Micro-grid distributed new energy data processing method and system

Abstract

The invention provides a micro-grid distributed new energy data processing method and system, and the system comprises the steps: obtaining user distribution images and wind driven generator distribution images corresponding to a plurality of household photovoltaic ends in a target region, and abnormal photovoltaic data and fan data in a preset time period; the user distribution image comprises a plurality of distribution slots corresponding to the household photovoltaic ends and the fan ends. And for the photovoltaic data, determining a distribution slot position of a corresponding household photovoltaic end as an abnormal slot position based on the abnormal photovoltaic, and obtaining a distribution relationship of the abnormal slot position in the user distribution image. According to fan data, abnormal slots are sorted based on distribution slots in a fan distribution image, an inspection plan is generated, and a collection and detection process is optimized in combination with multi-source data. According to the method, multi-source data cooperative processing of the photovoltaic power and the fan is realized, the efficiency and precision of anomaly detection and fault processing of the micro-grid are effectively improved, and technical support is provided for intelligent management of distributed new energy.

Description

Technical Field

[0001] The present invention relates to data processing technologies in the field of new energy, and particularly to a method and system for processing distributed new energy data in a microgrid, which can comprehensively process data of photovoltaic power generation and wind power generation, and realize efficient monitoring and intelligent management of distributed photovoltaic terminals and wind turbine terminals in the microgrid through the acquisition, analysis, and anomaly detection of multi-modal data such as images and audio. Background Art

[0002] A microgrid is an important part of a modern energy management system, known as a distributed energy island system, which mainly consists of modules such as distributed power sources, energy storage devices, energy conversion devices, loads, monitoring, and protection devices. This small-scale power generation and distribution system can efficiently provide electrical and thermal energy services to users, with the characteristics of strong flexibility and high controllability. In a microgrid, residential photovoltaic and wind power generation are typical forms of distributed new energy. Among them, residential photovoltaic uses photovoltaic panels to convert solar energy into electrical energy, usually installed on the top floor or courtyard of a family residence, while wind power generation relies on wind turbines to convert wind energy into electrical energy.

[0003] In the prior art, in order to ensure the reliability and operation efficiency of a microgrid, the operating states of photovoltaic modules and wind turbines need to be monitored regularly. However, due to long-term exposure to the outdoors, photovoltaic panels are easily affected by dust, dirt, and environmental factors, resulting in a decrease in power generation efficiency or equipment failure. Usually, the operating state of photovoltaic panels is detected by using drones to take pictures of them. Although this method has achieved a certain degree of automation, there are still the following problems in practical applications:

[0004] ● Insufficient customization: The prior art cannot perform targeted data collection and photography according to the personalized needs of users, and it is difficult for users to directly obtain efficient observation results for abnormal points.

[0005] ● Positioning error: Drone photography often relies on preset positioning information. Due to the inconsistency of the shooting positions, the matching degree between image acquisition and anomaly positioning is low, affecting the user's intuitive judgment and processing of the abnormal position.

[0006] ● Lack of multi-source collaborative processing: For the data processing of photovoltaic modules and wind turbines, most technical means are separated, lacking a collaborative processing mechanism, resulting in low efficiency of anomaly detection and fault diagnosis. Especially for wind turbines, anomaly detection usually only stays at the level of visual image analysis, ignoring the potential value of multi-modal data such as audio, and it is difficult to identify some complex fault modes.

[0007] In addition, there are similar deficiencies in the condition monitoring of wind turbines. Traditional inspection methods rely on manual or fixed sensors to collect data, which is not only inefficient but also may miss abnormal conditions due to the single data collection. In addition, the lack of a dynamically optimized inspection plan also increases the difficulty of equipment fault handling.

[0008] With the expansion of the scale of distributed new energy and the complexity of the operating environment, how to integrate multi-source data of photovoltaic and wind power generation, achieve collaborative processing, and further improve the efficiency and accuracy of anomaly detection and fault handling has become a key issue that urgently needs to be solved in the field of microgrid intelligent management. Summary of the Invention

[0009] The embodiments of the present invention provide a method and system for processing distributed new energy data in a microgrid, which can comprehensively process multi-source data of household photovoltaic terminals and wind power generation terminals, achieve efficient anomaly detection and fault management, and at the same time improve the consistency of data collection and display, facilitating users to intuitively observe and locate anomalies.

[0010] In the first aspect of the implementation of the present invention, a method for processing distributed new energy data in a microgrid is provided, including:

[0011] Obtain the user distribution images corresponding to multiple household photovoltaic terminals and wind power generation terminals in the target area, as well as the abnormal photovoltaic data and fan data within a preset time period; the user distribution image includes distribution slots corresponding to multiple household photovoltaic terminals and fan terminals;

[0012] For the photovoltaic data, determine the distribution slots of the corresponding household photovoltaic terminals as abnormal slots based on the abnormal photovoltaic data, obtain the distribution relationship of the abnormal slots in the user distribution image, sort the abnormal slots based on the distribution relationship, and generate a collection sequence;

[0013] Receive the customized collection requirements for each abnormal slot, control the collection device to collect based on the collection sequence to obtain a collection image, and process the collection image in combination with the point calibration strategy to generate a control signal;

[0014] Control the collection device to perform image collection according to the control signal to generate an abnormal contour image corresponding to the abnormal slot; update the user distribution image based on the abnormal contour image to generate an abnormal distribution image and send it to the management end;

[0015] For the fan data, obtain multi-modal data including audio through drone inspection, process the audio data using large model technology, detect abnormal noise features, and identify potential faults; sort the abnormal slots based on the distribution slots in the fan distribution image to generate an inspection plan, and optimize the collection and detection processes in combination with multi-source data, and finally generate a fan anomaly diagnosis report.

[0016] Through the above steps, the present invention realizes the collaborative processing of photovoltaic and wind turbines, integrates the anomaly detection results into a unified new energy anomaly distribution image, dynamically updates and sends it to the management terminal for comprehensive monitoring and control, and further improves the efficiency and accuracy of microgrid anomaly detection and fault handling.

[0017] Optionally, in a possible implementation manner of the first aspect, the obtaining of the user distribution image corresponding to the target area and multiple household photovoltaic terminals, as well as the abnormal photovoltaic data and wind turbine data within a preset time period, includes:

[0018] Obtain the area image corresponding to the target area, as well as the arrangement information of multiple household photovoltaic terminals and wind turbine terminals, where the arrangement information includes row information and column information;

[0019] Based on the arrangement information, construct corresponding distribution slots in the area image to obtain the user distribution image;

[0020] Statistically calculate the target power generation of each household photovoltaic terminal in the target area within a preset time period to obtain the total power generation, and calculate the average power generation in combination with the total number of target photovoltaics in the target area; determine the target photovoltaics with a target power generation lower than the average power generation as abnormal photovoltaics;

[0021] Obtain the operation data of the wind turbines in the target area, including audio features, vibration features, and other operation indicators, identify abnormal wind turbines and mark the corresponding slots as abnormal slots.

[0022] In this way, the present invention further combines user needs, realizes customized collection, precise positioning, and multi-source data collaborative processing, and provides efficient technical support for the intelligent management of distributed new energy.

[0023] Optionally, in a possible implementation manner of the first aspect, the determining the distribution slot of the corresponding household photovoltaic terminal as an abnormal slot based on the abnormal photovoltaic data, obtaining the distribution relationship of the abnormal slot in the user distribution image, and generating a collection sequence by sorting the abnormal slots based on this distribution relationship includes the following steps:

[0024] Determine the distribution slot of the corresponding household photovoltaic terminal as an abnormal slot based on the abnormal photovoltaic data, and obtain the arrangement information of the abnormal slot in the user distribution image, where the arrangement information includes the abnormal row and the abnormal column;

[0025] Sort the abnormal slots in ascending order according to the abnormal column to obtain an abnormal sequence, and count the abnormal slots with the same abnormal column in the abnormal sequence to generate an abnormal column set sorted in sequence;

[0026] Sort the abnormal slots in the abnormal column set in ascending order according to the abnormal row to obtain an abnormal path sequence sorted in sequence, and sequentially count the abnormal path sequence to generate a collection path set;

[0027] Sort the abnormal slots based on the collection path set to obtain the final collection sequence.

[0028] Optionally, in a possible implementation of the first aspect, the specific method for sorting the abnormal slots based on the collection path set to generate a collection sequence includes:

[0029] Obtain the first abnormal path sequence in the collection path set as the initial collection sequence, and update the collection path set based on the collection sequence;

[0030] Obtain the arrangement information of the last abnormal slot in the current collection sequence as the first arrangement information;

[0031] Obtain the first abnormal path sequence in the collection path set as the first path sequence, and extract the arrangement information of its starting abnormal slot and the last abnormal slot as the second arrangement information and the third arrangement information respectively;

[0032] Calculate the first distance between the first arrangement information and the second arrangement information, and the second distance between the first arrangement information and the third arrangement information;

[0033] Select the abnormal slot corresponding to the minimum distance in the first path sequence as the selected slot, and determine the selection order according to the position of the selected slot in the first path sequence;

[0034] Select the abnormal slots from the first path sequence in turn based on the selection order and add them to the current collection sequence to obtain the updated collection sequence, and at the same time update the collection path set;

[0035] Repeat the above steps until the collection path set is an empty set, and finally complete the generation of the collection sequence.

[0036] Optionally, in a possible implementation of the first aspect, the specific method for determining the selection order based on the position of the selected slot includes:

[0037] Obtain the slot position of the selected slot in the first path sequence, where the slot position includes the starting slot position and the last slot position;

[0038] When the position of the selected slot is the starting slot position, generate a forward selection order, and select the abnormal slots in the first path sequence in turn and add them to the collection sequence;

[0039] When the position of the selected slot is the last slot position, generate a reverse selection order, and select the abnormal slots in the first path sequence in turn and add them to the collection sequence.

[0040] Optionally, in a possible implementation of the first aspect, the specific method for receiving the customized acquisition requirements of the abnormal slots and controlling the acquisition device to perform acquisition, process the acquired images, and generate control signals based on the acquisition sequence includes:

[0041] Receive the customized acquisition requirements of the abnormal slots, where the customized acquisition requirements include individual acquisition requirements and combined acquisition requirements;

[0042] Sequentially obtain the geographical location information of each abnormal slot in the acquisition sequence, and based on this information, control the acquisition device to perform image acquisition on the target abnormal slot to generate acquired images;

[0043] Extract the image center point of the acquired image, and determine the contour center point in the image according to the customized acquisition requirements;

[0044] Generate a control signal corresponding to the acquisition device according to the relative position relationship between the image center point and the contour center point, and optimize the acquisition process of the acquisition device.

[0045] Through the above improvements, the present invention can dynamically optimize the generation method of the acquisition sequence, effectively reduce the time cost of data acquisition, and improve the accuracy of customized acquisition in combination with user requirements, further enhancing the efficiency of anomaly detection and fault management.

[0046] Optionally, in a possible implementation of the first aspect, the method for obtaining the image center point of the acquired image and determining the contour center point in the acquired image according to the customized acquisition requirements includes the following steps:

[0047] Obtain the image center point of the acquired image, and construct a coordinate system with the image center point as the origin;

[0048] Determine the type of the customized acquisition requirements:

[0049] If it is an individual acquisition requirement, obtain the photovoltaic pixel values of the target photovoltaic, and determine the individual contour center point in the acquired image based on the photovoltaic pixel values;

[0050] If it is a combined acquisition requirement, obtain the photovoltaic pixel values of the target photovoltaic and the heat collection pixel values of the target collector, and determine the combined contour center point in the acquired image based on the photovoltaic pixel values and the heat collection pixel values;

[0051] Use the individual contour center point or the combined contour center point as the contour center point of the acquired image.

[0052] Optionally, the specific method for determining the individual contour center point in the acquired image based on the photovoltaic pixel values includes:

[0053] Extract the photovoltaic pixel points in the acquired image based on the photovoltaic pixel values, and count the coordinates of the photovoltaic pixel points to generate a photovoltaic coordinate set;

[0054] Calculate the first abscissa according to the maximum and minimum values of the abscissas in the photovoltaic coordinate set;

[0055] Calculate the first ordinate according to the maximum and minimum values of the ordinates in the photovoltaic coordinate set;

[0056] Determine the center point of the single contour in the acquired image according to the first abscissa and the first ordinate.

[0057] Optionally, the specific method for determining the center point of the combined contour in the acquired image based on the photovoltaic pixel value and the heat collection pixel value includes:

[0058] Extract the combined pixel points in the acquired image based on the photovoltaic pixel value and the heat collection pixel value, and count the coordinates of the combined pixel points to generate a combined coordinate set;

[0059] Calculate the second abscissa according to the maximum and minimum values of the abscissas in the combined coordinate set;

[0060] Calculate the second ordinate according to the maximum and minimum values of the ordinates in the combined coordinate set;

[0061] Determine the center point of the combined contour in the acquired image according to the second abscissa and the second ordinate.

[0062] Optionally, in a possible implementation manner of the first aspect, the specific method for generating a control signal corresponding to the acquisition device according to the positional relationship between the image center point and the contour center point includes:

[0063] Determine the horizontal movement direction and the horizontal movement distance according to the image abscissa of the image center point and the contour abscissa of the contour center point;

[0064] Determine the vertical movement direction and the vertical movement distance according to the image ordinate of the image center point and the contour ordinate of the contour center point;

[0065] Generate a control signal corresponding to the acquisition device based on the following parameters:

[0066] The horizontal movement direction and the horizontal movement distance;

[0067] The vertical movement direction and the vertical movement distance.

[0068] Through the above improvements, the present invention can accurately determine the center point of the target contour in the acquired image, optimize the positioning and adjustment process of the acquisition device, significantly improve the accuracy and efficiency of customized acquisition, and at the same time enhance the processing ability of multi-source data of photovoltaic and wind turbines, providing strong support for distributed new energy anomaly detection.

[0069] In the second aspect of the embodiments of the present invention, a microgrid distributed new energy data processing system is provided, which is used to comprehensively process multi-source data of photovoltaic and wind turbines, and realize anomaly detection, fault diagnosis and dynamic management. The system includes:

[0070] An acquisition module, configured to acquire user distribution images corresponding to multiple household photovoltaic terminals and wind power generation terminals in a target area, as well as abnormal photovoltaic data and wind turbine data within a preset duration; the user distribution image includes distribution slots corresponding to multiple household photovoltaic terminals and wind turbine terminals.

[0071] A sorting module, configured to determine the distribution slots of the corresponding household photovoltaic terminals as abnormal slots based on the abnormal photovoltaic data; obtain the distribution relationship of the abnormal slots in the user distribution image, and sort the abnormal slots based on this distribution relationship to generate an acquisition sequence; for the wind turbine data, combine the distribution slots in the wind turbine distribution image, use multi-modal data analysis to detect anomalies, and sort the abnormal slots to generate an inspection plan.

[0072] A customization module, configured to receive the customization acquisition requirements of each abnormal slot, where the requirements include individual acquisition requirements and combined acquisition requirements; control the acquisition device to perform data acquisition on the target abnormal slots of the photovoltaic terminal and the wind turbine terminal based on the acquisition sequence to obtain an acquisition image; process the acquisition image in combination with the point calibration strategy, obtain the image center point and the contour center point, and generate a control signal according to their position relationship.

[0073] A control module, configured to control the acquisition device to perform image acquisition or inspection operations according to the control signal, and obtain abnormal contour images or audio data corresponding to the abnormal slots; for the photovoltaic data, update the user distribution image based on the abnormal contour image to generate an abnormal distribution image; for the wind turbine data, generate a wind turbine anomaly diagnosis report in combination with audio processing technology; integrate the abnormal distribution images of the photovoltaic and wind turbines into a unified new energy abnormal distribution image, and send it to the management end for comprehensive monitoring and control.

[0074] Through the collaborative work of the above modules, the system of the present invention realizes the dynamic integration and collaborative processing of multi-source data of photovoltaic and wind turbines, improves the accuracy of anomaly detection and the efficiency of fault management, and provides comprehensive technical support for the intelligent management of distributed new energy.

[0075] The beneficial effects of the present invention are as follows:

[0076] 1. Optimize the acquisition path and improve the acquisition efficiency. Based on the distribution relationship between abnormal photovoltaic and fan slots, the present invention automatically generates an acquisition sequence, significantly shortening the acquisition path of the acquisition device and improving the data acquisition efficiency. At the same time, according to the customized acquisition requirements of users, control signals for the acquisition device are generated through a point calibration strategy to ensure that the acquired images are presented consistently, facilitating user observation and positioning. Through abnormal photovoltaic or fan data within a preset duration, an optimized acquisition sequence is automatically generated, reducing repeated acquisitions and path redundancy, and further improving the overall data acquisition efficiency and system response speed.

[0077] 2. Intelligent sorting and path optimization. The present invention automatically generates an acquisition sequence through an intelligent sorting algorithm. First, based on the distribution relationship of abnormal slots (including abnormal rows and abnormal columns), the slots in the abnormal columns are sorted in ascending order, the abnormal slots in the same column are counted, and an abnormal column set is generated; then, based on the abnormal rows, the abnormal column set is sorted to generate an abnormal path sequence, and an acquisition path set is further constructed; during the path selection process, by comparing the arrangement information (including the first, second, and third arrangement information) of the current acquisition sequence and the first abnormal path sequence in the acquisition path set, the optimal selection slots are determined, and the acquisition sequence for forward or reverse selection is generated in sequence; finally, an acquisition sequence with the shortest path is formed, thereby realizing efficient and accurate acquisition path planning and further improving the data acquisition efficiency.

[0078] 3. Support customized acquisition and ensure data consistency. The present invention can perform customized acquisition of abnormal slots according to user needs and dynamically adjust the positioning of the acquisition device in combination with the acquired images: according to the user's customized requirements (single acquisition or combined acquisition), the acquisition device is controlled to acquire the target image, and the image center point and contour center point in the acquired image are extracted; based on the positional relationship between the image center point and the contour center point, a control signal is generated to adjust the position of the acquisition device to ensure the accuracy and consistency of the acquired image; the generated abnormal contour image is used to update the user distribution image, optimizing the data display effect, facilitating the administrator to intuitively view and locate the abnormal position, and completing maintenance and management in a timely manner.

[0079] 4. Collaborative processing of multi-source data and realization of intelligent management. The present invention integrates multi-source data of photovoltaic and fans, combines the distribution relationship of abnormal slots and the inspection requirements, dynamically generates a unified new energy abnormal distribution image, and sends it to the management end for comprehensive monitoring and control. Through automated acquisition path optimization, customized acquisition, and image processing technologies, not only the efficiency and accuracy of microgrid abnormal detection and fault diagnosis are improved, but also technical support for the intelligent management of distributed new energy is provided, with broad application value and promotion prospects. Description of the Drawings

[0080] Figure 1Flow chart of a method for processing distributed new energy data of a microgrid provided by the present invention;

[0081] Figure 2 Schematic structural diagram of a system for processing distributed new energy data of a microgrid provided by the present invention. Specific embodiments

[0082] The technical solutions of the present invention will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0083] The present invention provides a method for processing distributed new energy data of a microgrid, as Figure 1 shown, including steps S1 - S6:

[0084] (S1) Data acquisition

[0085] (S1.1) Obtain the regional image and arrangement information. The server obtains the regional image within the target area and obtains the arrangement information of the photovoltaic end and the wind turbine end, and constructs the basic data of their spatial distribution.

[0086] Photovoltaic part: Regional image: A white background image representing the target area (such as a community, street, etc.), used to identify the spatial distribution of photovoltaic and wind turbines. Arrangement information: The position of the photovoltaic end in the image, usually including row and column coordinates. For example, photovoltaic end A is located at row 1 and column 1.

[0087] Wind turbine part: Wind turbines are usually distributed more dispersedly, and the arrangement information can be marked according to geographical location information. For example, wind turbine X is located at (longitude 120.5, latitude 30.7). In addition, physical parameters of the wind turbine, such as height, installation direction, etc., can be obtained to assist subsequent abnormal data analysis.

[0088] For example, in the target area, the photovoltaic ends are distributed in 3 rows and 4 columns: Photovoltaic end A: row 1 and column 1, Photovoltaic end B: row 1 and column 2. There are 3 wind turbine ends: Wind turbine X: longitude 120.5, latitude 30.7; Wind turbine Y: longitude 121.0, latitude 30.9

[0089] (S1.2) Construct distribution slots. According to the arrangement information of the photovoltaic and wind turbines, corresponding distribution slots are constructed in the regional image to form a user distribution image. Photovoltaic distribution slots: According to the arrangement of buildings, distribution slots similar to a matrix are constructed, and each slot corresponds to a photovoltaic end one by one. For example, a 3 - row and 4 - column slot can be used to represent the photovoltaic distribution in a community. Wind turbine distribution slots: Mark the positions of the wind turbines according to longitude and latitude coordinates or actual geographical distribution. This process can generate an accurate distribution map in combination with a GIS map.

[0090] (S1.3) Statistically calculate the power generation and operation data. Statistically calculate the power generation or operation data of photovoltaic and wind turbines, and calculate the total data and quantity of the region.

[0091] Photovoltaic part: The target power generation is to statistically calculate the actual power generation of the photovoltaic side within a preset time period, and obtain weekly or daily data. Aggregate the total power generation and total number of the photovoltaic side within the target area. For example, photovoltaic power generation: A: 90 kWh, B: 110 kWh, total 200 kWh, total quantity 2.

[0092] Wind turbine part: Obtain operation data: including vibration characteristics, audio features (such as noise), rotational speed and other data, perform multi-modal data integration, and combine the wind turbine operation data with its geographical distribution information to provide support for subsequent analysis. For example: Wind turbine operation data: Wind turbine X: vibration amplitude 0.5 mm, noise frequency 100 Hz. Wind turbine Y: vibration amplitude 0.8 mm, noise frequency 110 Hz.

[0093] (S1.4) Determine abnormal photovoltaics and abnormal wind turbines. Analyze the statistical data of photovoltaics and wind turbines to determine abnormal points. For the photovoltaic part, calculate the average power generation of the photovoltaics (total power generation ÷ total quantity), and the photovoltaic side with power generation lower than the average value is marked as abnormal photovoltaic, for example, it may be caused by reasons such as dust and shadow resulting in insufficient power generation. For example: Total power generation: 200 kWh, total quantity: 2, average power generation: 100 kWh. Photovoltaic A: 90 kWh (abnormal), Photovoltaic B: 110 kWh (normal). For the wind turbine part, detect the abnormal operation characteristics of the wind turbine, such as abnormal vibration or noise. Combine multi-modal data (such as audio and vibration characteristics) to identify potential mechanical failures, and mark the wind turbine as an abnormal wind turbine. For example: Wind turbine X: vibration amplitude and noise frequency are within the normal range (normal). Wind turbine Y: vibration amplitude is on the high side (abnormal).

[0094] Step S2: Determine and sort abnormal slots

[0095] (S2.1) Determine and sort the abnormal photovoltaic slots. For photovoltaic data, determine the position of the abnormal slots based on the abnormal power generation and sort them. According to the abnormal photovoltaics marked in step S1, determine their positions (slots) in the user distribution image. The row and column information of each abnormal photovoltaic slot is clearly marked in the distribution map. Analyze the arrangement information of the abnormal photovoltaic slots in the distribution map, and extract the row and column coordinates of the abnormal slots (such as "abnormal row" and "abnormal column"). The slot sorting rules are as follows:

[0096] ● Sort in ascending order according to the abnormal column coordinates.

[0097] ● If the abnormal columns are the same, then sort in ascending order according to the abnormal row coordinates.

[0098] ● Generate an abnormal path sequence sorted in turn for optimizing the acquisition path.

[0099] For example, assume the abnormal photovoltaic slots are as follows: Slot A: row 1, column 1; Slot C: row 2, column 3; Slot D: row 1, column 4.

[0100] The sorting process is as follows: 1. Sort by column: [A(1,1), D(1,4), C(2,3)]; 2. Sort the slots in the same column by row to generate the final path: [A(1,1), C(2,3), D(1,4)].

[0101] (S2.2) Determination and sorting of abnormal fan slots. For fan data, use abnormal characteristics (such as vibration or noise) to determine the positions of abnormal slots and generate an inspection plan. The abnormal slots are determined according to the marking information in the fan distribution image. Mark the detected abnormal fan ends as abnormal slots and obtain their distribution coordinates (such as longitude and latitude). The slot sorting rules are as follows:

[0102] ● Calculate the inspection path between abnormal fans based on geographical distance.

[0103] ● Optimize the path order to reduce the inspection time and travel of inspection equipment (such as drones).

[0104] For example, assume the abnormal fans are: Fan X: longitude 120.5, latitude 30.7; Fan Y: longitude 121.0, latitude 30.9

[0105] The inspection path is optimized as follows: 1. Calculate the geographical distance between Fan X and Fan Y. 2. Plan the shortest inspection path, for example: [Fan X → Fan Y].

[0106] (S2.3) Generate the acquisition sequence and inspection plan. Based on the sorting results of the abnormal slots of photovoltaic and fans, generate the acquisition sequence and inspection plan. Among them, for the acquisition sequence of photovoltaic, generate a set of acquisition paths (including multiple acquisition path sequences) according to the sorting results, and at the same time optimize the image acquisition efficiency by dynamically adjusting the acquisition order. The inspection plan for fans should optimize the inspection process by combining multi-modal data (such as audio features), and key areas and fault features of the inspection should be marked in advance to improve the inspection efficiency.

[0107] For example: Photovoltaic acquisition sequence: Acquire images in the order of [A → C → D]. Fan inspection plan: First check Fan Y with abnormal vibration amplitude, and then check Fan X with abnormal noise.

[0108] Step S3: Customize data acquisition

[0109] (S3.1) Receive customized acquisition requirements. For abnormal slots, the server receives the customized acquisition requirements specified by the user or the system, and differentiates them into individual acquisition or combined acquisition. Some fixed-value requirements are to collect data for a single photovoltaic or wind turbine slot. For example, obtaining a detailed image of a certain photovoltaic module or vibration data of an abnormal point of a certain wind turbine is called individual acquisition. Some customized requirements need to simultaneously collect data for multiple abnormal slots or regions. For example, simultaneously obtaining image and audio data for a wind turbine module, or simultaneously collecting comprehensive data for multiple photovoltaic slots.

[0110] For example, the user specifies to individually collect data for photovoltaic slot A (row 1, column 1). Or, the system issues an instruction to collect images and audio for wind turbine X as a combined acquisition task.

[0111] (S3.2) Collect images and data. Control the acquisition device (such as a drone or a fixed camera) to complete the data acquisition of abnormal slots, including images, audio, and other operating data.

[0112] For the photovoltaic side, image acquisition requires controlling the device to move to the target slot and obtaining high-resolution images of the photovoltaic modules according to the acquisition requirements. Focus on shooting abnormal points such as dust and occlusion. The acquisition process needs to be optimized. Based on the acquisition sequence, adjust the position and angle of the acquisition device to ensure image clarity and coverage.

[0113] For the wind turbine side, use drone patrol to obtain external images of the wind turbine and simultaneously record audio data during operation. Take close-up shots of abnormal points, such as detecting blade cracks or bearing damage. Achieve multi-modal data synchronization, synchronously collect images, audio, and vibration data, and ensure the temporal and spatial consistency of the data.

[0114] For example, the drone collects high-resolution images of photovoltaic slot A to detect whether there is dust or damage on the photovoltaic panel. Obtain the appearance image of wind turbine X, record the audio, and collect vibration data, etc.

[0115] (S3.3) Point calibration and image processing. Perform point calibration and processing on the collected images to optimize the data quality and generate control signals.

[0116] First, it is necessary to extract the center point (image geometric center) of the collected image and construct a coordinate system. For the center point of the contour, it needs to be located according to the requirements. If it is individual acquisition, extract the photovoltaic pixel values of the target photovoltaic module to determine the center point of the photovoltaic module contour; if it is combined acquisition, extract the photovoltaic pixel values and the pixel values of the collector, and comprehensively calculate the combined contour center point.

[0117] Then, according to the relative positions of the image center point and the contour center point, control signals for the acquisition device are generated to optimize the position and angle of the device. For example: for the photovoltaic slot A, extract the photovoltaic pixel points in the acquired image, determine the contour center point, and use it to optimize the shooting angle of the device. Compare the image center point of the fan X with the position of the blade anomaly point to generate a device adjustment signal.

[0118] (S3.4) Data calibration and optimization. According to the point calibration results, adjust the acquisition path and optimize the acquisition sequence to ensure an efficient acquisition process. Dynamically adjust the acquisition sequence according to the actual acquisition situation (such as blurred images or device offsets). At the same time, it is necessary to optimize the acquisition path, adjust the acquisition order in the photovoltaic module to reduce the device movement distance; optimize the UAV inspection path during fan inspection to reduce repeated flights.

[0119] For example, if the acquired image of the photovoltaic slot A is blurred, the system will prioritize re-acquiring it. If it is found that the anomaly of the fan Y is serious after the UAV inspects the fan X, the acquisition of the fan Y will be prioritized.

[0120] Step S4: Image / Audio Processing and Analysis

[0121] (S4.1) Image processing and analysis. Process the image data acquired from the photovoltaic and the fan, extract the key feature points, and generate optimized control signals.

[0122] For photovoltaic images, first, it is necessary to extract the image center point from the acquired photovoltaic image, construct a two-dimensional coordinate system, and use the image center point as the reference point for adjusting the device positioning. Secondly, it is necessary to determine the contour center point. For a single photovoltaic slot, extract the photovoltaic pixel values in the image, analyze the regional boundaries of the photovoltaic module, and calculate the contour center point coordinates based on the maximum and minimum values of the boundaries; for a combined photovoltaic slot, analyze the pixel values of the photovoltaic and other components (such as collectors) simultaneously to calculate the comprehensive contour center point. Finally, it is necessary to generate control signals. Based on the relative positions of the image center point and the contour center point, generate control signals for the acquisition device to optimize the device position or adjust the acquisition angle.

[0123] For fan images, first, it is necessary to perform appearance anomaly detection, analyze the external image of the fan, and extract the key areas (such as blades, nacelles); detect anomaly points such as blade cracks and wear, and mark their specific positions. It is also necessary to analyze the blade dynamics, perform dynamic analysis on multiple frames of images, observe the blade movement trajectory, and identify problems such as eccentricity and deformation.

[0124] For example, for the photovoltaic slot A, extract the image center point (50, 50) and the contour center point (45, 48), and generate a signal to adjust the device angle to ensure shooting clarity. For the fan X, analyze the edge crack of the blade in the image and mark the crack length and position.

[0125] (S4.2) Audio processing and analysis. For the audio data collected by the fan, use audio signal processing technology and large model analysis methods to identify potential faults.

[0126] Extract key features in the fan operation audio, such as frequency, amplitude, and vibration noise. Analyze low-frequency vibration and high-frequency abnormal noise to preliminarily judge the type of potential fault.

[0127] Use the trained large model to deeply analyze the audio features for abnormal pattern and detect abnormal noise features. Common fault patterns include bearing abnormalities, gear wear, or motor failures.

[0128] Combine multi-modal data, integrate the audio analysis results with other data such as vibration and temperature, and comprehensively identify the specific fault location and type of the fan.

[0129] For example, an abnormal noise of 120 Hz is found in the audio of Fan X, which matches the bearing wear pattern and is identified as a potential bearing fault. The abnormal amplitude of low-frequency vibration in the audio of Fan Y, combined with vibration data, confirms the problem of blade eccentricity.

[0130] (S4.3) Data integration and fault classification. Integrate the multi-modal data analysis results of photovoltaic and fan, classify abnormal information, and generate a diagnostic report.

[0131] According to the image analysis results, classify photovoltaic faults into occlusion, contamination, or equipment damage, and mark the specific location. For the fan, combine audio and image analysis, classify fan faults into mechanical, blade, or electrical problems, and determine their severity. Prioritize the abnormal information according to the scope of influence of the abnormality and the severity of the fault, providing a decision basis for subsequent processing.

[0132] For example, Photovoltaic Slot A is classified as a contamination problem, and it is recommended to clean the surface. Fan X is classified as a bearing fault, and it is recommended to arrange maintenance first.

[0133] Step S5: Abnormal image generation and update

[0134] (S5.1) Abnormal contour image generation. Generate an abnormal contour image for each slot according to the abnormal slot data of photovoltaic and fan, showing specific abnormal information.

[0135] For photovoltaic abnormal slots, combine image data to mark the contour of the abnormal area, such as the dust occlusion area or the damaged area, highlight the abnormal location, and overlay text to describe the type of abnormality (such as "occlusion" or "contamination"). At the same time, combine the abnormal contour image with the user distribution image to ensure that the position of the abnormal area in the distribution map is clearly visible.

[0136] Overlay the abnormal areas (such as blade cracks) in the fan image with the abnormal audio data (such as noise characteristics) to generate an abnormal contour image of the fan. Display the specific abnormal parts of the fan and related characteristic parameters (such as crack length or noise frequency). The fan also supports dynamic marking to label the positions and types of abnormal components (such as bearings or blades), providing support for maintenance decisions.

[0137] For example: The abnormal contour image of the photovoltaic slot A marks the occlusion area in the upper left corner of the component and is accompanied by a text description "Occlusion rate 30%". The abnormal contour image of the fan X marks the position of the blade crack and shows that the crack length is 5 cm.

[0138] (S5.2) Generate abnormal distribution images. Based on the data of all abnormal slots, generate an abnormal distribution image of the target area to display the overall abnormal situation.

[0139] Overlay the photovoltaic abnormal slots and their corresponding abnormal contours on the user distribution image. Each abnormal slot is marked with a color to indicate the type of abnormality (such as red for severe abnormality and yellow for mild abnormality).

[0140] Combine the fan abnormality with the position of the fan slot in the distribution map and overlay the fan abnormality information. Use different icons or colors to distinguish the types of abnormalities (such as bearing abnormalities or blade abnormalities).

[0141] Integrate the abnormal distributions of the photovoltaic and the fan into one distribution map to dynamically display all abnormal slots and their distributions in the target area.

[0142] For example: The photovoltaic slot A is marked yellow (occlusion abnormality), and the slot B is marked red (component damage). The fan X is marked red (bearing abnormality), and the fan Y is marked orange (noise abnormality).

[0143] (S5.3) Dynamically update the abnormal distribution image. According to the real-time data input, dynamically update the abnormal distribution image to ensure that the management end can monitor the abnormal situation in real time.

[0144] For real-time data, through the real-time input of the acquisition device and the inspection system, the abnormal distribution image generation module automatically updates the new abnormal slots and abnormal information.

[0145] If the abnormal state changes, if the abnormal slot has returned to the normal state (such as the occlusion disappears after cleaning), the abnormal mark of this slot is automatically removed from the distribution image.

[0146] All abnormalities are uniformly managed, and the updated abnormal distribution image is sent to the management end for comprehensive monitoring and control. The management end can directly view the latest abnormal situation and its change trend in the target area.

[0147] For example, after the cleaning of the photovoltaic slot C is completed, the abnormal state disappears, and the abnormal distribution image removes the abnormal mark of this slot. After the inspection of the fan Y, a new noise abnormality is found, and the distribution image automatically adds relevant marks.

[0148] Step S6: Comprehensive monitoring and management

[0149] (S6.1) Upload and display of abnormal distribution images. The uniformly generated abnormal distribution images of photovoltaic and wind turbines are transmitted to the management end digitally to display the abnormal conditions of the target area in real time. When transmitting data, it includes the location information, abnormal type, and specific parameters (such as abnormal level, fault type) of each abnormal slot. The management end visually displays the abnormal distribution in the form of a unified image, and different types of abnormalities are marked by colors and icons. For example, red indicates a serious abnormality, and yellow indicates a minor abnormality.

[0150] For example, in the management interface, the photovoltaic slot A (with shading abnormality) is marked yellow, and the fan X (with bearing abnormality) is marked red. Managers can clearly view the distribution of abnormal points and their severity levels throughout the area.

[0151] (S6.2) Comprehensive analysis and classification of abnormal information. At the management end, classify and analyze by combining the abnormal distribution image data to assist managers in making decisions. The abnormal information is classified according to types (such as shading, blade crack) and severity levels, and a statistical report is generated, including the number of abnormalities, distribution areas, and change trends. Comprehensively analyze historical and real-time data, explore the causes, development trends, and influence ranges of abnormalities, and support multi-dimensional displays. For example, display the time change trend of the number of abnormalities or the distribution of regional abnormal hotspots in the form of a curve graph.

[0152] For example, the management end analyzes that there are many shading abnormalities in the photovoltaic of Community A, which may be related to the shading of nearby trees, and the bearing abnormality of fan X is developing rapidly and requires priority maintenance.

[0153] (S6.3) Fault handling and early warning. According to the abnormal distribution image and analysis results, the management end formulates a fault handling plan and issues an early warning notice. The system sorts by the severity and influence of the abnormalities, generates a priority handling list, and clarifies the plans for photovoltaic cleaning, component replacement, or fan maintenance, and assigns them to the maintenance team. At the same time, for serious abnormalities (such as excessive fan noise or damaged photovoltaic components), the system automatically sends a real-time early warning to relevant personnel, and the notice details the specific location, type, severity level, and recommended measures of the abnormality.

[0154] For example, prioritize cleaning the shading abnormality of photovoltaic slot A and arrange for the bearing inspection of fan X; if a new high-frequency noise abnormality is found in fan Y, the system automatically notifies the maintenance team to arrange an inspection.

[0155] (S6.4) Dynamic monitoring and adaptive optimization. The management terminal monitors the abnormal conditions of photovoltaic and wind turbines in real time, and dynamically adjusts the acquisition and inspection plans in combination with real-time data. The interface of the management terminal dynamically updates the abnormal distribution image, displays the changes of abnormalities in real time, and supports remote scheduling of acquisition devices and inspection tasks. The system can adaptively adjust the acquisition sequence and inspection frequency according to the changes of abnormalities. For example, it increases the number of acquisitions for high-frequency abnormal areas, while reducing the inspection frequency for areas that have been normal for a long time.

[0156] For example, when the management terminal detects an increase in the number of abnormal wind turbines in a certain area, it automatically adjusts the inspection plan to increase the inspection frequency; if there is no abnormality in photovoltaic slot B for several consecutive days, the acquisition frequency is reduced to save resources.

[0157] To better implement a method for processing distributed new energy data in a microgrid provided by the present invention, the present invention also provides a system for processing distributed new energy data in a microgrid, as Figure 2 shown, including:

[0158] An acquisition module, configured to acquire a user distribution image corresponding to a target area and multiple household photovoltaic terminals, and abnormal photovoltaics within a preset time period, where the user distribution image includes distribution slots corresponding to multiple household photovoltaic terminals;

[0159] A sorting module, configured to determine the distribution slots corresponding to the household photovoltaic terminals as abnormal slots based on the abnormal photovoltaics, acquire the distribution relationship of the abnormal slots in the user distribution image, and sort the abnormal slots based on the distribution relationship to obtain an acquisition sequence;

[0160] A customization module, configured to receive the customized acquisition requirements of each of the abnormal slots, control an acquisition device to perform acquisition based on the acquisition sequence to obtain an acquisition image, and process the acquisition image according to the customized acquisition requirements and a point calibration strategy to obtain a control signal;

[0161] A control module, configured to control the acquisition device to perform image acquisition according to the control signal, obtain an abnormal contour image corresponding to the abnormal slot, update the user distribution image based on the abnormal contour image, and send the abnormal distribution image to the management terminal.

[0162] The present invention also provides a readable storage medium, in which a computer program is stored, and when the computer program is executed by a processor, it is used to implement the methods provided by the above various embodiments.

[0163] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A microgrid distributed new energy data processing method, characterized in that: include: Obtaining a user distribution image corresponding to a target area and a plurality of household photovoltaic terminals and wind power generation terminals, as well as abnormal photovoltaic data and wind turbine data within a preset time period, wherein the user distribution image includes distribution slots corresponding to a plurality of household photovoltaic terminals and wind turbine terminals; Determine the distribution slots corresponding to the household photovoltaic terminal as photovoltaic abnormal slots based on the abnormal photovoltaic data, obtain the distribution relationship of the photovoltaic abnormal slots in the user distribution image, and sort the photovoltaic abnormal slots based on the distribution relationship to obtain a photovoltaic acquisition sequence; The audio data of the wind turbine end is obtained by using a drone, audio anomalies are detected and wind turbine failures are identified based on a large model technology, and the distribution slot corresponding to the wind turbine end is determined as the abnormal slot of the wind turbine; Sort the abnormal slots of the wind turbines based on the distribution relationship of the abnormal slots of the wind turbines in the user distribution image, and generate a wind turbine inspection plan; Receive customized collection requirements of each abnormal slot, control the collection equipment to collect image and audio data based on the collection sequence and inspection plan, and process the collected images in combination with the point calibration strategy to obtain control signals for photovoltaic and wind turbines; According to the control signal, the acquisition equipment is controlled to generate photovoltaic abnormal contour image and wind turbine fault diagnosis report respectively; based on the photovoltaic abnormal contour image and wind turbine fault diagnosis result, the user distribution image is updated, and a comprehensive abnormal distribution image is generated and sent to the management end.

2. The method according to claim 1, characterized in that The step of obtaining a user distribution image corresponding to a target area and a plurality of household photovoltaic terminals and wind turbine terminals includes: Acquire a regional image of the target area and arrangement information of a plurality of household photovoltaic terminals and wind turbine terminals, wherein the arrangement information includes row information and column information; Distribution slots are constructed in the region image based on the arrangement information to obtain a user distribution image. Counting the target power generation of each household photovoltaic terminal and wind turbine terminal in the target area within a preset time period to obtain the total power generation, and obtaining the number of the target photovoltaic and wind turbine terminals in the target area to obtain the total number; The average power generation is obtained according to the total power generation and the total number, and it is determined that the target power generation is less than the average power generation, and the corresponding target photovoltaic and wind turbine ends are regarded as abnormal devices.

3. The method according to claim 2, characterized in that The step of determining the distribution slot corresponding to the household device end as the abnormal slot based on the abnormal device, acquiring the distribution relationship of the abnormal slot in the user distribution image, and sorting the abnormal slot based on the distribution relationship to obtain a collection sequence includes: Determine the distribution slot corresponding to the household device end as the abnormal slot based on the abnormal device, and obtain arrangement information of the abnormal slot in the user distribution image as abnormal arrangement information, wherein the abnormal arrangement information includes abnormal rows and abnormal columns; According to the abnormal columns, the abnormal slots are sorted in ascending order to obtain an abnormal sequence, and the abnormal slots of the same abnormal column in the abnormal sequence are counted to obtain a sequentially sorted abnormal column set; Based on the abnormal row, the abnormal slots in each abnormal column set are sorted in ascending order to obtain a sequence of abnormal paths that are sorted in sequence, and the abnormal path sequences are counted in sequence to obtain a collection path set; The abnormal slots are sorted based on the collection path set to obtain a collection sequence.

4. The method according to claim 3, characterized in that The step of sorting the abnormal slots based on the collection path set to obtain a collection sequence includes: Acquire the first abnormal path sequence in the acquisition path set as the acquisition sequence, and delete and update the acquisition path set based on the acquisition sequence to obtain a current acquisition path set; Acquire abnormal arrangement information of the last abnormal slot in the acquisition sequence as first arrangement information; Acquire the first abnormal path sequence in the collection path set as the first sequence, acquire the abnormal arrangement information of the starting abnormal slot in the first sequence as the second arrangement information, and acquire the abnormal arrangement information of the last abnormal slot as the third arrangement information; Obtain a first distance according to the first arrangement information and the second arrangement information, and obtain a second distance according to the first arrangement information and the third arrangement information; Determine in the first sequence an abnormal slot corresponding to a minimum value between the first distance and the second distance as a selection slot; Determine a selection order based on the slot position of the selected slot in the first sequence, select abnormal slots in the first sequence in turn based on the selection order and add them to the collection sequence to obtain a second sequence, and delete and update the collection path set based on the first sequence to obtain a current collection path set; The second sequence is used as the current acquisition sequence, and the above steps are repeated until the current acquisition path set is an empty set.

5. The method according to claim 4, characterized in that The determining of a selection order based on the slot positions of the selected slots, and sequentially selecting abnormal slots in the first sequence based on the selection order and adding them to a collection sequence to obtain a second sequence, comprises: Obtaining a slot position of the selected slot in the first sequence, wherein the slot position includes an end slot position and a start slot position; Determine the slot position of the selected slot as the starting slot position, generate a forward selection sequence, and sequentially select abnormal slots in the first sequence based on the forward selection sequence and add them to the collection sequence to obtain a second sequence; The slot position of the selected slot is determined to be the last slot position, a reverse selection order is generated, and the abnormal slots in the first sequence are selected in sequence based on the reverse selection order and added to the acquisition sequence to obtain a second sequence.

6. The method according to claim 1, characterized in that The receiving of the customized acquisition requirements of each abnormal slot, controlling the acquisition equipment to collect image and audio data based on the acquisition sequence and the inspection plan, and processing the collected data according to the customized acquisition requirements and the point calibration strategy to obtain a control signal include: Receive customized collection requirements for each abnormal photovoltaic and wind turbine slot, wherein the customized collection requirements include individual collection requirements and combined collection requirements; For the photovoltaic abnormal slot, the geographical location information of each slot in the acquisition sequence is acquired in sequence, and the acquisition device is controlled to perform image acquisition based on the geographical location information to obtain the acquired image; For the abnormal slot of the wind turbine, the geographical location information of each slot in the inspection plan is obtained in turn, and the drone is controlled to collect audio data based on the geographical location information to obtain the collected audio; Acquire the image center point of the acquired image, and determine the contour center point in the acquired image according to the customized acquisition requirement; Using large model technology to perform abnormality detection on the collected audio, and determining the abnormal location and type of the fan according to the detection result; A control signal corresponding to the acquisition device is generated according to the positional relationship between the center point of the image and the center point of the contour, and the fan abnormality detection result.

7. The method according to claim 6, characterized in that The acquiring the image center point of the acquired image and determining the contour center point in the acquired image according to the customized acquisition requirement includes: Acquire the image center point of the acquired image, and construct a coordinate system with the image center point as the origin; Determining the customized acquisition requirement as a separate acquisition requirement, and acquiring the photovoltaic pixel value of the target photovoltaic; Determine a separate contour center point in the corresponding acquired image based on the photovoltaic pixel value; Determining the customized acquisition requirement as a combined acquisition requirement, and acquiring the photovoltaic pixel value of the target photovoltaic and the thermal pixel value of the target thermal collector; Determine a center point of a combined contour in the corresponding acquired image based on the photovoltaic pixel value and the thermal collection pixel value; Taking the single contour center point or the combined contour center point as the contour center point of the corresponding collected image; For the fan audio data, extract the audio feature values ​​and perform anomaly detection based on the large model; Extract key audio feature points based on the detected abnormal audio features, and determine the location of abnormal components based on the fan distribution slot information; Based on the positional relationship between the center point of the image contour and the audio feature points, a comprehensive diagnostic result is generated to provide support for the joint control of photovoltaic images and wind turbine audio data.

8. The method according to claim 7, characterized in that The method of determining the positions of the individual contour center points and the audio abnormality feature points in the collected image based on the photovoltaic pixel values ​​and the wind turbine audio feature values ​​respectively includes: For photovoltaic image data: extract photovoltaic pixel points in the collected image based on the photovoltaic pixel value, and count the coordinates of the photovoltaic pixel points to obtain a photovoltaic coordinate set; determine a first horizontal coordinate according to the maximum and minimum values ​​of the horizontal coordinates in the photovoltaic coordinate set, and determine a first vertical coordinate according to the maximum and minimum values ​​of the vertical coordinates in the photovoltaic coordinate set; calculate and determine a separate contour center point in the collected image according to the first horizontal coordinate and the first vertical coordinate; For the combined data of photovoltaic and thermal collectors: extract the combined pixel points in the collected image based on the photovoltaic pixel value and the thermal collector pixel value, and count the coordinates of the combined pixel points to obtain a combined coordinate set; determine the second horizontal coordinate according to the maximum and minimum values ​​of the horizontal coordinate in the combined coordinate set, and determine the second vertical coordinate according to the maximum and minimum values ​​of the vertical coordinate in the combined coordinate set; calculate and determine the center point of the combined contour in the collected image according to the second horizontal coordinate and the second vertical coordinate; For the fan audio data: extract the spectral feature points in the audio data, and count the distribution of abnormal spectral feature points; determine the coordinates of the abnormal audio feature points according to the amplitude range and frequency range of the spectral feature points; map the abnormal audio feature points to the spatial coordinates based on the geometric information of the fan distribution slots. The individual contour center point, combined contour center point and audio abnormal feature point are used as key data, integrated and used for subsequent analysis and abnormal location.

9. The method according to claim 7, characterized in that: The step of generating a control signal corresponding to the acquisition device according to the positional relationship between the image center point, the contour center point, and the audio abnormality feature point comprises: For photovoltaic image data: according to the image horizontal coordinate of the center point of the image and the contour horizontal coordinate of the center point of the contour, determine the lateral movement direction and lateral movement distance of the photovoltaic collection device; according to the image vertical coordinate of the center point of the image and the contour vertical coordinate of the center point of the contour, determine the longitudinal movement direction and longitudinal movement distance of the photovoltaic collection device; based on the lateral movement direction, lateral movement distance, longitudinal movement direction and longitudinal movement distance, generate a control signal corresponding to the photovoltaic collection device; For wind turbine audio data: determine the flight path direction and distance of the drone based on the positional relationship between the audio abnormal feature points and the wind turbine distribution slots; adjust the drone's hovering position and collection angle based on the position change trend of the abnormal feature points; generate a control signal corresponding to the wind turbine collection device based on the flight path direction, flight distance and hovering position; Integrated control signal: Integrate the control signals of photovoltaic collection equipment and wind turbine collection equipment to achieve multi-device collaborative operation and ensure the synchronization and efficiency of image and audio collection.

10. A microgrid distributed new energy data processing system, characterized in that: include: Acquisition module: used to acquire the user distribution image corresponding to the target area and multiple household photovoltaic terminals and wind power generation terminals, as well as abnormal photovoltaic data and wind turbine audio data within a preset time period; the user distribution image includes the distribution slots corresponding to multiple household photovoltaic terminals and wind turbine terminals; Sorting module: used to determine the distribution slot corresponding to the household photovoltaic terminal as the photovoltaic abnormal slot based on the abnormal photovoltaic data, and obtain the distribution relationship of the photovoltaic abnormal slot in the user distribution image, and sort the photovoltaic abnormal slot based on the distribution relationship to generate a collection sequence; at the same time, detect abnormal features based on the fan audio data and determine the corresponding fan abnormal slot, and generate a fan inspection plan; Customization module: used to receive customized collection requirements of each abnormal photovoltaic and wind turbine slot, and control the collection equipment to collect image and audio data based on the collection sequence and inspection plan; Processing the collected data according to the customized collection requirements and point calibration strategy to obtain control signals for photovoltaic and wind turbines; Control module: used to control the acquisition equipment to perform photovoltaic image acquisition and fan audio acquisition according to the control signal; generate photovoltaic abnormal contour image and fan fault diagnosis report; update user distribution image based on the photovoltaic abnormal contour image and fan diagnosis result; Update module: used to integrate the updated photovoltaic and wind turbine abnormal distribution images into a comprehensive abnormal distribution image, and send it to the management end for comprehensive monitoring and fault handling.

Images