Distributed photovoltaic characteristic clustering analysis method based on multi-source data

The method addresses inefficiencies in managing distributed photovoltaic systems by clustering devices based on power generation and humidity data, enabling proactive maintenance and reducing downtime.

CN119066460AActive Publication Date: 2024-12-03STATE GRID SHANDONG ELECTRIC POWER CO
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202411561753.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-05
Publication Date
2024-12-03
Estimated Expiration
2044-11-05

AI Technical Summary

Technical Problem

Distributed photovoltaic systems in urban areas face challenges due to varying environmental conditions leading to dirt and pollution on panels, reducing efficiency, and scattered locations make it difficult for management personnel to promptly identify and maintain abnormal conditions.

Method used

A distributed photovoltaic characteristics clustering analysis method using multi-source data to collect and analyze power generation curves, humidity data, and perform cluster analysis to identify and predict abnormal conditions, facilitating targeted maintenance via email alerts.

Benefits of technology

Enhances management efficiency by grouping similar devices, predicting potential failures, and reducing downtime through centralized and intelligent maintenance strategies.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119066460A_ABST
    Figure CN119066460A_ABST
Patent Text Reader

Abstract

The invention discloses a distributed photovoltaic characteristic clustering analysis method based on multi-source data, and relates to the technical field of photovoltaic power generation operation and maintenance management, and the method comprises the following steps: collecting N generated power curves of different photovoltaic devices in every day in sunny weather for distributed photovoltaic devices under the jurisdiction of a city, and combining the characteristics of the photovoltaic devices to obtain a clustering result; performing clustering analysis on the distributed photovoltaic equipment under jurisdiction; and according to the characteristics of the distributed photovoltaic equipment after clustering analysis, respectively establishing databases, collecting equipment generation power data of different photovoltaic equipment in different databases every day, and setting a humidity sensor based on the geographic position of the equipment to collect humidity data. According to the method, the data science technology and the energy management technology are comprehensively applied, the distributed photovoltaic equipment in the city is managed in a refined and intelligent mode, abnormity of the distributed photovoltaic equipment can be found conveniently through the unified variables, and the management effect and the maintenance effect of the distributed photovoltaic equipment in the city are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power generation operation and maintenance management, and in particular to a distributed photovoltaic characteristic clustering analysis method based on multi-source data. Background Art

[0002] Distributed photovoltaic system refers to a photovoltaic power generation system deployed at the user's site or near the electricity consumption site. This system is usually used for self-generation and self-use, and can meet part or all of the electricity demand. It is usually composed of solar panels, inverters, brackets, cables, etc., and is often installed on the roofs, walls, and ground of urban buildings. However, the environmental factors in cities are relatively complex. The dust and pollution conditions in different areas of the city are different, which can easily lead to dirt on the exposed surface of photovoltaic panels, thereby reducing the light absorption capacity of photovoltaic panels and affecting power generation efficiency. In addition, since the locations of various photovoltaic devices in the distributed photovoltaic system are different and relatively scattered, management personnel cannot promptly determine the abnormal conditions of photovoltaic panels and cannot promptly carry out targeted maintenance on abnormal photovoltaic devices, resulting in low practicality and functionality. Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0003] In view of the problems in the related art, the present invention proposes a distributed photovoltaic characteristics clustering analysis method based on multi-source data to overcome the above technical problems existing in the existing related art.

[0004] To this end, the specific technical solution adopted by the present invention is as follows: A distributed photovoltaic characteristics clustering analysis method based on multi-source data, the method comprising the following steps: S1. For the distributed photovoltaic equipment under the jurisdiction of the city, collect the power generation curves of different photovoltaic equipment every day in N sunny days, and perform cluster analysis on the distributed photovoltaic equipment under the jurisdiction based on the characteristics of the photovoltaic equipment; S2. According to the characteristics of distributed photovoltaic equipment after cluster analysis, databases are established to collect the equipment power generation data of different photovoltaic equipment in different databases every day, and humidity sensors are set based on the geographical location of the equipment to collect humidity data; S3. Based on the power generation of different photovoltaic devices in n cycles in different databases, determine and analyze the abnormal causes of abnormal photovoltaic devices in the same database, including dust accumulation and humidity effects; S4. Mark the abnormal photovoltaic equipment in the database and transmit the abnormal information to the mailbox of the person in charge of the relevant photovoltaic equipment via email.

[0005] As a preferred implementation, S11, for distributed photovoltaic devices under the jurisdiction of the city, collect power generation data of different photovoltaic devices every hour every day in N sunny weather; S12, averaging the power generation data of the same photovoltaic device for N hours, and drawing representative power generation curves of different photovoltaic devices based on the hourly average power generation. The specific steps are as follows: ; in, represents the power generation at hour h on day d, Represents the average power generation of the current photovoltaic device in the hth hour over all N days. The value range of h is 1 to 24, and the value range of d is 1 to N. The current photovoltaic equipment every hour Connect them in sequence to form the daily power generation curve: ; in, That is the representative power generation curve; S13. Based on the representative power generation curves of each photovoltaic device and the model of the photovoltaic device, cluster analysis and grouping of the distributed photovoltaic devices under jurisdiction are performed.

[0006] As a preferred embodiment, the S13 includes the following sub-steps: S131. Collect relevant data of each photovoltaic device, including representative power generation curve and device model, use the representative power generation curve as a time series feature, and extract the daily maximum power generation, minimum power generation, average power generation, and power generation curve slope; S132, preliminarily grouping the photovoltaic devices under jurisdiction by the model of the photovoltaic devices, classifying the photovoltaic devices of the same model into the same group, and then combining the extracted photovoltaic device features into a feature vector as a comprehensive feature representation of each photovoltaic device, and the algorithm formula is: ; in, represents the eigenvector of the ith photovoltaic device, Respectively represent average power, minimum power, maximum power, and slope; S133, grouping all photovoltaic devices under the jurisdiction by K-means clustering and combining the feature vectors of the photovoltaic devices, the specific steps are as follows: For photovoltaic devices in different types of groups, K devices are randomly selected as the initial cluster centers. , for each device , calculate the Euclidean distance to the K cluster centers, and assign them to the nearest cluster center according to the Euclidean distance. The algorithm formula is: ; in, Represents a specific data point The process of assigning to the kth cluster, Representative data points To cluster center distance; S134. According to the allocation result, the centers of the K clusters are recalculated and updated. The algorithm formula is: ; Repeat the steps of assigning and updating until the cluster center no longer changes significantly or the maximum number of iterations is reached. , for each device in different preliminary groups , assigned to the updated nearest cluster center according to the Euclidean distance, and secondary groupings are established based on the number of cluster centers.

[0007] As a preferred embodiment, S2 comprises the following steps: S21. For the first-level groupings obtained by the equipment type, a first-level classification database is established through MySQL. According to the second-level groupings under the first-level groupings, a second-level classification database is established in the first-level classification database. The photovoltaic equipment under the jurisdiction is classified into the second-level classification database based on the second-level groupings, and a photovoltaic equipment file is established separately, which includes the photovoltaic equipment number and the email address of the person in charge; S22, collecting the hourly power generation data of photovoltaic equipment from different secondary classification databases every day, and recording the power generation data with a timestamp in the relevant photovoltaic equipment files; S23, based on the photovoltaic equipment under the secondary classification, the longitude and latitude information of the photovoltaic equipment is obtained through GPS, the location information of the photovoltaic equipment is clustered using a clustering algorithm, the center point of each cluster is selected as a representative location, a third-level classification database is established in the second-level classification database, and the photovoltaic equipment files that meet the clustering are divided into the third-level classification database; S24. A humidity sensor is set at the center point of the cluster to collect the humidity every hour every day in real time and record it in the three-level classification database.

[0008] As a preferred embodiment, S3 comprises the following steps: S31, collecting data on the power generation data of photovoltaic devices in the three-level classification database within n cycles, performing a horizontal comparison on the power generation data of the same hour in the n cycles, and determining the photovoltaic power generation devices to be determined to be abnormal; S32. For the photovoltaic power generation equipment with pending abnormalities, abnormality determination and analysis are performed in combination with the humidity data in the three-level classification database to determine the cause of the abnormality of the photovoltaic equipment.

[0009] As a preferred implementation, the S31 includes the following sub-steps: S311. Count the power generation data of photovoltaic devices in the three-level classification database within n cycles, collect the photovoltaic power generation data of each hour of each day within n cycles in the three-level classification, and calculate the mean and standard deviation of the power generation data of all devices within each hour. The algorithm formula is: ; in, represents the mean, is the standard deviation, m is the total number of photovoltaic devices in the current three-level classification database, Represents the power generated by the jth device in the current hour; S312, calculate the Z-score of each device at the same day and hour in the n-cycle, and make an initial determination of the anomaly based on the Z-score. The algorithm formula is: ; Initialize the abnormality determination coefficient based on cycle n , The initial value of is 0. Greater than threshold When , it means that the photovoltaic power generation of the relevant equipment in the current hour is abnormal, the abnormality determination coefficient is increased by 1, and the abnormal value is recorded as ,right Perform initial abnormality determination every hour to obtain the periodic abnormality determination coefficient ,when When , it means that there is an abnormality in the relevant photovoltaic equipment, which is marked as pending abnormal photovoltaic power generation equipment. h is the determination coefficient, and its value is 8n.

[0010] As a preferred embodiment, the S32 includes the following sub-steps: S321, collect the abnormal hourly humidity values ​​S in the cycle n of the relevant abnormal photovoltaic equipment in the three-level classification database, based on the humidity threshold Make a determination; Set the initialization abnormality determination coefficient , The initial value is 0. When S is greater than the threshold When , it means that the humidity of the relevant equipment is abnormal at the same time, the abnormal determination coefficient is increased by 1, and the abnormal value is recorded as , perform abnormality determination on each abnormal hour of the photovoltaic equipment and obtain the periodic abnormality determination coefficient ; when When , it means that the relevant undetermined abnormal photovoltaic power generation equipment has abnormal humidity, h is the determination coefficient, and its value is two-thirds of the total humidity determination times; S322, when When , it means that the related photovoltaic power generation equipment with pending abnormalities has abnormal dust accumulation.

[0011] As a preferred embodiment, the S4 includes the following sub-steps: S41. Mark the humidity abnormal devices in the third-level database in blue, mark the dust accumulation abnormal data in the third-level database in red, and extract the number of the relevant abnormal devices and the email address of the person in charge; S42. Use the email sending library to send the preset email of the abnormal report to the email address of the person in charge of the relevant abnormal photovoltaic equipment.

[0012] The beneficial effects of the present invention are: The present invention clusters and groups distributed photovoltaic equipment under the jurisdiction of the city through cluster analysis, and groups equipment with similar photovoltaic power generation curves under the same model into the same secondary group, which can achieve more centralized and efficient management and facilitate management personnel to plan and maintain distributed photovoltaic equipment under the jurisdiction of the city; The present invention divides the distributed photovoltaic devices in the secondary group into three levels based on humidity, collects the power generation of each photovoltaic device in the three-level group in the same hour according to the cycle, identifies the photovoltaic devices with abnormal power generation, and can determine the abnormal photovoltaic devices under the same variable to predict the potential failure risk; In the present invention, by collecting humidity data and performing three-level grouping, the humidity abnormality and dust accumulation abnormality of the photovoltaic device can be determined when the photovoltaic device power generation is abnormal. The power generation abnormality caused by humidity or dust accumulation can be quickly identified through unified variables, thereby enhancing the functionality and practicality of the method. Through long-term data accumulation and analysis, the present invention can predict potential humidity anomalies and dust accumulation risks, take preventive maintenance measures in advance, reduce downtime and power generation efficiency losses caused by anomalies, and send the information to the mailbox of the person in charge of the relevant photovoltaic equipment via email, so as to facilitate targeted maintenance of abnormal photovoltaic equipment; The present invention discloses a distributed photovoltaic characteristic clustering analysis method based on multi-source data, which comprehensively applies data science technology and energy management technology to realize refined and intelligent management of distributed photovoltaic equipment in cities, and facilitates the discovery of abnormalities in distributed photovoltaic equipment through unified variables, thereby improving the management and maintenance effects of distributed photovoltaic equipment in cities. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 The present invention is a flowchart of a distributed photovoltaic characteristic clustering analysis method based on multi-source data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0015] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in the field should be able to understand other possible implementations and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0016] According to an embodiment of the present invention, a distributed photovoltaic characteristic clustering analysis method based on multi-source data is provided.

[0017] The present invention is further described with reference to the accompanying drawings and specific embodiments: Example 1: Figure 1 As shown, according to a distributed photovoltaic characteristic clustering analysis method based on multi-source data according to an embodiment of the present invention, the method comprises the following steps: S1. For the distributed photovoltaic equipment under the jurisdiction of the city, collect the power generation curves of different photovoltaic equipment every day in N sunny days, and perform cluster analysis on the distributed photovoltaic equipment under the jurisdiction based on the characteristics of the photovoltaic equipment; S11. For the distributed photovoltaic equipment under the jurisdiction of the city, collect the power generation data of different photovoltaic equipment every hour every day in N sunny weather; S12, averaging the power generation data of the same photovoltaic device for N hours, and drawing representative power generation curves of different photovoltaic devices based on the hourly average power generation. The specific steps are as follows: ; in, represents the power generation at hour h on day d, Represents the average power generation of the current photovoltaic device in the hth hour over all N days. The value range of h is 1 to 24, and the value range of d is 1 to N. The current photovoltaic equipment every hour Connect them in sequence to form the daily power generation curve: ; in, That is the representative power generation curve; It should be noted that the value of N is usually set to 15, and N can be increased or decreased according to actual conditions. Clear weather can be determined based on the weather forecast results in the meteorological forecast. By averaging, the impact of random fluctuations can be effectively reduced when the amount of data is large, and the overall trend of the curve can be extracted to avoid a single anomaly affecting the fitting of the final curve.

[0018] S13, combining the representative power generation curves of each photovoltaic device and the model of the photovoltaic device, performing cluster analysis and grouping of the distributed photovoltaic devices under its jurisdiction; S131. Collect relevant data of each photovoltaic device, including representative power generation curve and device model, use the representative power generation curve as a time series feature, and extract the daily maximum power generation, minimum power generation, average power generation, and power generation curve slope; S132, preliminarily grouping the photovoltaic devices under jurisdiction by the model of the photovoltaic devices, classifying the photovoltaic devices of the same model into the same group, and then combining the extracted photovoltaic device features into a feature vector as a comprehensive feature representation of each photovoltaic device, and the algorithm formula is: ; in, represents the eigenvector of the ith photovoltaic device, Respectively represent average power, minimum power, maximum power, and slope; S133, grouping all photovoltaic devices under the jurisdiction by K-means clustering and combining the feature vectors of the photovoltaic devices, the specific steps are as follows: For photovoltaic devices in different types of groups, K devices are randomly selected as the initial cluster centers. , for each device , calculate the Euclidean distance to the K cluster centers, and assign them to the nearest cluster center according to the Euclidean distance. The algorithm formula is: ; in, Represents a specific data point The process of assigning to the kth cluster, Representative data points To cluster center distance; It should be noted that It is used to find the k value that can minimize the following distance, where the value of k ranges from 1 to K. The Euclidean distance is used for calculation, and the algorithm formula is: ; in, For data points The jth feature of is the cluster center The jth feature of .

[0019] S134. According to the allocation result, the centers of the K clusters are recalculated and updated. The algorithm formula is: ; Repeat the steps of assigning and updating until the cluster center no longer changes significantly or the maximum number of iterations is reached. , for each device in different preliminary groups , assigned to the updated nearest cluster center according to the Euclidean distance, and secondary groupings are established based on the number of cluster centers.

[0020] It should be noted that after the K-means clustering algorithm converges, the subordinate devices are grouped according to the existing clusters, that is, the devices are assigned to the nearest clusters according to the distance between the devices and the cluster center. The final grouping results are used for further data analysis or decision support. The photovoltaic devices after the secondary classification can be understood as having similar working environments, and the power generation values ​​in the same hours under sufficient sunlight are similar.

[0021] S2. According to the characteristics of distributed photovoltaic equipment after cluster analysis, databases are established to collect the equipment power generation data of different photovoltaic equipment in different databases every day, and humidity sensors are set based on the geographical location of the equipment to collect humidity data; S21. For the first-level groupings obtained by the equipment type, a first-level classification database is established through MySQL. According to the second-level groupings under the first-level groupings, a second-level classification database is established in the first-level classification database. The photovoltaic equipment under the jurisdiction is classified into the second-level classification database based on the second-level groupings, and a photovoltaic equipment file is established separately, which includes the photovoltaic equipment number and the email address of the person in charge; S22, collecting the hourly power generation data of photovoltaic equipment from different secondary classification databases every day, and recording the power generation data with a timestamp in the relevant photovoltaic equipment files; S23, based on the photovoltaic equipment under the secondary classification, the longitude and latitude information of the photovoltaic equipment is obtained through GPS, the location information of the photovoltaic equipment is clustered using a clustering algorithm, the center point of each cluster is selected as a representative location, a third-level classification database is established in the second-level classification database, and the photovoltaic equipment files that meet the clustering are divided into the third-level classification database; S24. A humidity sensor is set at the center point of the cluster to collect the humidity every hour every day in real time and record it in the three-level classification database.

[0022] It should be noted that by clustering the geographic information of photovoltaic equipment and selecting representative centers to set up humidity sensors, real-time humidity information of photovoltaic equipment at the cluster can be obtained, which can be used to perform fault judgment and analysis on abnormal photovoltaic equipment during subsequent fault judgment.

[0023] Embodiment 2: S3, based on the power generation of different photovoltaic devices in n cycles in different databases, determining and analyzing the abnormal causes of abnormal photovoltaic devices in the same database, including dust accumulation and humidity effects; S31, collecting power generation data of photovoltaic equipment in the three-level classification database within n cycles, performing horizontal comparison on the power generation data of the same hour in n cycles, and determining the photovoltaic power generation equipment to be determined to be abnormal; S311. Count the power generation data of photovoltaic devices in the three-level classification database within n cycles, collect the photovoltaic power generation data of each hour of each day within n cycles in the three-level classification, and calculate the mean and standard deviation of the power generation data of all devices within each hour. The algorithm formula is: ; in, represents the mean, is the standard deviation, m is the total number of photovoltaic devices in the current three-level classification database, Represents the power generated by the jth device in the current hour; S312, calculate the Z-score of each device at the same day and hour in the n-cycle, and make an initial determination of the anomaly based on the Z-score. The algorithm formula is: ; Initialize the abnormality determination coefficient based on cycle n , The initial value of is 0. Greater than threshold When , it means that the photovoltaic power generation of the relevant equipment in the current hour is abnormal, the abnormality determination coefficient is increased by 1, and the abnormal value is recorded as ,right Perform initial abnormality determination every hour to obtain the periodic abnormality determination coefficient ,when When , it means that there is an abnormality in the relevant photovoltaic equipment, which is marked as pending abnormal photovoltaic power generation equipment. h is the determination coefficient, and its value is 8n.

[0024] It should be noted that The value of can be set by consulting experts in related fields, usually set to 3, n is often set to 7, and can also be adjusted according to actual conditions. h is the determination coefficient, and its value is 8n. By taking the average direct sunlight time per day as 12 hours, minus And multiply It is obtained that due to the longer sunshine hours in summer, the sunshine time is about 14 hours a day, and the shorter sunshine hours in winter, the sunshine time is about 10 hours a day. The data of h can be flexibly adjusted according to the actual conditions of different regions to meet the usage needs of different regions.

[0025] S32, for the abnormal photovoltaic power generation equipment to be determined, the abnormality judgment and analysis are performed in combination with the humidity data in the three-level classification database to determine the abnormal cause of the photovoltaic equipment; S321, collect the abnormal hourly humidity values ​​S in the cycle n of the relevant abnormal photovoltaic equipment in the three-level classification database, based on the humidity threshold Make a determination; Set the initialization abnormality determination coefficient , The initial value is 0. When S is greater than the threshold When , it means that the humidity of the relevant equipment is abnormal at the same time, the abnormal determination coefficient is increased by 1, and the abnormal value is recorded as , perform abnormality determination on each abnormal hour of the photovoltaic equipment and obtain the periodic abnormality determination coefficient ; when When , it means that the relevant undetermined abnormal photovoltaic power generation equipment has abnormal humidity, h is the determination coefficient, and its value is two-thirds of the total humidity determination times; S322, when When , it means that the related photovoltaic power generation equipment with pending abnormalities has abnormal dust accumulation.

[0026] S4. Mark the abnormal photovoltaic equipment in the database, and transmit the abnormal information to the mailbox of the person in charge of the relevant photovoltaic equipment via email; S41. Mark the humidity abnormal devices in the third-level database in blue, mark the dust accumulation abnormal data in the third-level database in red, and extract the number of the relevant abnormal devices and the email address of the person in charge; S42. Use the email sending library to send the preset email of the abnormal report to the email address of the person in charge of the relevant abnormal photovoltaic equipment.

[0027] It should be noted that, according to the different marking colors, two preset emails are set respectively, and the corresponding preset emails are sent to the mailbox of the person in charge through the email module in Python.

[0028] In summary, the present invention clusters and groups distributed photovoltaic equipment under the jurisdiction of the city through cluster analysis, and groups equipment with similar photovoltaic power generation curves under the same model into the same secondary group, which can achieve more centralized and efficient management, and facilitate managers to plan and maintain distributed photovoltaic equipment under the jurisdiction of the city. By grouping the distributed photovoltaic equipment within the secondary group into three levels based on humidity, and collecting the power generation of each photovoltaic equipment in the three-level group in the same hour according to the cycle, photovoltaic equipment with abnormal power generation is identified, and abnormal photovoltaic equipment can be determined under the same variable to predict potential failure risks; By collecting humidity data and grouping them into three levels, the humidity anomaly and dust accumulation anomaly of the photovoltaic equipment can be judged when the photovoltaic equipment has abnormal power generation. The unified variable can be used to quickly identify whether the power generation anomaly is caused by humidity or dust accumulation, thereby enhancing the functionality and practicality of the method. The comprehensive application of data science and energy management technology can realize refined and intelligent management of distributed photovoltaic equipment in the city. The unified variable can be used to facilitate the discovery of anomalies in distributed photovoltaic equipment, thereby improving the management and maintenance effects of distributed photovoltaic equipment in the city.

[0029] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A distributed photovoltaic characteristics clustering analysis method based on multi-source data, It is characterized in that The method comprises the following steps: S1. For the distributed photovoltaic equipment under the jurisdiction of the city, collect the power generation curves of different photovoltaic equipment every day in N sunny days, and perform cluster analysis on the distributed photovoltaic equipment under the jurisdiction based on the characteristics of the photovoltaic equipment; S2. According to the characteristics of distributed photovoltaic equipment after cluster analysis, databases are established to collect the equipment power generation data of different photovoltaic equipment in different databases every day, and humidity sensors are set based on the geographical location of the equipment to collect humidity data; S3. Based on the power generation of different photovoltaic devices in n cycles in different databases, determine and analyze the abnormal causes of abnormal photovoltaic devices in the same database, including dust accumulation and humidity effects; S4. Mark the abnormal photovoltaic equipment in the database and transmit the abnormal information to the mailbox of the person in charge of the relevant photovoltaic equipment via email.

2. The distributed photovoltaic characteristics clustering analysis method based on multi-source data according to claim 1, It is characterized in that The S1 comprises the following sub-steps: S11. For the distributed photovoltaic equipment under the jurisdiction of the city, collect the power generation data of different photovoltaic equipment every hour every day in N sunny weather; S12, averaging the power generation data of the same photovoltaic device for N hours, and drawing representative power generation curves of different photovoltaic devices based on the hourly average power generation. The specific steps are as follows: ; in, represents the power generation at hour h on day d, Represents the average power generation of the current photovoltaic device in the hth hour over all N days. The value range of h is 1 to 24, and the value range of d is 1 to N. The current photovoltaic equipment every hour Connect them in sequence to form the daily power generation curve: ; in, That is the representative power generation curve; S13. Based on the representative power generation curves of each photovoltaic device and the model of the photovoltaic device, cluster analysis and grouping of the distributed photovoltaic devices under jurisdiction are performed.

3. The distributed photovoltaic characteristics clustering analysis method based on multi-source data according to claim 2, It is characterized in that The S13 comprises the following sub-steps: S131. Collect relevant data of each photovoltaic device, including representative power generation curve and device model, use the representative power generation curve as a time series feature, and extract the daily maximum power generation, minimum power generation, average power generation, and power generation curve slope; S132, preliminarily grouping the photovoltaic devices under jurisdiction by the model of the photovoltaic devices, classifying the photovoltaic devices of the same model into the same group, and then combining the extracted photovoltaic device features into a feature vector as a comprehensive feature representation of each photovoltaic device, and the algorithm formula is: ; in, represents the eigenvector of the ith photovoltaic device, Respectively represent average power, minimum power, maximum power, and slope; S133, grouping all photovoltaic devices under the jurisdiction by K-means clustering and combining the feature vectors of the photovoltaic devices, the specific steps are as follows: For photovoltaic devices in different types of groups, K devices are randomly selected as the initial cluster centers. , for each device , calculate the Euclidean distance to the K cluster centers, and assign them to the nearest cluster center according to the Euclidean distance. The algorithm formula is: ; in, Represents a specific data point The process of assigning to the kth cluster, Representative data points To cluster center distance; S134. According to the allocation result, the centers of the K clusters are recalculated and updated. The algorithm formula is: ; Repeat the steps of assigning and updating until the cluster center no longer changes significantly or the maximum number of iterations is reached. , for each device in different preliminary groups , assigned to the updated nearest cluster center according to the Euclidean distance, and secondary groupings are established based on the number of cluster centers.

4. The distributed photovoltaic characteristics clustering analysis method based on multi-source data according to claim 1, It is characterized in that The S2 comprises the following steps: S21. For the first-level groupings obtained by the equipment type, a first-level classification database is established through MySQL. According to the second-level groupings under the first-level groupings, a second-level classification database is established in the first-level classification database. The photovoltaic equipment under the jurisdiction is classified into the second-level classification database based on the second-level groupings, and a photovoltaic equipment file is established separately, which includes the photovoltaic equipment number and the email address of the person in charge; S22, collecting the hourly power generation data of photovoltaic equipment from different secondary classification databases every day, and recording the power generation data with a timestamp in the relevant photovoltaic equipment files; S23, based on the photovoltaic equipment under the secondary classification, the longitude and latitude information of the photovoltaic equipment is obtained through GPS, the location information of the photovoltaic equipment is clustered using a clustering algorithm, the center point of each cluster is selected as a representative location, a third-level classification database is established in the second-level classification database, and the photovoltaic equipment files that meet the clustering are divided into the third-level classification database; S24. A humidity sensor is set at the center point of the cluster to collect the humidity every hour every day in real time and record it in the three-level classification database.

5. The distributed photovoltaic characteristics clustering analysis method based on multi-source data according to claim 1, It is characterized in that The S3 comprises the following steps: S31, collecting data on the power generation data of photovoltaic devices in the three-level classification database within n cycles, performing a horizontal comparison on the power generation data of the same hour in the n cycles, and determining the photovoltaic power generation devices to be determined to be abnormal; S32. For the photovoltaic power generation equipment with pending abnormalities, abnormality determination and analysis are performed in combination with the humidity data in the three-level classification database to determine the cause of the abnormality of the photovoltaic equipment.

6. The distributed photovoltaic characteristics clustering analysis method based on multi-source data according to claim 5, It is characterized in that The S31 includes the following sub-steps: S311. Count the power generation data of photovoltaic devices in the three-level classification database within n cycles, collect the photovoltaic power generation data of each hour of each day within n cycles in the three-level classification, and calculate the mean and standard deviation of the power generation data of all devices within each hour. The algorithm formula is: ; in, represents the mean, is the standard deviation, m is the total number of photovoltaic devices in the current three-level classification database, Represents the power generated by the jth device in the current hour; S312, calculate the Z-score of each device at the same day and hour in the n-cycle, and make an initial determination of the anomaly based on the Z-score. The algorithm formula is: ; Initialize the abnormality determination coefficient based on cycle n , The initial value of is 0. Greater than threshold When , it means that the photovoltaic power generation of the relevant equipment in the current hour is abnormal, the abnormality determination coefficient is increased by 1, and the abnormal value is recorded as ,right Perform initial abnormality determination every hour to obtain the periodic abnormality determination coefficient ,when When , it means that there is an abnormality in the relevant photovoltaic equipment, which is marked as pending abnormal photovoltaic power generation equipment. h is the determination coefficient, and its value is 8n.

7. The distributed photovoltaic characteristics clustering analysis method based on multi-source data according to claim 6, It is characterized in that The S32 includes the following sub-steps: S321, collect the abnormal hourly humidity values ​​S in the cycle n of the relevant abnormal photovoltaic equipment in the three-level classification database, based on the humidity threshold Make a determination; Set the initialization abnormality determination coefficient , The initial value is 0. When S is greater than the threshold When , it means that the humidity of the relevant equipment is abnormal at the same time, the abnormal determination coefficient is increased by 1, and the abnormal value is recorded as , perform abnormality determination on each abnormal hour of the photovoltaic equipment and obtain the periodic abnormality determination coefficient ; when When , it means that the relevant undetermined abnormal photovoltaic power generation equipment has abnormal humidity, h is the determination coefficient, and its value is two-thirds of the total humidity determination times; S322, when When , it means that the related photovoltaic power generation equipment with pending abnormalities has abnormal dust accumulation.

8. The distributed photovoltaic characteristics clustering analysis method based on multi-source data according to claim 1, It is characterized in that The S4 comprises the following sub-steps: S41. Mark the humidity abnormal devices in the third-level database in blue, mark the dust accumulation abnormal data in the third-level database in red, and extract the number of the relevant abnormal devices and the email address of the person in charge; S42. Use the email sending library to send the preset email of the abnormal report to the email address of the person in charge of the relevant abnormal photovoltaic equipment.

Citation Information

Patent Citations

  • Meteorological monitoring point selection method based on photovoltaic power generation output clustering

    CN110008982A

  • Photovoltaic power generation user group division method based on multi-factor mapping

    CN110009385A

  • Photovoltaic output power prediction method and system based on curve characteristic index clustering

    CN114792156A

  • Photovoltaic system current anomaly identification method based on similar day typical curve

    CN117955428A

  • Photovoltaic short-term generation power prediction method based on similar daily clustering and deep learning

    CN117977573A