Photovoltaic power station centralized monitoring operation and maintenance system and method based on multi-network cooperation and intelligent early warning

Abstract

The application relates to a photovoltaic power station centralized monitoring operation and maintenance system and method based on multi-network cooperation and intelligent early warning, which comprises the following steps: continuously acquiring environmental monitoring data and communication data of a photovoltaic power station, determining the energy efficiency ratio of the photovoltaic power station according to the communication data and the environmental monitoring data, wherein the communication data comprises photovoltaic power station operation data and communication parameters; predicting an environmental change sequence according to the synchronous change rate of the energy efficiency ratio associated with the environmental monitoring data, determining a switchable target communication network based on a neural network prediction model; determining a dynamic threshold value of fault early warning based on the energy efficiency ratio and the environmental monitoring data, performing a first judgment on real-time communication data based on the dynamic threshold value, outputting a first operation and maintenance result to perform communication network switching; after the communication network is switched, performing a second judgment on the real-time communication data based on the synchronous change rate, secondly determining the target communication network and updating the dynamic threshold value, and outputting a second operation and maintenance result.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power plant monitoring technology, specifically to a centralized monitoring and operation and maintenance system and method for photovoltaic power plants based on multi-network collaboration and intelligent early warning. Background Technology

[0002] With the rapid development of renewable energy, the scale and number of photovoltaic power plants are constantly increasing. Photovoltaic power plants are usually distributed in remote areas. In order to remotely and efficiently monitor the operating status of photovoltaic power plants in real time, the existing photovoltaic power plant operation and maintenance system is becoming increasingly dependent on communication networks. Therefore, the stability of communication networks is particularly important for the operation and maintenance of photovoltaic power plants.

[0003] Existing technologies typically employ simple switching rules to switch communication links. However, the environment in photovoltaic power plant areas is complex, and communication signals are affected by factors such as weather and environmental changes. If communication is interrupted or transmission quality degrades, it will lead to missing or abnormal real-time monitoring data, failing to meet the actual needs of high-reliability remote communication in photovoltaic power plants. In particular, the monitoring and maintenance process of photovoltaic power plants requires consideration of the dynamic response speed and steady-state accuracy of intelligent early warning systems; otherwise, delayed, false, or missed reports are easily generated, seriously affecting the real-time performance, accuracy, and reliability of photovoltaic power plant operation and maintenance. Therefore, existing technologies have shortcomings that urgently need to be addressed. Summary of the Invention

[0004] The primary objective of this application is to address at least one of the aforementioned problems by providing a centralized monitoring and operation and maintenance system and method for photovoltaic power plants based on multi-network collaboration and intelligent early warning.

[0005] To achieve the objectives of this application, the following technical solution is adopted:

[0006] A centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning, provided to meet one of the purposes of this application, includes the following steps:

[0007] Continuously acquire environmental monitoring data and communication data of photovoltaic power plants, and determine the energy efficiency ratio of photovoltaic power plants based on communication data and environmental monitoring data. The communication data includes photovoltaic power plant operation data and communication parameters.

[0008] Based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, an environmental change sequence is predicted, and a switchable target communication network is determined based on a neural network prediction model.

[0009] Based on energy efficiency ratio and environmental monitoring data, a dynamic threshold for fault early warning is determined. Based on the dynamic threshold, a first judgment is made on real-time communication data, and a first operation and maintenance result is output to execute the switching of the communication network.

[0010] After switching the communication network, a second judgment is made on the real-time communication data based on the synchronization change rate to determine the target communication network for the second time and update the dynamic threshold, and output the second operation and maintenance result.

[0011] In an optional embodiment, environmental monitoring data and communication data of the photovoltaic power plant are continuously acquired, specifically including:

[0012] The operation data of the photovoltaic power station is continuously acquired through sensors. Deviation thresholds are set for the acquired operation data according to the data type. The operation data includes the voltage, current, and power of the photovoltaic equipment when it is working, as well as the power generation of the photovoltaic power station. The environmental monitoring data includes weather type, irradiance, and temperature.

[0013] Determine whether the deviation between the operating data and the historical benchmark data of the photovoltaic power station is greater than the deviation threshold. Set the operating data that is less than the deviation threshold as the communication data to be uploaded, and supplement the communication data to be uploaded with the operating data that is greater than the deviation threshold by continuous interpolation.

[0014] The average value of the communication data to be uploaded obtained from multiple consecutive samples is taken as the valid communication data and uploaded.

[0015] In an optional embodiment, determining the energy efficiency ratio of the photovoltaic power plant based on communication data and environmental monitoring data includes:

[0016] The corresponding irradiance and temperature data for the communication data at the corresponding time are determined based on the synchronization time stamp.

[0017] Analyze the communication data to determine the maximum power point corresponding to the irradiance data and temperature data. The maximum power point represents the predicted maximum power of the photovoltaic power station under the corresponding irradiance and temperature.

[0018] The theoretical maximum power generation of the photovoltaic power station is calculated based on the predicted maximum power. The energy efficiency ratio of the photovoltaic power station is determined based on the theoretical maximum power generation and the corresponding actual power generation. The energy efficiency ratio is used to characterize the ratio between the actual power generation of the photovoltaic power station and the theoretical maximum power generation determined based on the predicted maximum power.

[0019] In an optional embodiment, an environmental change sequence is predicted based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, including:

[0020] The first change parameter is obtained by calculating the month-on-month change rate of environmental monitoring data in two adjacent sampling periods, and the second change parameter is obtained by calculating the month-on-month change rate of energy efficiency ratio in two adjacent sampling periods. Based on the ratio of the first change parameter and the second change parameter, the synchronous change rate of energy efficiency ratio associated with environmental monitoring data is obtained.

[0021] The environmental monitoring data closest to the current time and the corresponding synchronous change rate are input into the trained environmental prediction model to obtain the output future environmental monitoring data. The environmental prediction model is obtained by training multiple training environmental change data with synchronous change rate labels.

[0022] Based on the time points of environmental monitoring data collection, the future environmental monitoring data are sorted according to the time point order to obtain the environmental change sequence.

[0023] In an optional embodiment, determining the switchable target communication network based on a neural network prediction model includes:

[0024] The environmental change sequence is input into the neural network prediction model to obtain the switchable communication network and the corresponding baseline communication parameters. Based on the preset communication switching rules, the target communication network is determined based on the switchable communication network and the baseline communication parameters.

[0025] The neural network prediction model is a composite network, including a prediction network, a discriminant network, and a verification network. The neural network prediction model is obtained by training multiple environmental change training sequences, multiple communication network training sets corresponding to multiple environmental change sequences, and multiple communication data training sets corresponding to multiple communication networks.

[0026] The prediction network is used to predict the environmental changes in the next cycle based on the environmental change sequence. The discrimination network is used to classify the environmental changes into preset communication networks according to the corresponding communication parameters to obtain switchable communication networks. The verification network is used to verify the matching degree between the switchable communication networks and the corresponding communication data.

[0027] In an optional embodiment, the target communication network is determined based on a switchable communication network and baseline communication parameters according to a preset communication switching rule, including:

[0028] For each switchable communication network, retrieve the historical communication parameters of successful communication from the historical communication records of the communication network, and select the benchmark communication parameters that meet the similarity threshold based on the similarity between the historical communication parameters and the benchmark communication parameters as the reference communication parameters.

[0029] Calculate the average similarity between all referenced communication parameters and the communication parameters corresponding to the currently switchable communication network to obtain the first communication similarity.

[0030] The second communication similarity is obtained by calculating the similarity between the average communication quality corresponding to all reference communication parameters and the communication quality corresponding to the currently switchable communication network. The communication quality is evaluated based on the year-on-year rate of data dispersion between the communication data in the historical communication records and the environmental monitoring data.

[0031] Calculate the weighted sum of the first communication similarity and the second communication similarity to obtain the network selectable parameters of the switchable communication network;

[0032] The switchable communication networks are sorted from highest to lowest according to the network selectable parameters, and the switchable communication network with the highest network selectable parameter is selected as the target communication network.

[0033] In an optional embodiment, a dynamic threshold for fault warning is determined based on energy efficiency ratio and environmental monitoring data, including:

[0034] Obtain the energy efficiency ratio of the photovoltaic power station for multiple sampling periods before switching the communication network, and calculate the mean, standard deviation and data dispersion of the energy efficiency ratio;

[0035] The environmental monitoring data of the photovoltaic power station corresponding to multiple sampling periods before switching the communication network is obtained, the data dispersion rate of the environmental monitoring data is calculated, and the ratio of the energy efficiency ratio to the data dispersion rate of the environmental monitoring data is set as the confidence coefficient of the standard deviation of the energy efficiency ratio.

[0036] The dynamic threshold for fault warning is obtained by summing the product of the confidence coefficient and the standard deviation of the energy efficiency ratio with the mean of the energy efficiency ratio. The dynamic threshold is dynamically adjusted based on the environmental monitoring data and the confidence coefficient calculated from the energy efficiency ratio.

[0037] In an optional embodiment, a first judgment is made on real-time communication data based on a dynamic threshold, and a first operation and maintenance result is output to perform a switching of the communication network, including:

[0038] Calculate the real-time energy efficiency ratio of real-time communication data. When the real-time energy efficiency ratio is greater than the dynamic threshold, the preset first threshold value is incremented once to obtain the real-time energy efficiency ratio at the next moment and continue to compare it with the dynamic threshold for judgment. If the first threshold value is greater than the accumulated threshold multiple times, and the real-time energy efficiency ratios in the continuous sampling period are still greater than the dynamic threshold, the first threshold value is reset and the communication network switching is not performed.

[0039] When the real-time energy efficiency ratio is less than the dynamic threshold, the real-time energy efficiency ratio of the next sampling period is obtained for judgment. If the real-time energy efficiency ratio of two consecutive sampling periods is less than the dynamic threshold, the preset second threshold value is accumulated once, and the real-time energy efficiency ratio of the next two consecutive sampling periods is obtained for judgment. This continues until the second threshold value is greater than the accumulated threshold, then the target communication network is switched and the fault warning record is cached.

[0040] In an optional embodiment, after switching communication networks, a second judgment is made on real-time communication data based on the synchronization rate of change to determine the target communication network a second time and update the dynamic threshold, outputting a second operation and maintenance result, including:

[0041] The system acquires environmental monitoring data and energy efficiency ratios within multiple consecutive sampling periods before and after switching communication networks, obtains the synchronous change rate between environmental monitoring data and energy efficiency ratios, and makes a judgment based on the linear relationship of the preset synchronous change rate. If the synchronous change rate of data within multiple consecutive sampling periods before and after switching communication networks is non-linear, the system continues to select the target communication network corresponding to the second highest network optional parameter according to the sorting of network optional parameters.

[0042] After switching communication networks, if the rate of synchronous change of data at consecutive moments before and after is linear, the target communication network is determined for the second time, and the theoretical maximum power generation for calculating the energy efficiency ratio is updated according to the linear relationship of the rate of synchronous change. The correction value of the environmental monitoring data is determined according to the linear relationship of the rate of synchronous change, and the ratio of data dispersion is recalculated based on the corrected environmental monitoring data to update the confidence coefficient of the dynamic threshold.

[0043] Based on the updated theoretical maximum power generation and confidence coefficient, the dynamic threshold is redefined, and the photovoltaic power station is continuously monitored based on the redefined dynamic threshold. The operation and maintenance results based on fault warning records and communication network switching are output in real time.

[0044] On the other hand, a centralized monitoring and operation and maintenance system for photovoltaic power plants based on multi-network collaboration and intelligent early warning, provided to meet one of the purposes of this application, includes:

[0045] The data acquisition module is used to continuously acquire environmental monitoring data and communication data of the photovoltaic power station, and determine the energy efficiency ratio of the photovoltaic power station based on the communication data and environmental monitoring data. The communication data includes photovoltaic power station operation data and communication parameters.

[0046] The network selection module is used to predict the environmental change sequence based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, and determine the switchable target communication network based on the neural network prediction model.

[0047] The first operation and maintenance module is used to determine the dynamic threshold for fault warning based on energy efficiency ratio and environmental monitoring data, make a first judgment on real-time communication data based on the dynamic threshold, and output the first operation and maintenance result to execute the switching of communication network.

[0048] The second operation and maintenance module is used to make a second judgment on real-time communication data based on the synchronization change rate after switching communication networks, so as to determine the target communication network for the second time, update the dynamic threshold, and output the second operation and maintenance result.

[0049] On another note, a centralized monitoring and maintenance device for photovoltaic power plants based on multi-network collaboration and intelligent early warning, provided to meet one of the purposes of this application, includes a central processing unit and a memory. The central processing unit is used to call and run a computer program stored in the memory to execute the steps of the centralized monitoring and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning described in this application.

[0050] In another aspect, a computer-readable storage medium is provided to suit one of the purposes of this application, the computer-readable storage medium storing computer-executable instructions for causing a computer to perform a centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as disclosed in any of the first aspects of this invention.

[0051] The technical solution of this application has many advantages, including but not limited to the following aspects:

[0052] This application firstly filters continuously acquired operational data by setting a deviation threshold, then corrects data that does not meet the threshold using continuous interpolation, and uploads the average of the communication data to be uploaded based on the sampling time. It calculates the energy efficiency ratio (EER) using continuously acquired operational data and environmental monitoring data. The EER represents the ratio between the actual power generation of the photovoltaic power station and the theoretical maximum power generation determined based on the predicted maximum power. This EER can serve as data support for initially judging the current operational status of the photovoltaic power station, improving the real-time accuracy of reflecting the photovoltaic power station's status and providing accurate basis for subsequent communication network switching and dynamic early warning, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0053] Next, the synchronous change rate within two adjacent sampling periods is calculated to characterize the change in the energy efficiency ratio associated with environmental monitoring data. Future environmental monitoring data is obtained through an environmental prediction model, thereby constructing an environmental change sequence. The environmental change sequence predicted by the synchronous change rate is input into a neural network prediction model. First, the environmental change sequence is used to predict the environmental change situation in the next period. Then, a discriminant network is used to classify the communication parameters and communication networks of the environmental change situation. Finally, a verification network judges the matching degree of the communication data, thereby selecting switchable communication networks and their benchmark communication parameters as the basis for subsequently determining the target communication network. This improves the accuracy of selecting switchable communication networks and thus enhances the real-time performance, accuracy, and reliability of photovoltaic power plant operation and maintenance.

[0054] Secondly, by calculating the mean, standard deviation, and data dispersion rate of the energy efficiency ratio (EER), environmental monitoring data corresponding to the sampling period of the photovoltaic power station are introduced. The ratio of the EER to the data dispersion rate of the environmental monitoring data is set as the confidence coefficient of the EER standard deviation. The environmental monitoring data and the EER are correlated, and abnormal fluctuations in communication transmission are judged by the degree of data dispersion. Multiple judgments are made based on the real-time EER and dynamic thresholds. At the same time, an accumulative first threshold value is set to ensure that the first judgment process is not affected by a single anomaly, thus preventing misjudgment or omission. This meets the actual needs of high-reliability remote communication for photovoltaic power stations. Meanwhile, the dynamic response speed and steady-state accuracy of intelligent early warning are considered, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0055] Finally, by using the preset linear relationship of the synchronous change rate, the linear relationship of the synchronous change rate of the energy efficiency ratio and environmental monitoring data in the communication data received before and after switching networks is determined, thereby ensuring the stability of communication and the response speed when switching to other communication networks. Based on the determination of the target communication network, the theoretical maximum power generation and confidence coefficient are recalculated through the linear relationship of the synchronous change rate, thereby updating the dynamic threshold and meeting the actual needs of high-reliability remote communication for photovoltaic power plants. At the same time, the dynamic response speed and steady-state accuracy of intelligent early warning are considered, thereby improving the real-time performance, accuracy and reliability of photovoltaic power plant operation and maintenance. Attached Figure Description

[0056] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0057] Figure 1 This is a flowchart of an embodiment of the centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning according to this application;

[0058] Figure 2 This is a schematic diagram of the centralized monitoring and operation and maintenance system for photovoltaic power plants based on multi-network collaboration and intelligent early warning, which is adopted in this application. Detailed Implementation

[0059] The technical solution of this application is applicable to the field of photovoltaic power plant monitoring technology, and is particularly applicable to scenarios where multiple photovoltaic power plants are centrally monitored and maintained through multi-network collaboration and intelligent early warning. In this context, with the rapid development of renewable energy, the scale and number of photovoltaic power plants are constantly increasing. Photovoltaic power plants are usually distributed in remote areas. In order to remotely and efficiently monitor the operating status of photovoltaic power plants in real time, the existing photovoltaic power plant operation and maintenance system is becoming increasingly dependent on communication networks. Therefore, the stability of communication networks is particularly important for the operation and maintenance of photovoltaic power plants.

[0060] Existing technologies typically employ simple switching rules to switch communication links. However, the environment in photovoltaic power plant areas is complex, and communication signals are affected by factors such as weather and environmental changes. If communication is interrupted or transmission quality degrades, it will lead to missing or abnormal real-time monitoring data, failing to meet the actual needs of high-reliability remote communication in photovoltaic power plants. In particular, the monitoring and maintenance process of photovoltaic power plants requires consideration of the dynamic response speed and steady-state accuracy of intelligent early warning systems; otherwise, delayed, false, or missed reports are easily generated, seriously affecting the real-time performance, accuracy, and reliability of photovoltaic power plant operation and maintenance. Therefore, existing technologies have shortcomings that urgently need to be addressed.

[0061] To address the issue that communication stability is affected by environmental conditions, thus impacting the dynamic response speed and steady-state accuracy of intelligent early warning systems, embodiments of this application provide a technical solution, the overall concept of which is as follows:

[0062] The system continuously acquires real-time environmental monitoring and communication data from photovoltaic power plants, calculates the energy efficiency ratio (EER) characterizing the operating status of the photovoltaic power plant, and uploads it to the operation and maintenance platform. It generates an environmental change sequence based on the synchronous change rate of the EER associated with the environmental monitoring data to characterize the relationship between environmental changes and changes in operating status. After a neural network prediction model outputs switchable target communication networks for consideration, and the system determines the dynamic threshold for fault warnings based on the EER and environmental monitoring data, it first assesses whether to switch communication networks and cache warning information. Then, it performs a second assessment based on the linear relationship of the synchronous change rate to determine the communication network and update the dynamic threshold. Finally, it confirms the fault warning based on the updated dynamic threshold and outputs the operation and maintenance results.

[0063] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments.

[0064] The specific embodiments described below can be combined with each other, and the same or similar concepts or processes will not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0065] See Figure 1 This application discloses a centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning, including the following steps:

[0066] 110. Continuously acquire environmental monitoring data and communication data of the photovoltaic power station, and determine the energy efficiency ratio of the photovoltaic power station based on the communication data and environmental monitoring data. The communication data includes photovoltaic power station operation data and communication parameters.

[0067] Optionally, environmental monitoring data can include weather type data, irradiance, temperature data, humidity data, air pressure data, altitude, and equipment dust accumulation data. The data is mainly obtained by continuously acquiring relevant environmental monitoring data that affect the operation of the photovoltaic power station through sensors.

[0068] Optionally, operational data can be power data, voltage data, current data, temperature data, or power generation data of equipment within the photovoltaic power station. This data is primarily collected through sensors to monitor the operational status of the photovoltaic power station. Communication parameters include the sender address, communication data compliance information, sender end, communication equipment parameters, communication protocol, communication purpose, and communication permissions. It is understood that photovoltaic power stations are typically located in remote, high-altitude areas. To achieve monitoring and maintenance of multiple different photovoltaic power stations, environmental monitoring data and operational data collected by sensors at local photovoltaic power stations can be uniformly sent to an operation and maintenance platform for centralized monitoring and early warning. This provides data support for subsequent operation and maintenance work. The communication data includes real-time environmental monitoring data reflecting the current weather and equipment operating environment of the photovoltaic power station, as well as operational data on the working status of the photovoltaic equipment. The communication parameters are parameters that must be accurately set when uploading communication data, which can improve the accuracy of the data required for operation and maintenance.

[0069] Optionally, the target communication network can be a radio link, 4G network, 5G network, fiber optic network, Ethernet, wide area network, Bluetooth network, LoRa or WIFI network. It is understandable that since photovoltaic power plants are usually built in remote areas, communication interruptions or transmission quality degradation when using a single communication network will lead to the loss or abnormality of real-time monitoring data. Using a multi-network collaborative approach can integrate various network resources, such as 5G, 4G, fiber optic, WIFI or LoRa, thereby achieving efficient collaboration between power equipment, systems and maintenance personnel, improving maintenance efficiency, reducing fault risks, and ensuring the stable operation of the power grid.

[0070] In one specific embodiment, the continuous acquisition of environmental monitoring data and communication data from the photovoltaic power plant includes:

[0071] The operation data of the photovoltaic power station is continuously acquired through sensors. Deviation thresholds are set for the acquired operation data according to the data type. The operation data includes the voltage, current, and power of the photovoltaic equipment when it is working, as well as the power generation of the photovoltaic power station. The environmental monitoring data includes weather type, irradiance, and temperature.

[0072] Determine whether the deviation between the operating data and the historical benchmark data of the photovoltaic power station is greater than the deviation threshold. Set the operating data that is less than the deviation threshold as the communication data to be uploaded, and supplement the communication data to be uploaded with the operating data that is greater than the deviation threshold by continuous interpolation.

[0073] The average value of the communication data to be uploaded obtained from multiple consecutive samples is taken as the valid communication data and uploaded.

[0074] Optionally, when continuously acquiring operational data, sensors corresponding to the operational data type can be used for collection. Environmental monitoring data can also be collected using sensors corresponding to the data type. During the collection process, operational data can be continuously sampled at a certain sampling frequency. The sampling frequency can be set to several to tens of times per second, for example, it can be selected between 5 times per second and 100 times per second to meet the needs of different application scenarios. Especially in scenarios with rapidly changing weather, high-frequency collection can effectively reflect the status of the photovoltaic power station and provide accurate basis for subsequent calculations and early warnings.

[0075] Optionally, for operational data such as voltage and current, an analog-to-digital converter can be used to convert the corresponding voltage and current signals into digital signals to obtain the corresponding operational data. This facilitates the subsequent uploading of communication data from the local photovoltaic power station to the operation and maintenance platform for further processing. Especially when environmental factors cause network communication instability, high-frequency continuous data collection is required to ensure whether there are any abnormalities in the communication data before and after switching communication networks, providing an accurate basis for subsequent operation and maintenance.

[0076] In the actual monitoring process of photovoltaic power plants, after the sensors or analog-to-digital converters acquire real-time operating data and environmental monitoring data, they preprocess these data through the local photovoltaic power plant, such as filtering and noise reduction, to obtain more accurate real-time data. This preprocessed data serves as the data basis for calculating the synchronization change ratio and dynamic threshold in subsequent steps, thereby achieving precise control of multi-grid coordination and intelligent early warning of photovoltaic power plants. The deviation threshold can be calculated based on historical data, set through prior experience, or solved by setting a mathematical model. This application does not limit the calculation and solution of the deviation threshold.

[0077] As can be seen, through the above embodiments, by setting a deviation threshold to initially screen the continuously acquired operating data, then correcting the data that does not meet the threshold using continuous interpolation, and finally uploading the average value of the communication data to be uploaded based on the sampling time, the real-time accuracy of reflecting the status of the photovoltaic power station is improved, providing an accurate basis for subsequent communication network switching and dynamic early warning, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0078] In practical implementation, the energy efficiency ratio of the photovoltaic power station is determined based on communication data and environmental monitoring data, including:

[0079] The corresponding irradiance and temperature data for the communication data at the corresponding time are determined based on the synchronization time stamp.

[0080] Analyze the communication data to determine the maximum power point corresponding to the irradiance data and temperature data. The maximum power point represents the predicted maximum power of the photovoltaic power station under the corresponding irradiance and temperature.

[0081] The theoretical maximum power generation of the photovoltaic power station is calculated based on the predicted maximum power. The energy efficiency ratio of the photovoltaic power station is determined based on the theoretical maximum power generation and the corresponding actual power generation. The energy efficiency ratio is used to characterize the ratio between the actual power generation of the photovoltaic power station and the theoretical maximum power generation determined based on the predicted maximum power.

[0082] Specifically, since power generation is affected by irradiance and equipment temperature, theoretical power generation can be calculated using irradiance, the operating temperature of photovoltaic modules, nominal power, and temperature coefficient. Therefore, the maximum power point can also be calculated by inversely using varying irradiance and temperature to obtain the relationship between maximum power changes caused by environmental monitoring data. Then, the theoretical maximum power generation can be calculated from the predicted maximum power, simulating the maximum feasible power generation under extreme conditions. This ensures that the monitoring and maintenance process guarantees that the photovoltaic power station operates close to its maximum power generation even when environmental monitoring data changes. The actual power generation is then compared with the theoretical maximum power generation to calculate the energy efficiency ratio. Since both actual and theoretical maximum power generation are affected by changes in environmental monitoring data, and actual power generation will be less than the theoretical maximum power generation, in the absence of communication anomalies... Before a fault warning is issued, the ratio of actual power generation to theoretical maximum power generation will be in a relatively stable fluctuation state. For example, when the weather changes suddenly, irradiance and temperature decrease, and the numerator of the ratio, which is the actual power generation, and the denominator, which is the theoretical maximum power generation, will decrease simultaneously. Only when communication is abnormal or the photovoltaic power station is about to experience a fault will the numerator and denominator of the energy efficiency ratio change significantly, leading to an abnormal energy efficiency ratio. This can be used to initially determine that there is an operation and maintenance problem in the photovoltaic power station, and further investigation is needed to determine whether the communication problem is caused by environmental factors, which affects data transmission and causes false alarms or missed alarms in the fault warning, or whether the real-time data of the photovoltaic power station is accurate and equipment failure is about to occur. This provides an accurate basis for subsequent communication network switching and dynamic warning, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0083] As can be seen, through the above embodiments, the energy efficiency ratio is calculated by continuously acquiring operational data and environmental monitoring data. The energy efficiency ratio is used to characterize the ratio between the actual power generation of the photovoltaic power station and the theoretical maximum power generation determined based on the predicted maximum power. This ratio can serve as data support for initially judging the current operating status of the photovoltaic power station, thereby providing an accurate basis for subsequent communication network switching and dynamic early warning, and thus improving the real-time performance, accuracy and reliability of photovoltaic power station operation and maintenance.

[0084] 210. Based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, predict the environmental change sequence, and determine the switchable target communication network based on the neural network prediction model.

[0085] Optionally, the environmental change sequence is obtained by predicting through a neural network. In this application, the neural network can be an LSTM network, a CNN network, an RNN network, a GNN network, a Transformer network, or a composite network composed of multiple neural networks. This application does not limit the type of predictive neural network.

[0086] Specifically, both environmental monitoring data and energy efficiency ratio (EER) are dynamically changing data. The dynamic change of EER is affected by both environmental changes and the operating status of the photovoltaic power station. The synchronous change rate between the two also reflects the real-time status of the local photovoltaic power station and the data change status after uploading. For example, if the environmental status of the local photovoltaic power station changes abruptly within a certain period (usually due to weather changes), and the synchronous change rate of environmental monitoring data and EER fluctuates within a stable range, it can be considered that the environmental change has not affected communication transmission. If the synchronous change rate of environmental change and EER exceeds the stable range, it can be determined whether the currently received communication data is affected by environmental changes and causes communication data failure based on the instantaneous change of environmental monitoring data or EER. This allows for a determination of whether to switch to other target communication networks, thereby reducing the impact of communication instability on operation and maintenance.

[0087] In one specific embodiment, an environmental change sequence is predicted based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, including:

[0088] The first change parameter is obtained by calculating the month-on-month change rate of environmental monitoring data in two adjacent sampling periods, and the second change parameter is obtained by calculating the month-on-month change rate of energy efficiency ratio in two adjacent sampling periods. Based on the ratio of the first change parameter and the second change parameter, the synchronous change rate of energy efficiency ratio associated with environmental monitoring data is obtained.

[0089] The environmental monitoring data closest to the current time and the corresponding synchronous change rate are input into the trained environmental prediction model to obtain the output future environmental monitoring data. The environmental prediction model is obtained by training multiple training environmental change data with synchronous change rate labels.

[0090] Based on the time points of environmental monitoring data collection, the future environmental monitoring data are sorted according to the time point order to obtain the environmental change sequence.

[0091] As can be seen, through the above embodiments, the month-on-month change rate of environmental monitoring data is calculated as the first change parameter, and the month-on-month change rate of energy efficiency ratio is calculated as the second change parameter. Thus, the synchronous change rate within two adjacent sampling periods is calculated to characterize the change of energy efficiency ratio associated with environmental monitoring data. Future environmental monitoring data is obtained through the environmental prediction model, thereby constructing an environmental change sequence as the data basis for predicting subsequent switchable target communication networks. This provides an accurate basis for subsequent communication network switching and dynamic early warning, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power plant operation and maintenance.

[0092] In one specific embodiment, determining the switchable target communication network based on a neural network prediction model includes:

[0093] The environmental change sequence is input into the neural network prediction model to obtain the switchable communication network and the corresponding baseline communication parameters. Based on the preset communication switching rules, the target communication network is determined based on the switchable communication network and the baseline communication parameters.

[0094] The neural network prediction model is a composite network, including a prediction network, a discriminant network, and a verification network. The neural network prediction model is obtained by training multiple environmental change training sequences, multiple communication network training sets corresponding to multiple environmental change sequences, and multiple communication data training sets corresponding to multiple communication networks.

[0095] The prediction network is used to predict the environmental changes in the next cycle based on the environmental change sequence. The discrimination network is used to classify the environmental changes into preset communication networks according to the corresponding communication parameters to obtain switchable communication networks. The verification network is used to verify the matching degree between the switchable communication networks and the corresponding communication data.

[0096] Specifically, the preset communication switching rules calculate the address similarity between sending target addresses and the information similarity between sending user terminals based on the baseline communication parameters corresponding to the switchable network and the communication parameters in the real-time communication data. Furthermore, the association representation value between communication data is determined based on the weighted sum of address similarity and information similarity. This allows the comprehensive impact of address integrity and the richness of user terminal data to be fully considered in the calculation of the association representation value, thereby improving the accuracy of selecting switchable communication networks and ultimately enhancing the real-time performance, accuracy, and reliability of photovoltaic power plant operation and maintenance.

[0097] As can be seen, through the above embodiments, the environmental change sequence obtained by synchronous change rate prediction is input into the neural network prediction model. First, the environmental change sequence is used to predict the periodic environmental change situation. Then, the discriminant network classifies the communication parameters and communication networks of the environmental change situation. Finally, the verification network judges the matching degree of the communication data, thereby selecting switchable communication networks and the benchmark communication parameters of the communication networks as the basis for subsequent determination of target communication networks. This improves the accuracy of selecting switchable communication networks, thereby improving the real-time performance, accuracy and reliability of photovoltaic power plant operation and maintenance.

[0098] 310. Determine the dynamic threshold for fault early warning based on energy efficiency ratio and environmental monitoring data, make a first judgment on real-time communication data based on the dynamic threshold, and output the first operation and maintenance result to execute the switching of communication network;

[0099] Specifically, dynamic thresholds serve as the basis for fault warning. Due to communication instability leading to abnormal data transmission, dynamic thresholds need to simultaneously assess communication network stability and issue fault warnings. First, the impact of communication network anomalies must be eliminated. Suspicious fault warnings should be recorded first to reduce the probability of false alarms or missed alarms. If communication anomalies are present, a second assessment should be conducted after switching networks to verify the accuracy of the fault warning and the feasibility of network switching. This multi-network collaboration and fault warning system enable centralized monitoring and maintenance of photovoltaic power plants, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power plant operation and maintenance.

[0100] In one specific embodiment, determining the dynamic threshold for fault early warning based on energy efficiency ratio and environmental monitoring data includes:

[0101] Obtain the energy efficiency ratio of the photovoltaic power station for multiple sampling periods before switching the communication network, and calculate the mean, standard deviation and data dispersion of the energy efficiency ratio;

[0102] The environmental monitoring data of the photovoltaic power station corresponding to multiple sampling periods before switching the communication network is obtained, the data dispersion rate of the environmental monitoring data is calculated, and the ratio of the energy efficiency ratio to the data dispersion rate of the environmental monitoring data is set as the confidence coefficient of the standard deviation of the energy efficiency ratio.

[0103] The dynamic threshold for fault warning is obtained by summing the product of the confidence coefficient and the standard deviation of the energy efficiency ratio with the mean of the energy efficiency ratio. The dynamic threshold is dynamically adjusted based on the environmental monitoring data and the confidence coefficient calculated from the energy efficiency ratio.

[0104] As can be seen, through the above embodiments, by calculating the mean, standard deviation, and data dispersion of the energy efficiency ratio, and introducing environmental monitoring data corresponding to the sampling period of the photovoltaic power station, the ratio of the energy efficiency ratio and the data dispersion of the environmental monitoring data is set as the confidence coefficient of the standard deviation of the energy efficiency ratio. The environmental monitoring data and the energy efficiency ratio are correlated, and abnormal fluctuations in communication transmission are judged by the degree of data dispersion. When the abnormal fluctuation is too large and exceeds the dynamic threshold, it is determined that there is an anomaly in the communication data and network switching is required. It is understood that it is necessary to quickly obtain multiple consecutive real-time energy efficiency ratios for judgment to improve response speed and steady-state accuracy. The time interval between multiple real-time energy efficiency ratios can be adjusted according to the requirements of response sensitivity and early warning accuracy, thereby improving the accuracy of communication network switching and dynamic early warning, and thus improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0105] In one specific embodiment, a first judgment is made on real-time communication data based on a dynamic threshold, and a first operation and maintenance result is output to execute a communication network switch, including:

[0106] Calculate the real-time energy efficiency ratio of real-time communication data. When the real-time energy efficiency ratio is greater than the dynamic threshold, the preset first threshold value is incremented once to obtain the real-time energy efficiency ratio at the next moment and continue to compare it with the dynamic threshold for judgment. If the first threshold value is greater than the accumulated threshold multiple times, and the real-time energy efficiency ratios in the continuous sampling period are still greater than the dynamic threshold, the first threshold value is reset and the communication network switching is not performed.

[0107] When the real-time energy efficiency ratio is less than the dynamic threshold, the real-time energy efficiency ratio of the next sampling period is obtained for judgment. If the real-time energy efficiency ratio of two consecutive sampling periods is less than the dynamic threshold, the preset second threshold value is accumulated once, and the real-time energy efficiency ratio of the next two consecutive sampling periods is obtained for judgment. This continues until the second threshold value is greater than the accumulated threshold, then the target communication network is switched and the fault warning record is cached.

[0108] As can be seen, through the above embodiments, multiple judgments are made based on the real-time energy efficiency ratio and a dynamic threshold. Simultaneously, an accumulative first threshold is set to ensure that the first judgment process is not affected by a single anomaly, preventing misjudgment or missed judgment. The accumulative threshold can be set based on prior experience or sensitivity requirements. Only when multiple consecutive energy efficiency ratios are less than the dynamic threshold is a network switch determined, and fault warning records are cached. It is understood that, to achieve accurate warnings, the fault warning threshold set by the dynamic threshold is a dynamic interval between the threshold corresponding to the actual fault occurrence. Compared to traditional fault warnings, the length of the dynamic interval is larger to allow for early prediction of potential faults and secondary verification through network switching. The dynamic threshold includes a confidence coefficient representing the dispersion of environmental monitoring data. When environmental monitoring data changes significantly, the corresponding dynamic range increases, making the judgment process more sensitive. When environmental monitoring data changes slightly, the corresponding dynamic range shrinks, making the judgment process more accurate. Only when environmental monitoring data changes significantly does it reflect that the real-time environmental changes of the photovoltaic power station will affect the communication between the local photovoltaic power station and the centralized operation and maintenance platform. Only then will it consider whether to switch the communication network, thereby meeting the actual needs of high-reliability remote communication for photovoltaic power stations. At the same time, it considers the dynamic response speed and steady-state accuracy of intelligent early warning, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0109] 410. After switching the communication network, a second judgment is made on the real-time communication data based on the synchronization change rate to determine the target communication network for the second time and update the dynamic threshold, and output the second operation and maintenance result.

[0110] In practical implementation, after switching communication networks, a second judgment is made on real-time communication data based on the synchronization rate of change to determine the target communication network a second time and update the dynamic threshold, outputting a second operation and maintenance result, including:

[0111] The system acquires environmental monitoring data and energy efficiency ratios within multiple consecutive sampling periods before and after switching communication networks, obtains the synchronous change rate between environmental monitoring data and energy efficiency ratios, and makes a judgment based on the linear relationship of the preset synchronous change rate. If the synchronous change rate of data within multiple consecutive sampling periods before and after switching communication networks is non-linear, the system continues to select the target communication network corresponding to the second highest network optional parameter according to the sorting of network optional parameters.

[0112] After switching communication networks, if the rate of synchronous change of data at consecutive moments before and after is linear, the target communication network is determined for the second time, and the theoretical maximum power generation for calculating the energy efficiency ratio is updated according to the linear relationship of the rate of synchronous change. The correction value of the environmental monitoring data is determined according to the linear relationship of the rate of synchronous change, and the ratio of data dispersion is recalculated based on the corrected environmental monitoring data to update the confidence coefficient of the dynamic threshold.

[0113] Based on the updated theoretical maximum power generation and confidence coefficient, the dynamic threshold is redefined, and the photovoltaic power station is continuously monitored based on the redefined dynamic threshold. The operation and maintenance results based on fault warning records and communication network switching are output in real time.

[0114] As can be seen from the above embodiments, by using the linear relationship of the preset synchronous change rate, the linear relationship of the synchronous change rate of the energy efficiency ratio and environmental monitoring data in the communication data received before and after switching networks is determined, thereby ensuring the stability of communication and the response speed of switching other communication networks. Based on the determination of the target communication network, the theoretical maximum power generation and confidence coefficient are recalculated through the linear relationship of the synchronous change rate, thereby updating the dynamic threshold, thus meeting the actual needs of high-reliability remote communication of photovoltaic power plants. At the same time, the dynamic response speed and steady-state accuracy of intelligent early warning are considered, thereby improving the real-time performance, accuracy and reliability of photovoltaic power plant operation and maintenance.

[0115] The unique technical advantage of this application lies in the following: First, by setting a deviation threshold, the continuously acquired operational data is initially screened. Then, data that does not meet the threshold is corrected using continuous interpolation. The average value of the communication data to be uploaded is taken based on the sampling time and then uploaded. The energy efficiency ratio is calculated using the continuously acquired operational data and environmental monitoring data. The energy efficiency ratio is used to characterize the ratio between the actual power generation of the photovoltaic power station and the theoretical maximum power generation determined based on the predicted maximum power. This ratio can serve as data support for initially judging the current operating status of the photovoltaic power station, improving the real-time accuracy of reflecting the status of the photovoltaic power station. It provides an accurate basis for subsequent communication network switching and dynamic early warning, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0116] Next, the synchronous change rate within two adjacent sampling periods is calculated to characterize the change in the energy efficiency ratio associated with environmental monitoring data. Future environmental monitoring data is obtained through an environmental prediction model, thereby constructing an environmental change sequence. The environmental change sequence predicted by the synchronous change rate is input into a neural network prediction model. First, the environmental change sequence is used to predict the environmental change situation in the next period. Then, a discriminant network is used to classify the communication parameters and communication networks of the environmental change situation. Finally, a verification network judges the matching degree of the communication data, thereby selecting switchable communication networks and their benchmark communication parameters as the basis for subsequently determining the target communication network. This improves the accuracy of selecting switchable communication networks and thus enhances the real-time performance, accuracy, and reliability of photovoltaic power plant operation and maintenance.

[0117] Secondly, by calculating the mean, standard deviation, and data dispersion of the energy efficiency ratio (EER), environmental monitoring data from the corresponding sampling period of the photovoltaic power station is introduced. The ratio of the EER to the data dispersion of the environmental monitoring data is set as the confidence coefficient of the EER standard deviation. The environmental monitoring data and EER are correlated, and abnormal fluctuations in communication transmission are judged by the degree of data dispersion. Multiple judgments are made based on the real-time EER and a dynamic threshold. Simultaneously, an accumulative first threshold is set to ensure that the first judgment process is not affected by a single anomaly, leading to misjudgment or missed judgment. The accumulative threshold can be set based on prior experience or sensitivity requirements. Only when multiple consecutive EERs are less than the dynamic threshold is a network switch required, and fault warning records are cached. It can be understood that, to achieve accurate warnings, the fault warning threshold set by the dynamic threshold is a dynamic range from the threshold corresponding to the actual occurrence of the fault. Compared to traditional fault early warning systems, the dynamic range is longer to allow for early prediction of potential faults. This is followed by secondary verification via network switching. The dynamic threshold includes a confidence coefficient representing the dispersion of environmental monitoring data. Significant changes in environmental monitoring data lead to a larger dynamic range, making the judgment process more sensitive. Conversely, minor changes lead to a smaller dynamic range, making the judgment process more accurate. Only when environmental monitoring data changes significantly does the real-time environmental changes of the photovoltaic power station affect communication between the local photovoltaic power station and the centralized operation and maintenance platform, prompting consideration of whether to switch communication networks. This meets the practical needs of high-reliability remote communication for photovoltaic power stations, while also considering the dynamic response speed and steady-state accuracy of intelligent early warning systems, thereby improving the real-time performance, accuracy, and reliability of photovoltaic power station operation and maintenance.

[0118] Finally, by using the preset linear relationship of the synchronous change rate, the linear relationship of the synchronous change rate of the energy efficiency ratio and environmental monitoring data in the communication data received before and after switching networks is determined, thereby ensuring the stability of communication and the response speed when switching to other communication networks. Based on the determination of the target communication network, the theoretical maximum power generation and confidence coefficient are recalculated through the linear relationship of the synchronous change rate, thereby updating the dynamic threshold and meeting the actual needs of high-reliability remote communication for photovoltaic power plants. At the same time, the dynamic response speed and steady-state accuracy of intelligent early warning are considered, thereby improving the real-time performance, accuracy and reliability of photovoltaic power plant operation and maintenance.

[0119] Please see Figure 2 According to one aspect of this application, a centralized monitoring and operation and maintenance system for photovoltaic power plants based on multi-network collaboration and intelligent early warning is provided, the system comprising:

[0120] The data acquisition module is used to continuously acquire environmental monitoring data and communication data of the photovoltaic power station, and determine the energy efficiency ratio of the photovoltaic power station based on the communication data and environmental monitoring data. The communication data includes photovoltaic power station operation data and communication parameters.

[0121] The network selection module is used to predict the environmental change sequence based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, and determine the switchable target communication network based on the neural network prediction model.

[0122] The first operation and maintenance module is used to determine the dynamic threshold for fault warning based on energy efficiency ratio and environmental monitoring data, make a first judgment on real-time communication data based on the dynamic threshold, and output the first operation and maintenance result to execute the switching of communication network.

[0123] The second operation and maintenance module is used to make a second judgment on real-time communication data based on the synchronization change rate after switching communication networks, so as to determine the target communication network for the second time, update the dynamic threshold, and output the second operation and maintenance result.

[0124] Based on any embodiment of the system in this application, the data acquisition module of this application further includes: a data acquisition and upload module, configured to continuously acquire operating data of the photovoltaic power station through sensors, set a deviation threshold for the acquired operating data according to the data type, wherein the operating data includes the voltage, current, power of the photovoltaic equipment during operation and the power generation of the photovoltaic power station, and the environmental monitoring data includes weather type, irradiance, and temperature; determine whether the deviation between the operating data and the historical benchmark data of the photovoltaic power station is greater than the deviation threshold, set the operating data less than the deviation threshold as communication data to be uploaded, and supplement the communication data to be uploaded with the operating data greater than the deviation threshold through continuous interpolation; and take the average value of the communication data to be uploaded obtained by multiple continuous samplings according to the sampling time as valid communication data for uploading.

[0125] Based on any embodiment of the system in this application, the data acquisition module of the system in this application further includes: an energy efficiency ratio calculation module, configured to determine the irradiance data and temperature data corresponding to the communication data at the corresponding time according to the synchronization time scale; analyze the communication data to determine the maximum power point corresponding to the irradiance data and temperature data, the maximum power point representing the predicted maximum power of the photovoltaic power station under the corresponding irradiance and temperature; calculate the theoretical maximum power generation of the photovoltaic power station based on the predicted maximum power; and determine the energy efficiency ratio of the photovoltaic power station based on the theoretical maximum power generation and the corresponding actual power generation, the energy efficiency ratio being used to characterize the proportional relationship between the actual power generation of the photovoltaic power station and the theoretical maximum power generation determined based on the predicted maximum power.

[0126] Based on any embodiment of the system in this application, the network selection module of this application further includes: a sequence construction module, configured to calculate the month-on-month change rate of environmental monitoring data in two adjacent sampling periods to obtain a first change parameter, and calculate the month-on-month change rate of energy efficiency ratio in two adjacent sampling periods to obtain a second change parameter; and obtain the synchronous change rate of energy efficiency ratio associated with environmental monitoring data based on the ratio of the first change parameter and the second change parameter; input the environmental monitoring data closest to the current time and the synchronous change rate at the corresponding time into a trained environmental prediction model to obtain the output future environmental monitoring data, wherein the environmental prediction model is obtained by training multiple training environmental change data with synchronous change rate labels; and sort the future environmental monitoring data according to the time point of environmental monitoring data collection to obtain an environmental change sequence.

[0127] Based on any embodiment of the system in this application, the network selection module of this application further includes: a prediction module, configured to input an environmental change sequence into a neural network prediction model to obtain switchable communication networks and corresponding baseline communication parameters, and determine a target communication network based on the switchable communication networks and baseline communication parameters according to a preset communication switching rule; the neural network prediction model is a composite network, including a prediction network, a discriminant network, and a verification network, and is obtained by training a neural network prediction model using multiple environmental change training sequences, multiple communication network training sets corresponding to multiple environmental change sequences, and multiple communication data training sets of corresponding communication networks; the prediction network is used to predict the environmental change situation in the next cycle based on the environmental change sequence, the discriminant network is used to classify the preset communication networks according to the corresponding communication parameters to obtain switchable communication networks, and the verification network is used to verify the matching degree between the switchable communication networks and the corresponding communication data.

[0128] Based on any embodiment of the system in this application, the network selection module of this application further includes: a pre-switching module, configured to, for each switchable communication network, obtain historical communication parameters of successful communication from the historical communication records of the communication network, and filter benchmark communication parameters that meet the similarity threshold as reference communication parameters based on the similarity between the historical communication parameters and the benchmark communication parameters; calculate the average similarity between all reference communication parameters and the communication parameters corresponding to the current switchable communication network to obtain a first communication similarity; calculate the similarity between the average communication quality corresponding to all reference communication parameters and the communication quality corresponding to the current switchable communication network to obtain a second communication similarity, wherein the communication quality is evaluated based on the year-on-year rate of data dispersion between the communication data in the historical communication records and the environmental monitoring data of the communication network; calculate the weighted sum of the first communication similarity and the second communication similarity to obtain the network selectable parameters of the switchable communication network; sort the switchable communication networks from high to low according to the network selectable parameters, and select the switchable communication network corresponding to the highest network selectable parameters as the target communication network.

[0129] Based on any embodiment of the system in this application, the first operation and maintenance module of the system in this application further includes: a threshold determination module, configured to acquire the energy efficiency ratio of the photovoltaic power station corresponding to multiple sampling periods before switching the communication network, and calculate the mean, standard deviation and data dispersion rate of the energy efficiency ratio; acquire environmental monitoring data of the photovoltaic power station corresponding to multiple sampling periods before switching the communication network, calculate the data dispersion rate of the environmental monitoring data, and set the ratio of the energy efficiency ratio and the data dispersion rate of the environmental monitoring data as the confidence coefficient of the standard deviation of the energy efficiency ratio; sum the product between the confidence coefficient and the standard deviation of the energy efficiency ratio with the mean of the energy efficiency ratio to obtain the dynamic threshold for fault warning, wherein the dynamic threshold is dynamically adjusted according to the confidence coefficient calculated from the environmental monitoring data and the energy efficiency ratio.

[0130] Based on any embodiment of the system in this application, the first operation and maintenance module of the system in this application further includes: a first judgment module, configured to calculate the real-time energy efficiency ratio of real-time communication data. When the real-time energy efficiency ratio is greater than the dynamic threshold, the preset first threshold value is incremented once to obtain the real-time energy efficiency ratio at the next moment and continue to compare it with the dynamic threshold for judgment. If the first threshold value is greater than the accumulated threshold multiple times, and the real-time energy efficiency ratios in the continuous sampling period are still greater than the dynamic threshold, the first threshold value is reset and no communication network switching is performed. When the real-time energy efficiency ratio is less than the dynamic threshold, the real-time energy efficiency ratio of the next sampling period is obtained for judgment. If the real-time energy efficiency ratio of two consecutive sampling periods is less than the dynamic threshold, the preset second threshold value is incremented once to obtain the real-time energy efficiency ratio of the next two consecutive sampling periods for judgment. This continues until the second threshold value is greater than the accumulated threshold, then the system switches to the target communication network and caches the fault warning record.

[0131] Based on any embodiment of the system in this application, the second operation and maintenance module of this application further includes: a second judgment module, configured to acquire environmental monitoring data and energy efficiency ratio corresponding to multiple consecutive sampling periods before and after switching the communication network, obtain the synchronous change rate of the data between the environmental monitoring data and the energy efficiency ratio, and make a judgment based on the linear relationship of the preset synchronous change rate. If the synchronous change rate of the data in multiple consecutive sampling periods before and after switching the communication network is non-linear, then the target communication network corresponding to the second highest network optional parameter is selected for switching according to the sorting of network optional parameters. If the synchronous change rate of the data in consecutive moments before and after switching the communication network is linear, then the target communication network is determined a second time, and the theoretical maximum power generation of the calculated energy efficiency ratio is updated according to the linear relationship of the synchronous change rate, and the correction value of the environmental monitoring data is determined according to the linear relationship of the synchronous change rate. Based on the corrected environmental monitoring data, the ratio of data dispersion rate is recalculated to update the confidence coefficient of the dynamic threshold. Based on the updated theoretical maximum power generation and confidence coefficient, the dynamic threshold is re-determined, and the photovoltaic power station is continuously monitored based on the re-determined dynamic threshold, and the operation and maintenance results based on fault warning records and communication network switching are output in real time.

[0132] Another embodiment of this application provides a centralized monitoring and maintenance device for photovoltaic power plants based on multi-network collaboration and intelligent early warning. This device includes a processor, a computer-readable storage medium, a memory, and a network interface connected via a system bus. The computer-readable, non-volatile storage medium stores an operating system, a database, and computer-readable instructions. The database can store information sequences. When the computer-readable instructions are executed by the processor, the processor can implement a centralized monitoring and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning.

[0133] The processor of this centralized monitoring and maintenance equipment for photovoltaic power plants based on multi-network collaboration and intelligent early warning provides computing and control capabilities, supporting the operation of the entire equipment. The memory of this equipment can store computer-readable instructions, which, when executed by the processor, cause the processor to perform the centralized monitoring and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in this application. The network interface of this equipment is used for communication with terminals.

[0134] In this embodiment, the processor is used to execute... Figure 2The system defines the specific functions of each module, and the memory stores the program code and various data required to execute these modules or submodules. The network interface is used to enable data transmission between user terminals or servers.

[0135] The non-volatile readable storage medium in this embodiment stores the program code and data required to execute all modules in the photovoltaic power plant centralized monitoring and operation and maintenance system based on multi-network collaboration and intelligent early warning of this application. The server can call the program code and data of the server to execute the functions of all modules.

[0136] This application also provides a non-volatile readable storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors cause the one or more processors to perform the steps of the centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning according to any embodiment of this application.

[0137] This application also provides a computer program product, including a computer program / instructions that, when executed by one or more processors, implement the steps of the method described in any embodiment of this application.

Claims

1. A centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning, characterized in that, The steps include the following: Continuously acquire environmental monitoring data and communication data of photovoltaic power plants, and determine the energy efficiency ratio of photovoltaic power plants based on communication data and environmental monitoring data. The communication data includes photovoltaic power plant operation data and communication parameters. Based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, an environmental change sequence is predicted, and a switchable target communication network is determined based on a neural network prediction model. Based on energy efficiency ratio and environmental monitoring data, a dynamic threshold for fault early warning is determined. Based on the dynamic threshold, a first judgment is made on real-time communication data, and a first operation and maintenance result is output to execute the switching of the communication network. After switching the communication network, a second judgment is made on the real-time communication data based on the synchronization change rate to determine the target communication network for the second time and update the dynamic threshold, and output the second operation and maintenance result; Based on a neural network prediction model, switchable target communication networks are identified, including: The environmental change sequence is input into the neural network prediction model to obtain the switchable communication network and the corresponding baseline communication parameters. Based on the preset communication switching rules, the target communication network is determined based on the switchable communication network and the baseline communication parameters. The neural network prediction model is a composite network, including a prediction network, a discriminant network, and a verification network. The neural network prediction model is obtained by training multiple environmental change training sequences, multiple communication network training sets corresponding to multiple environmental change sequences, and multiple communication data training sets corresponding to multiple communication networks. The prediction network is used to predict the environmental changes in the next cycle based on the environmental change sequence. The discrimination network is used to classify the environmental changes into preset communication networks according to the corresponding communication parameters to obtain switchable communication networks. The verification network is used to verify the matching degree between the switchable communication networks and the corresponding communication data. Based on preset communication switching rules, the target communication network is determined using switchable communication networks and baseline communication parameters, including: For each switchable communication network, retrieve the historical communication parameters of successful communication from the historical communication records of the communication network, and select the benchmark communication parameters that meet the similarity threshold based on the similarity between the historical communication parameters and the benchmark communication parameters as the reference communication parameters. Calculate the average similarity between all referenced communication parameters and the communication parameters corresponding to the currently switchable communication network to obtain the first communication similarity. The second communication similarity is obtained by calculating the similarity between the average communication quality corresponding to all reference communication parameters and the communication quality corresponding to the currently switchable communication network. The communication quality is evaluated based on the year-on-year rate of data dispersion between the communication data in the historical communication records and the environmental monitoring data. Calculate the weighted sum of the first communication similarity and the second communication similarity to obtain the network selectable parameters of the switchable communication network; The switchable communication networks are sorted from highest to lowest according to the network selectable parameters, and the switchable communication network with the highest network selectable parameter is selected as the target communication network.

2. The centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in claim 1, characterized in that, Continuously acquire environmental monitoring data and communication data from photovoltaic power plants, specifically including: The operation data of the photovoltaic power station is continuously acquired through sensors. Deviation thresholds are set for the acquired operation data according to the data type. The operation data includes the voltage, current, and power of the photovoltaic equipment when it is working, as well as the power generation data of the photovoltaic power station. The environmental monitoring data includes irradiance data and temperature data. Determine whether the deviation between the operating data and the historical benchmark data of the photovoltaic power station is greater than the deviation threshold. Set the operating data that is less than the deviation threshold as the communication data to be uploaded, and supplement the communication data to be uploaded with the operating data that is greater than the deviation threshold by continuous interpolation. The average value of the communication data to be uploaded obtained from multiple consecutive samples is taken as the valid communication data and uploaded.

3. The centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in claim 2, characterized in that, The energy efficiency ratio of a photovoltaic power plant is determined based on communication data and environmental monitoring data, including: The corresponding irradiance and temperature data for the communication data at the corresponding time are determined based on the synchronization time stamp. Analyze the communication data to determine the maximum power point corresponding to the irradiance data and temperature data. The maximum power point represents the predicted maximum power of the photovoltaic power station under the corresponding irradiance and temperature. The theoretical maximum power generation of the photovoltaic power station is calculated based on the predicted maximum power. The energy efficiency ratio of the photovoltaic power station is determined based on the theoretical maximum power generation and the corresponding actual power generation. The energy efficiency ratio is used to characterize the ratio between the actual power generation of the photovoltaic power station and the theoretical maximum power generation determined based on the predicted maximum power.

4. The centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in claim 1, characterized in that, Based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, an environmental change sequence is predicted, including: The first change parameter is obtained by calculating the month-on-month change rate of environmental monitoring data in two adjacent sampling periods, and the second change parameter is obtained by calculating the month-on-month change rate of energy efficiency ratio in two adjacent sampling periods. Based on the ratio of the first change parameter and the second change parameter, the synchronous change rate of energy efficiency ratio associated with environmental monitoring data is obtained. The environmental monitoring data closest to the current time and the corresponding synchronous change rate are input into the trained environmental prediction model to obtain the output future environmental monitoring data. The environmental prediction model is obtained by training multiple training environmental change data with synchronous change rate labels. Based on the time points of environmental monitoring data collection, the future environmental monitoring data are sorted according to the time point order to obtain the environmental change sequence.

5. The centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in claim 1, characterized in that, Dynamic thresholds for fault early warning are determined based on energy efficiency ratio and environmental monitoring data, including: Obtain the energy efficiency ratio of the photovoltaic power station for multiple sampling periods before switching the communication network, and calculate the mean, standard deviation and data dispersion of the energy efficiency ratio; The environmental monitoring data of the photovoltaic power station corresponding to multiple sampling periods before switching the communication network is obtained, the data dispersion rate of the environmental monitoring data is calculated, and the ratio of the energy efficiency ratio to the data dispersion rate of the environmental monitoring data is set as the confidence coefficient of the standard deviation of the energy efficiency ratio. The dynamic threshold for fault warning is obtained by summing the product of the confidence coefficient and the standard deviation of the energy efficiency ratio with the mean of the energy efficiency ratio. The dynamic threshold is dynamically adjusted based on the environmental monitoring data and the confidence coefficient calculated from the energy efficiency ratio.

6. The centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in claim 5, characterized in that, Based on dynamic thresholds, a first judgment is made on real-time communication data, and a first operation and maintenance result is output to execute the switching of the communication network, including: Calculate the real-time energy efficiency ratio of real-time communication data. When the real-time energy efficiency ratio is greater than the dynamic threshold, the preset first threshold value is incremented once to obtain the real-time energy efficiency ratio at the next moment and continue to compare it with the dynamic threshold for judgment. If the first threshold value is greater than the accumulated threshold multiple times, and the real-time energy efficiency ratios in the continuous sampling period are still greater than the dynamic threshold, the first threshold value is reset and the communication network switching is not performed. When the real-time energy efficiency ratio is less than the dynamic threshold, the real-time energy efficiency ratio of the next sampling period is obtained for judgment. If the real-time energy efficiency ratio of two consecutive sampling periods is less than the dynamic threshold, the preset second threshold value is accumulated once, and the real-time energy efficiency ratio of the next two consecutive sampling periods is obtained for judgment. This continues until the second threshold value is greater than the accumulated threshold, then the target communication network is switched and the fault warning record is cached.

7. The centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in claim 6, characterized in that, After switching communication networks, a second judgment is made on the real-time communication data based on the synchronization rate of change to determine the target communication network a second time and update the dynamic threshold. A second maintenance result is then output, including: The system acquires environmental monitoring data and energy efficiency ratios within multiple consecutive sampling periods before and after switching communication networks, obtains the synchronous change rate between environmental monitoring data and energy efficiency ratios, and makes a judgment based on the linear relationship of the preset synchronous change rate. If the synchronous change rate of data within multiple consecutive sampling periods before and after switching communication networks is non-linear, the system continues to select the target communication network corresponding to the second highest network optional parameter according to the sorting of network optional parameters. After switching communication networks, if the rate of synchronous change of data at consecutive moments before and after is linear, the target communication network is determined for the second time, and the theoretical maximum power generation for calculating the energy efficiency ratio is updated according to the linear relationship of the rate of synchronous change. The correction value of the environmental monitoring data is determined according to the linear relationship of the rate of synchronous change, and the ratio of data dispersion is recalculated based on the corrected environmental monitoring data to update the confidence coefficient of the dynamic threshold. Based on the updated theoretical maximum power generation and confidence coefficient, the dynamic threshold is redefined, and the photovoltaic power station is continuously monitored based on the redefined dynamic threshold. The operation and maintenance results based on fault warning records and communication network switching are output in real time.

8. A centralized monitoring and operation and maintenance system for photovoltaic power plants based on multi-network collaboration and intelligent early warning, characterized in that, The system is used to execute the centralized monitoring and operation and maintenance method for photovoltaic power plants based on multi-network collaboration and intelligent early warning as described in any one of claims 1-7, and the system includes: The data acquisition module is used to continuously acquire environmental monitoring data and communication data of the photovoltaic power station, and determine the energy efficiency ratio of the photovoltaic power station based on the communication data and environmental monitoring data. The communication data includes photovoltaic power station operation data and communication parameters. The network selection module is used to predict the environmental change sequence based on the synchronous change rate of energy efficiency ratio associated with environmental monitoring data, and determine the switchable target communication network based on the neural network prediction model. The first operation and maintenance module is used to determine the dynamic threshold for fault warning based on energy efficiency ratio and environmental monitoring data, make a first judgment on real-time communication data based on the dynamic threshold, and output the first operation and maintenance result to execute the switching of communication network. The second operation and maintenance module is used to make a second judgment on real-time communication data based on the synchronization change rate after switching communication networks, so as to determine the target communication network for the second time, update the dynamic threshold, and output the second operation and maintenance result.

Images