power generation efficiency; the β n The specific value is calculated by uploading the data information of previous inspections to the Probit model.

[0041] Preferably, the specific calculation method for calculating the total score of each demolition path is:

[0042] G=ξ(E1Z1+E2η1Z2…+E n η (n-1) Z n ),

[0043] Where G is the total score of the path, E n is the probability of photovoltaic panel removal, η(n-1) is the position parameter, specifically the position parameter for removing the adjacent photovoltaic panels centered on the photovoltaic panel at step n-1, Z n is the virtual parameter for removing the photovoltaic panel, ξ is the overcapacity factor, and the specific calculation method is: Among them, U max is the maximum capacity of the photovoltaic panel under the current path, Q n is the total capacity of the current path, Q c is the total excess capacity;

[0044] Calculate the total demolition score of all paths, compare them and export the path with the highest total demolition score.

[0045] Preferably, the generation of a specific demolition plan is specifically to generate a specific demolition plan based on the optimal demolition path calculated by the ground receiving terminal, and send the specific demolition plan to the relevant personnel port through the cloud management system.

[0046] A distributed photovoltaic super-capacity intelligent detection method system based on drone, comprising:

[0047] Information collection unit: used to collect image information of the photovoltaic panels under inspection using the camera carried by the drone, and pre-process the captured video information to obtain image information to be analyzed;

[0048] The model training unit is used to construct a photovoltaic panel specific information recognition model and train the photovoltaic panel model recognition model using sample data in the sample library;

[0049] The over-capacity detection unit is used to calculate the over-capacity situation, calculate the removal method under the over-capacity situation, and generate a specific detection report.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1. High efficiency: UAVs can quickly cover a large area of ​​distributed photovoltaics and conduct efficient and continuous inspections, which greatly shortens the inspection time and improves work efficiency. This is more obvious for photovoltaic power generation households with a large number of photovoltaic panels.

[0052] 2. Higher accuracy: Drones are equipped with high-resolution cameras and other sensors that can accurately capture the details of photovoltaic panels, including component models, quantities, etc., as well as the layout and capacity information of photovoltaic arrays, so as to more accurately determine whether there is overcapacity.

[0053] 3. Good safety: Drones do not require manual climbing or working at heights, reducing the safety risks of workers. Drone inspections are more advantageous, especially in severe weather or terrain conditions.

[0054] 4. Data recording and analysis: UAV inspections can record and store image data in real time, which is convenient for later analysis and comparison, forming historical archives, and helping to track and evaluate the long-term operating status of photovoltaic power generators.

[0055] 5. Reduce labor costs: UAV inspections reduce a lot of manpower input and reduce labor costs. At the same time, if there is an overcapacity situation, it can automatically generate the best demolition plan according to the status of the photovoltaic panels, which is conducive to improving the standardization and intelligence level of photovoltaic inspections. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0057] Figure 1 This is a flow chart of a distributed photovoltaic super-capacitor intelligent detection method based on drones in the present invention.

[0058] Figure 2 The present invention discloses a method for calculating the optimal removal path in a distributed photovoltaic super-capacitor intelligent detection method based on a drone.

[0059] Figure 3 The present invention is a flowchart of an implementation method of a distributed photovoltaic super-capacity intelligent detection method based on a drone.

[0060] Figure 4 This is an information transmission relationship diagram of a distributed photovoltaic super-capacity intelligent detection method based on drones in the present invention. DETAILED DESCRIPTION

[0061] For example, certain words are used in the specification and claims to refer to specific components. Those skilled in the art should understand that hardware manufacturers may use different terms to refer to the same component. This specification and claims do not use differences in names as a way to distinguish components, but use differences in the functions of components as the criteria for distinction. For example, "including" mentioned throughout the specification and claims is an open term, so it should be interpreted as "including but not limited to". "Approximately" means that within an acceptable error range, those skilled in the art can solve the technical problem within a certain error range and basically achieve the technical effect.

[0062] The following is combined with Figure 1-4 The present invention further describes in detail a distributed photovoltaic super-capacity intelligent detection method and system based on a drone, but this is not intended to limit the present invention.

[0063] A distributed photovoltaic overcapacity intelligent detection method based on drone, specifically:

[0064] Step S1, obtaining specific information of photovoltaic panels through drone collection;

[0065] Step S2, the ground receiving equipment extracts information from the video information collected by the drone, and extracts and analyzes the photovoltaic panoramic image to obtain the specific photovoltaic panel component number and information;

[0066] Step S3, based on the specific photovoltaic panel information and the registered data information in the system, the ground receiving device calculates the virtual overcapacity and obtains the total overcapacity;

[0067] Step S4, the ground receiving device calculates and obtains all removal paths, and calculates the best removal path;

[0068] Step S5, generating a specific demolition plan based on the optimal demolition path and sending it to the relevant personnel port through the cloud management system.

[0069] Each step is described in detail below.

[0070] Get photovoltaic panel information. Photograph the photovoltaic panels using the camera device of the drone to take a panoramic view of the photovoltaic panels in the current area. Get the current drone's location using the drone's positioning module, such as a GPS module, to get the specific location of the photovoltaic panel currently being tested, and send the location information to the ground receiving end. The ground receiving end reads the registration information of the cloud management system and obtains the photovoltaic panel information in the registration information through keyword search technology. The photovoltaic panel information includes brand, model, service life, product power, operating voltage, operating current and other information.

[0071] Extract the video information and extract and analyze the photovoltaic panorama to obtain the specific photovoltaic panel number and information. The drone sends the image information to the ground receiving device through the communication module, and the ground receiving device processes the image information on the controller of the ground receiving device. The specific processing process is as follows:

[0072] First, the FasterRCNN network model is constructed to identify the photovoltaic panel information by extracting feature points, including the photovoltaic panel nameplate, the component structure inside the photovoltaic panel, etc.

[0073] The photovoltaic module recognition image datasets of various brands are set as training sets, and the FasterRCNN network model is trained through the training sets.

[0074] Secondly, the panoramic area of ​​the photovoltaic panel in the video information is captured using small target image enhancement technology. The photovoltaic panel area image is dedistorted using the remap algorithm of OpenCV, and the feature points of the panoramic image are matched with the feature points of the photovoltaic panel plane map through the surf algorithm. The matched area is marked, and the image is segmented according to the matching situation to obtain the image information of a single photovoltaic module, and numbered according to the location to facilitate subsequent demolition calculations.

[0075] Finally, the trained network model is used to identify a single photovoltaic module, and the recognition results and photovoltaic panel image label information are output through TXT text content.

[0076] Obtain the specific brand information of the photovoltaic panels registered in the current area from the cloud management system for matching. If there is a mismatch, mark it and generate an early warning message. Obtain the capacity data information of each photovoltaic panel component based on the photovoltaic panel brand information identified by the model. Obtain the false label overcapacity and the total overcapacity based on the specific photovoltaic panel information and system data.

[0077] The drone is provided with a standard capacity photovoltaic panel, and the photovoltaic panel image captured in the image information is calculated to obtain the installation inclination angle of the ground photovoltaic panel. The current state of the drone is obtained through the IMU module of the drone, and the drone is controlled to make the inclination angle of the photovoltaic panel installed on the drone consistent with the inclination angle of the photovoltaic panel on the ground, and the power generation data of the photovoltaic panel of the standard capacity per unit time is obtained.

[0078] In a specific embodiment of adjusting the tilt angle of the photovoltaic panel on the aircraft, a flat photovoltaic panel is installed on the drone. The tilt angle of the photovoltaic panel is adjusted by adjusting the flight attitude of the drone, adjusting the rotor state of the drone, obtaining the current flight state of the drone through the IMU module, and adjusting it to be consistent with the tilt of the ground photovoltaic panel.

[0079] In the second embodiment of adjusting the tilt angle of the photovoltaic panel on the aircraft, the photovoltaic panel is arranged on the drone, and a height adjustment device such as a linear motor is arranged under the photovoltaic panel to adjust the angle of the photovoltaic panel. After the drone obtains the installation angle information of the photovoltaic panel on the ground, the drone maintains a hovering state and adjusts the angle of the photovoltaic panel to make the tilt angle of the photovoltaic panel on the aircraft consistent with that of the photovoltaic panel on the ground.

[0080] Read the on-site power generation of the power generation customers in the same period and space in the system to determine whether there is a false mark. If there is a false mark in the photovoltaic panel, the false mark parameter Z n Configure the value.

[0081] According to the identified specific photovoltaic panel parameter information, the total capacitance of the photovoltaic components and the total capacitance of the capacitor panels registered in the database are calculated, and the above two capacitances are calculated to obtain the specific value of the excess capacity.

[0082] Get all demolition paths and calculate the best demolition path. Specifically:

[0083] Step S41, obtaining the removal path of all photovoltaic panels.

[0084] Step S42, calculating the total score of all demolition paths.

[0085] Step S43, calculating and obtaining the best removal path.

[0086] Get all the photovoltaic panel removal paths. The specific steps are:

[0087] Step S411, starting with number 1, sets the excess capacity of the photovoltaic panel numbered m to a m .

[0088] Step S412, find the adjacent photovoltaic panel b n , and calculate the total capacity of the photovoltaic panels under all current paths and place the photovoltaic panel capacity value in the array U n middle.

[0089] Step S413, sequentially transform the array U n The sum of the current capacity values ​​of each path in Q n With the total excess capacity Q c Compare, capacity value Q n When it is less than the total excess capacity, jump to the next step. The excess capacity value Q n Greater than or equal to the total excess capacity Q c Export the current path.

[0090] Step S414, search for all photovoltaic panels around the photovoltaic panel in the above step, and calculate the total capacity Q of the photovoltaic panels under all current paths n And place the array U n middle.

[0091] Step S415, repeat the above steps until the total capacity of all paths is Q n Are greater than or equal to the total excess capacity Q c , then all the paths starting with number 1 are obtained.

[0092] Calculate the total score of all demolition paths.

[0093] G=ξ(E1Z1+E2η1Z2…+E n η (n-1) Z n ),

[0094] Where G is the total score of the path, E n is the probability of photovoltaic panel removal, η (n-1) is the position parameter, specifically the position parameter for the removal of the adjacent photovoltaic panels with the photovoltaic panel at the center in the n-1th step, with a value range of 0 to 1. The specific value is determined by the installation position of the photovoltaic panel. The photovoltaic panels at the periphery are convenient for removal, so the position parameter is set to a higher value, which decreases from the periphery to the inside. ξ is the overcapacity factor. The specific calculation method is: Among them, U max is the maximum capacity of the photovoltaic panel under the current path, Q nis the total capacity of the current path, Q c is the total excess capacity, Z n is the virtual marking parameter for removing the photovoltaic panel. If there is a virtual marking, a parameter value less than 1 is assigned, otherwise the initial value is 1. n The specific calculation method is:

[0095] The calculation of the impact factor is specifically through E = Φ (β0 + β1x1 + ... + β n x n ) to build the model, where E is the probability of choosing to remove the photovoltaic panel, β n is the Probit regression coefficient under the nth independent variable, x n is the nth independent variable, which is a parameter that affects the use of photovoltaic panels, including service life, photovoltaic brand, corresponding brand usage years, past maintenance status, and past power generation efficiency. n The specific value is obtained by uploading the data information of previous inspections to the Probit model to calculate the specific regression coefficient.

[0096] The best demolition path is calculated and compared to obtain the path with the highest total demolition score for export. The relevant path information is uploaded to the system through the processor of the ground receiving device.

[0097] The optimal demolition path is used to generate a specific demolition plan, which includes relevant path information and corresponding photovoltaic panels. The specific demolition plan includes information such as the photovoltaic panels to be demolished and the specific locations of the photovoltaic panels, and the demolition plan is sent to the relevant person in charge through the cloud management system.

[0098] A distributed photovoltaic super-capacity intelligent detection method system based on drone, comprising:

[0099] The information collection unit is used to collect image information of the photovoltaic panel under inspection using the camera carried by the drone, and pre-process the captured video information to obtain image information to be analyzed.

[0100] The model training unit is used to build a photovoltaic panel specific information recognition model and use the sample data in the sample library to train the photovoltaic panel model recognition model.

[0101] The over-capacity detection unit is used to calculate the over-capacity situation, calculate the removal method under the over-capacity situation, and generate a specific detection report.

[0102] The specific implementation method of this embodiment is:

[0103] Auditors screen suspected power generation customers with overcapacity and generate marketing audit anomalies in the system. After the cloud management system receives the marketing audit, the drone receiving signal of the nearby base station is activated. The drone goes to the target address through GPS positioning, and adjusts the flight status after reaching the target location. Take a panoramic picture of the photovoltaic panel at the current location, send the captured image information to the ground receiving device through the communication module, and identify the photovoltaic panel through the controller to obtain the installation angle of the photovoltaic panel and the model and parameters of the photovoltaic panel.

[0104] The calculated installation angle of the ground photovoltaic panel is sent to the drone through the communication module. The drone makes the airborne photovoltaic panel reach the same tilt angle, hovers, and obtains the power generation of the photovoltaic panel per unit capacity. The data information is sent to the ground receiving end through the drone communication module to control the value calculation to see if there is overcapacity and false labeling.

[0105] If there is overcapacity and false bid, the best demolition path will be calculated, and a demolition plan will be generated and uploaded to the cloud management system, which will then send the demolition plan to the relevant staff port. If there is no overcapacity and false bid, an inspection result report will be generated and uploaded to the cloud management system for record keeping.

[0106] The embodiments of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of ordinary technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A distributed photovoltaic super-capacity intelligent detection method based on drones, characterized in that: include: Step S1, obtaining specific information of photovoltaic panels through drone collection; Step S2, the ground receiving equipment extracts information from the video information collected by the drone, and extracts and analyzes the photovoltaic panoramic image to obtain the specific photovoltaic panel component number and information; Step S3, based on the specific photovoltaic panel information and the registered data information in the system, the ground receiving device calculates the virtual label situation and obtains the total excess capacity; Step S4, the ground receiving device calculates and obtains all removal paths, and calculates the best removal path; Step S5, generating a specific demolition plan based on the optimal demolition path and sending it to the relevant personnel port through the cloud management system.

2. According to the UAV-based distributed photovoltaic super-capacitor intelligent detection method of claim 1, it is characterized in that: The method of obtaining the photovoltaic panel information specifically includes installing a camera and a positioning device on the drone, collecting panoramic video information of the photovoltaic panels through the camera of the drone, obtaining the location information of the photovoltaic panels through the positioning device of the drone, and querying the photovoltaic panel information registered in the cloud management system through the location information.

3. The distributed photovoltaic super-capacity intelligent detection method based on drone according to claim 1 is characterized in that: The specific steps of extracting information from the video information and extracting and analyzing the photovoltaic panoramic image to obtain the specific photovoltaic panel number and information are as follows: Step S21, constructing a FasterRCNN network model to identify photovoltaic panel information by extracting feature points, setting the photovoltaic module recognition image data set of each brand as a training set, and training the FasterRCNN network model through the training set; Step S22, extracting information from the video information to obtain photovoltaic panel information, specifically: Step S221, using a small target image enhancement technology to capture a panoramic area of ​​the photovoltaic panel in the video information; Step S222, using the remap algorithm of OpenCV to perform dedistortion processing on the photovoltaic panel area image; Step S223, matching the feature points of the panoramic image with the feature points of the photovoltaic panel plan view by using the surf algorithm, marking the matching area, performing image segmentation according to the matching situation, and obtaining image information of a single photovoltaic module; Step S224, numbering according to the position; Step S225, using the trained network model to identify a single photovoltaic module, and outputting the identification result and photovoltaic panel image label information via TXT text content; Step S226, obtaining specific brand information of photovoltaic panels registered in the current area from the cloud management system for matching, and obtaining the capacity data information of each photovoltaic panel component according to the photovoltaic panel brand information identified by the model.

4. The distributed photovoltaic super-capacity intelligent detection method based on drone according to claim 1 is characterized in that: The method of obtaining the virtual overcapacity and the total overcapacity is as follows: The UAV is provided with a standard capacity photovoltaic panel. The tilt angle of the ground photovoltaic panel is obtained by processing and analyzing the video image information. The angle of the photovoltaic panel carried by the UAV is adjusted to make the tilt angle of the UAV photovoltaic panel the same as the tilt angle of the ground photovoltaic panel. The power generation data per unit time of the standard capacity photovoltaic panel is obtained. According to the photovoltaic panel information in the image information, the false marking situation is obtained. If there is a false marking situation, the false marking coefficient Z is configured. n ; Calculate the total capacitance of the photovoltaic modules and the total capacitance of the capacitor plates registered in the database, and compare the two capacitances to obtain the total excess capacitance.

5. The distributed photovoltaic super-capacity intelligent detection method based on drone according to claim 1 is characterized in that: The steps of obtaining all removal paths and calculating the best removal path are as follows: Step S41, obtaining the removal path of all photovoltaic panels; Step S42, calculating the impact factor and the total score of each demolition path; Step S43, calculating and obtaining the best removal path.

6. The method for intelligent detection of distributed photovoltaic supercapacitors based on drones according to claim 5 is characterized in that: The specific path for obtaining the removal of all photovoltaic panels is as follows: Step S411, starting with number 1, sets the excess capacity of the photovoltaic panel numbered m to a m ; Step S412, find the adjacent photovoltaic panel b n , and calculate the total capacity of the photovoltaic panels under all current paths and place the photovoltaic panel capacity value in the array U n middle; Step S413, sequentially n The sum of the current capacity values ​​of each path in Q n With the total excess capacity Q c Compare, capacity value Q n Less than the total excess capacity Q c Jump to the next step when the excess capacity value Q n Greater than or equal to the total excess capacity Q c Export the current path when Step S414, search for all photovoltaic panels around the photovoltaic panel in the above step, and calculate the total capacity Q of the photovoltaic panels under all current paths n And place the array U n middle; Step S415, repeat the above steps until the total capacity of all paths is Q n Are greater than or equal to the total excess capacity Q c , then all the paths starting with number 1 are obtained.

7. The method for intelligent detection of distributed photovoltaic supercapacitors based on drones according to claim 5 is characterized in that: The calculation of the impact factor is specifically performed by E=Φ(β0+β1x1+…+β n x n ) The model is constructed, where E is the probability of the photovoltaic panel being removed, Φ is the probability density function of the standard normal distribution, β n is the Probit regression coefficient under the nth independent variable, x n is the nth independent variable, which is a parameter that affects the use of photovoltaic panels, including service life, photovoltaic brand, corresponding brand usage years, past maintenance status, and past power generation efficiency; the β n The specific value is calculated by uploading the data information of previous inspections to the Probit model.

8. The method for intelligent detection of distributed photovoltaic supercapacitors based on drones according to claim 5 is characterized in that: The specific calculation method for calculating the total score of each removal path is: G=ξ(E1Z1+E2η1Z2…+E n or (n-1) Z n ), Where G is the total score of the path, E n is the probability of photovoltaic panel removal, η (n-1) is the position parameter, specifically the position parameter for removing the adjacent photovoltaic panels centered on the photovoltaic panel at step n-1, Z n is the virtual parameter for removing the photovoltaic panel, ξ is the overcapacity factor, and the specific calculation method is: Among them, U max is the maximum capacity of the photovoltaic panel under the current path, Q n is the total capacity of the current path, Q c is the total excess capacity; Calculate the total demolition score of all paths, compare them and export the path with the highest total demolition score.

9. The distributed photovoltaic super-capacity intelligent detection method based on drone according to claim 1 is characterized in that: The specific demolition plan is generated by calculating the best demolition path from the ground receiving end to generate a specific demolition plan, and sending the specific demolition plan to the relevant personnel port through the cloud management system.

10. A distributed photovoltaic super-capacity intelligent detection system based on drones, characterized in that: include: Information collection unit: used to collect image information of the photovoltaic panels under inspection using the camera carried by the drone, and pre-process the captured video information to obtain image information to be analyzed; Model training unit: used to build a photovoltaic panel specific information recognition model, and use the sample data in the sample library to train the photovoltaic panel model recognition model; Overcapacity detection unit: used to calculate the overcapacity situation, calculate the removal method under the overcapacity situation, and generate a specific detection report.

Images