Distributed photovoltaic module power generation management method and system based on Internet of Things

Abstract

The invention discloses a distributed photovoltaic module power generation management method and system based on the Internet of Things. The method is applied to a control module of the distributed photovoltaic module power generation management system. Comprising the following steps: acquiring operation data of a plurality of photovoltaic modules acquired by a multi-mode sensor array; acquiring a fused satellite cloud picture corresponding to the plurality of photovoltaic modules and unmanned aerial vehicle aerial data; inputting the fused satellite cloud picture and unmanned aerial vehicle aerial data into a preset space-time attention prediction network, and outputting a future shadow distribution thermodynamic diagram; obtaining a node contribution degree corresponding to each photovoltaic module, and constructing a hybrid topological structure corresponding to the plurality of photovoltaic modules according to each node contribution degree; generating management information corresponding to each photovoltaic module according to the operation data corresponding to the plurality of photovoltaic modules, the future shadow distribution thermodynamic diagram and the hybrid topological structure; the management information at least comprises inclination angle adjustment information of the photovoltaic module; and distributed management of the plurality of photovoltaic modules is completed.

Description

Technical Field

[0001] The present application relates to the technical field of power network systems, and in particular to a distributed photovoltaic component power generation management method and system based on the Internet of Things. Background Art

[0002] With the continuous growth of global energy demand and the increasingly severe environmental problems, photovoltaic power generation has received widespread attention as a clean and renewable energy form. Distributed photovoltaic power generation systems are widely used in rooftops, building exterior walls and other scenarios due to their flexibility and adaptability. However, the existing distributed photovoltaic management methods still have many shortcomings in terms of operating efficiency, data collection and analysis, and system optimization. It is urgent to propose an innovative management method to improve system performance.

[0003] Traditional photovoltaic management methods mainly rely on a single data source (such as meteorological data or local sensor data) for power generation forecasting and system management. It is difficult to fully reflect the operating status of photovoltaic modules in complex environments, making it difficult to further improve the operating efficiency and reliability of photovoltaic power generation systems.

[0004] Therefore, there is an urgent need for a method to improve the operating efficiency and reliability of photovoltaic power generation systems. Summary of the invention

[0005] The present application provides a distributed photovoltaic module power generation management method and system based on the Internet of Things, which aims to solve the problem that traditional photovoltaic management methods mainly rely on a single data source (such as meteorological data or local sensor data) for power generation forecasting and system management, which is difficult to fully reflect the operating status of photovoltaic modules in complex environments, resulting in difficulty in further improving the operating efficiency and reliability of photovoltaic power generation systems.

[0006] In a first aspect, the present application provides a distributed photovoltaic assembly power generation management method, which is applied to a control module of a distributed photovoltaic assembly power generation management system; the method comprises: Acquiring operation data of a plurality of photovoltaic modules collected by a multimodal sensor array; the multimodal sensor array comprises a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager; Obtain fused satellite cloud images and drone aerial photography data corresponding to multiple photovoltaic modules; Input the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network, and output a future shadow distribution heat map; Acquire a node contribution corresponding to each of the photovoltaic components, and construct a hybrid topology structure corresponding to a plurality of the photovoltaic components according to each of the node contributions; Generate management information corresponding to each photovoltaic component according to operation data corresponding to the plurality of photovoltaic components, a future shadow distribution thermal map and a hybrid topology structure; The management information at least includes the tilt adjustment information of the photovoltaic module; and the distributed management of the plurality of photovoltaic modules is completed.

[0007] In some embodiments, the inputting of the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network includes: respectively obtaining the geographic coordinate systems corresponding to the fused satellite cloud image and the drone aerial photography data; spatially and temporally aligning the fused satellite cloud image and the drone aerial photography data according to the geographic coordinate system; adjusting the resolution of the fused satellite cloud image and the drone aerial photography data based on interpolation or downsampling technology; respectively extracting satellite feature information and aerial photography feature information from the adjusted fused satellite cloud image and drone aerial photography data; the satellite feature information includes at least cloud distribution and cloud thickness; the aerial photography feature information includes at least terrain features and obstruction features; and inputting the satellite feature information and aerial photography feature information into the spatiotemporal attention prediction network to output the future shadow distribution heat map.

[0008] Exemplarily, the spatiotemporal attention prediction network includes a convolution layer, a long short-term memory layer and an attention module; the inputting the satellite feature information and the aerial photography feature information into the spatiotemporal attention prediction network to output the future shadow distribution heat map includes: inputting the satellite feature information and the aerial photography feature information into the convolution layer and the long short-term memory layer respectively, the convolution layer outputs the spatial features corresponding to the satellite feature information and the aerial photography feature information, and the long short-term memory layer outputs the time feature sequence corresponding to the satellite feature information and the aerial photography feature information; the spatial features and the time feature sequence are input into the attention module respectively, and the attention module outputs the time attention information and the spatial attention information respectively; the time attention information is used to determine the key change moments in the time feature sequence; the spatial attention information is used to determine the occlusion area features corresponding to the spatial features; and the future shadow distribution heat map is output according to the occlusion area features and the key change moments.

[0009] It should be noted that, in some embodiments, outputting the future shadow distribution heat map according to the occlusion area characteristics and the key change moment includes: obtaining the shadow distribution influence weight corresponding to the occlusion area characteristics; obtaining the environmental change information corresponding to the key change moment; the environmental change information includes at least cloud movement and change in the sun's incidence angle; adjusting the shadow distribution influence weight according to the environmental change information; and generating the future shadow distribution heat map according to the adjusted shadow distribution influence weight.

[0010] In some embodiments, obtaining the node contribution corresponding to each of the photovoltaic components includes: obtaining operating data corresponding to each of the photovoltaic components; the operating data knowledge includes power output information, current information and voltage information; obtaining topological structure information corresponding to each of the photovoltaic components; the topological structure information is used to determine the connection relationship between each of the photovoltaic components and the remaining photovoltaic components; obtaining environmental data corresponding to each of the photovoltaic components; the environmental data includes at least light intensity, temperature and wind speed; and calculating the node contribution corresponding to each of the photovoltaic components based on the operating data, topological structure information and environmental data.

[0011] Exemplarily, the calculating of the node contribution corresponding to each of the photovoltaic components according to the operating data, topological structure information and environmental data includes: calculating the energy contribution corresponding to each of the photovoltaic components according to the operating data; calculating the reliability information corresponding to each of the photovoltaic components according to the operating data and environmental data; the reliability information is used to determine the stability of each photovoltaic component under abnormal conditions; calculating the network influence coefficient of each of the photovoltaic components according to the topological structure information and the operating data; the network influence coefficient is used to determine the influence of the location of each of the photovoltaic components on the communication and energy transmission of multiple photovoltaic components shown.

[0012] In some embodiments, constructing a hybrid topology structure corresponding to multiple photovoltaic components according to the contribution of each node includes: dividing the multiple photovoltaic components into any one of a perception layer, a transmission layer and a management layer according to the node contribution; wherein each photovoltaic component in the perception layer adopts a mesh topology, each photovoltaic component in the transmission layer adopts a tree topology, and the photovoltaic components in the management layer adopt a star topology; constructing the hybrid topology structure according to the perception layer, the transmission layer and the management layer.

[0013] Exemplarily, constructing the hybrid topology structure according to the perception layer, transmission layer and management layer includes: obtaining a preset topology optimization target; the topology optimization target includes at least a target energy transmission efficiency, a target communication delay and a target reliability; adjusting the connection relationship between the perception layer, transmission layer and management layer according to the topology optimization target to complete the construction of the topology structure.

[0014] In some embodiments, the management information corresponding to each photovoltaic component is generated according to the operating data corresponding to the multiple photovoltaic components, the future shadow distribution heat map and the hybrid topology structure, including: parsing the operating data, obtaining the efficiency impact information of each photovoltaic component according to the light intensity information and the temperature information, and obtaining the component health information of each photovoltaic component according to the IV curve information and the hot spot information; parsing the future shadow distribution heat map, obtaining the shading area and the shading time corresponding to the shading area, and calculating the shadow impact information corresponding to each photovoltaic component according to the shading area and the corresponding shading time; obtaining the importance information corresponding to each photovoltaic component according to the hybrid topology structure; based on a preset optimization algorithm, outputting the management information corresponding to each photovoltaic component according to the efficiency impact information, component health information, shadow impact information and importance information corresponding to each photovoltaic component, so as to maximize the power generation corresponding to the multiple photovoltaic components.

[0015] In the second aspect, the present application provides a distributed photovoltaic component power generation management system based on the Internet of Things, including: Multiple photovoltaic panels; A multimodal sensor array; the multimodal sensor array comprises at least a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager; A control module, the control module comprising a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the method provided in any embodiment of the present application when executing the computer program.

[0016] In a third aspect, the present application provides a distributed photovoltaic assembly power generation management device, comprising: A data acquisition unit, used to acquire operating data of multiple photovoltaic modules collected by a multimodal sensor array; the multimodal sensor array includes a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager; A cloud image acquisition unit, used to acquire fused satellite cloud images and drone aerial photography data corresponding to multiple photovoltaic modules; A data fusion unit, used to input the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network, and output a future shadow distribution heat map; A contribution acquisition unit, used to acquire a node contribution corresponding to each of the photovoltaic components, and construct a hybrid topology structure corresponding to a plurality of the photovoltaic components according to each of the node contributions; A management completion unit is used to generate management information corresponding to each photovoltaic component based on the corresponding operating data of multiple photovoltaic components, future shadow distribution thermal map and hybrid topology structure; the management information at least includes the tilt adjustment information of the photovoltaic component; and complete the distributed management of multiple photovoltaic components.

[0017] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer-readable instructions are executed by the processor, one or more processors execute a method as provided in any embodiment of the present application.

[0018] This application provides a distributed photovoltaic assembly power generation management method and system based on the Internet of Things. The specific technical contents are as follows: Multimodal sensor data acquisition: Use a multimodal sensor array (including light intensity sensor, temperature sensor, IV curve detector and hot spot imager) to collect the operating data of multiple photovoltaic modules. These data can fully reflect the operating status of photovoltaic modules under different environmental conditions; Fusion of satellite cloud images and drone aerial photography data: Obtain fusion satellite cloud images and drone aerial photography data corresponding to multiple photovoltaic modules. These data provide a wider range of environmental information, which helps to more accurately predict the future operating status of photovoltaic modules; Spatiotemporal attention prediction network: The fused satellite cloud image and drone aerial photography data are input into the preset spatiotemporal attention prediction network to output the future shadow distribution heat map. The network can capture the changes in time and space and predict the future shadow distribution, thus providing a basis for the management of photovoltaic modules; Node contribution and hybrid topology: Obtain the node contribution corresponding to each photovoltaic module, and build a hybrid topology corresponding to multiple photovoltaic modules based on each node contribution. This helps to optimize the layout and connection method of photovoltaic modules and improve the efficiency and reliability of the overall system; Management information generation: Generate management information corresponding to each photovoltaic module based on the operation data corresponding to multiple photovoltaic modules, future shadow distribution thermal map and hybrid topology structure. The management information at least includes the tilt adjustment information of the photovoltaic module, thereby realizing distributed management of multiple photovoltaic modules.

[0019] The provided method has at least the following beneficial effects: Comprehensive data collection and analysis: Through multi-modal sensor arrays and the fusion of satellite cloud images and drone aerial data, the operation data of photovoltaic modules can be comprehensively collected and analyzed, overcoming the limitation of traditional methods that rely on a single data source; Accurate prediction and optimization: Using the spatiotemporal attention prediction network, the future shadow distribution can be accurately predicted, providing a scientific basis for the management of photovoltaic modules. Combining node contribution and hybrid topology, the layout and connection of photovoltaic modules can be optimized to improve the overall efficiency and reliability of the system. Distributed management: By generating management information corresponding to each photovoltaic module, distributed management of multiple photovoltaic modules can be achieved. This management method can be adjusted according to the specific conditions of each photovoltaic module, further improving the operating efficiency and reliability of the system; Improve the operating efficiency and reliability of photovoltaic power generation systems: Through the above technical means, the operating status of photovoltaic modules in complex environments can be fully reflected, and the management of photovoltaic modules can be optimized, thereby improving the operating efficiency and reliability of photovoltaic power generation systems; In summary, the method and system provided in this application effectively solve the limitations of traditional photovoltaic management methods through comprehensive data collection, accurate prediction and optimization, distributed management and other technical means, and significantly improve the operating efficiency and reliability of photovoltaic power generation systems.

[0020] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 This is a schematic block diagram of the structure of a distributed photovoltaic assembly power generation management system provided by an embodiment of the present application; Figure 2 This is a schematic flow chart of the steps of a distributed photovoltaic assembly power generation management method based on the Internet of Things provided by an embodiment of the present application; Figure 3 It is a schematic block diagram of the structure of a control module provided in one embodiment of the present application.

[0023] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0025] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0026] It should be understood that, in order to facilitate the clear description of the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, the words "first", "second", etc. are used to distinguish the same items or similar items with substantially the same functions and effects. Those skilled in the art can understand that the words "first", "second", etc. do not limit the quantity and execution order, and the words "first", "second", etc. do not necessarily limit the difference.

[0027] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0028] It should also be understood that the term “and / or” used in the specification and appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0029] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0030] With the continuous growth of global energy demand and the increasingly severe environmental problems, photovoltaic power generation has received widespread attention as a clean and renewable energy form. Distributed photovoltaic power generation systems are widely used in rooftops, building exterior walls and other scenarios due to their flexibility and adaptability. However, the existing distributed photovoltaic management methods still have many shortcomings in terms of operating efficiency, data collection and analysis, and system optimization. It is urgent to propose an innovative management method to improve system performance.

[0031] Traditional photovoltaic management methods mainly rely on a single data source (such as meteorological data or local sensor data) for power generation forecasting and system management. It is difficult to fully reflect the operating status of photovoltaic modules in complex environments, making it difficult to further improve the operating efficiency and reliability of photovoltaic power generation systems.

[0032] Therefore, there is an urgent need for a method to improve the operating efficiency and reliability of photovoltaic power generation systems.

[0033] To solve the above problems, please refer to Figure 1The present application provides a distributed photovoltaic component power generation management system, including: a plurality of photovoltaic components ( Figure 1 The three are only examples, and the present application does not limit the number of photovoltaic modules); a multimodal sensor array; the multimodal sensor array includes at least a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager; a control module, the control module includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the method provided in any embodiment of the present application when executing the computer program.

[0034] Specifically, the distributed photovoltaic module power generation management system proposed in this application aims to improve the operating efficiency and reliability of the photovoltaic power generation system through multimodal data fusion and advanced algorithms. The system mainly includes the following parts: Multimodal sensor array: including light intensity sensor, temperature sensor, IV curve detector and hot spot imager, used to comprehensively collect operating status data of photovoltaic modules; Control module: includes memory and processor, used to store and execute computer programs to realize the intelligent management function of the system; Data acquisition and processing: The system collects the operating data of photovoltaic modules through a multimodal sensor array, integrates satellite cloud images and drone aerial photography data, and inputs them into the spatiotemporal attention prediction network to predict the future shadow distribution heat map; Node contribution and hybrid topology: Build a hybrid topology based on the node contribution of each PV module to optimize the system layout; Management information generation: Generate management information for each PV module based on operating data, future shadow distribution thermal maps and hybrid topology, including tilt adjustment information, to achieve distributed management of multiple PV modules.

[0035] By installing a multimodal sensor array on the photovoltaic module, light intensity, temperature, IV curve and hot spot imaging data are collected in real time. The sensor data is combined with fused satellite cloud images and drone aerial photography data, and the future shadow distribution is predicted through spatiotemporal attention prediction network analysis. According to the prediction results and the node contribution of the photovoltaic module, a hybrid topology is constructed to optimize the layout and operation strategy of the photovoltaic module. According to the optimization results, the management information of each photovoltaic module, such as tilt adjustment, is generated, and automatic management is achieved through the control module.

[0036] Through multimodal data fusion and advanced algorithms, the system can more accurately predict and adjust the operating status of photovoltaic modules, thereby improving overall power generation efficiency. The system can monitor the health status of photovoltaic modules in real time, detect and deal with potential problems in a timely manner, reduce failure rates, and improve system reliability. The distributed management method enables the system to flexibly adapt to different environmental conditions, such as weather changes, shadows, etc., to maintain efficient operation. Through automated management and intelligent optimization, manual intervention is reduced, operation and maintenance costs are reduced, and management efficiency is improved.

[0037] To sum up, the distributed photovoltaic component power generation management system provided in this application effectively improves the operating efficiency and reliability of the photovoltaic power generation system through multimodal data fusion, advanced algorithms and intelligent management, and has significant technical advantages and application value.

[0038] See also Figure 2 , Figure 2 This is a schematic flow chart of a distributed photovoltaic assembly power generation management method based on the Internet of Things provided by an embodiment of the present application. The execution device of the method is a control module of a distributed photovoltaic assembly power generation management system provided by any embodiment of the present application.

[0039] like Figure 2 As shown, the provided method includes steps S101 to S105. The control module may be a handheld terminal, a notebook computer, a wearable device or a robot, etc., for implementing steps S101 to S105 and their corresponding embodiments.

[0040] Step S101. Acquire operating data of multiple photovoltaic modules collected by a multimodal sensor array; the multimodal sensor array includes a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager.

[0041] Specifically, the core of step S101 is to collect the operating data of the photovoltaic module through a multimodal sensor array. The multimodal sensor array includes a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager. These sensors monitor the operating status of the photovoltaic module from different dimensions: Light intensity sensor: real-time monitoring of the light intensity received by the photovoltaic modules, providing basic data for power generation efficiency evaluation; Temperature sensor: monitors the surface temperature of photovoltaic modules. Too high temperature will reduce power generation efficiency and accelerate module aging. IV curve detector: Evaluate the electrical performance of PV modules by measuring the current-voltage curve to detect potential failures or performance degradation; Hot spot imager: Use infrared imaging technology to detect whether there are hot spots on photovoltaic modules. Hot spots will reduce the life of the modules and cause safety hazards.

[0042] A multimodal sensor array is installed on each photovoltaic module to ensure that the sensor can accurately collect data. The sensor data is transmitted to the control module through wired or wireless communication technology (such as LoRa, ZigBee or 5G). The control module pre-processes the data, including denoising, formatting and storage, to provide high-quality data for subsequent analysis.

[0043] Multimodal sensors can comprehensively monitor the status of photovoltaic modules from multiple dimensions such as light, temperature, electrical performance and thermal imaging, avoiding the limitations of a single data source. Real-time data collection can promptly detect abnormal conditions of photovoltaic modules, such as hot spots or performance degradation, facilitating rapid response and processing. It provides reliable data support for subsequent shadow prediction, topology optimization and management information generation.

[0044] Step S102: Obtain fused satellite cloud images and drone aerial photography data corresponding to multiple photovoltaic modules.

[0045] Specifically, the core of step S102 is to obtain environmental information of the area where the photovoltaic components are located by integrating satellite cloud images and drone aerial photography data.

[0046] Satellite cloud images: provide large-scale meteorological data, such as cloud distribution, light intensity and weather trends.

[0047] Drone aerial photography data: Provides high-precision local environmental information, such as the location and height of obstructions such as buildings and trees.

[0048] Obtain satellite cloud image data of the area where the photovoltaic modules are located from meteorological satellites. Use drones to take aerial photos of the area where the photovoltaic modules are located to obtain high-resolution environmental data. Fuse satellite cloud images and drone aerial data to generate comprehensive environmental data containing large-scale meteorological information and local occlusion information.

[0049] By integrating data, we can fully understand the environmental characteristics of the area where the photovoltaic modules are located, including weather changes and the distribution of obstructions. This provides high-precision input data for subsequent shadow distribution predictions and improves prediction accuracy. The system can dynamically adjust the operation strategy of photovoltaic modules according to environmental changes to improve overall adaptability.

[0050] Step S103: Input the fused satellite cloud image and drone aerial photography data into the preset spatiotemporal attention prediction network, and output the future shadow distribution heat map.

[0051] Specifically, the core of step S103 is to use the spatiotemporal attention prediction network to predict future shadow distribution based on fused data. The spatiotemporal attention prediction network is a deep learning model that can capture complex relationships in time and space dimensions and is suitable for shadow prediction in the area where photovoltaic modules are located.

[0052] The fused satellite cloud image and drone aerial photography data are input into the spatiotemporal attention prediction network. The model predicts the shadow distribution in the future by analyzing historical data and current environmental characteristics. The heat map of future shadow distribution is output, which visualizes the location, range and intensity of shadows.

[0053] The spatiotemporal attention prediction network can capture complex environmental changes and provide high-precision shadow distribution prediction. The prediction results can provide early warning of the impact of shadows on photovoltaic modules, making it easier to take adjustment measures. It provides key input for subsequent topology optimization and management information generation, improving system operation efficiency.

[0054] Step S104: Obtain the node contribution corresponding to each photovoltaic component, and construct a hybrid topology structure corresponding to multiple photovoltaic components according to each node contribution.

[0055] Specifically, the core of step S104 is to construct a hybrid topology structure according to the node contribution of the photovoltaic modules. The node contribution refers to the importance of each photovoltaic module in the overall system, which is usually evaluated based on factors such as power generation efficiency, location and environmental impact.

[0056] The node contribution of each photovoltaic module is calculated based on historical operating data and environmental characteristics. Based on the node contribution, the photovoltaic modules are divided into different groups to build a hybrid topology. The hybrid topology can be a star, mesh or hierarchical structure, and the specific form is determined according to actual needs.

[0057] By building a hybrid topology, we can optimize the layout and connection of photovoltaic modules and improve the overall efficiency of the system. We can rationally allocate resources according to the contribution of nodes and ensure that important components are managed first. The hybrid topology can adapt to different environments and operating requirements and improve the flexibility of the system.

[0058] Step S105. Generate management information corresponding to each photovoltaic module according to the operation data corresponding to the multiple photovoltaic modules, the future shadow distribution heat map and the hybrid topology structure; the management information at least includes the tilt adjustment information of the photovoltaic module; and complete the distributed management of the multiple photovoltaic modules.

[0059] Specifically, the core of step S105 is to generate management information of each photovoltaic module based on the operation data, the shadow distribution heat map and the hybrid topology structure. The management information includes tilt adjustment information, cleaning and maintenance suggestions and fault handling solutions.

[0060] Comprehensively analyze operating data, shadow distribution heat map and hybrid topology to evaluate the operating status of each PV module. Generate management information for each PV module, such as adjusting the tilt angle to optimize light reception or arranging cleaning maintenance to remove obstructions. Send management information to execution equipment (such as robots or automatic adjustment devices) through the control module to complete distributed management.

[0061] Management information generated based on multi-dimensional data can accurately optimize the operating status of each photovoltaic module. Through automated equipment to execute management information, manual intervention is reduced, and operation and maintenance costs are reduced. Through dynamic adjustment and optimization, the overall efficiency and reliability of the photovoltaic power generation system are improved.

[0062] Steps S101 to S105 together constitute a distributed photovoltaic module power generation management method based on the Internet of Things. Through multimodal data collection, environmental perception, shadow prediction, topology optimization and intelligent management, the system can significantly improve the operating efficiency and reliability of the photovoltaic power generation system and provide strong technical support for the efficient use of clean energy.

[0063] In some embodiments, the inputting of the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network includes: respectively obtaining the geographic coordinate systems corresponding to the fused satellite cloud image and the drone aerial photography data; spatially and temporally aligning the fused satellite cloud image and the drone aerial photography data according to the geographic coordinate system; adjusting the resolution of the fused satellite cloud image and the drone aerial photography data based on interpolation or downsampling technology; respectively extracting satellite feature information and aerial photography feature information from the adjusted fused satellite cloud image and drone aerial photography data; the satellite feature information includes at least cloud distribution and cloud thickness; the aerial photography feature information includes at least terrain features and obstruction features; and inputting the satellite feature information and aerial photography feature information into the spatiotemporal attention prediction network to output the future shadow distribution heat map.

[0064] The core of this embodiment is to preprocess and extract features from the fused satellite cloud image and drone aerial photography data to improve the input data quality of the spatiotemporal attention prediction network, thereby improving the prediction accuracy of the future shadow distribution heat map. The specific steps are as follows: Extract geographic coordinate system information, such as longitude, latitude, altitude, etc., from satellite cloud images and drone aerial photography data respectively. Ensure that the geographic coordinate systems of the two data are consistent to facilitate subsequent spatial alignment.

[0065] According to the geographic coordinate system, the satellite cloud image and drone aerial photography data are spatially aligned to ensure that they match at the same geographic location. The timestamps are aligned to ensure that the time dimensions of the two data are consistent (such as the same day, the same time or the same time period).

[0066] Use interpolation or downsampling techniques to adjust the resolution of satellite cloud images and drone aerial photography data to make the spatial resolution of the two consistent. For example, downsample high-resolution drone aerial photography data to the same resolution as satellite cloud images, or interpolate low-resolution satellite cloud images to high resolution.

[0067] Extract satellite feature information from the adjusted satellite cloud images, including cloud distribution, cloud thickness, cloud movement speed, etc. Extract aerial photography feature information from the adjusted drone aerial photography data, including terrain features (such as slope and slope direction), obstruction features (such as the height and position of buildings and trees), etc.

[0068] The extracted satellite feature information and aerial photography feature information are input into the spatiotemporal attention prediction network. The network analyzes these feature information, captures the impact of clouds and obstructions on photovoltaic modules, and outputs a future shadow distribution heat map.

[0069] By aligning the geographic coordinate system and adjusting the resolution, the consistency of satellite cloud images and drone aerial photography data in spatial and temporal dimensions is ensured, providing high-quality input data for subsequent analysis.

[0070] Feature information extracted from satellite cloud images and drone aerial photography data (such as cloud distribution and obstruction characteristics) can fully reflect the environmental characteristics of the area where the photovoltaic modules are located, and improve the richness of the input information of the spatiotemporal attention prediction network. The spatiotemporal attention prediction network can more accurately predict future shadow distribution based on high-precision feature information, reduce prediction errors, and provide a reliable basis for the operation optimization of photovoltaic modules. The embodiment can effectively handle shadow prediction problems in complex environments, such as cloudy weather, undulating terrain, and uneven distribution of obstructions, and improve the adaptability and robustness of the system. The resolution is adjusted by interpolation or downsampling technology to reduce the amount of data, reduce the computational complexity of the spatiotemporal attention prediction network, and improve the prediction efficiency.

[0071] In cloudy weather, the rapid movement of clouds can cause frequent changes in the light intensity received by the photovoltaic modules. Example 1 can accurately predict the shadow effect of clouds on photovoltaic modules by extracting cloud distribution and thickness information, making it easier to adjust the module inclination or operation strategy in advance.

[0072] In mountainous or urban environments, terrain undulations and obstructions such as buildings and trees can have a significant impact on the power generation efficiency of photovoltaic modules. By extracting terrain features and obstruction features, the embodiment can accurately predict the shadow distribution caused by these factors and optimize the layout and operation of photovoltaic modules. It can monitor environmental changes in real time (such as cloud movement, addition or removal of obstructions), dynamically update the shadow distribution heat map, and ensure that photovoltaic modules are always in the best operating state.

[0073] The embodiment significantly improves the input data quality of the spatiotemporal attention prediction network by preprocessing and extracting features from the fusion of satellite cloud images and drone aerial photography data, thereby improving the prediction accuracy of future shadow distribution heat maps. This method has significant advantages in scenarios such as cloudy weather, complex terrain, and dynamic environments, and can provide reliable environmental perception and operation optimization support for distributed photovoltaic component power generation management systems, ultimately improving the overall efficiency and reliability of the system.

[0074] Exemplarily, the spatiotemporal attention prediction network includes a convolution layer, a long short-term memory layer and an attention module; the inputting the satellite feature information and the aerial photography feature information into the spatiotemporal attention prediction network to output the future shadow distribution heat map includes: inputting the satellite feature information and the aerial photography feature information into the convolution layer and the long short-term memory layer respectively, the convolution layer outputs the spatial features corresponding to the satellite feature information and the aerial photography feature information, and the long short-term memory layer outputs the time feature sequence corresponding to the satellite feature information and the aerial photography feature information; the spatial features and the time feature sequence are input into the attention module respectively, and the attention module outputs the time attention information and the spatial attention information respectively; the time attention information is used to determine the key change moments in the time feature sequence; the spatial attention information is used to determine the occlusion area features corresponding to the spatial features; and the future shadow distribution heat map is output according to the occlusion area features and the key change moments.

[0075] This example describes in detail the structure and workflow of the spatiotemporal attention prediction network. The network consists of convolutional layers, long short-term memory layers (LSTM), and attention modules. It aims to extract spatial and temporal features from satellite feature information and aerial photography feature information, and capture key change moments and occluded area features through the attention mechanism, and finally output the future shadow distribution heat map. The specific steps are as follows: The satellite feature information (such as cloud distribution and cloud thickness) and aerial photography feature information (such as terrain features and obstruction features) are respectively input into the spatiotemporal attention prediction network.

[0076] The convolution layer is used to extract the spatial features of the input feature information. The satellite feature information and the aerial feature information are convoluted separately to output the corresponding spatial features. The spatial features can reflect the environmental characteristics of the area where the photovoltaic modules are located, such as the spatial distribution of clouds and the location of obstructions.

[0077] The long short-term memory layer (LSTM) is used to extract the time feature sequence of the input feature information. The satellite feature information and the aerial photography feature information are processed by LSTM respectively, and the corresponding time feature sequence is output. The time feature sequence can reflect the changing trend of environmental features over time, such as the moving speed of clouds and the dynamic changes of obstructions.

[0078] The spatial features output by the convolution layer and the temporal feature sequence output by the LSTM layer are input to the attention module respectively. The attention module includes a temporal attention mechanism and a spatial attention mechanism. Temporal attention mechanism: Analyze the temporal feature sequence to determine the key change moments (such as the moment when the clouds move quickly or the occluders are added). Spatial attention mechanism: Analyze the spatial features to determine the occluded area features (such as the location and range of the occluders). The attention module outputs temporal attention information and spatial attention information respectively. Based on the temporal attention information and spatial attention information, combined with the occluded area features and the key change moments, a heat map of future shadow distribution is generated. The heat map visualizes the location, range and intensity of shadows over a period of time in the future.

[0079] The convolutional layer and the long short-term memory layer extract spatial features and temporal features respectively, which can fully reflect the environmental changes in the area where the photovoltaic modules are located. The temporal attention mechanism can capture the key change moments in the temporal feature sequence, such as the moment when the clouds move quickly or the obstructions are added, and provide key time nodes for shadow prediction. The spatial attention mechanism can identify the occlusion area features in the spatial features, such as the location and range of the obstructions, and provide key spatial information for shadow prediction. By combining temporal attention information and spatial attention information, the spatiotemporal attention prediction network can more accurately predict future shadow distribution and reduce prediction errors. The method can handle dynamic environmental changes (such as cloud movement, addition or removal of obstructions) in real time, dynamically update the shadow distribution heat map, and ensure the timeliness and accuracy of the prediction results. The attention mechanism can focus on key change moments and occlusion area features, reduce unnecessary calculations, and improve prediction efficiency.

[0080] In cloudy weather, the rapid movement of clouds will cause frequent changes in the light intensity received by photovoltaic modules. The key moments of cloud movement are captured by the temporal attention mechanism, and the spatial attention mechanism is combined to identify the spatial distribution of clouds, which can accurately predict the shadow effect of clouds on photovoltaic modules.

[0081] In mountainous or urban environments, terrain undulations and obstructions such as buildings and trees can have a significant impact on the power generation efficiency of photovoltaic modules. The spatial attention mechanism is used to identify the location and range of obstructions, and the temporal attention mechanism is used to capture the dynamic changes of obstructions, which can accurately predict the shadow distribution caused by these factors. The example can monitor environmental changes in real time (such as cloud movement, addition or removal of obstructions), dynamically update the shadow distribution heat map, and ensure that photovoltaic modules are always in the best operating state.

[0082] Through the combination of convolutional layers, long short-term memory layers and attention modules, the spatial and temporal characteristics of the area where the photovoltaic components are located are fully captured, and the key change moments and occluded area characteristics are accurately located through the attention mechanism, and finally a high-precision future shadow distribution heat map is output. This method has significant advantages in scenes such as cloudy weather, complex terrain and dynamic environments, and can provide reliable environmental perception and operation optimization support for distributed photovoltaic component power generation management systems, ultimately improving the overall efficiency and reliability of the system.

[0083] It should be noted that, in some embodiments, outputting the future shadow distribution heat map according to the occlusion area characteristics and the key change moment includes: obtaining the shadow distribution influence weight corresponding to the occlusion area characteristics; obtaining the environmental change information corresponding to the key change moment; the environmental change information includes at least cloud movement and change in the sun's incidence angle; adjusting the shadow distribution influence weight according to the environmental change information; and generating the future shadow distribution heat map according to the adjusted shadow distribution influence weight.

[0084] The process of generating future shadow distribution heat maps based on the characteristics of the occluded area and key change moments is further refined. This method dynamically adjusts the shadow distribution prediction results by introducing shadow distribution influence weights and environmental change information, thereby improving the accuracy and practicality of the heat map. The specific steps are as follows: According to the characteristics of the shading area (such as the location, height and range of the shading object), the weight of its influence on the shadow distribution of the photovoltaic module is calculated. The influence weight reflects the degree of influence of the shading object on the power generation efficiency of the photovoltaic module. For example, a tall building may have a larger influence weight.

[0085] Extract key change moments from the time attention information and obtain corresponding environmental change information. Environmental change information includes cloud movement speed, cloud thickness change, solar incident angle change, etc. This information can reflect the dynamic impact of the environment on the shadow distribution of photovoltaic modules.

[0086] Dynamically adjust the influence weights of the features of the shading area based on the environmental change information. For example, when the clouds move quickly, the influence weight of the shading object may decrease; when the sun's incident angle changes, the shadow range of the shading object may expand or shrink. Generate a future shadow distribution heat map based on the adjusted shadow distribution influence weights. The heat map visualizes the location, range, and intensity of the shadow over a period of time in the future, which facilitates the optimization of PV module operation.

[0087] By introducing environmental change information and dynamically adjusting the weight of shadow distribution, the dynamic impact of the environment on the shadow distribution of photovoltaic modules can be more accurately reflected. Combining the characteristics of the blocked area and environmental change information, a more accurate future shadow distribution heat map can be generated to reduce prediction errors. It can effectively handle shadow prediction problems in complex environments, such as cloudy weather, undulating terrain, and uneven distribution of obstructions, and improve the adaptability and robustness of the system. High-precision future shadow distribution heat maps provide a reliable basis for the operation optimization of photovoltaic modules, such as adjusting the inclination of modules or cleaning and maintenance arrangements to improve the overall power generation efficiency. By accurately predicting shadow distribution, unnecessary operation and maintenance operations can be reduced, and the operation and maintenance costs of photovoltaic power generation systems can be reduced.

[0088] In some embodiments, obtaining the node contribution corresponding to each of the photovoltaic components includes: obtaining operating data corresponding to each of the photovoltaic components; the operating data knowledge includes power output information, current information and voltage information; obtaining topological structure information corresponding to each of the photovoltaic components; the topological structure information is used to determine the connection relationship between each of the photovoltaic components and the remaining photovoltaic components; obtaining environmental data corresponding to each of the photovoltaic components; the environmental data includes at least light intensity, temperature and wind speed; and calculating the node contribution corresponding to each of the photovoltaic components based on the operating data, topological structure information and environmental data.

[0089] By comprehensively considering the operation data, topology information and environmental data of the PV modules, the node contribution of each PV module is calculated. The node contribution reflects the importance of the PV module in the overall system and is a key indicator for building a hybrid topology and optimizing the operation strategy. The specific steps are as follows: The operation data of each photovoltaic module is obtained from the multimodal sensor array, including power output information, current information and voltage information. These data reflect the power generation performance and operation status of the photovoltaic module. The topological structure information of each photovoltaic module is obtained, including its connection relationship with other photovoltaic modules (such as series, parallel or hybrid connection). The topological structure information is used to evaluate the position and role of the photovoltaic module in the system. The environmental data of the area where each photovoltaic module is located is obtained from the environmental sensor, including light intensity, temperature and wind speed. These data reflect the impact of the environment on the power generation efficiency of the photovoltaic module. The node contribution of each photovoltaic module is calculated based on the operation data, topological structure information and environmental data. By comprehensively considering the operation data, topological structure information and environmental data, the importance of each photovoltaic module in the system can be comprehensively evaluated to avoid the limitations of a single indicator. The node contribution provides a key basis for constructing a hybrid topological structure, ensuring that important components are managed and optimized in the system with priority. The optimization strategy based on node contribution can improve the overall efficiency of the photovoltaic power generation system and reduce resource waste. The method can effectively handle the node contribution calculation problem in complex environments, such as cloudy weather, undulating terrain and uneven distribution of obstructions, and improve the adaptability and robustness of the system. According to the node contribution, the operation strategy of the photovoltaic modules (such as tilt adjustment, cleaning and maintenance) is dynamically adjusted to ensure that the system is always in the best operating state.

[0090] Exemplarily, the calculating of the node contribution corresponding to each of the photovoltaic components according to the operating data, topological structure information and environmental data includes: calculating the energy contribution corresponding to each of the photovoltaic components according to the operating data; calculating the reliability information corresponding to each of the photovoltaic components according to the operating data and environmental data; the reliability information is used to determine the stability of each photovoltaic component under abnormal conditions; calculating the network influence coefficient of each of the photovoltaic components according to the topological structure information and the operating data; the network influence coefficient is used to determine the influence of the location of each of the photovoltaic components on the communication and energy transmission of multiple photovoltaic components shown.

[0091] By comprehensively considering energy contribution, reliability information and network impact coefficient, the importance of each photovoltaic module in the system can be fully evaluated to avoid the limitations of a single indicator. Node contribution provides a key basis for building a hybrid topology, ensuring that important components are managed and optimized in the system. The optimization strategy based on node contribution can improve the overall efficiency of the photovoltaic power generation system and reduce resource waste. It can effectively handle node contribution calculation problems in complex environments, such as cloudy weather, undulating terrain and uneven distribution of obstructions, and improve the adaptability and robustness of the system. According to the node contribution, the operation strategy of the photovoltaic module (such as tilt adjustment, cleaning and maintenance) is dynamically adjusted to ensure that the system is always in the best operating state.

[0092] In cloudy weather, the light intensity received by different photovoltaic modules may vary greatly. By introducing reliability information, the node contribution of each module can be accurately evaluated, which facilitates the optimization of operation strategies. In mountainous or urban environments, the undulating terrain and obstructions such as buildings and trees will have a significant impact on the power generation efficiency of photovoltaic modules. By comprehensively considering environmental data, the node contribution of each module can be accurately calculated to optimize the system layout. It can monitor environmental changes in real time (such as cloud movement, addition or removal of obstructions), dynamically update node contributions, and ensure that the system is always in the best operating state.

[0093] By introducing energy contribution, reliability information and network impact coefficient, the node contribution of each photovoltaic module is comprehensively evaluated, providing a key basis for building a hybrid topology and optimizing operation strategies. This method has significant advantages in scenarios such as cloudy weather, complex terrain and dynamic environments, and can provide reliable environmental perception and operation optimization support for distributed photovoltaic module power generation management systems, ultimately improving the overall efficiency and reliability of the system.

[0094] It should be noted that, in some embodiments, the node contribution is a key indicator for evaluating the importance of photovoltaic modules in the system. The following is a calculation formula for the node contribution: Ci=α⋅Ei+β⋅Hi+γ⋅Si+δ⋅Ii; Ci is the node contribution of the i-th photovoltaic module. Ei is the efficiency impact information of the i-th photovoltaic module, which describes the impact of environmental factors (such as light intensity and temperature) on the power generation efficiency of the photovoltaic module. Hi is the component health information of the i-th photovoltaic module, which describes the operating stability and potential failure risk of the photovoltaic module. Si is the shadow impact information of the i-th photovoltaic module, which describes the impact of obstructions on the power generation efficiency of the photovoltaic module. Ii is the importance information of the i-th photovoltaic module, which describes the position and role of the photovoltaic module in the system. The importance information is related to the level (perception layer, transmission layer, management layer) of the photovoltaic module in the hybrid topology structure, and can be assigned by the following rules: Perception layer: Ii=0.3; transmission layer: Ii=0.5; management layer: Ii=0.8. α, β, γ⋅δ are weight coefficients used to adjust the weights of efficiency impact information, component health information, shadow impact information, and importance information in node contribution. This application takes α, β, γ⋅δ as 0.4, 0.3, 0.2, and 0.1 respectively.

[0095] In some embodiments, constructing a hybrid topology structure corresponding to multiple photovoltaic components according to the contribution of each node includes: dividing the multiple photovoltaic components into any one of a perception layer, a transmission layer and a management layer according to the node contribution; wherein each photovoltaic component in the perception layer adopts a mesh topology, each photovoltaic component in the transmission layer adopts a tree topology, and the photovoltaic components in the management layer adopt a star topology; constructing the hybrid topology structure according to the perception layer, the transmission layer and the management layer.

[0096] The core of this embodiment is to divide the photovoltaic modules into the perception layer, transmission layer and management layer according to their node contribution, and adopt different topological structures (mesh topology, tree topology and star topology) respectively, and finally build a hybrid topological structure. This method improves the efficiency and reliability of the photovoltaic power generation system through hierarchical design and optimization of the topological structure. The specific steps are as follows: According to the node contribution of each PV module, it is divided into perception layer, transmission layer or management layer.

[0097] The division rules are as follows: Perception layer: PV modules with low node contribution are mainly responsible for data collection and environmental perception; Transmission layer: PV modules with medium node contribution are mainly responsible for data transmission and energy transmission; Management level: PV modules with high node contribution are mainly responsible for system management and optimization decisions; Perception layer: adopts mesh topology.

[0098] The mesh topology has high redundancy and high reliability, and is suitable for data collection and environmental perception tasks. Each photovoltaic module is directly connected to multiple adjacent modules to ensure the stability and fault tolerance of data transmission. Transmission layer: adopts a tree topology. The tree topology has an efficient data transmission path, which is suitable for energy transmission and data relay tasks. Photovoltaic modules are connected according to the hierarchical relationship to form a transmission path from the perception layer to the management layer. Management layer: adopts a star topology. The star topology has the characteristics of centralized management and is suitable for system optimization and decision-making tasks. All photovoltaic modules are directly connected to the central management node to facilitate data aggregation and command issuance.

[0099] The perception layer, transmission layer and management layer are integrated according to their functions and topological structures to form a hybrid topology. The hybrid topology combines the advantages of mesh topology, tree topology and star topology, and can meet the needs of data collection, transmission and management at the same time.

[0100] Through hierarchical design and hybrid topology, the architecture of photovoltaic power generation system is optimized to improve the overall efficiency and reliability of the system. The combination of mesh topology and tree topology can ensure the efficient transmission of data in the perception layer and transmission layer, reducing delay and packet loss rate. The centralized management characteristics of star topology facilitate the management layer to monitor the system in real time and optimize decisions, thereby improving the operation efficiency of the system. The high redundancy of mesh topology and the efficient transmission path of tree topology can improve the fault tolerance of the system and ensure that the system can still operate normally when some components fail. The hybrid topology can effectively handle system architecture problems in complex environments, such as cloudy weather, undulating terrain, and uneven distribution of obstructions, and improve the adaptability and robustness of the system. By optimizing the system architecture and improving the operation efficiency, unnecessary operation and maintenance operations can be reduced, and the operation and maintenance costs of photovoltaic power generation systems can be reduced.

[0101] Exemplarily, constructing the hybrid topology structure according to the perception layer, transmission layer and management layer includes: obtaining a preset topology optimization target; the topology optimization target includes at least a target energy transmission efficiency, a target communication delay and a target reliability; adjusting the connection relationship between the perception layer, transmission layer and management layer according to the topology optimization target to complete the construction of the topology structure.

[0102] Preset topology optimization goals, including but not limited to: Target energy transmission efficiency: efficiency requirements of the system during energy transmission. Target communication delay: delay requirements of the system during data transmission. Target reliability: stability requirements of the system under abnormal conditions.

[0103] According to the target communication delay and target reliability, adjust the connection relationship of photovoltaic modules in the perception layer. Adopt a mesh topology to increase redundant connections to ensure the stability and fault tolerance of data transmission. According to the target energy transmission efficiency and target communication delay, adjust the connection relationship of photovoltaic modules in the transmission layer. Adopt a tree topology to optimize the data transmission path and reduce energy loss and communication delay. According to the target reliability and target communication delay, adjust the connection relationship of photovoltaic modules in the management layer. Adopt a star topology to centrally manage data aggregation and command issuance to ensure the reliability of system operation.

[0104] According to the topology optimization goal, the perception layer, transmission layer and management layer are integrated according to their functions and connection relationships to form a hybrid topology. The hybrid topology combines the advantages of mesh topology, tree topology and star topology, and can meet the needs of energy transmission, data transmission and system management at the same time.

[0105] By introducing topology optimization objectives, the connection relationship between each layer is dynamically adjusted, the architecture of the photovoltaic power generation system is optimized, and the overall efficiency and reliability of the system are improved. The tree topology of the transmission layer can optimize the energy transmission path, reduce energy loss, and improve the energy transmission efficiency of the system. The mesh topology of the perception layer and the tree topology of the transmission layer can optimize the data transmission path, reduce communication delays, and ensure the real-time data transmission. The mesh topology of the perception layer and the star topology of the management layer can improve the redundancy and fault tolerance of the system, ensuring that the system can still operate normally when some components fail. The hybrid topology can effectively handle system architecture problems in complex environments, such as cloudy weather, undulating terrain, and uneven distribution of obstructions, and improve the adaptability and robustness of the system. By optimizing the system architecture and improving operation efficiency, unnecessary operation and maintenance operations can be reduced, and the operation and maintenance costs of the photovoltaic power generation system can be reduced.

[0106] In some embodiments, the management information corresponding to each photovoltaic component is generated according to the operating data corresponding to the multiple photovoltaic components, the future shadow distribution heat map and the hybrid topology structure, including: parsing the operating data, obtaining the efficiency impact information of each photovoltaic component according to the light intensity information and the temperature information, and obtaining the component health information of each photovoltaic component according to the IV curve information and the hot spot information; parsing the future shadow distribution heat map, obtaining the shading area and the shading time corresponding to the shading area, and calculating the shadow impact information corresponding to each photovoltaic component according to the shading area and the corresponding shading time; obtaining the importance information corresponding to each photovoltaic component according to the hybrid topology structure; based on a preset optimization algorithm, outputting the management information corresponding to each photovoltaic component according to the efficiency impact information, component health information, shadow impact information and importance information corresponding to each photovoltaic component, so as to maximize the power generation corresponding to the multiple photovoltaic components.

[0107] By extracting light intensity information and temperature information from the operating data, the efficiency impact information of each photovoltaic module is calculated. The efficiency impact information reflects the impact of environmental factors on the power generation efficiency of photovoltaic modules. The IV curve information and hot spot information are extracted from the operating data to evaluate the health status of each photovoltaic module. The module health information reflects the operating stability and potential failure risk of the photovoltaic module. The shaded area and its corresponding shaded time are extracted from the future shadow distribution heat map to calculate the shadow impact information of each photovoltaic module. The shadow impact information reflects the impact of the shade on the power generation efficiency of the photovoltaic module. The importance information of each photovoltaic module is extracted from the hybrid topology to reflect its position and role in the system. According to the efficiency impact information, module health information, shadow impact information and importance information, the preset optimization algorithm (such as genetic algorithm, particle swarm optimization algorithm, etc.) is used to generate the management information corresponding to each photovoltaic module. The management information includes operation strategies (such as tilt adjustment, cleaning and maintenance) and optimization goals (such as maximizing power generation). The optimization strategy based on management information can improve the overall efficiency of the photovoltaic power generation system and reduce resource waste. Through the evaluation of module health information and shadow impact information, potential failure risks can be discovered in time to ensure the stable operation of the system.

[0108] The embodiments of the present application also provide a distributed photovoltaic assembly power generation management device. The distributed photovoltaic assembly power generation management device is used to execute the steps of the distributed photovoltaic assembly power generation management method based on the Internet of Things shown in the above embodiments. The distributed photovoltaic assembly power generation management device can be a single server or a server cluster, or the distributed photovoltaic assembly power generation management device can be a terminal, which can be a handheld terminal, a laptop computer, a wearable device or a robot, etc.

[0109] The distributed photovoltaic module power generation management device includes: An information acquisition unit, used to acquire user power usage information of a user terminal corresponding to the smart socket system, and construct power usage behavior information corresponding to the user terminal according to the user power usage information; A type acquisition unit, used to acquire a device type corresponding to the electrical device connected to the smart socket system; A sensing acquisition unit, used to acquire circuit sensing information collected by a sensor array of each independent control circuit; A timing generation unit, configured to generate timing control information corresponding to each of the independent control circuits according to the plurality of circuit sensing information, power usage behavior information and the device type corresponding to each of the independent control circuits; A power supply control unit is used to control the relay of each independent control circuit to supply power to the electrical equipment according to the timing control information.

[0110] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working processes of the distributed photovoltaic component power generation management device and each unit described above can refer to the corresponding processes in the embodiments of the distributed photovoltaic component power generation management method based on the Internet of Things described in the above embodiments, and will not be repeated here.

[0111] The above-mentioned distributed photovoltaic assembly power generation management method is implemented in the form of a computer program, and the computer program can be run on the above-mentioned device.

[0112] See also Figure 3 , Figure 3 : is a schematic block diagram of the structure of a control module provided in an embodiment of the present application. The control module includes a processor, a memory and a network interface connected via a device bus, wherein the memory may include a storage medium and an internal memory.

[0113] The storage medium can store an operating device and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any embodiment of the distributed photovoltaic assembly power generation management method.

[0114] The processor is used to provide computing and control capabilities and support the operation of the entire control module.

[0115] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any method of the distributed photovoltaic component power generation management system.

[0116] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the terminal to which the scheme of the present application is applied. The specific control module may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0117] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0118] In one embodiment, the processor is used to run a computer program stored in the memory to obtain the operating data of multiple photovoltaic modules collected by a multi-modal sensor array; the multi-modal sensor array includes a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager; Obtain fused satellite cloud images and drone aerial photography data corresponding to multiple photovoltaic modules; Input the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network, and output a future shadow distribution heat map; Acquire a node contribution corresponding to each of the photovoltaic components, and construct a hybrid topology structure corresponding to a plurality of the photovoltaic components according to each of the node contributions; Management information corresponding to each photovoltaic component is generated according to the corresponding operating data of multiple photovoltaic components, future shadow distribution thermal map and hybrid topology structure; the management information at least includes the tilt adjustment information of the photovoltaic component; and the distributed management of the multiple photovoltaic components is completed.

[0119] In some embodiments, the inputting of the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network includes: respectively obtaining the geographic coordinate systems corresponding to the fused satellite cloud image and the drone aerial photography data; spatially and temporally aligning the fused satellite cloud image and the drone aerial photography data according to the geographic coordinate system; adjusting the resolution of the fused satellite cloud image and the drone aerial photography data based on interpolation or downsampling technology; respectively extracting satellite feature information and aerial photography feature information from the adjusted fused satellite cloud image and drone aerial photography data; the satellite feature information includes at least cloud distribution and cloud thickness; the aerial photography feature information includes at least terrain features and obstruction features; and inputting the satellite feature information and aerial photography feature information into the spatiotemporal attention prediction network to output the future shadow distribution heat map.

[0120] Exemplarily, the spatiotemporal attention prediction network includes a convolution layer, a long short-term memory layer and an attention module; the inputting the satellite feature information and the aerial photography feature information into the spatiotemporal attention prediction network to output the future shadow distribution heat map includes: inputting the satellite feature information and the aerial photography feature information into the convolution layer and the long short-term memory layer respectively, the convolution layer outputs the spatial features corresponding to the satellite feature information and the aerial photography feature information, and the long short-term memory layer outputs the time feature sequence corresponding to the satellite feature information and the aerial photography feature information; the spatial features and the time feature sequence are input into the attention module respectively, and the attention module outputs the time attention information and the spatial attention information respectively; the time attention information is used to determine the key change moments in the time feature sequence; the spatial attention information is used to determine the occlusion area features corresponding to the spatial features; and the future shadow distribution heat map is output according to the occlusion area features and the key change moments.

[0121] It should be noted that, in some embodiments, outputting the future shadow distribution heat map according to the occlusion area characteristics and the key change moment includes: obtaining the shadow distribution influence weight corresponding to the occlusion area characteristics; obtaining the environmental change information corresponding to the key change moment; the environmental change information includes at least cloud movement and change in the sun's incidence angle; adjusting the shadow distribution influence weight according to the environmental change information; and generating the future shadow distribution heat map according to the adjusted shadow distribution influence weight.

[0122] In some embodiments, obtaining the node contribution corresponding to each of the photovoltaic components includes: obtaining operating data corresponding to each of the photovoltaic components; the operating data knowledge includes power output information, current information and voltage information; obtaining topological structure information corresponding to each of the photovoltaic components; the topological structure information is used to determine the connection relationship between each of the photovoltaic components and the remaining photovoltaic components; obtaining environmental data corresponding to each of the photovoltaic components; the environmental data includes at least light intensity, temperature and wind speed; and calculating the node contribution corresponding to each of the photovoltaic components based on the operating data, topological structure information and environmental data.

[0123] Exemplarily, the calculating of the node contribution corresponding to each of the photovoltaic components according to the operating data, topological structure information and environmental data includes: calculating the energy contribution corresponding to each of the photovoltaic components according to the operating data; calculating the reliability information corresponding to each of the photovoltaic components according to the operating data and environmental data; the reliability information is used to determine the stability of each photovoltaic component under abnormal conditions; calculating the network influence coefficient of each of the photovoltaic components according to the topological structure information and the operating data; the network influence coefficient is used to determine the influence of the location of each of the photovoltaic components on the communication and energy transmission of multiple photovoltaic components shown.

[0124] In some embodiments, constructing a hybrid topology structure corresponding to multiple photovoltaic components according to the contribution of each node includes: dividing the multiple photovoltaic components into any one of a perception layer, a transmission layer and a management layer according to the node contribution; wherein each photovoltaic component in the perception layer adopts a mesh topology, each photovoltaic component in the transmission layer adopts a tree topology, and the photovoltaic components in the management layer adopt a star topology; constructing the hybrid topology structure according to the perception layer, the transmission layer and the management layer.

[0125] Exemplarily, constructing the hybrid topology structure according to the perception layer, transmission layer and management layer includes: obtaining a preset topology optimization target; the topology optimization target includes at least a target energy transmission efficiency, a target communication delay and a target reliability; adjusting the connection relationship between the perception layer, transmission layer and management layer according to the topology optimization target to complete the construction of the topology structure.

[0126] In some embodiments, the management information corresponding to each photovoltaic component is generated according to the operating data corresponding to the multiple photovoltaic components, the future shadow distribution heat map and the hybrid topology structure, including: parsing the operating data, obtaining the efficiency impact information of each photovoltaic component according to the light intensity information and the temperature information, and obtaining the component health information of each photovoltaic component according to the IV curve information and the hot spot information; parsing the future shadow distribution heat map, obtaining the shading area and the shading time corresponding to the shading area, and calculating the shadow impact information corresponding to each photovoltaic component according to the shading area and the corresponding shading time; obtaining the importance information corresponding to each photovoltaic component according to the hybrid topology structure; based on a preset optimization algorithm, outputting the management information corresponding to each photovoltaic component according to the efficiency impact information, component health information, shadow impact information and importance information corresponding to each photovoltaic component, so as to maximize the power generation corresponding to the multiple photovoltaic components.

[0127] It should be noted that technicians in the relevant field can clearly understand that for the convenience and brevity of description, the specific working process of the processor described above can refer to the corresponding process in the method embodiments described in the above embodiments, and will not be repeated here.

[0128] A computer-readable storage medium is also provided in an embodiment of the present application, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and the processor executes the program instructions to implement the steps of the distributed photovoltaic component power generation management method based on the Internet of Things provided in the above-mentioned embodiments of the present application.

[0129] The computer-readable storage medium may be an internal storage unit of the control module described in the above embodiment, such as a hard disk or memory of the control module. The computer-readable storage medium may also be an external storage device of the control module, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the control module.

[0130] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A distributed photovoltaic assembly power generation management method based on the Internet of Things, characterized in that: A control module applied to a distributed photovoltaic assembly power generation management system; the method comprises: Acquiring operation data of a plurality of photovoltaic modules collected by a multimodal sensor array; the multimodal sensor array comprises a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager; Obtain fused satellite cloud images and drone aerial photography data corresponding to multiple photovoltaic modules; Input the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network, and output a future shadow distribution heat map; Acquire a node contribution corresponding to each of the photovoltaic components, and construct a hybrid topology structure corresponding to a plurality of the photovoltaic components according to each of the node contributions; Management information corresponding to each photovoltaic component is generated according to the corresponding operating data of multiple photovoltaic components, future shadow distribution thermal map and hybrid topology structure; the management information at least includes the tilt adjustment information of the photovoltaic component; and the distributed management of the multiple photovoltaic components is completed.

2. The method according to claim 1, characterized in that The step of inputting the fused satellite cloud image and drone aerial photography data into a preset spatiotemporal attention prediction network includes: Respectively obtain the geographic coordinate systems corresponding to the fused satellite cloud image and the drone aerial photography data; Aligning the fused satellite cloud image with the drone aerial photography data in space and time according to the geographic coordinate system; Adjusting the resolution of the fused satellite cloud image and the drone aerial photography data based on interpolation or downsampling technology; Extract satellite feature information and aerial photography feature information from the adjusted fused satellite cloud image and drone aerial photography data respectively; the satellite feature information at least includes cloud layer distribution and cloud layer thickness; the aerial photography feature information at least includes terrain features and obstruction features; The satellite feature information and the aerial photography feature information are input into the spatiotemporal attention prediction network to output the future shadow distribution heat map.

3. The method according to claim 2, characterized in that The spatiotemporal attention prediction network includes a convolutional layer, a long short-term memory layer and an attention module; the inputting of the satellite feature information and the aerial photography feature information into the spatiotemporal attention prediction network to output the future shadow distribution heat map includes: The satellite feature information and the aerial photography feature information are input into the convolution layer and the long short-term memory layer respectively, the convolution layer outputs the spatial features corresponding to the satellite feature information and the aerial photography feature information, and the long short-term memory layer outputs the temporal feature sequence corresponding to the satellite feature information and the aerial photography feature information; The spatial feature and the temporal feature sequence are respectively input into the attention module, and the attention module outputs temporal attention information and spatial attention information respectively; the temporal attention information is used to determine the key change moment in the temporal feature sequence; the spatial attention information is used to determine the occlusion area feature corresponding to the spatial feature; The future shadow distribution heat map is output according to the occlusion area characteristics and key change moments.

4. The method according to claim 3, characterized in that Outputting the future shadow distribution heat map according to the occlusion area features and key change moments includes: Obtaining the shadow distribution influence weight corresponding to the occlusion area feature; Acquire environmental change information corresponding to the key change moment; the environmental change information at least includes cloud movement and change in the sun's incident angle; Adjusting the shadow distribution influence weight according to the environmental change information; The future shadow distribution heat map is generated according to the adjusted shadow distribution influence weight.

5. The method according to claim 1, characterized in that The obtaining of the node contribution corresponding to each photovoltaic component includes: Acquire operation data corresponding to each photovoltaic module; the operation data includes power output information, current information and voltage information; Acquire topological structure information corresponding to each photovoltaic component; the topological structure information is used to determine the connection relationship between each photovoltaic component and the remaining photovoltaic components; Acquire environmental data corresponding to each photovoltaic module; the environmental data at least includes light intensity, temperature and wind speed; The node contribution corresponding to each photovoltaic component is calculated according to the operating data, topological structure information and environmental data.

6. The method according to claim 5, characterized in that The calculating the node contribution corresponding to each photovoltaic component according to the operation data, topology information and environmental data includes: Calculate the energy contribution corresponding to each photovoltaic component according to the operating data; Calculating reliability information corresponding to each photovoltaic module according to the operating data and the environmental data; the reliability information is used to determine the stability of each photovoltaic module under abnormal conditions; A network influence coefficient of each photovoltaic component is calculated according to the topology information and the operation data; the network influence coefficient is used to determine the influence of the position of each photovoltaic component on the communication and energy transmission of the plurality of photovoltaic components.

7. The method according to claim 1, characterized in that The step of constructing a hybrid topology structure corresponding to a plurality of photovoltaic components according to the contribution of each node includes: Divide the plurality of photovoltaic components into any one of a perception layer, a transmission layer and a management layer according to the node contribution; wherein each photovoltaic component in the perception layer adopts a mesh topology, each photovoltaic component in the transmission layer adopts a tree topology, and the photovoltaic components in the management layer adopt a star topology; The hybrid topology is constructed according to the perception layer, the transmission layer and the management layer.

8. The method according to claim 7, characterized in that The step of constructing the hybrid topology structure according to the perception layer, the transmission layer and the management layer includes: Obtaining a preset topology optimization target; the topology optimization target at least includes a target energy transmission efficiency, a target communication delay, and a target reliability; The connection relationship between the perception layer, the transmission layer and the management layer is adjusted according to the topology optimization target to complete the construction of the topology structure.

9. The method according to claim 1, characterized in that: The generating of management information corresponding to each photovoltaic component according to the operation data corresponding to the plurality of photovoltaic components, the future shadow distribution thermal map and the hybrid topology structure includes: Parsing the operating data, obtaining efficiency impact information of each photovoltaic module according to light intensity information and temperature information, and obtaining component health information of each photovoltaic module according to IV curve information and hot spot information; Analyze the future shadow distribution heat map, obtain the shaded area and the shaded time corresponding to the shaded area, and calculate the shadow impact information corresponding to each photovoltaic module according to the shaded area and the corresponding shaded time; Acquire importance information corresponding to each photovoltaic component according to the hybrid topology structure; Based on a preset optimization algorithm, management information corresponding to each photovoltaic component is output according to the efficiency impact information, component health information, shadow impact information and importance information corresponding to each photovoltaic component, so as to maximize the power generation corresponding to the plurality of photovoltaic components.

10. A distributed photovoltaic module power generation management system, characterized in that ,include: Multiple photovoltaic panels; A multimodal sensor array; the multimodal sensor array comprises at least a light intensity sensor, a temperature sensor, an IV curve detector and a hot spot imager; A control module, the control module comprising a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the method as claimed in any one of claims 1 to 9 when executing the computer program.

Images