An AI-driven big data intelligent photovoltaic power generation monitoring method and system

Abstract

The present invention discloses an AI-driven big data intelligent photovoltaic power generation monitoring method and system, which relates to the technical field of photovoltaic systems. The steps of the method include: collecting historical data at each node in the historical audit rotation period of the photovoltaic power generation system through a measuring device; creating a power generation system safety monitoring strategy, calculating the power generation threat value in each historical audit rotation period; establishing a power generation threat prediction model to obtain the power generation threat value of the current audit rotation period; performing fault analysis, and outputting the fault monitoring results of photovoltaic power generation equipment; setting a task scheduling algorithm, and generating an operation and maintenance task list according to the power generation threat value output of the current audit rotation period, including equipment maintenance, cleaning, and replacement tasks, and tracking the execution of tasks. The present invention can accurately predict the power generation threat value of the current audit rotation period based on historical data, providing a scientific basis for operation and maintenance decisions; and adaptively adjusting the strategy according to the current situation to ensure the continuous optimization and improvement of operation and maintenance work.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic systems, and in particular to an AI-driven big data intelligent photovoltaic power generation monitoring method and system. Background Art

[0002] As the global demand for clean energy continues to increase, photovoltaic power generation, as an environmentally friendly, efficient and sustainable energy conversion method, has received widespread attention and rapid development in recent years. Photovoltaic power generation uses the photovoltaic effect to convert solar energy into electrical energy, and has the environmental advantages of low carbon and zero emissions. It is of great significance to alleviate energy pressure, reduce environmental pollution and achieve sustainable development.

[0003] However, photovoltaic power generation systems face many challenges in actual operation. First, the output power of photovoltaic power generation is significantly affected by weather conditions, such as light intensity, temperature, cloud cover, etc., which makes photovoltaic power generation volatile and uncertain, posing challenges to the stable operation of the power grid. Secondly, photovoltaic power stations are usually widely distributed and numerous, and it is difficult to monitor and maintain equipment such as crystal panels and energy storage components. In addition, the system architecture of photovoltaic power stations is huge and the data is complex, which places extremely high demands on the real-time nature of data collection, data processing efficiency and analysis capabilities.

[0004] In order to solve the above problems and improve the performance and stability of photovoltaic power generation systems, artificial intelligence technology has been introduced into photovoltaic power generation monitoring. Through intelligent algorithms and data analysis, real-time monitoring and precise control of photovoltaic power generation systems can be achieved, thereby increasing energy output, reducing maintenance costs and improving the overall efficiency of the system.

[0005] For example, an existing Chinese patent application with publication number CN115225032A discloses a distributed photovoltaic operation and maintenance system and operation and maintenance method. The method collects the current image on the photovoltaic panel, compares the current image with the preset image to determine the dirty point area, and further obtains the current optical power data of the photovoltaic panel. If the ratio of the sum of the current optical power data of all photovoltaic panels to the preset optical power data is higher than the critical optical power ratio, the local cleaning device is controlled to clean the dirty point area; if the ratio of the sum of the current optical power data of all photovoltaic panels to the preset optical power data is lower than or equal to the critical optical power ratio, the global cleaning device is controlled to clean all photovoltaic panels.

[0006] However, the optical power data designed above may fluctuate due to various factors, such as weather changes, equipment aging, etc. Therefore, relying solely on optical power data to determine whether cleaning is needed may not be accurate enough. In addition, the setting of the critical optical power ratio may also need to be adjusted according to actual conditions, which increases the complexity of operation and maintenance.

[0007] In the Chinese patent with the authorization announcement number CN109802634B, a smart operation and maintenance method and operation and maintenance system for photovoltaic power stations based on big data are disclosed, which include the following steps: S1: obtaining data of all dimensions of all photovoltaic power stations within a preset range; S2: segmenting and reorganizing the data of each dimension in S1 to reduce the data granularity and perform dimensionality-upgrading operation on the data; S3: performing association analysis on the dimensionality-upgraded data traversal and calculating the correlation; S4: extracting the data set with a correlation coefficient of not less than 0.6 in S3; S5: constructing an artificial neural network model; S6: based on the data set of S4, using the artificial neural network model in S5 to predict the operation and maintenance status of the photovoltaic power stations within the preset range in the future time period; S7: generating operation and maintenance suggestions based on the prediction results of S6 and the operating status of the photovoltaic power stations within the preset range. This makes up for the shortcomings to a certain extent.

[0008] However, the effectiveness of the method in generating operation and maintenance suggestions based on the prediction results may be affected by many factors, such as the execution ability of the operation and maintenance personnel, the availability of the equipment, etc. If the operation and maintenance suggestions cannot be effectively implemented or there is a fault in the equipment, the operation and maintenance effect may be affected. To this end, the present invention provides an AI-driven big data intelligent photovoltaic power generation monitoring method and system. Summary of the invention

[0009] The purpose of the present invention is to provide an AI-driven big data intelligent photovoltaic power generation monitoring method and system to solve the existing problems raised in the above background technology.

[0010] To achieve the above object, the present invention provides the following technical solution: an AI-driven big data intelligent photovoltaic power generation monitoring, comprising the following steps:

[0011] S1. Collect historical power generation data, grid load data and environmental threat data at each node during the historical audit rotation period of the photovoltaic power generation system through a measuring device;

[0012] S2. Create a power generation system security monitoring strategy and calculate the power generation threat value within each historical audit rotation period;

[0013] S3, establish a power generation threat prediction model, use the historical power generation data, grid load data and environmental threat data as input data, and obtain the power generation threat value of the current audit rotation period;

[0014] S4, performing fault analysis on the data obtained in step S3 and step S2, and outputting fault monitoring results of the photovoltaic power generation equipment;

[0015] S5. Set up a task scheduling algorithm to generate an operation and maintenance task list based on the power generation threat value output of the current audit rotation period, including equipment maintenance, cleaning, and replacement tasks, and track task execution status.

[0016] A further improvement of the present invention is that the power generation data includes current data and voltage data of photovoltaic power generation equipment; the grid load data includes peak power generation load, average load, operating efficiency and equipment temperature; and the environmental threat data includes light intensity, ambient temperature, ambient humidity and wind speed.

[0017] A further improvement of the present invention is that the creation process of the power generation threat prediction model includes: based on the U-NET neural network model, the historical audit rotation period number n, historical power generation data, power grid load data and environmental threat data are combined into the form of feature vectors, and the set of all feature vectors is used as the input of the power generation threat prediction model. The power generation threat prediction model uses the power generation threat value predicted by each group of feature vectors as output, and the actual power generation threat value corresponding to each group of feature vectors as the prediction target, and minimizes the sum of the prediction accuracies of all predicted power generation threat values ​​as the training target, and stops training when the sum of the prediction accuracies converges. The prediction accuracy is obtained by the square of the difference between the predicted value and the actual value of the power generation threat prediction model.

[0018] A further improvement of the present invention is that the power generation threat prediction model also includes recording the average of the power generation threat values ​​in the historical audit rotation period of the photovoltaic power generation system as the threat threshold, and comparing the threat threshold with the power generation threat value in the current audit rotation period; when the power generation threat value in the current audit rotation period is less than the threat threshold, a safety signal is issued; when the power generation threat value in the current audit rotation period is greater than or equal to the threat threshold, the recovery expectation value of the current audit rotation period is calculated, and the recovery expectation value is obtained by calculating the proportion of the number of times the power generation threat value is greater than or equal to the threat threshold in the historical audit rotation period of the photovoltaic power generation system to the total number of times the power generation threat value is calculated in the historical audit rotation period.

[0019] The present invention is further improved in that the power generation threat prediction model further includes: the power generation threat value of the current audit rotation period and the power generation threat value sequence Hpge of the photovoltaic power generation system in the historical audit rotation period are combined. i , respectively as bidders and commodities, and use the task bidding auction algorithm to solve the photovoltaic power generation system fault analysis results; first, Hpge i As an auction sequence, tasks participate in the auction one by one from the first round; in each round of auction, the current audit cycle period power generation threat value bids for the historical audit cycle period power generation threat value, and the bid result is the difference between the income of the current audit cycle period power generation threat value that meets the historical audit cycle period power generation threat value and the sum of the bids for the current audit cycle period power generation threat value that have won the bid; the highest bid result in the current audit cycle period power generation threat value is extracted to obtain the Hpge iThe execution right of one of the tasks participating in the auction; when there is still a power generation threat value in the current review rotation period that has not won the bid, the next round of auction will start in the auction order until all tasks are won, and the auction ends, and the power generation threat value of the current review rotation period is output.

[0020] A further improvement of the present invention is that the power generation system safety monitoring strategy obtains a power generation threat value by standardizing the historical power generation data, power grid load data and environmental threat data at each node in the historical audit rotation period and then performing weighted summation.

[0021] The present invention is further improved in that the specific steps of the task scheduling algorithm include:

[0022] S51. Sort all power generation equipment from high to low according to the power generation threat value of the current audit rotation period; set a first risk threshold and a second risk threshold, classify power generation equipment with a power generation threat value greater than the second risk threshold as high-risk equipment, classify power generation equipment with a power generation threat value greater than or equal to the first risk threshold and less than or equal to the second risk threshold as medium-risk equipment, and classify power generation equipment with a power generation threat value less than the first risk threshold as low-risk equipment;

[0023] S52. Create different task type lists for different risk levels;

[0024] S53, setting initial annealing parameters, including initial temperature, temperature drop rate and termination temperature;

[0025] S54, randomly generating an initial task scheduling plan, including task allocation, execution order and time arrangement of all devices;

[0026] S55. Design objective function (the objective function aims to minimize cost):

[0027]

[0028] Among them, M represents the total amount of tasks, t j represents the actual completion time of task j, d j represents the expected completion time of task j, p j represents the set priority of task j, μ j represents the set contribution after task j is completed, expressed as a numerical value, v j represents the power generation threat value of task j, w t 、w p 、w r and w v represents weight;

[0029] S56, if the objective function is less than the set threshold, a neighborhood search is performed on the current solution to generate a new candidate solution, and the objective function of the new solution is calculated again. If the objective function of the new solution is greater than or equal to the objective function of the old solution, the new solution is accepted;

[0030] S57, updating the current temperature according to the temperature drop rate, where the current temperature is expressed as the product of the temperature drop rate and the previous temperature;

[0031] S58. Repeat step S56 and step S57 until the temperature is lower than the termination temperature, and output the currently accepted best task scheduling solution.

[0032] On the other hand, the present invention provides an AI-driven big data intelligent photovoltaic power generation monitoring system, comprising:

[0033] The data collection module collects historical power generation data, grid load data and environmental threat data at each node during the historical audit rotation period of the photovoltaic power generation system through a measuring device;

[0034] The power generation security threat extraction module creates a power generation system security monitoring strategy and calculates the power generation threat value within each historical audit rotation period;

[0035] A power generation security threat calculation module is used to establish a power generation threat prediction model, and the historical power generation data, grid load data and environmental threat data are used as input data to obtain the power generation threat value of the current audit rotation period;

[0036] A detection result output module performs fault analysis on the data obtained in the power generation safety threat extraction module and the power generation safety threat calculation module, and outputs the photovoltaic power generation equipment fault monitoring result;

[0037] The task scheduling module sets the task scheduling algorithm, generates an operation and maintenance task list based on the power generation threat value output of the current audit rotation period, including equipment maintenance, cleaning, and replacement tasks, and tracks the task execution status.

[0038] A further improvement of the present invention is that the power generation safety threat calculation module includes a recovery expectation calculation unit and a power generation threat output unit. The recovery expectation calculation unit is used to calculate the recovery expectation value of the current review rotation period; the power generation threat output unit is used to use the task bidding auction algorithm to solve the photovoltaic power generation system fault analysis result and output the power generation threat value of the current review rotation period.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. The present invention firstly realizes real-time monitoring and evaluation of the safety status of the photovoltaic power generation system by creating a power generation system safety monitoring strategy and calculating the power generation threat value in each historical audit rotation period;

[0041] 2. Establish a power generation threat prediction model that can accurately predict the power generation threat value of the current audit rotation period based on historical data, providing a scientific basis for operation and maintenance decisions;

[0042] 3. By comparing the prediction results with historical data, the fault monitoring results of photovoltaic power generation equipment are output, which improves the accuracy and timeliness of fault detection;

[0043] 4. Set up a task scheduling algorithm to automatically generate an operation and maintenance task order based on the power generation threat value, and track the task execution status to ensure the efficient and orderly progress of the operation and maintenance work. Use the simulated annealing algorithm for task scheduling, and be able to adaptively adjust the strategy according to the current situation to ensure continuous optimization and improvement of the operation and maintenance work. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a flow chart of an AI-driven big data intelligent photovoltaic power generation monitoring method of the present invention;

[0045] Figure 2 This is a task scheduling algorithm flow chart of an AI-driven big data intelligent photovoltaic power generation monitoring method of the present invention;

[0046] Figure 3 This is a framework diagram of an AI-driven big data intelligent photovoltaic power generation monitoring system of the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0048] The term "and / or" is only a description of the association relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the related objects are in an "or" relationship.

[0049] Example 1

[0050] Figure 1 The flowchart of an AI-driven big data intelligent photovoltaic power generation monitoring method disclosed in this embodiment is shown, and the steps are as follows:

[0051] S1. Collect historical power generation data, grid load data and environmental threat data at each node during the historical audit rotation period of the photovoltaic power generation system through a measuring device;

[0052] S2. Create a power generation system security monitoring strategy and calculate the power generation threat value within each historical audit rotation period;

[0053] S3, establish a power generation threat prediction model, use the historical power generation data, grid load data and environmental threat data as input data, and obtain the power generation threat value of the current audit rotation period;

[0054] S4, performing fault analysis on the data obtained in step S3 and step S2, and outputting fault monitoring results of the photovoltaic power generation equipment;

[0055] S5. Set up a task scheduling algorithm to generate an operation and maintenance task list based on the power generation threat value output of the current audit rotation period, including equipment maintenance, cleaning, and replacement tasks, and track task execution status.

[0056] The power generation data includes current data and voltage data of photovoltaic power generation equipment; the grid load data includes peak power generation load, average load, operating efficiency and equipment temperature; the environmental threat data includes light intensity, ambient temperature, ambient humidity and wind speed.

[0057] The creation process of the power generation threat prediction model includes: based on the U-NET neural network model, the historical audit rotation period number n, historical power generation data, power grid load data and environmental threat data are combined into the form of feature vectors, and the set of all feature vectors is used as the input of the power generation threat prediction model. The power generation threat prediction model uses the power generation threat value predicted by each group of feature vectors as the output, and the actual power generation threat value corresponding to each group of feature vectors as the prediction target. Minimizing the sum of the prediction accuracies of all predicted power generation threat values ​​is used as the training target, and training is stopped until the sum of the prediction accuracies converges. The prediction accuracy is obtained by the square of the difference between the predicted value and the actual value of the power generation threat prediction model.

[0058] The power generation threat prediction model also includes recording the average of the power generation threat values ​​in the historical review rotation period of the photovoltaic power generation system as the threat threshold, and comparing the threat threshold with the power generation threat value in the current review rotation period; when the power generation threat value in the current review rotation period is less than the threat threshold, a safety signal is issued; when the power generation threat value in the current review rotation period is greater than or equal to the threat threshold, the expected recovery value of the current review rotation period is calculated, and the expected recovery value is obtained by calculating the proportion of the number of times the power generation threat value is greater than or equal to the threat threshold in the historical review rotation period of the photovoltaic power generation system to the total number of times the power generation threat value is calculated in the historical review rotation period.

[0059] The power generation threat prediction model also includes: combining the power generation threat value of the current audit rotation period and the power generation threat value sequence Hpge of the photovoltaic power generation system in the historical audit rotation period i, respectively as bidders and commodities, and use the task bidding auction algorithm to solve the photovoltaic power generation system fault analysis results; first, Hpge i As an auction sequence, tasks participate in the auction one by one from the first round; in each round of auction, the current audited rotation period power generation threat value bids for the historical audited rotation period power generation threat value, and the bid result is the difference between the revenue of the current audited rotation period power generation threat value that meets the historical audited rotation period power generation threat value and the sum of the winning bids for the current audited rotation period power generation threat value;

[0060] The grid load data is weighted and summed to obtain the grid load value grl, the environmental threat data is weighted and summed to obtain the environmental threat value ent, and the grid load standard value sgrl and the environmental threat standard value sent are set;

[0061] When the grid load value is greater than the grid load standard value and the environmental threat value is less than the environmental threat standard value, it indicates that the future weather conditions are good and the grid load is high. The bidding price of the photovoltaic power generation system can be appropriately increased by setting the bidding coefficient to multiply the original bidding price in order to obtain more power generation and revenue. The bidding coefficient calculation formula is:

[0062]

[0063] When the grid load value is less than or equal to the grid load standard value or the environmental threat value is greater than or equal to the environmental threat standard value, it indicates that the weather conditions are bad or the grid load is low. The bid should be appropriately reduced by multiplying the original bid by the bid coefficient to reduce the economic losses caused by insufficient power generation capacity or potential risks.

[0064] If a critical equipment failure or maintenance is detected, the bid should be immediately reduced to 50% of the original bid, and the bidding strategy should be adjusted based on the system recovery after the maintenance is completed.

[0065] Extract the highest bid result in the current audit rotation period power generation threat value to obtain the Hpge i The execution right of one of the tasks participating in the auction; when there is still a power generation threat value in the current review rotation period that has not won the bid, the next round of auction will start in the auction order until all tasks are won, and the auction ends, and the power generation threat value of the current review rotation period is output.

[0066] The benefit is usually a mathematical expression, which calculates a specific value or indicator based on the difference between the system safety value of the current period and the historical period to represent the economic or safety benefits brought about by this difference.

[0067] The power generation system safety monitoring strategy obtains the power generation threat value by standardizing the historical power generation data, power grid load data and environmental threat data at each node during the historical audit rotation period and then performing weighted summation.

[0068] Example 2

[0069] Figure 2 The task scheduling algorithm flow chart of an AI-driven big data intelligent photovoltaic power generation monitoring method disclosed in this embodiment is shown. The specific steps of the task scheduling algorithm include:

[0070] S51. Sort all power generation equipment from high to low according to the power generation threat value of the current audit rotation period; set a first risk threshold and a second risk threshold, classify power generation equipment with a power generation threat value greater than the second risk threshold as high-risk equipment, classify power generation equipment with a power generation threat value greater than or equal to the first risk threshold and less than or equal to the second risk threshold as medium-risk equipment, and classify power generation equipment with a power generation threat value less than the first risk threshold as low-risk equipment;

[0071] S52. Create different task type lists for different risk levels;

[0072] S53, setting initial annealing parameters, including initial temperature, temperature drop rate and termination temperature;

[0073] S54, randomly generating an initial task scheduling plan, including task allocation, execution order and time arrangement of all devices;

[0074] S55. Design the objective function, which aims to minimize the cost:

[0075]

[0076] Among them, M represents the total amount of tasks, t j represents the actual completion time of task j, d j represents the expected completion time of task j, p j represents the set priority of task j, μ j represents the set contribution after task j is completed, expressed as a numerical value, v j represents the power generation threat value of task j, w t 、w p 、w r and w v represents weight;

[0077] t j -d j | represents the time deviation for completing task j. The absolute value indicates that there is a cost whether it is early or late. Used to calculate the degree of violation of task priority; when a task is delayed (t j -d j >0), the higher the priority of the task (p j The smaller the priority, the higher the cost. If the task is completed ahead of time, this cost is not calculated.j Indicates the contribution of resource utilization. In this embodiment, it is assumed that μ j The larger the value, the higher the resource utilization rate (if it is a negative number, it means resource waste), so we use a negative sign to convert it into a cost item, v j Directly represents the impact of the power generation threat value. The larger the value, the higher the cost.

[0078] S56, if the objective function is less than the set threshold, a neighborhood search is performed on the current solution to generate a new candidate solution, and the objective function of the new solution is calculated again. If the objective function of the new solution is greater than or equal to the objective function of the old solution, the new solution is accepted;

[0079] S57, updating the current temperature according to the temperature drop rate, where the current temperature is expressed as the product of the temperature drop rate and the previous temperature;

[0080] S58. Repeat step S56 and step S57 until the temperature is lower than the termination temperature, and output the currently accepted best task scheduling solution.

[0081] The threshold and weight may be set by default according to the present invention, or may be set by an operator.

[0082] Example 3

[0083] Figure 3 The present invention shows an AI-driven big data intelligent photovoltaic power generation monitoring system framework diagram. Based on the same inventive concept as Example 1, the present invention provides an AI-driven big data intelligent photovoltaic power generation monitoring system, including:

[0084] The data collection module collects historical power generation data, grid load data and environmental threat data at each node during the historical audit rotation period of the photovoltaic power generation system through a measuring device;

[0085] The power generation security threat extraction module creates a power generation system security monitoring strategy and calculates the power generation threat value within each historical audit rotation period;

[0086] A power generation security threat calculation module is used to establish a power generation threat prediction model, and the historical power generation data, grid load data and environmental threat data are used as input data to obtain the power generation threat value of the current audit rotation period;

[0087] A detection result output module performs fault analysis on the data obtained in the power generation safety threat extraction module and the power generation safety threat calculation module, and outputs the photovoltaic power generation equipment fault monitoring result;

[0088] The task scheduling module sets the task scheduling algorithm, generates an operation and maintenance task list based on the power generation threat value output of the current audit rotation period, including equipment maintenance, cleaning, and replacement tasks, and tracks the task execution status.

[0089] The power generation security threat calculation module includes a recovery expectation calculation unit and a power generation threat output unit;

[0090] The recovery expectation calculation unit is used to calculate the recovery expectation value of the current audit rotation period;

[0091] The power generation threat output unit is used to use the task bidding auction algorithm to solve the photovoltaic power generation system fault analysis result and output the power generation threat value of the current audit rotation period.

[0092] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0093] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0094] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0095] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1The steps for the functions specified in one or more boxes.

[0096] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. An AI-driven big data intelligent photovoltaic power generation monitoring method, characterized in that: The following steps are involved: S1. Collect historical power generation data, grid load data and environmental threat data at each node during the historical audit rotation period of the photovoltaic power generation system through a measuring device; S2. Create a power generation system security monitoring strategy and calculate the power generation threat value within each historical audit rotation period; S3, establish a power generation threat prediction model, use the historical power generation data, grid load data and environmental threat data as input data, and obtain the power generation threat value of the current audit rotation period; S4, performing fault analysis on the data obtained in step S3 and step S2, and outputting fault monitoring results of the photovoltaic power generation equipment; S5. Set up a task scheduling algorithm, generate an operation and maintenance task list based on the power generation threat value output of the current audit rotation period, and track the task execution status; The specific steps of the task scheduling algorithm include: S51. Sort all power generation equipment from high to low according to the power generation threat value of the current audit rotation period; set a first risk threshold and a second risk threshold, classify power generation equipment with a power generation threat value greater than the second risk threshold as high-risk equipment, classify power generation equipment with a power generation threat value greater than or equal to the first risk threshold and less than or equal to the second risk threshold as medium-risk equipment, and classify power generation equipment with a power generation threat value less than the first risk threshold as low-risk equipment; S52. Create different task type lists for different risk levels; S53, setting initial annealing parameters, including initial temperature, temperature drop rate and termination temperature; S54, randomly generating an initial task scheduling plan, including task allocation, execution order and time arrangement of all devices; S55. Design objective function: Among them, M represents the total amount of tasks, t j represents the actual completion time of task j, d j represents the expected completion time of task j, p j represents the set priority of task j, μ j represents the set contribution after task j is completed, expressed as a numerical value, v j represents the power generation threat value of task j, w t 、w p 、w r and w v represents weight; S56, if the objective function is less than the set threshold, a neighborhood search is performed on the current solution to generate a new candidate solution, and the objective function of the new solution is calculated again. If the objective function of the new solution is greater than or equal to the objective function of the old solution, the new solution is accepted; S57, updating the current temperature according to the temperature drop rate, where the current temperature is expressed as the product of the temperature drop rate and the previous temperature; S58. Repeat step S56 and step S57 until the temperature is lower than the termination temperature, and output the currently accepted best task scheduling solution.

2. The AI-driven big data intelligent photovoltaic power generation monitoring method according to claim 1, characterized in that: The power generation data includes current data and voltage data of photovoltaic power generation equipment; the grid load data includes peak power generation load, average load, operating efficiency and equipment temperature; the environmental threat data includes light intensity, ambient temperature, ambient humidity and wind speed.

3. The AI-driven big data intelligent photovoltaic power generation monitoring method according to claim 2 is characterized in that: The creation process of the power generation threat prediction model includes: based on the U-NET neural network model, the historical audit rotation period number n, historical power generation data, power grid load data and environmental threat data are combined into the form of feature vectors, and the set of all feature vectors is used as the input of the power generation threat prediction model. The power generation threat prediction model uses the power generation threat value predicted by each group of feature vectors as the output, and the actual power generation threat value corresponding to each group of feature vectors as the prediction target. Minimizing the sum of the prediction accuracies of all predicted power generation threat values ​​is used as the training target, and training is stopped until the sum of the prediction accuracies converges. The prediction accuracy is obtained by the square of the difference between the predicted value and the actual value of the power generation threat prediction model.

4. The AI-driven big data intelligent photovoltaic power generation monitoring method according to claim 3 is characterized by: The power generation threat prediction model also includes recording the average of the power generation threat values ​​in the historical review rotation period of the photovoltaic power generation system as the threat threshold, and comparing the threat threshold with the power generation threat value in the current review rotation period; when the power generation threat value in the current review rotation period is less than the threat threshold, a safety signal is issued; when the power generation threat value in the current review rotation period is greater than or equal to the threat threshold, the expected recovery value of the current review rotation period is calculated, and the expected recovery value is obtained by calculating the proportion of the number of times the power generation threat value is greater than or equal to the threat threshold in the historical review rotation period of the photovoltaic power generation system to the total number of times the power generation threat value is calculated in the historical review rotation period.

5. The AI-driven big data intelligent photovoltaic power generation monitoring method according to claim 4 is characterized in that: The power generation threat prediction model also includes: combining the power generation threat value of the current audit rotation period and the power generation threat value sequence Hpge of the photovoltaic power generation system in the historical audit rotation period i , respectively as bidders and commodities, and use the task bidding auction algorithm to solve the photovoltaic power generation system fault analysis results; first, Hpge i As an auction sequence, tasks participate in the auction one by one from the first round; in each round of auction, the current audit cycle period power generation threat value bids for the historical audit cycle period power generation threat value, and the bid result is the difference between the income of the current audit cycle period power generation threat value that meets the historical audit cycle period power generation threat value and the sum of the bids for the current audit cycle period power generation threat value that have won the bid; the highest bid result in the current audit cycle period power generation threat value is extracted to obtain the Hpge i The execution right of one of the tasks participating in the auction; when there is still a power generation threat value in the current review rotation period that has not won the bid, the next round of auction will start in the auction order until all tasks are won, and the auction ends, and the power generation threat value of the current review rotation period is output.

6. The AI-driven big data intelligent photovoltaic power generation monitoring method according to claim 5, characterized in that: The power generation system safety monitoring strategy obtains the power generation threat value by standardizing the historical power generation data, power grid load data and environmental threat data at each node during the historical audit rotation period and then performing weighted summation.

7. An AI-driven big data intelligent photovoltaic power generation monitoring system, used to execute an AI-driven big data intelligent photovoltaic power generation monitoring method as claimed in any one of claims 1 to 6, characterized in that: include: The data collection module collects historical power generation data, grid load data and environmental threat data at each node during the historical audit rotation period of the photovoltaic power generation system through a measuring device; The power generation security threat extraction module creates a power generation system security monitoring strategy and calculates the power generation threat value within each historical audit rotation period; A power generation security threat calculation module is used to establish a power generation threat prediction model, and the historical power generation data, grid load data and environmental threat data are used as input data to obtain the power generation threat value of the current audit rotation period; A detection result output module performs fault analysis on the data obtained in the power generation safety threat extraction module and the power generation safety threat calculation module, and outputs the photovoltaic power generation equipment fault monitoring result; The task scheduling module sets the task scheduling algorithm, generates an operation and maintenance task order based on the power generation threat value output of the current audit rotation period, and tracks the task execution status.

8. The AI-driven big data intelligent photovoltaic power generation monitoring system according to claim 7, characterized in that: The power generation safety threat calculation module includes a recovery expectation calculation unit and a power generation threat output unit. The recovery expectation calculation unit is used to calculate the recovery expectation value of the current review rotation period; the power generation threat output unit is used to use the task bidding auction algorithm to solve the photovoltaic power generation system fault analysis result and output the power generation threat value of the current review rotation period.

Images