Diagnostic analysis method and system based on digital twinborn technology and intelligent management and control platform

Abstract

The invention discloses a diagnostic analysis method and system based on a digital twinborn technology and an intelligent management and control platform. The method comprises the following steps: acquiring environmental data of a photovoltaic power station in real time and matching an inspection scheme; calling a corresponding inspection device and completing data acquisition according to the inspection scheme by using the intelligent management and control platform; preprocessing the collected data, performing fusion arrangement on the data by utilizing a knowledge graph of the photovoltaic power station, and dividing the data by adopting a preset layering and blocking mode; mapping the divided data to a pre-constructed three-dimensional model of the photovoltaic power station according to a progressive verification and construction strategy by using a digital twinborn technology, and generating a digital twinborn body of the photovoltaic power station; and performing fault prediction and judgment on the photovoltaic power station by utilizing the digital twinborn body of the photovoltaic power station and combining a machine learning algorithm, and obtaining a prediction and judgment result and displaying the prediction and judgment result in the digital twinborn body. According to the invention, the fault of the photovoltaic power station can be diagnosed timely and accurately in a complex or extreme environment.

Description

Technical Field

[0001] The present application relates to the field of digital twin and intelligent management and control technology, and specifically to a diagnostic analysis method and system based on digital twin technology and intelligent management and control platform. Background Art

[0002] With the widespread application of new energy power systems, especially the deployment and application of photovoltaic power generation in remote high-altitude areas, new challenges have been brought to the inspection and operation and maintenance of photovoltaic power stations under complex or extreme conditions. In order to achieve accurate operation and maintenance of photovoltaic power stations, it is necessary to accurately judge the faults of photovoltaic power stations and detect abnormalities in time, especially in complex environmental scenarios.

[0003] In the existing technology, in order to solve the above problems, there are methods that use traditional manual inspections and conventional data analysis tools to realize fault diagnosis of photovoltaic power stations; there are also methods that collect and fuse multi-source data, combine digital twin technology with three-dimensional surveying and mapping technology, and try to build a digital twin of the photovoltaic power station to simulate the operation and analysis in different environments, so as to realize the digital twin three-dimensional monitoring of the photovoltaic power station with full-domain perception, and achieve the technical effect of effective supervision and operation and maintenance of the photovoltaic power station.

[0004] However, the above methods still have some shortcomings in practical applications. First, traditional technologies are inefficient, and a single inspection method and data source can hardly meet the inspection needs in complex environments, and cannot provide comprehensive equipment status information, which leads to errors in the final diagnosis. Secondly, although some systems have begun to adopt digital twin technology, the accuracy of the constructed digital twins is often insufficient due to insufficient accuracy of the collected data or insufficient calibration methods in model construction, resulting in low accuracy of fault judgment and prediction, which is difficult to meet the actual operation and maintenance needs. Therefore, how to diagnose photovoltaic power station faults in a complex environment in a more timely and accurate manner, and then achieve efficient and accurate operation and maintenance management, has become an urgent problem to be solved. Summary of the invention

[0005] In order to achieve more timely and accurate diagnosis of photovoltaic power station faults in complex or extreme environments, and then complete efficient and precise operation and maintenance management, the present application provides a diagnostic analysis method and system based on digital twin technology and intelligent management and control platform.

[0006] In the first aspect, the present application provides a diagnostic analysis method based on digital twin technology and intelligent management and control platform, including: Collect environmental data of photovoltaic power plants in real time and match preset inspection plans based on the collected environmental data; Using the smart control platform, the corresponding inspection device is called according to the inspection combination in the inspection plan, and multi-source data collection is performed on the photovoltaic power station according to the inspection strategy in the inspection plan; the smart control platform integrates multiple inspection plans and controls the operation of the corresponding inspection device according to the inspection plan; Preprocess the collected multi-source data, use the knowledge graph of the photovoltaic power station to fuse and organize the preprocessed multi-source data, and divide the fused and organized data by using preset layers and blocks; the preset layers include: basic environment layer, equipment layer, and system layer; the preset blocks divide each layer of data into multiple data blocks according to the functions of the photovoltaic power station based on the preset layers; Using digital twin technology, the divided data is mapped to a pre-built three-dimensional model of a photovoltaic power station according to a progressive verification and construction strategy to generate a digital twin of the photovoltaic power station; the progressive verification and construction strategy includes: building block by block and layer by layer according to a comprehensive preset sequence of preset layers and blocks, and continuing to build in sequence after verifying the accuracy of the constructed blocks and layers by comparing historical collection data with simulation data; The digital twin of the photovoltaic power station is used in combination with machine learning algorithms to predict and judge faults of the photovoltaic power station. The results of the prediction and judgment are displayed in the digital twin and presented in a visual interface.

[0007] By adopting the above solution, based on the intelligent management and control platform that integrates multiple inspection methods, the environmental data is used to select the appropriate inspection method to achieve accurate collection of photovoltaic power stations; the knowledge graph is used to realize the stratification and blocking of the collected data, and the progressive verification and construction strategy is combined to improve the accuracy of the construction of the digital twin of the photovoltaic power station; the constructed digital twin is used to simulate the operating status of the real photovoltaic power station, and potential risks are predicted through real-time data analysis to improve the accuracy of diagnosis.

[0008] Preferably, the real-time collection of environmental data of the photovoltaic power station and matching of a preset inspection plan according to the collected environmental data include: Obtain historical multi-source data collected by using different preset inspection schemes under different preset environmental data conditions; for each preset environmental data condition, calculate the similarity between the collected historical multi-source data and the data of the historical actual photovoltaic power station under the corresponding environmental data condition, and select the inspection combination scheme with the greatest similarity as the inspection scheme matching the preset environmental data condition; the inspection scheme includes different inspection combinations and their corresponding inspection strategies; the inspection combination includes: drone inspection, surveillance video inspection, robot inspection, and inspection combinations generated by at least two inspection methods in IV, CV, and ID multi-fusion diagnostic analysis; the inspection strategy includes: inspection routes, inspection priorities, and inspection frequencies; The real-time collected environmental data of the photovoltaic power station is compared with the preset environmental data to find the preset environmental data with the highest similarity, and the preset environmental data with the highest similarity is matched with the preset inspection plan as the collected environmental data matching the preset inspection plan.

[0009] By adopting the above solution, the preset optimal inspection plan is matched according to the real-time collected environmental data of the photovoltaic power station, ensuring that each inspection can conduct the most effective inspection based on the current actual environmental conditions, thereby improving the pertinence and efficiency of the inspection.

[0010] Preferably, the preprocessing of the collected multi-source data, fusing and collating the preprocessed multi-source data using the knowledge graph of the photovoltaic power station, and dividing the fused and collated data using a preset layering and block method includes: Divide the photovoltaic power station area and determine the key area of ​​the photovoltaic power station area, deploy a first edge computing node in each divided area, and deploy a second edge computing node and a plurality of third edge computing nodes in the determined key area; The multi-source data collected in each area is preprocessed by the first edge computing node, and the preprocessed multi-source data is transmitted to the second edge computing node. The second edge computing node is used to obtain the relationship data between entities in the knowledge graph of the photovoltaic power station. Based on the extracted relationship data between entities in the knowledge graph, the preprocessed multi-source data is fused and sorted, and the fused and sorted data is processed according to the preset layering, and the layered data is transmitted to the third edge computing node. The third edge computing node is used to process the layered data matching it according to the preset blocks; wherein, several third edge computing nodes are respectively preset to process the layered data matching it.

[0011] By adopting the above solution, multi-level edge computing nodes are set up to achieve collaborative work of multi-level edge computing nodes, quickly respond to data preprocessing, data fusion and overall data division in each area, and effectively reduce data transmission delays and pressure on central servers.

[0012] Preferably, the process of using the digital twin technology to map the divided data to a pre-built three-dimensional model of a photovoltaic power station according to a progressive verification and construction strategy to generate a digital twin of the photovoltaic power station includes: When constructing layer by layer and block by block in accordance with the comprehensive preset order of preset stratification and blocking, and if there is a verification failure in the constructed stratification or blocking, record the digital twin version corresponding to the current constructed stratification or blocking, and based on the number of recorded digital twin versions, when the number is less than the first preset number, use the improved divided data mapping method to re-construct the stratification or blocking; when the number is greater than or equal to the first preset number and less than the second preset number, use the historically collected data to update the pre-constructed three-dimensional model of the photovoltaic power station to re-construct the stratification or blocking; when the number is greater than or equal to the second preset number, use the multi-source data re-acquisition and division method to re-construct the stratification or blocking.

[0013] By adopting the above solution, taking into account the verification failure of the constructed layering or blocking, the versions that failed the verification are recorded and the number of times is counted. The optimization method of the layering or blocking process of different constructions is selected by counting the number of times, so as to improve the high reliability and accuracy of the digital twin.

[0014] Preferably, the method of using the digital twin of the photovoltaic power station in combination with a machine learning algorithm to predict and judge faults of the photovoltaic power station includes: A prediction algorithm is matched from a machine learning algorithm library in combination with environmental data in a basic environment layer in a digital twin of a photovoltaic power station and data characteristics of different blocks divided at a device layer or a system layer in the digital twin of a photovoltaic power station; the machine learning algorithm library stores several prediction algorithms, and each prediction algorithm is designed with block data corresponding to preset environmental data conditions and preset data characteristics that match it; using the matched prediction algorithm, fault prediction and judgment are performed on data of different blocks divided at a device layer or a system layer in a digital twin of a photovoltaic power station.

[0015] By adopting the above solution, the most appropriate prediction algorithm is matched according to different environmental data conditions and data characteristics, so as to improve the accuracy and reliability of fault prediction, thereby achieving more accurate fault diagnosis.

[0016] Preferably, the method of using the digital twin of the photovoltaic power station in combination with a machine learning algorithm to predict and judge the fault of the photovoltaic power station, obtaining the prediction and judgment results and displaying them in the digital twin and presenting them in a visual interface also includes: Regularly predict and judge faults of photovoltaic power stations, use different colors to distinguish the parts that have changed from the judgment results of the previous cycle, and prioritize the fault contents with continuously changing colors at the top of the fault detection list when generating statistics on fault detection lists.

[0017] By adopting the above solution, fault prediction and judgment of photovoltaic power stations are carried out regularly, and the changing trend of equipment status is discovered in time. Different colors are used to distinguish the parts that have changed from the judgment results of the previous cycle, so that operation and maintenance personnel can intuitively identify new or persistent faults, thereby improving the pertinence and efficiency of fault handling; fault contents with continuously changing colors are prioritized at the front of the fault detection list, which helps to prioritize the handling of important faults.

[0018] Preferably, an instruction output by a user to turn on or off the display of periodic fault results is received, and the display of the periodic fault results is adjusted according to the display instruction.

[0019] By adopting the above solution, the display of periodic fault results can be turned on or off at any time according to actual needs, which facilitates flexible management of fault information in different scenarios, avoids interference from irrelevant information, and improves operation and maintenance efficiency.

[0020] In the second aspect, the present application provides a diagnostic analysis system based on digital twin technology and intelligent management and control platform, including: Inspection plan acquisition module, used to collect environmental data of photovoltaic power stations in real time and match preset inspection plans according to the collected environmental data; The inspection data acquisition module is used to use the intelligent management and control platform to call the corresponding inspection device according to the inspection combination in the inspection plan, and collect multi-source data for the photovoltaic power station according to the inspection strategy in the inspection plan; the intelligent management and control platform integrates multiple inspection plans and controls the operation of the corresponding inspection device according to the inspection plan; The inspection data division module is used to pre-process the collected multi-source data, fuse and organize the pre-processed multi-source data using the knowledge graph of the photovoltaic power station, and divide the fused and organized data using a preset layering and block method; the preset layering includes: basic environment layer, equipment layer, and system layer; the preset block is based on the preset layering, and divides each layer of data into multiple data blocks according to the function of the photovoltaic power station; A digital twin generation module is used to map the divided data to a pre-built three-dimensional model of a photovoltaic power station using digital twin technology according to a progressive verification and construction strategy to generate a digital twin of the photovoltaic power station; the progressive verification and construction strategy includes: building block by block and layer by layer according to a comprehensive preset sequence of preset layers and blocks, and continuing to build in sequence after verifying the accuracy of the constructed blocks and layers by comparing historical collection data with simulation data; The fault prediction and judgment module is used to use the digital twin of the photovoltaic power station and combine it with the machine learning algorithm to predict and judge the faults of the photovoltaic power station. The results of the prediction and judgment are displayed in the digital twin and presented in a visual interface.

[0021] By adopting the above solution, more timely and accurate diagnosis of photovoltaic power station faults can be achieved in complex environments.

[0022] In a third aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method as described above.

[0023] In a fourth aspect, the present application provides a computer device, comprising a memory, a processor, and a program stored and executable on the memory, wherein the program implements the steps of the above method when executed by the processor.

[0024] In summary, this application has the following beneficial effects: 1. By utilizing an intelligent management and control platform that integrates multiple inspection methods and matching the appropriate inspection methods with the actual environmental data collected, efficient and accurate collection of multi-source data is achieved. Multi-source data is processed by preset stratification and block methods, and the digital twin of the photovoltaic power station is constructed in combination with progressive verification and construction strategies, achieving more accurate construction of the digital twin. Through digital twin technology, the operating status of a real photovoltaic power station is simulated, and the fault situation of the photovoltaic power station is accurately predicted and judged in combination with machine learning algorithms, which improves the accuracy of equipment status assessment, helps to formulate preventive maintenance strategies, reduce sudden failure rates, and extend equipment service life.

[0025] 2. Before the digital twin construction process, in order to ensure the accuracy of the construction data, multi-level edge computing nodes are designed. According to the collaborative work of multi-level edge computing nodes, the needs of data processing, fusion and division can be quickly responded to; during the data twin construction process, in order to ensure the accuracy of verification and construction, the construction method is adaptively adjusted according to the number of verifications, thereby achieving more accurate digital twin modeling of photovoltaic power stations; 3. When using the digital twin of a photovoltaic power station to diagnose a photovoltaic power station, the most appropriate prediction algorithm is matched according to different environmental data conditions and data characteristics to improve the accuracy and reliability of fault prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the diagnostic analysis method based on digital twin technology and intelligent management and control platform described in a specific embodiment; Figure 2 It is a structural diagram of the diagnostic and analysis system based on digital twin technology and intelligent management and control platform described in the specific embodiment. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0028] Example 1 like Figure 1 As shown, the embodiment of the present application discloses a diagnostic analysis method based on digital twin technology and intelligent management and control platform, and the specific steps include: S1. Collect environmental data of the photovoltaic power station in real time, and match the preset inspection plan according to the collected environmental data.

[0029] Specifically, the environmental data of the photovoltaic power station is collected in real time, including but not limited to temperature, humidity, wind speed, light intensity, pressure, etc. The collection equipment can be various sensors, such as temperature and humidity sensors, anemometers, light sensors, etc. These sensors are distributed in different areas of the photovoltaic power station to ensure the comprehensiveness and accuracy of the data.

[0030] Matching the preset inspection scheme according to the collected environmental data can be achieved through a preset rule base, which includes the preset inspection schemes that need to be adopted within the preset environmental data range. For example, when the ambient temperature is lower than a certain threshold, it will automatically increase the frequency of drone inspections to check whether there is icing or use enhanced protective materials and technologies, such as using a drone with a nano-insulation coating on the shell to improve its thermal insulation performance in low temperature environments; when the ambient humidity is lower than a certain threshold, a robot with an airtight packaging technology shell will be selected to prevent external moisture from entering the internal circuit; when the light intensity is lower than a certain threshold, it will automatically increase the number of video inspections to check whether the photovoltaic modules are blocked; when the ambient wind speed is greater than a certain threshold, it will refuse to select drone inspections or robot inspections to reduce the probability of damage to the inspection device, etc.

[0031] S2. Collect multi-source data of the photovoltaic power station according to the preset inspection plan.

[0032] Specifically, the preset inspection plan includes multiple inspection methods, thereby generating a collection instruction containing a matching preset inspection plan and sending it to the smart management and control platform of the photovoltaic power station. The smart management and control platform integrates multiple inspection plans and controls the corresponding inspection devices to operate according to the inspection plan; the smart management and control platform responds to the collection instruction and calls the corresponding inspection device according to the inspection combination in the inspection plan, and controls the corresponding inspection device to perform multi-source data collection on the photovoltaic power station according to the inspection strategy in the inspection plan.

[0033] The multi-source data of the photovoltaic power station may include: environmental data of the photovoltaic power station, equipment operation data of the photovoltaic power station, image data of the photovoltaic power station, and management data of the photovoltaic power station.

[0034] S3: Preprocess, merge and sort the multi-source data collected from the photovoltaic power station, and perform data division processing.

[0035] Specifically, the multi-source data of the collected photovoltaic power station is first preprocessed, and the preprocessing includes steps such as data cleaning, denoising, and standardization to ensure the quality of the data; for example, a filtering algorithm is used to remove noise, and a normalization method is used to convert the data into the same dimension.

[0036] Secondly, the knowledge graph of the photovoltaic power station is used to fuse and organize the pre-processed multi-source data. The knowledge graph of the photovoltaic power station corresponding to the photovoltaic power station to be diagnosed is obtained, and the entities (equipment or system) of the photovoltaic power station and the relationship between the entities are determined according to the knowledge graph of the photovoltaic power station. The pre-processed multi-source data is associated and mapped with the entities in the knowledge graph to realize the semantic representation of the data (such as attribute information). The association relationship and attribute information in the knowledge graph are used to deeply fuse and organize the multi-source data; for example, the data of the photovoltaic panel can be associated and integrated with the data of the connected inverter to form a fusion data of equipment components.

[0037] Finally, in order to construct the digital twin of the photovoltaic power station layer by layer and block by block in the future, the fused and sorted data is divided by preset stratification and block. Among them, considering that the environment is the basis for the operation of the photovoltaic power station, the basic environment layer is set in advance; after the basic environment is determined, the equipment is the core of the photovoltaic power station, such as photovoltaic panels, inverters and other key equipment, and the equipment layer is set accordingly; finally, in addition to the equipment, the photovoltaic power station also has several system layers, which are responsible for the energy flow and data flow analysis and system optimization of the entire photovoltaic power station, and the system layer is set accordingly. In order to further refine, on the basis of stratification, the functional block data is divided. Specifically, the preset block is to divide each layer of data into multiple data blocks according to the function of the photovoltaic power station; for example, the basic environment layer can be divided into temperature blocks, humidity blocks, wind speed blocks, light intensity blocks and other environmental data blocks; the equipment layer can be divided into photovoltaic panel blocks, inverter blocks, grid-connected equipment blocks, protection equipment blocks and other equipment blocks; the system layer can be divided into power grid interface blocks, energy storage system blocks and other system blocks.

[0038] S4. Using digital twin technology, map the divided data to the pre-built three-dimensional model of the photovoltaic power station according to the progressive verification and construction strategy to generate a digital twin of the photovoltaic power station.

[0039] Specifically, a three-dimensional model of a photovoltaic power station is pre-built using advanced 3D modeling technology and the actual layout and design parameters of the photovoltaic power station. This model not only includes static information such as the building structure and equipment layout of the photovoltaic power station, but also reflects dynamic information such as the operating status and performance of the power station. In order to improve the accuracy and authenticity of the model, historical data collected based on advanced technologies such as oblique photography technology and lidar scanning can be used to assist in modeling and construction.

[0040] After the real-time data collection and processing of the photovoltaic power station is completed and the pre-built three-dimensional model of the photovoltaic power station is completed, the data is mapped to the three-dimensional model through data mapping algorithms and technologies; the entire mapping process involves steps such as spatial positioning, time synchronization and attribute matching of the data. In order to further ensure the accuracy of the generated digital twin, the construction of the digital twin of the photovoltaic power station is realized by adopting a progressive verification and construction strategy; wherein the progressive verification and construction strategy includes: building block by block and layer by layer according to the comprehensive preset order of preset stratification and block, and continuing the sequential construction after verifying the accuracy of the constructed blocks and stratification by comparing the historical collection data with the simulation data.

[0041] The specific process of constructing the digital twin of the photovoltaic power station by using the progressive verification and construction strategy includes: predetermining the comprehensive preset order of the preset layers and blocks of the digital twin of the photovoltaic power station, and the comprehensive preset order of the preset layers and blocks can be set from large to small according to artificial settings or according to the importance of the preset layers and blocks, such as: the basic environment layer can be constructed first, and then the core equipment layer and the system layer can be gradually added. After constructing one or more blocks of the same order in the preset order, the constructed blocks are verified. The specific verification method can be simulated according to the constructed blocks, and various faults can be simulated, and the data collected corresponding to the actual faults in history are compared. If the similarity of the comparison result is greater than the preset similarity threshold, it is determined that the block is constructed successfully, that is, the constructed block is accurate; continue to construct the remaining content in order; otherwise, it is determined that the verification fails and reconstructs, and improvement measures can be taken according to the specific situation, such as changing the mapping method, re-dividing the data, updating the three-dimensional model, etc.; until all the content is constructed, the digital twin of the photovoltaic power station is generated.

[0042] S5. Use the digital twin of the photovoltaic power station in combination with the machine learning algorithm to predict and judge the faults of the photovoltaic power station, obtain the prediction and judgment results and display them.

[0043] Specifically, machine learning algorithms such as decision tree algorithms and neural network algorithms can be used to analyze the data in the digital twin to directly determine whether there is a fault and predict possible faults or give corresponding suggestions. The obtained results are displayed in the digital twin on a visual interface, and the operating status and potential risks of each device can be intuitively seen, which facilitates operation and maintenance personnel to make decisions. Among them, it supports device group-level fault rendering, allowing users to view the fault type by clicking on the group string.

[0044] In addition, in order to facilitate users to view fault information and timely operate and maintain faulty equipment, fault prediction and judgment are performed on photovoltaic power stations regularly, and different colors are used to distinguish the parts that have changed from the judgment results of the previous cycle. When generating fault detection forms based on fault prediction and judgment statistics, the fault contents with continuously changing colors are prioritized at the top of the fault detection forms. In order to prevent a large amount of fault information bombardment, it is convenient for operation and maintenance personnel to choose to view the fault conditions of specific equipment or systems according to their own operation and maintenance content, and then allow users to input the display instructions for turning on or off the periodic fault results, and receive the display instructions for turning on or off the periodic fault results output by the user, and adjust the display of the periodic fault results according to the display instructions.

[0045] Example 2 The difference from the above embodiment is that, in order to further obtain the optimal inspection plan that is more suitable for the environmental data, the historical data of different inspection combinations under different environmental data can be analyzed and verified to determine the optimal inspection plan as the matching preset inspection plan; it also includes: According to the historical environmental data of the photovoltaic power station collected in recent years, a number of different preset environmental data are generated through statistical analysis and corresponding division. Each preset environmental data includes a corresponding preset temperature, a corresponding preset humidity, a corresponding preset wind speed, a corresponding preset light intensity, a corresponding preset pressure, etc., such as: the first preset environmental data, the second preset environmental data and the third preset environmental data.

[0046] Based on historical experience, a variety of preset inspection plans are set up, and the various preset inspection plans correspond to different inspection combinations and their corresponding inspection strategies; the inspection combinations include: drone inspection, surveillance video inspection, robot inspection, and inspection combinations generated by at least two inspection methods in IV, CV, and ID multi-fusion diagnostic analysis; the inspection strategies include: inspection routes, inspection priorities, and inspection frequencies, etc.; such as: a first preset inspection plan including a first inspection combination and a first inspection strategy, a second preset inspection plan including a second inspection combination and a second inspection strategy, and a second preset inspection plan including a second inspection combination and a second inspection strategy.

[0047] The historical multi-source data of the photovoltaic power station collected by using different preset inspection schemes under several different preset environmental data conditions are obtained; for each preset environmental data condition, the similarity between the historical multi-source data of the photovoltaic power station collected and the data of the historical actual photovoltaic power station under the corresponding environmental data condition is calculated, and the inspection combination scheme with the greatest similarity is selected as the preset inspection scheme matching the preset environmental data condition.

[0048] The real-time collected environmental data of the photovoltaic power station is compared with the preset environmental data to find the preset environmental data with the highest similarity, and the preset environmental data with the highest similarity is matched with the preset inspection plan as the collected environmental data matching the preset inspection plan.

[0049] Example 3 The difference from the above embodiment is that in order to further quickly and orderly complete the processing, fusion and division of the multi-source data of the photovoltaic power station collected, combined with edge computing technology, multiple types of edge computing nodes are set to complete data processing in layers and blocks, and also include: The photovoltaic power station area is divided and the key areas of the photovoltaic power station area are determined. A first edge computing node is deployed in each divided area, and a second edge computing node and several third edge computing nodes are deployed in the determined key areas; the multi-source data collected in each area is preprocessed by the first edge computing node, and the preprocessed multi-source data is transmitted to the second edge computing node, and the second edge computing node is used to obtain the relationship data between entities in the knowledge graph of the photovoltaic power station, and the preprocessed multi-source data is fused and sorted based on the extracted relationship data between entities in the knowledge graph, and the fused and sorted data is processed according to preset layers, and the layered data is transmitted to the third edge computing node, and the layered data matching it is processed according to preset blocks by the third edge computing node; wherein, several third edge computing nodes are respectively preset to process the layered data matching it.

[0050] Example 4 The difference from the above embodiment is that, considering that there is a verification failure in the construction of the digital twin technology of the photovoltaic power station, the main reason for the failure may be that there is an error in the mapping data process, the pre-built three-dimensional model is not accurate, or there is a problem with the mapping data itself. In order to ensure the accuracy of the construction of the digital twin technology and improve the construction efficiency, it also includes: It is constructed layer by layer according to the comprehensive preset order of preset layers and blocks, and in the event of verification failure of the constructed layers or blocks, the digital twin version corresponding to the current constructed layers or blocks is recorded, such as: the digital twin versions 1.0, 2.0 and 3.0 corresponding to the photovoltaic panel block.

[0051] Taking into account the multi-source data collected by matching the optimal inspection plan, the data mapping is adjusted first when reconstructing the digital twin, and then the three-dimensional model and the partition data are considered to be updated; specifically, based on the number of recorded digital twin versions, when the number is less than the first preset number, the method of improving the partitioned data mapping is used to re-layer or block the structure; wherein, the method of improving the partitioned data mapping includes: adjusting the elements and attributes that need to be mapped in the digital twin; when the number is greater than or equal to the first preset number and less than the second preset number, the method of updating the pre-built three-dimensional model of the photovoltaic power station using the historically collected data is used to re-layer or block the structure; when the number is greater than or equal to the second preset number, the multi-source data is re-acquired and divided to re-layer or block the structure; wherein, the first preset number is less than the second preset number.

[0052] Example 5 The difference from the above embodiment is that, in order to further improve the accurate judgment of whether the photovoltaic power station is faulty or not, it also includes: combining the environmental data in the basic environment layer in the digital twin of the photovoltaic power station and the data characteristics of different blocks divided by the device layer or system layer in the digital twin of the photovoltaic power station to match the prediction algorithm from the machine learning algorithm library; wherein the different data characteristics include: data characteristics with linear relationships, data characteristics with time series, and data characteristics with classification problems; The machine learning algorithm library stores several prediction algorithms, including: LSTM, SVM, random forest algorithm, etc. with different model parameters; each prediction algorithm is designed with a preset environmental data condition that matches it and block data corresponding to the preset data characteristics; for example: the LSTM algorithm of the first model parameter matches the first environmental data condition and the block data with the data characteristics of a time series; using the matched prediction algorithm, fault prediction and judgment are performed on the data of different blocks divided at the equipment layer or system layer in the digital twin of the photovoltaic power station.

[0053] like Figure 2 As shown, the present application provides a diagnostic analysis system based on digital twin technology and intelligent management and control platform, which specifically includes: Inspection scheme acquisition module 101, used to collect environmental data of the photovoltaic power station in real time, and match a preset inspection scheme according to the collected environmental data; The inspection data acquisition module 102 is used to use the smart control platform to call the corresponding inspection device according to the inspection combination in the inspection plan, and collect multi-source data for the photovoltaic power station according to the inspection strategy in the inspection plan; the smart control platform integrates multiple inspection plans and controls the operation of the corresponding inspection device according to the inspection plan; The inspection data division module 103 is used to pre-process the collected multi-source data, fuse and organize the pre-processed multi-source data using the knowledge graph of the photovoltaic power station, and divide the fused and organized data using a preset layering and block method; the preset layering includes: a basic environment layer, a device layer, and a system layer; the preset block is based on the preset layering, and divides each layer of data into multiple data blocks according to the function of the photovoltaic power station; The digital twin generation module 104 is used to map the divided data to a pre-built three-dimensional model of the photovoltaic power station using the digital twin technology according to a progressive verification and construction strategy to generate a digital twin of the photovoltaic power station; the progressive verification and construction strategy includes: building block by block and layer by layer according to a comprehensive preset sequence of preset layers and blocks, and continuing to build in sequence after verifying the accuracy of the constructed blocks and layers by comparing historical collection data with simulation data; The fault prediction and judgment module 105 is used to use the digital twin of the photovoltaic power station in combination with a machine learning algorithm to predict and judge faults of the photovoltaic power station, obtain the prediction and judgment results, display them in the digital twin and display them in a visual interface.

[0054] The embodiment of the present application also discloses a computer-readable storage medium.

[0055] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executes the diagnostic and analysis method based on the digital twin technology and the intelligent management and control platform as mentioned above. The computer-readable storage medium includes, for example: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0056] The embodiment of the present application also discloses a computer device.

[0057] Specifically, the computer device includes a memory and a processor, and the memory stores a computer program that can be loaded by the processor and execute the above-mentioned diagnostic and analysis method based on digital twin technology and intelligent management and control platform.

[0058] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in this specification (including the abstract and drawings), unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

Claims

1. A diagnostic analysis method based on digital twin technology and intelligent management and control platform, characterized in that: include: Collect environmental data of photovoltaic power plants in real time and match preset inspection plans based on the collected environmental data; Using the smart management and control platform, the corresponding inspection device is called according to the inspection combination in the inspection plan, and multi-source data collection of the photovoltaic power station is carried out according to the inspection strategy in the inspection plan; The smart control platform integrates multiple inspection schemes and controls the operation of corresponding inspection devices according to the inspection schemes; Preprocess the collected multi-source data, use the knowledge graph of the photovoltaic power station to integrate and organize the preprocessed multi-source data, and divide the integrated and organized data using preset stratification and block methods; The preset layers include: basic environment layer, equipment layer, and system layer; the preset blocks divide each layer of data into multiple data blocks based on the preset layers according to the functions of the photovoltaic power station; Using digital twin technology, the divided data is mapped to a pre-built three-dimensional model of a photovoltaic power station according to a progressive verification and construction strategy to generate a digital twin of the photovoltaic power station; the progressive verification and construction strategy includes: building block by block and layer by layer according to a comprehensive preset sequence of preset layers and blocks, and continuing to build in sequence after verifying the accuracy of the constructed blocks and layers by comparing historical collection data with simulation data; The digital twin of the photovoltaic power station is used in combination with machine learning algorithms to predict and judge faults of the photovoltaic power station. The results of the prediction and judgment are displayed in the digital twin and presented in a visual interface.

2. The diagnostic analysis method based on digital twin technology and intelligent management and control platform according to claim 1 is characterized in that: The real-time collection of environmental data of the photovoltaic power station and matching of a preset inspection plan according to the collected environmental data include: Obtain historical multi-source data collected by using different preset inspection schemes under different preset environmental data conditions; for each preset environmental data condition, calculate the similarity between the collected historical multi-source data and the data of the historical actual photovoltaic power station under the corresponding environmental data condition, and select the inspection combination scheme with the greatest similarity as the inspection scheme matching the preset environmental data condition; the inspection scheme includes different inspection combinations and their corresponding inspection strategies; the inspection combination includes: drone inspection, surveillance video inspection, robot inspection, and inspection combinations generated by at least two inspection methods in IV, CV, and ID multi-fusion diagnostic analysis; the inspection strategy includes: inspection routes, inspection priorities, and inspection frequencies; The real-time collected environmental data of the photovoltaic power station is compared with the preset environmental data to find the preset environmental data with the highest similarity, and the preset environmental data with the highest similarity is matched with the preset inspection plan as the collected environmental data matching the preset inspection plan.

3. The diagnostic analysis method based on digital twin technology and intelligent management and control platform according to claim 1 is characterized in that: The preprocessing of the collected multi-source data, fusing and collating the preprocessed multi-source data using the knowledge graph of the photovoltaic power station, and dividing the fused and sorted data using a preset layering and block method include: Divide the photovoltaic power station area and determine the key area of ​​the photovoltaic power station area, deploy a first edge computing node in each divided area, and deploy a second edge computing node and a plurality of third edge computing nodes in the determined key area; The multi-source data collected in each area is preprocessed by the first edge computing node, and the preprocessed multi-source data is transmitted to the second edge computing node. The second edge computing node is used to obtain the relationship data between entities in the knowledge graph of the photovoltaic power station. Based on the extracted relationship data between entities in the knowledge graph, the preprocessed multi-source data is fused and sorted, and the fused and sorted data is processed according to the preset layering, and the layered data is transmitted to the third edge computing node. The third edge computing node is used to process the layered data matching it according to the preset blocks; wherein, several third edge computing nodes are respectively preset to process the layered data matching it.

4. The diagnostic analysis method based on digital twin technology and intelligent management and control platform according to claim 1 is characterized in that: The process of using the digital twin technology to map the divided data to the pre-built three-dimensional model of the photovoltaic power station according to the progressive verification and construction strategy to generate the digital twin of the photovoltaic power station includes: When constructing layer by layer and block by block in accordance with the comprehensive preset order of preset stratification and blocking, and if there is a verification failure in the constructed stratification or blocking, record the digital twin version corresponding to the current constructed stratification or blocking, and based on the number of recorded digital twin versions, when the number is less than the first preset number, use the improved divided data mapping method to re-construct the stratification or blocking; when the number is greater than or equal to the first preset number and less than the second preset number, use the historically collected data to update the pre-constructed three-dimensional model of the photovoltaic power station to re-construct the stratification or blocking; when the number is greater than or equal to the second preset number, use the multi-source data re-acquisition and division method to re-construct the stratification or blocking.

5. The diagnostic analysis method based on digital twin technology and intelligent management and control platform according to claim 1 is characterized in that: The use of the digital twin of the photovoltaic power station in combination with the machine learning algorithm to predict and judge the fault of the photovoltaic power station includes: A prediction algorithm is matched from a machine learning algorithm library in combination with environmental data in a basic environment layer in a digital twin of a photovoltaic power station and data characteristics of different blocks divided at a device layer or a system layer in the digital twin of a photovoltaic power station; the machine learning algorithm library stores several prediction algorithms, and each prediction algorithm is designed with block data corresponding to preset environmental data conditions and preset data characteristics that match it; using the matched prediction algorithm, fault prediction and judgment are performed on data of different blocks divided at a device layer or a system layer in a digital twin of a photovoltaic power station.

6. The diagnostic analysis method based on digital twin technology and intelligent management and control platform according to claim 1 is characterized in that: The method of using the digital twin of the photovoltaic power station in combination with the machine learning algorithm to predict and judge the fault of the photovoltaic power station, obtaining the prediction and judgment results and displaying them in the digital twin and displaying them in a visual interface also includes: Regularly predict and judge faults of photovoltaic power stations, use different colors to distinguish the parts that have changed from the judgment results of the previous cycle, and prioritize the fault contents with continuously changing colors at the top of the fault detection list when generating statistics on fault detection lists.

7. The diagnostic analysis method based on digital twin technology and intelligent management and control platform according to claim 6 is characterized in that: Receive a user-inputted instruction to turn on or off the display of periodic fault results, and adjust the display of periodic fault results according to the display instruction.

8. A diagnostic analysis system based on digital twin technology and intelligent management and control platform, characterized in that: include: Inspection plan acquisition module, used to collect environmental data of photovoltaic power stations in real time and match preset inspection plans according to the collected environmental data; The inspection data acquisition module is used to use the intelligent management and control platform to call the corresponding inspection device according to the inspection combination in the inspection plan, and collect multi-source data for the photovoltaic power station according to the inspection strategy in the inspection plan; The smart control platform integrates multiple inspection schemes and controls the operation of corresponding inspection devices according to the inspection schemes; The inspection data division module is used to pre-process the collected multi-source data, integrate and organize the pre-processed multi-source data using the knowledge graph of the photovoltaic power station, and divide the integrated and organized data using preset stratification and block methods; The preset layers include: basic environment layer, equipment layer, and system layer; the preset blocks divide each layer of data into multiple data blocks based on the preset layers according to the functions of the photovoltaic power station; A digital twin generation module is used to map the divided data to a pre-built three-dimensional model of a photovoltaic power station using digital twin technology according to a progressive verification and construction strategy to generate a digital twin of the photovoltaic power station; the progressive verification and construction strategy includes: building block by block and layer by layer according to a comprehensive preset sequence of preset layers and blocks, and continuing to build in sequence after verifying the accuracy of the constructed blocks and layers by comparing historical collection data with simulation data; The fault prediction and judgment module is used to use the digital twin of the photovoltaic power station and combine it with the machine learning algorithm to predict and judge the faults of the photovoltaic power station. The results of the prediction and judgment are displayed in the digital twin and presented in a visual interface.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A computer device, characterized in that: The computer device comprises a memory, a processor and a program stored and executable on the memory, and the program implements the steps of the method according to any one of claims 1 to 7 when executed by the processor.

Images