Method and system for monitoring running state of photovoltaic power station in real time

Abstract

The invention provides a method and system for monitoring the running state of a photovoltaic power station in real time, and relates to the technical field of photovoltaic power station monitoring, and the method comprises the steps: integrating a multi-mode sensor array in a photovoltaic module junction box, and collecting the sensor data of the photovoltaic power station in real time; performing localization preprocessing on the sensor data through an edge computing node, and transmitting different priority data to a cloud based on a dynamic hybrid communication protocol; sensor data are fused at the cloud, a four-dimensional digital twinborn model is constructed, and time-space continuous meteorological prediction, component aging and dust retention evolution data are fused; recognizing the fault mode of the photovoltaic power station, and dynamically optimizing the string topological structure of the photovoltaic power station in combination with the real-time operation state of the photovoltaic power station, the equipment health data and the fault mode recognition result; by integrating the distributed sensor, the edge calculation module and the intelligent alarm mechanism, the problems of delay, insufficient intelligent analysis and response lag in the prior art can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power station monitoring, and particularly to a method and system for real-time monitoring of the operating state of a photovoltaic power station. Background Art

[0002] With the rapid development of renewable energy, photovoltaic power generation, as an important technology among them, is widely used in various scenarios. However, most of the existing photovoltaic power station monitoring systems rely on centralized data processing platforms, and most systems rely on data acquisition modules and data transmission networks at the photovoltaic power station site.

[0003] However, the existing photovoltaic power station monitoring systems have problems such as single perception dimension, lag in decision-making response, and insufficient adaptability. They mainly rely on electrical parameter monitoring and cannot capture physical state changes such as component microcracks and dust accumulation. The abnormal monitoring and early warning mechanisms often rely on preset rules and fail to fully utilize data mining technology to identify potential fault patterns from massive data. The fixed networking topology used is difficult to cope with dynamic scenarios such as shadow movement and component aging, and cannot automatically adjust the operation strategy according to real-time environmental changes or system load. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to propose a method and system for real-time monitoring of the operating state of a photovoltaic power station based on a distributed sensor network and edge computing, aiming to improve data transmission efficiency, enhance intelligent analysis capabilities, and achieve dynamic optimization management of the power station. By integrating distributed sensors, edge computing modules, and intelligent alarm mechanisms, the present invention can effectively solve the problems of delay, insufficient intelligent analysis, and lag in response in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solutions: Based on the above purpose, in a first aspect, the present invention provides a method for real-time monitoring of the operating state of a photovoltaic power station, including the following steps: Integrate a multi-modal sensor array in the junction box of the photovoltaic module to collect sensor data of the photovoltaic power station in real time; Perform local preprocessing on the sensor data through edge computing nodes deployed at the photovoltaic power station site, and transmit data with different priorities to the cloud based on a dynamic hybrid communication protocol; Fuse the sensor data in the cloud, construct a four-dimensional digital twin model, and fuse spatio-temporally continuous meteorological prediction, component aging, and dust accumulation evolution data; Adopt a transfer learning framework to learn historical fault data, identify the fault patterns of the photovoltaic power station, and dynamically optimize the string topology structure of the photovoltaic power station in combination with the real-time operating state, equipment health data, and fault pattern recognition results of the photovoltaic power station; Present the optimized string topology structure of the PV power station through a visualization interface, and real-time feedback the health status and control instructions of the power station. The control instructions are sent to the edge computing nodes on-site in real time.

[0006] As a further solution of the present invention, the sensor data of the PV power station is collected in real time, including electrical parameters, physical states, and environmental data. When performing local preprocessing on the sensor data, it includes anomaly detection, data compression, and priority marking. The data with different priorities is transmitted to the cloud through LoRa Mesh, 5G slicing, or satellite links.

[0007] As a further solution of the present invention, the multi-modal sensor array includes graphene quantum dot sensors, piezoelectric vibration sensors, and polarized light sensors, which synchronously detect the working voltage, leakage current, microcracks, dust accumulation thickness, surface temperature, polarization angle offset, and hidden crack depth of the component.

[0008] As a further solution of the present invention, the control instructions are sent to the edge computing nodes on-site in real time. The edge computing nodes perform the following steps: Run the lightweight ResNet-8 model to complete real-time fault classification within a preset time. Use sparse coding to compress the multi-dimensional sensing data, and divide the data priority according to the fault confidence. Among them, the emergency warning data is marked as the priority level, and the trigger data is directly transmitted to the cloud through LoRa Mesh, 5G slicing, or satellite links.

[0009] As a further solution of the present invention, based on the dynamic hybrid communication protocol, when transmitting data with different priorities to the cloud, the dynamic hybrid communication protocol satisfies: The coverage radius of the LoRa Mesh network ≥ 1 km, and the power consumption of a single node ≤ 0.3 W. The end-to-end delay of the 5G slicing network ≤ 8 ms, supporting 1000 terminal concurrent accesses per square kilometer. The satellite link adopts the dual-mode redundancy of Beidou-3 and Starlink, automatically switches when the communication is interrupted, and the switching delay ≤ 3 s.

[0010] As a further solution of the present invention, at the cloud, the sensor data is fused to construct a four-dimensional digital twin model, including the following steps: Establish a three-dimensional power station model according to the spatial dimension, integrating the physical parameters and electrical connection relationships of each component. Fuse the meteorological forecast data of a preset time period according to the time dimension, including irradiance fluctuations, wind speed changes, and precipitation probability. According to the state dimension, calculate the component aging index through EL image degradation analysis and predict the remaining life. According to the environmental dimension, a dust accumulation growth model is constructed based on the PM2.5 concentration and rainfall frequency to predict the cleaning cycle.

[0011] As a further solution of the present invention, when using a transfer learning framework to learn historical fault data and identify the fault modes of a photovoltaic power station, the implementation of the transfer learning framework includes: Construct a feature vector knowledge base containing several historical fault modes; When an unknown anomaly is detected, zero-shot detection is achieved by matching topological similarity through a graph convolutional network (GCN); Aggregate the data of each power station through federated learning every week to update the global model parameters while retaining local privacy data.

[0012] As a further solution of the present invention, dynamically optimizing the string topology structure of a photovoltaic power station includes the following steps: Real-time scan the irradiance distribution map of the power station to identify the areas where the power generation efficiency is lower than the power generation threshold; Switch the components in the shaded area to an independent MPPT channel through a matrix switch array to form a dynamic sub-array; Achieve Pareto optimality among power generation efficiency, heat loss, and equipment life based on deep deterministic policy gradient; Recalculate the optimization strategy every 60 minutes, and the network reconstruction response time ≤ 30s.

[0013] As a further solution of the present invention, the method for real-time monitoring of the operating state of a photovoltaic power station further includes a self-calibration mechanism: Eliminate temperature drift errors through the mutual verification algorithm of adjacent sensor data every 20ms; Perform offline calibration every day at dawn, and use a laser interferometer to correct the reference value for microcrack detection; When the dust thickness > 3g / m², automatically trigger a reservation work order for the drone cleaning system.

[0014] In a second aspect, the present invention also provides a system for real-time monitoring of the operating state of a photovoltaic power station, including the following components: A data acquisition module for real-time collecting the sensor data of the photovoltaic power station by integrating a multi-modal sensor array in the junction box of the photovoltaic module; An edge computing module for locally preprocessing the sensor data through edge computing nodes deployed at the photovoltaic power station site and transmitting data with different priorities to the cloud based on a dynamic hybrid communication protocol; A model construction module for fusing sensor data in the cloud to construct a four-dimensional digital twin model, fusing spatio-temporally continuous meteorological prediction, component aging, and dust accumulation evolution data; A fault identification module, which is used to learn from historical fault data by adopting a transfer learning framework, identify the fault modes of the photovoltaic power station, and dynamically optimize the string topology structure of the photovoltaic power station in combination with the real-time operation status, equipment health data and fault mode identification results of the photovoltaic power station; A visualization display module, which is used to present the optimized string topology structure of the photovoltaic power station through a visualization interface, and real-time feedback the health status and control instructions of the power station, and the control instructions are sent to the edge computing nodes on-site in real time.

[0015] As a further solution of the present invention, the model construction module further includes a cloud data processing platform that receives data from edge computing nodes, fuses sensor data and constructs a four-dimensional digital twin model. The cloud data processing platform includes the following functional modules: A spatial dimension modeling module: establish a three-dimensional power station model, and integrate the physical parameters and electrical connection relationships of each component; A time dimension modeling module: fuse meteorological forecast data, including irradiance fluctuations, wind speed changes and precipitation probability; A state dimension modeling module: calculate the component aging index through EL image degradation analysis and predict the remaining life; An environment dimension modeling module: construct a dust accumulation growth model based on the PM2.5 concentration and rainfall frequency to predict the cleaning cycle.

[0016] As a further solution of the present invention, the fault identification module includes a string topology optimization unit that dynamically optimizes the string topology structure of the photovoltaic power station according to the real-time operation status, equipment health data and fault mode identification results of the power station. The string topology optimization unit includes: Scan the irradiance distribution map of the power station and identify the areas where the power generation efficiency is lower than the power generation threshold; Switch the components in the shadow area to an independent MPPT channel through a matrix switch array to form a dynamic sub-array; Based on the Deep Deterministic Policy Gradient (DDPG) algorithm, achieve Pareto optimality among power generation efficiency, heat loss and equipment life; Recalculate the optimization strategy every 60 minutes, and the networking reconstruction response time ≤ 30s.

[0017] As a further solution of the present invention, the system for real-time monitoring of the operation status of the photovoltaic power station further includes a self-calibration mechanism, and the self-calibration mechanism performs the following operations: Eliminate the temperature drift error every 20ms through the cross-check algorithm of adjacent sensor data; Perform offline calibration every day at midnight, and use a laser interferometer to correct the reference value of micro-crack detection; When the dust accumulation thickness > 3g / m², automatically trigger a reservation work order for the drone cleaning system.

[0018] Compared with the prior art, a method and system for real-time monitoring of the operating state of a photovoltaic power station proposed by the present invention have the following beneficial effects: 1. The present invention can monitor the electrical parameters, physical states, and environmental data of a photovoltaic power station in real time through an integrated multi-modal sensor array, such as voltage, leakage current, microcracks, dust thickness, surface temperature, polarization angle offset, and hidden crack depth. Through multi-dimensional data acquisition, it provides comprehensive monitoring of the operating state of the photovoltaic power station, which helps to accurately identify potential faults or anomalies. By deploying edge computing nodes to perform local data preprocessing and real-time fault classification on-site, it can not only effectively reduce data transmission delay but also avoid bandwidth pressure caused by uploading large-scale data. The edge nodes perform hierarchical transmission according to the priority of the data to ensure that key data (such as emergency alarm data) can be quickly transmitted to the cloud, improving the fault response speed and enhancing the intelligence and reliability of the system.

[0019] 2. The present invention can identify the historical fault modes of a photovoltaic power station and achieve zero-sample detection. Through intelligent identification of fault modes, it can predict potential faults based on historical data and operating states, issue early warnings, reduce the occurrence of sudden faults, and ensure the long-term stable operation of the power station. Based on real-time power station operation data and fault mode identification results, the present invention can dynamically optimize the string topology structure of the power station, improving the power generation efficiency of the photovoltaic power station. By using deep deterministic policy gradients, it can find the optimal balance among power generation efficiency, heat loss, and equipment life, achieving a Pareto-optimal string topology configuration, thereby further improving the overall economic benefits of the photovoltaic power station.

[0020] 3. The present invention can also provide a visual interface with real-time feedback, intuitively presenting the optimized string topology structure, operating health status, and real-time control instructions of the photovoltaic power station, facilitating users to easily view the power station status and issue real-time control instructions, improving the operability and management convenience of the system. By real-time monitoring the operation data of the photovoltaic power station, it triggers an alarm in a timely manner when a fault or anomaly occurs, and ensures that the alarm information can be quickly transmitted to relevant personnel or systems through priority sorting, ensuring that the photovoltaic power station can be processed as early as possible when a fault occurs, reducing the impact of the fault on the power station operation, reducing system downtime, and improving the reliability of the power station.

[0021] 4. The string topology optimization and fault mode recognition mechanism of the present invention can not only improve the power generation efficiency of a photovoltaic power station, but also reduce the thermal loss of equipment and extend the service life of equipment. Through intelligent optimization and management, while improving the power generation benefit of the photovoltaic power station, the system reduces the loss of equipment, thereby extending the service life of the power station equipment and having high long-term economic value; combined with edge computing and cloud big data processing, it can realize real-time data processing and control decision-making at the site of the photovoltaic power station, and the self-calibration mechanism and fault recognition module of the system enable the power station to automatically optimize the system and give early warnings of faults without human intervention, greatly reducing the labor cost and improving the intelligent level of system operation.

[0022] In summary, the method and system for real-time monitoring of the operation status of a photovoltaic power station provided by the present invention can efficiently, accurately and intelligently monitor the operation status of the power station, predict and prevent faults, optimize the operation configuration of the power station, ensure the stable and efficient operation of the photovoltaic power station, and improve the economic and environmental benefits of photovoltaic power generation.

[0023] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. It should be understood that the above general description and the following detailed description are only exemplary and explanatory and cannot limit the present application. Brief Description of the Drawings

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the following will briefly introduce the drawings required for use in the description of the exemplary embodiments or related technologies. The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings: Figure 1 is a flowchart of a method for real-time monitoring of the operation status of a photovoltaic power station according to an embodiment of the present invention.

[0025] Figure 2 is a flowchart of constructing a four-dimensional digital twin model in a method for real-time monitoring of the operation status of a photovoltaic power station according to an embodiment of the present invention.

[0026] Figure 3 is a flowchart of the implementation of the transfer learning framework in a method for real-time monitoring of the operation status of a photovoltaic power station according to an embodiment of the present invention.

[0027] Figure 4 is a flowchart of dynamically optimizing the string topology structure of a photovoltaic power station in a method for real-time monitoring of the operation status of a photovoltaic power station according to an embodiment of the present invention.

[0028] Figure 5 is a flowchart of the execution of an edge computing node in a method for real-time monitoring of the operation status of a photovoltaic power station according to an embodiment of the present invention. Detailed implementation manners

[0029] Next, in combination with the accompanying drawings and specific implementation manners, the present application will be further described. It should be noted that, on the premise of no conflict, the following-described embodiments or technical features can be arbitrarily combined to form new embodiments with each other.

[0030] For the purpose, technical solutions and advantages of the present invention to be more clearly understood, the following further describes the embodiments of the present invention in detail with reference to specific embodiments and the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application.

[0031] It should be noted that all the expressions using "first" and "second" in the embodiments of the present invention are used to distinguish two non-identical entities or non-identical parameters with the same name. It can be seen that "first" and "second" are only for the convenience of expression and should not be construed as a limitation on the embodiments of the present invention. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units inherently includes other steps or units.

[0032] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in combination with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0033] The flowcharts shown in the accompanying drawings are only illustrative examples, not necessarily including all contents and operations / steps, nor necessarily executed in the described order. For example, some operations / steps can also be decomposed, combined or partially merged. Therefore, the actual execution order may be changed according to the actual situation.

[0034] Next, in combination with the accompanying drawings, some implementation manners of the present application will be described in detail. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0035] Aiming at the problems of delay, insufficient intelligent analysis and lagging response existing in the existing photovoltaic power station monitoring system, the present invention proposes a method for real-time monitoring of the operation status of a photovoltaic power station, aiming to improve the data transmission efficiency, enhance the intelligent analysis ability, and realize the dynamic optimization management of the power station. The above problems are solved by integrating distributed sensors, edge computing modules and intelligent warning mechanisms.

[0036] See Figure 1As shown in the figure, an embodiment of the present invention provides a method for real-time monitoring of the operating status of a photovoltaic power station, and the method includes the following steps: Step S10: Integrate a multi-modal sensor array in the junction box of the photovoltaic module to collect the sensor data of the photovoltaic power station in real time.

[0037] In this step, collecting the sensor data of the photovoltaic power station in real time includes electrical parameters, physical states, and environmental data.

[0038] Among them, the multi-modal sensor array includes a graphene quantum dot sensor, a piezoelectric vibration sensor, and a polarized light sensor, which synchronously detect the working voltage, leakage current, micro-cracks, dust accumulation thickness, surface temperature, polarization angle offset, and hidden crack depth of the component. Specifically, in the integrated sensor array of the photovoltaic module junction box, the functions of each sensor are as follows: Graphene quantum dot sensor: Used to detect the electrical parameters of the photovoltaic module with high precision, including current, voltage, etc.

[0039] Piezoelectric vibration sensor: Detect physical states such as micro-cracks, component structure vibration, and mechanical stress.

[0040] Polarized light sensor: Used to detect the physical properties of the surface of the photovoltaic module, such as dust accumulation thickness, surface temperature, and polarization angle offset.

[0041] When synchronously detecting the working state of the photovoltaic module, the working voltage and leakage current, micro-cracks, dust accumulation thickness, surface temperature, polarization angle offset, and hidden crack depth of the component are synchronously detected through the graphene quantum dot sensor, piezoelectric vibration sensor, and polarized light sensor in the above multi-modal sensor array. Among them, micro-cracks are used to judge whether the component is damaged due to external stress, and the dust accumulation thickness affects the light absorption efficiency of the component; the surface temperature is used to detect the overheating situation of the component; the polarization angle offset is used to detect the influence of external environmental factors on the photovoltaic module; the hidden crack depth is used to evaluate the health status of the component by detecting the depth of micro-cracks. The sensor array provides multi-dimensional data on component health, power generation efficiency, and environmental impact for power station management personnel by real-time monitoring the above parameters of the photovoltaic module, ensuring that potential faults or performance degradation can be detected in a timely manner. For example: when the voltage of the photovoltaic module is 30V during normal operation, if the voltage of a certain component is lower than the preset threshold, the system will judge that the component may have a fault and issue an alarm.

[0042] Step S20: Locally preprocess the sensor data through the edge computing nodes deployed on-site in the photovoltaic power station, and transmit data with different priorities to the cloud based on the dynamic hybrid communication protocol.

[0043] In this step, when performing local preprocessing on sensor data, it includes anomaly detection, data compression, and priority marking; different priority data is transmitted to the cloud via LoRa Mesh, 5G slices, or satellite links. Among them, during anomaly detection, the edge node determines whether the collected electrical parameters and physical states are abnormal through a model. For example, an alarm is triggered when the voltage fluctuation exceeds the threshold; during data compression, since the amount of data generated by the sensor is huge, the edge computing node compresses the data, only retains important information, reduces the amount of data transmission, and reduces bandwidth consumption; during priority marking, by marking different data with priorities, the edge node can ensure that emergency data (such as device failure alerts) is uploaded to the cloud first to ensure timely response.

[0044] In this embodiment, when transmitting different priority data to the cloud based on the dynamic hybrid communication protocol, the dynamic hybrid communication protocol satisfies: The coverage radius of the LoRa Mesh network ≥ 1 km, and the power consumption of a single node ≤ 0.3 W; The end-to-end delay of the 5G slice network ≤ 8 ms, and it supports 1000 terminals to access concurrently per square kilometer; The satellite link adopts the dual-mode redundancy of Beidou-3 and Starlink, and automatically switches when the communication is interrupted, and the switching delay ≤ 3 s.

[0045] In this embodiment, local processing and preliminary analysis of data are achieved through the edge computing node, which can reduce the amount of data transmission and ensure the priority transmission of key information. Using the dynamic hybrid communication protocol, the system can select an appropriate transmission method according to the priority of different data, thereby optimizing the use of network resources. For example: when a device fails, the edge node detects the fault data and immediately marks it as high priority, and uploads this data to the cloud through the 5G slice network. Other data (such as environmental temperature) is uploaded regularly through the LoRa Mesh network. The edge computing node can perform real-time data preprocessing on-site, reduce the data transmission load, improve the response speed, ensure that the fault can be quickly processed and alarmed when it occurs, adopt the dynamic hybrid communication protocol, select an appropriate transmission method according to the priority of the data, maximize the communication efficiency, ensure that important data can be quickly transmitted to the cloud, and through the combination of LoRa Mesh, 5G slices, and satellite links, even in an unstable network environment, the system can still ensure stable data transmission, improving the reliability and stability of the system.

[0046] Step S30: Fuse the sensor data in the cloud to construct a four-dimensional digital twin model, and fuse the meteorological prediction, component aging, and dust accumulation evolution data that are spatio-temporally continuous.

[0047] In this step, refer to Figure 2As shown in the figure, sensor data is fused in the cloud to build a four-dimensional digital twin model, including the following steps: Step S301: Establish a three-dimensional power station model according to the spatial dimension, and integrate the physical parameters and electrical connection relationships of each component.

[0048] In this embodiment, a three-dimensional model of the power station is created in the cloud, which includes the physical locations and spatial distributions of all components in the photovoltaic power station; the physical parameters (dimensions, models, installation angles, and directions) of each photovoltaic component are data modeled according to the actual installation situation; through the electrical connection relationships, the electrical circuits and transmission paths between the power station components are established to simulate the electrical connections and data flows of each component inside the power station. The three-dimensional power station model provides a comprehensive and three-dimensional view of the power station, which helps to understand the physical layout of the power station and the electrical interconnection relationships between components. For example: Taking a certain photovoltaic power station as an example, the power station contains 100 photovoltaic components, and the installation angles and directions of each component are different. The three-dimensional model draws the accurate positions of the components by obtaining the actual data of the components and marks the electrical connection relationships between each component and the power station busbar box. Through this model, the electrical performance and spatial distribution of each component can be monitored in real time.

[0049] Step S302: Fuse the meteorological forecast data for a preset time period according to the time dimension, including irradiance fluctuations, wind speed changes, and precipitation probability.

[0050] In this embodiment, the meteorological data is integrated into the digital twin model. By cooperating with a meteorological service provider, the meteorological data for a preset time period is obtained, including: Irradiance fluctuations: The expected radiation intensity per hour, which affects the power generation efficiency of photovoltaic components.

[0051] Wind speed changes: The changes in wind speed have an impact on the heat dissipation performance and wind damage of photovoltaic components.

[0052] Precipitation probability: Precipitation affects the cleanliness and power generation efficiency of photovoltaic components.

[0053] Among them, the meteorological data in the time dimension can help predict the impact of different weather conditions on the power station. For example, if the wind speed is high during a certain time period, the system can predict possible wind damage or insufficient heat dissipation of the photovoltaic components, and then make necessary warnings or maintenance arrangements.

[0054] Exemplarily, assume that the meteorological data within the next 24 hours is: the irradiance fluctuation range is 800 - 1200 W / m², the wind speed change range is 5 - 15 m / s, and the precipitation probability is 20%. By integrating these data, the system can predict the power generation situation in the next 24 hours and timely adjust the cleaning plan or alarm the areas with high wind speed.

[0055] Step S303: Calculate the component aging index according to the status dimension through EL image degradation analysis to predict the remaining life.

[0056] In this embodiment, by obtaining the electroluminescence (EL) image of the photovoltaic module, the degradation state of the module is analyzed; image processing technology is used to perform degradation analysis on the electroluminescence (EL) image to detect problems such as microcracks, hot spots, and defects inside the module; according to these degradation indicators, the aging index of the module is calculated, and combined with historical data and degradation trends, the remaining life is predicted. Among them, EL image degradation analysis can identify degradation phenomena such as microcracks and hot spots by analyzing the electro-optical response of the photovoltaic module. Through this degradation information, combined with the design life and operating status of the module, the system can provide an accurate prediction of the remaining service life.

[0057] For example: If the EL image analysis of the photovoltaic module shows the existence of multiple microcracks, after degradation analysis, the system calculates that the aging index of this module is 0.85, and its remaining life is predicted to be 5 years. Based on this information, the operation and maintenance personnel can plan the maintenance or replacement work of the module in advance.

[0058] Step S304: According to the environmental dimension, construct a dust accumulation growth model based on the PM2.5 concentration and rainfall frequency to predict the cleaning cycle.

[0059] In this embodiment, by monitoring the PM2.5 concentration and rainfall frequency in the air, a dust accumulation growth model is constructed to estimate the impact of dust accumulation on the performance of the photovoltaic module. Among them, the PM2.5 concentration directly affects the surface dust accumulation of the photovoltaic module, and a higher PM2.5 concentration means more dust accumulation; the precipitation frequency affects the cleaning cycle of the dust accumulation. Precipitation can help remove some dust and reduce the cleaning frequency. Through the above data, the system can calculate the optimal cleaning time window. Combining the environmental dimension data with the dust accumulation growth model, through the prediction of the surface dust accumulation of the photovoltaic module, the system can dynamically adjust the cleaning strategy. For example, when the PM2.5 concentration rises, the system can adjust the cleaning plan in advance.

[0060] Exemplarily, if the monitored PM2.5 concentration is 150 μg / m³ and the rainfall probability is low, the system predicts that the dust accumulation on the photovoltaic modules in this area will reach the preset cleaning standard and recommends cleaning next month. If the precipitation is frequent, the cleaning plan can be postponed to reduce the maintenance cost.

[0061] In the embodiment of the present invention, by fusing sensor data in the cloud, a four-dimensional digital twin model is constructed to make the operation of the photovoltaic power station more intelligent and automated. By establishing a three-dimensional power station model, the physical structure and electrical connection relationship of the photovoltaic power station are comprehensively described. By integrating meteorological forecast data, the system can predict the impact of weather on the power generation efficiency and safety of the power station and adjust the operation strategy in a timely manner. Based on the EL image degradation analysis, the system can accurately predict the aging degree and remaining life of the components, which is helpful for the long-term operation and maintenance management of the power station. Through the dust accumulation growth model, combined with environmental data, the cleaning cycle is predicted to optimize the cleaning plan of the power station and reduce unnecessary maintenance expenses, providing a more efficient and intelligent operation and maintenance management solution for the photovoltaic power station, which can significantly improve the operation efficiency and reliability of the power station.

[0062] Step S40: Use a transfer learning framework to learn historical fault data, identify the fault modes of the photovoltaic power station, and dynamically optimize the string topology structure of the photovoltaic power station in combination with the real-time operation status, equipment health data, and fault mode identification results of the photovoltaic power station.

[0063] In this step, as shown in Figure 3 When using a transfer learning framework to learn historical fault data and identify the fault modes of the photovoltaic power station, the implementation of the transfer learning framework includes: Step S401: Construct a feature vector knowledge base containing several historical fault modes.

[0064] In this embodiment, by collecting a large amount of historical fault data, including different types of fault modes (such as short circuit, poor contact, component aging, system failure, etc.), feature vectors are extracted for each fault mode. The features include abnormal fluctuations in real-time monitoring data such as voltage, current, temperature, and irradiance. Then, the fault data is preprocessed to generate feature vectors of different fault modes and construct a knowledge base. Among them, the feature vector of each fault mode represents the change trend of different parameters (such as temperature, power, etc.) under this mode. By constructing a fault mode feature library, it is possible to quickly match historical fault modes when an anomaly is detected, improving the accuracy of fault diagnosis.

[0065] Step S402: When an unknown anomaly is detected, match the topological similarity through a Graph Convolutional Network (GCN) to achieve zero-shot detection.

[0066] Among them, when an unknown anomaly occurs in the power station, the current state of the power station is first captured in real time through sensor data and the monitoring system. Then, a graph convolutional network (GCN) is used to model the topological structure of the power station, and the current state is matched with historical fault patterns to find potential fault patterns. By learning the topological structure between power station components, the GCN model can effectively associate anomalies with fault patterns, especially in the case of insufficient labeled data, and still perform zero-shot detection.

[0067] For example: when the power station detects abnormal current fluctuations in a certain component, the GCN model matches according to the topological position of the component (for example, which other components it is connected to) and historical fault data, and finds that this is a zero-shot detection result of a "poor contact" fault pattern, and promptly triggers a fault alarm.

[0068] Step S403: Aggregate the data of each power station through federated learning every week to update the global model parameters while retaining local privacy data.

[0069] In this embodiment, the photovoltaic power station adopts a federated learning mechanism, which allows different power stations to perform data training locally and transmit the updated model parameters to the cloud for aggregation; federated learning data aggregation is performed regularly every week to ensure the continuous optimization of the models of each power station while protecting local data privacy and without the need to upload sensitive data to the cloud. The federated learning framework can, on the premise of ensuring data privacy, realize the update and optimization of the model by sharing and aggregating local model parameters, and improve the accuracy and robustness of the global model.

[0070] For example: 5 photovoltaic power stations participate in federated learning. Each power station trains a fault diagnosis model using local sensor data and uploads the parameters to the central server for aggregation. The server optimizes the global fault detection model through the aggregated model parameters, thereby improving the ability to identify unknown fault patterns.

[0071] In this embodiment, refer to Figure 4 As shown, dynamically optimize the string topology of the photovoltaic power station, including the following steps: Step S411: Real-time scan the irradiance distribution map of the power station and identify the areas where the power generation efficiency is lower than the power generation threshold.

[0072] Among them, the irradiance distribution map of the power station is monitored in real time through the sensors of the photovoltaic power station. By using image processing or sensor data analysis technology, the areas with power generation efficiency lower than the preset threshold are identified. The power generation efficiency in these areas decreases due to shading, faults or component damage. Through the irradiance distribution map, the power generation situation of each area of the power station can be monitored in real time. When the power generation efficiency of a certain area is lower than the threshold, the system can quickly identify and give a fault warning. For example, if the sensors of a power station detect that the irradiance in a certain area is low, resulting in the power generation efficiency of this area being only 70%, which is lower than the set threshold of 90%. After the system identifies this anomaly, it automatically triggers a maintenance plan to check whether there are faults or shading in the photovoltaic components in this area.

[0073] Step S412: Switch the components in the shadow area to an independent MPPT channel through the matrix switch array to form a dynamic sub-array.

[0074] Among them, for the area with uneven irradiance distribution, the matrix switch array technology is used to switch the components in this area to independent Maximum Power Point Tracking (MPPT) channels, which can avoid the influence of the shadow area on the power generation efficiency of the entire power station and maximize the power generation performance of each area. Through the matrix switch array, the power station can dynamically switch the components in different areas to independent MPPT channels to ensure that the components in the shadow area do not affect the power generation efficiency of the entire power station.

[0075] Step S413: Achieve Pareto optimality among power generation efficiency, heat loss and equipment life based on deep deterministic policy gradient.

[0076] Step S414: Recalculate the optimization strategy every 60 minutes, and the network reconstruction response time ≤ 30s.

[0077] Among them, by recalculating and adjusting the optimization strategy of the power station by the system every 60 minutes, it can ensure that the topological structure of the power station is optimized according to the latest environmental changes, equipment status and power generation data. The response time requirement for network reconstruction does not exceed 30 seconds to ensure that the power station can adjust and adapt to external condition changes in a timely manner.

[0078] For example, when there are external environmental changes (such as cloudy weather) during the operation of the power station, the system will recalculate and optimize the topological structure every 60 minutes to ensure that the reconstruction response is completed within 30 seconds and continue to maintain the best power generation efficiency.

[0079] The present invention can identify the historical fault modes of a photovoltaic power station and achieve zero-sample detection. Through the intelligent identification of fault modes, it can predict potential faults based on historical data and operating status, give early warnings, reduce the occurrence of sudden faults, and ensure the long-term stable operation of the power station. Based on the real-time power station operation data and the fault mode recognition results, the present invention can dynamically optimize the string topology of the power station, improving the power generation efficiency of the photovoltaic power station. By using deep deterministic policy gradients, it can find the optimal balance among power generation efficiency, heat loss, and equipment life, achieving a Pareto-optimal string topology configuration, thereby further improving the overall economic benefits of the photovoltaic power station.

[0080] Step S50: Present the optimized string topology of the photovoltaic power station through a visualization interface, and real-time feedback the power station health status and control instructions. The control instructions are sent to the on-site edge computing nodes in real time.

[0081] In this step, as shown in Figure 5 shown, the control instructions are sent to the on-site edge computing nodes in real time, and the edge computing nodes perform the following steps: Step S501: Run the lightweight ResNet-8 model to complete real-time fault classification within a preset time; Step S502: Compress the multi-dimensional sensing data using sparse coding and divide the data priority according to the fault confidence. Among them, the emergency alarm data is marked as the priority level, and the trigger data is directly transmitted to the cloud through LoRa Mesh, 5G slicing, or satellite links.

[0082] The present invention can also provide a visualization interface with real-time feedback, intuitively presenting the optimized string topology of the photovoltaic power station, the operating health status, and real-time control instructions, facilitating users to easily view the power station status and send control instructions in real time, improving the operability and management convenience of the system. By real-time monitoring the operation data of the photovoltaic power station, alarms are triggered in a timely manner when a fault or anomaly occurs, and through priority sorting, it is ensured that the alarm information can be quickly transmitted to relevant personnel or systems, ensuring that the photovoltaic power station can be processed as early as possible when a fault occurs, reducing the impact of the fault on the power station operation, reducing the system downtime, and improving the reliability of the power station.

[0083] In some embodiments, the method for real-time monitoring of the operating status of the photovoltaic power station further includes a self-calibration mechanism: Every 20 ms, eliminate the temperature drift error through the cross-check algorithm of adjacent sensor data; Perform offline calibration at 0:00 every day, and use a laser interferometer to correct the reference value of micro-crack detection; When the dust accumulation thickness > 3 g / m², automatically trigger a reservation work order for the drone cleaning system.

[0084] The method for real-time monitoring of the operating status of a photovoltaic power station according to the present invention can, by integrating a multi-modal sensor array, monitor in real time the electrical parameters, physical states, and environmental data of the photovoltaic power station, such as voltage, leakage current, microcracks, dust accumulation thickness, surface temperature, polarization angle offset, and hidden crack depth. Through multi-dimensional data acquisition, it provides comprehensive monitoring of the operating status of the photovoltaic power station, which helps to accurately identify potential faults or anomalies. By deploying edge computing nodes, local data preprocessing and real-time fault classification are carried out on-site, which can not only effectively reduce data transmission delay, but also avoid the bandwidth pressure caused by uploading large-scale data. The edge nodes perform hierarchical transmission according to the priority of the data to ensure that critical data (such as emergency alarm data) can be quickly transmitted to the cloud, improving the fault response speed and enhancing the intelligence and reliability of the system.

[0085] It should be noted that the above-mentioned drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present invention, rather than for limiting purposes. It is easy to understand that the processes shown in the above-mentioned drawings do not indicate or limit the time sequence of these processes. Additionally, it is also easy to understand that these processes can be executed synchronously or asynchronously, for example, in multiple modules.

[0086] It should be understood that although the above is described in a certain order, these steps are not necessarily executed in the above order successively. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, a part of the steps of this embodiment may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same moment, but can be executed at different moments. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0087] An embodiment of the present invention also provides a system for real-time monitoring of the operating status of a photovoltaic power station, which is used to execute the aforementioned method for real-time monitoring of the operating status of a photovoltaic power station. The system for real-time monitoring of the operating status of a photovoltaic power station includes: A data acquisition module, which is used to integrate a multi-modal sensor array in the junction box of a photovoltaic module to collect in real time the sensor data of the photovoltaic power station; An edge computing module, which is used to perform local preprocessing on the sensor data through edge computing nodes deployed at the photovoltaic power station site, and transmit data of different priorities to the cloud based on a dynamic hybrid communication protocol; A model construction module, which is used to fuse the sensor data in the cloud to construct a four-dimensional digital twin model, and fuse meteorological prediction, component aging, and dust accumulation evolution data that are spatio-temporally continuous; A fault identification module, which is used to learn from historical fault data using a transfer learning framework, identify the fault modes of a photovoltaic power station, and dynamically optimize the string topology structure of the photovoltaic power station by combining the real-time operation status, equipment health data, and fault mode identification results of the photovoltaic power station; A visualization display module, which is used to present the string topology structure of the optimized photovoltaic power station through a visualization interface, and real-time feedback the health status and control instructions of the power station. The control instructions are sent to the edge computing nodes on-site in real time.

[0088] Among them, the model construction module also includes a cloud data processing platform that receives data from edge computing nodes, fuses sensor data, and constructs a four-dimensional digital twin model. The cloud data processing platform includes the following functional modules: A spatial dimension modeling module: establishing a three-dimensional power station model and integrating the physical parameters and electrical connection relationships of each component; A time dimension modeling module: fusing meteorological forecast data, including irradiance fluctuations, wind speed changes, and precipitation probability; A state dimension modeling module: calculating the component aging index through EL image degradation analysis and predicting the remaining life; An environment dimension modeling module: constructing a dust accumulation growth model based on the PM2.5 concentration and rainfall frequency to predict the cleaning cycle.

[0089] Among them, the fault identification module includes a string topology optimization unit that dynamically optimizes the string topology structure of a photovoltaic power station according to the real-time operation status, equipment health data, and fault mode identification results of the power station. The string topology optimization unit includes: Scanning the irradiance distribution map of the power station to identify areas where the power generation efficiency is lower than the power generation threshold; Switching the components in the shaded area to an independent MPPT channel through a matrix switch array to form a dynamic sub-array; Based on the Deep Deterministic Policy Gradient (DDPG) algorithm, achieving Pareto optimality among power generation efficiency, heat loss, and equipment life; Recalculating the optimization strategy every 60 minutes, and the network reconstruction response time ≤ 30s.

[0090] Among them, the system for real-time monitoring of the operation status of a photovoltaic power station also includes a self-calibration mechanism, and the self-calibration mechanism performs the following operations: Eliminating temperature drift errors through a mutual verification algorithm of adjacent sensor data every 20ms; Performing offline calibration at 0:00 every day, and using a laser interferometer to correct the reference value for micro-crack detection; When the dust accumulation thickness > 3g / m², automatically triggering a reservation work order for the drone cleaning system.

[0091] The series string topology optimization and fault mode recognition mechanism of the present invention can not only improve the power generation efficiency of a photovoltaic power station, but also reduce the thermal loss of equipment and extend the service life of the equipment. Through intelligent optimization and management, while enhancing the power generation benefit of the photovoltaic power station, the system reduces the loss of equipment, thereby extending the service life of the power station equipment, and has high long-term economic value; combined with edge computing and cloud big data processing, it can realize real-time data processing and control decision-making at the site of the photovoltaic power station, and the self-calibration mechanism and fault recognition module of the system enable the power station to automatically optimize the system and give early warnings of faults without human intervention, greatly reducing the labor cost and improving the intelligent level of system operation.

[0092] Through the above detailed steps, the system for real-time monitoring of the operation status of a photovoltaic power station according to the present invention is used to execute the steps of the method for real-time monitoring of the operation status of a photovoltaic power station in the above embodiments, which will not be elaborated here. The method and system for real-time monitoring of the operation status of a photovoltaic power station provided by the present invention can efficiently, accurately and intelligently monitor the operation status of the power station, predict and prevent faults, optimize the operation configuration of the power station, ensure the stable and efficient operation of the photovoltaic power station, and enhance the economic and environmental benefits of photovoltaic power generation.

[0093] The above are exemplary embodiments disclosed by the present invention. However, it should be noted that various changes and modifications can be made without departing from the scope of the embodiments disclosed by the present invention as defined by the claims. The functions, steps, and / or actions of the method claims according to the disclosed embodiments herein do not need to be executed in any specific order. In addition, although the elements disclosed by the embodiments of the present invention can be described or claimed in individual form, they can also be understood as plural unless clearly limited to the singular.

[0094] It should be understood that, as used herein, unless the context clearly supports an exception, the singular form "a" is also intended to include the plural form. It should also be understood that the "and / or" used herein refers to any and all possible combinations of one or more of the related listed items. The above serial numbers of the disclosed embodiments of the present invention are only for description and do not represent the superiority or inferiority of the embodiments.

[0095] Those of ordinary skill in the art should understand that: the discussion of any of the above embodiments is only exemplary, and is not intended to imply that the scope of the embodiments disclosed by the present invention (including the claims) is limited to these examples; under the concept of the embodiments of the present invention, the technical features between the above embodiments or different embodiments can also be combined, and there are many other variations in different aspects of the above embodiments of the present invention, which are not provided in detail for the sake of brevity. Therefore, any omission, modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the present invention shall be included in the protection scope of the embodiments of the present invention.

Claims

1. A method for real-time monitoring of the operating status of a photovoltaic power station, characterized in that, The method includes the following steps: Integrate a multi-modal sensor array in the junction box of the photovoltaic module to collect sensor data of the photovoltaic power station in real time; Perform local preprocessing on the sensor data through edge computing nodes deployed on-site in the photovoltaic power station, and transmit data with different priorities to the cloud based on a dynamic hybrid communication protocol; Fuse the sensor data in the cloud, build a four-dimensional digital twin model, and fuse meteorological predictions, component aging, and dust accumulation evolution data with continuous space-time; Use a transfer learning framework to learn historical fault data, identify the fault modes of the photovoltaic power station, and dynamically optimize the string topology of the photovoltaic power station in combination with the real-time operating status, equipment health data, and fault mode identification results of the photovoltaic power station; Present the optimized string topology of the photovoltaic power station through a visualization interface, and real-time feedback the health status and control instructions of the power station. The control instructions are sent to the on-site edge computing nodes in real time.

2. The method for real-time monitoring of the operating state of a photovoltaic power station according to claim 1, wherein Collecting sensor data of the photovoltaic power station in real time includes electrical parameters, physical states, and environmental data; when performing local preprocessing on the sensor data, it includes anomaly detection, data compression, and priority marking; transmit data with different priorities to the cloud through LoRa Mesh, 5G slices, or satellite links.

3. The method for real-time monitoring of the operation status of a photovoltaic power station according to claim 2, wherein, The multi-modal sensor array includes graphene quantum dot sensors, piezoelectric vibration sensors, and polarized light sensors, which synchronously detect the working voltage, leakage current, microcracks, dust accumulation thickness, surface temperature, polarization angle offset, and hidden crack depth of the components.

4. The method for real-time monitoring of the operation status of a photovoltaic power station according to claim 1, wherein, The control instructions are sent to the on-site edge computing nodes in real time, and the edge computing nodes perform the following steps: Run a lightweight ResNet-8 model to complete real-time fault classification within a preset time; Use sparse coding to compress multi-dimensional sensing data, and divide data priorities according to fault confidence levels. Among them, mark the emergency warning data as the priority level, and trigger the direct transmission of data to the cloud through LoRa Mesh, 5G slices, or satellite links.

5. The method for real-time monitoring of the operating status of a photovoltaic power station according to claim 4, wherein, When transmitting data with different priorities to the cloud based on the dynamic hybrid communication protocol, the dynamic hybrid communication protocol satisfies: The coverage radius of the LoRa Mesh network ≥ 1 km, and the power consumption of a single node ≤ 0.3 W; The end-to-end delay of the 5G slice network ≤ 8 ms, supporting 1000 terminal concurrent accesses per square kilometer; The satellite link uses the Beidou-3 and Starlink dual-mode redundancy, automatically switches when the communication is interrupted, and the switching delay ≤ 3 s.

6. The method for real-time monitoring of the operation status of a photovoltaic power station according to claim 5, wherein, Fuse the sensor data in the cloud and build a four-dimensional digital twin model, including the following steps: Establish a three-dimensional power station model according to the spatial dimension, integrating the physical parameters and electrical connection relationships of each component; Fuse meteorological forecast data for a preset time period according to the time dimension, including irradiance fluctuations, wind speed changes, and precipitation probability; According to the state dimension, calculate the component aging index through EL image degradation analysis and predict the remaining life; According to the environmental dimension, build a dust accumulation growth model based on the PM2.5 concentration and rainfall frequency to predict the cleaning cycle.

7. The method for real-time monitoring of the operating state of a photovoltaic power station according to claim 6, characterized in that, When using a transfer learning framework to learn historical fault data and identify the fault modes of the photovoltaic power station, the implementation of the transfer learning framework includes: Build a feature vector knowledge base containing several historical fault modes; When an unknown anomaly is detected, zero-shot detection is achieved by matching topological similarity through a graph convolutional network; Aggregate the data of each power station through federated learning every week to update the global model parameters while preserving local privacy data.

8. The method for real-time monitoring of the operating state of a photovoltaic power station according to claim 7, characterized in that, Dynamically optimize the string topology structure of a photovoltaic power station, including the following steps: Real-time scan the irradiance distribution map of the power station to identify areas where the power generation efficiency is lower than the power generation threshold; Switch the components in the shaded area to an independent MPPT channel through a matrix switch array to form a dynamic sub-array; Achieve Pareto optimality among power generation efficiency, heat loss, and equipment life based on deep deterministic policy gradient; Recalculate the optimization strategy every 60 minutes, and the network reconstruction response time ≤ 30s.

9. The method for real-time monitoring of the operating status of a photovoltaic power station according to claim 8, characterized in that, The method for real-time monitoring the operation status of a photovoltaic power station further includes a self-calibration mechanism: Eliminate temperature drift errors through a cross-check algorithm of adjacent sensor data every 20ms; Perform offline calibration every day at dawn, and use a laser interferometer to correct the reference value for micro-crack detection; When the dust accumulation thickness > 3g / m², automatically trigger a reservation work order for the drone cleaning system.

10. A system for real-time monitoring of the operating status of a photovoltaic power station, characterized in that, For implementing the method for real-time monitoring the operation status of a photovoltaic power station according to any one of claims 1-9, the system for real-time monitoring the operation status of a photovoltaic power station includes: A data acquisition module for real-time collecting the sensor data of a photovoltaic power station by integrating a multi-modal sensor array in the junction box of a photovoltaic module; An edge computing module for locally preprocessing the sensor data through edge computing nodes deployed at the site of a photovoltaic power station and transmitting data with different priorities to the cloud based on a dynamic hybrid communication protocol; A model construction module for fusing sensor data in the cloud to construct a four-dimensional digital twin model, fusing meteorological prediction, component aging, and dust accumulation evolution data that are spatio-temporally continuous; A fault identification module for learning historical fault data using a transfer learning framework to identify the fault modes of a photovoltaic power station, and dynamically optimizing the string topology structure of the photovoltaic power station in combination with the real-time operation status, equipment health data, and fault mode identification results of the photovoltaic power station; A visualization display module for presenting the optimized string topology structure of a photovoltaic power station through a visualization interface, and real-time feedback on the health status and control instructions of the power station, and the control instructions are sent to the edge computing nodes at the site in real-time.

Images