Monitoring method and system based on direct communication of photovoltaic modules, terminal and medium

Abstract

The invention relates to a monitoring method and system based on direct communication of a photovoltaic module, a terminal and a medium, and relates to the field of photovoltaic technology, and the method comprises the steps: obtaining the operation parameters of a photovoltaic intelligent module through a data detection module, and the operation parameters comprise the operation voltage, the operation current and the operation temperature; the operation parameters are transmitted to the data concentrator through the communication module; transmitting the operation parameters to a cloud server through a data concentrator; analyzing the operation parameters through the cloud server to obtain a fault weight score; judging whether the fault weight score meets a preset condition or not; if not, obtaining a control instruction according to the fault weight score; and the control instruction is sent to the photovoltaic intelligent assembly through the cloud server. The method has the effect of improving the troubleshooting efficiency of the photovoltaic module.

Description

Technical Field

[0001] This application relates to the field of photovoltaic technology, and in particular to a monitoring method, system, terminal and medium based on direct communication between photovoltaic modules. Background Technology

[0002] With the large-scale construction and widespread application of photovoltaic power generation systems, the number of photovoltaic modules inside photovoltaic power plants is constantly increasing. The ability to monitor their operating status in real time and diagnose faults is directly related to the overall power generation efficiency and operational safety of the power plant.

[0003] In the current photovoltaic monitoring technology system, both mainstream solutions face insurmountable bottlenecks: First, the monitoring architecture centered on string inverters is limited by the data acquisition granularity, which can only cover the "string level." This makes it impossible to accurately locate faults in individual photovoltaic modules (such as hot spots, partial shading, individual module attenuation, wiring abnormalities, etc.), resulting in fault diagnosis relying on manual inspection of each module, which is inefficient and prone to overlooking hidden problems. Second, the module-level monitoring solution based on microinverters (micro-inverters) or power optimizers, although achieving module-level data acquisition, requires the additional deployment of a large number of micro-inverter / optimizer hardware. This not only increases the system deployment cost but also increases the installation and commissioning difficulty and subsequent operation and maintenance complexity in complex scenarios (such as distributed rooftops and BIPV building-integrated photovoltaics) due to the redundancy of access device nodes. More importantly, both of the above solutions belong to the "indirect communication paradigm"—the photovoltaic modules themselves do not have active communication capabilities and need to rely on third-party intermediate devices such as string inverters, micro-inverters, and power optimizers to relay and collect data. This not only introduces data transmission delay and information loss, but also makes the reliability of the monitoring system highly dependent on the stable operation of the intermediate devices, posing a risk of "intermediate device failure leading to module data disconnection". Summary of the Invention

[0004] To improve the efficiency of troubleshooting photovoltaic modules, this application provides a monitoring method, system, terminal, and medium based on direct communication between photovoltaic modules.

[0005] In a first aspect, this application provides a monitoring method, system, terminal, and medium based on direct communication between photovoltaic modules, employing the following technical solution: A monitoring method based on direct communication between photovoltaic modules includes: The operating parameters of the photovoltaic smart module are obtained through the data detection module. These operating parameters include operating voltage, operating current, and operating temperature. The operating parameters are transmitted to the data concentrator via the communication module; The runtime parameters are transmitted to the cloud server via a data concentrator; By analyzing the operating parameters through a cloud server, a fault weight score is obtained. Determine whether the fault weight score meets the preset conditions; If not, then control commands are obtained based on the fault weight score; Control commands are sent to the photovoltaic smart modules via a cloud server.

[0006] By adopting the above technical solution, the direct collection and uploading of operating parameters of photovoltaic smart modules is achieved, and the operating parameters are analyzed on the cloud server side. Fine-grained quantitative analysis of the operating deviations between the module itself and its neighbors is possible. This eliminates reliance on string-level or third-party intermediate devices for indirect diagnosis, reducing problems such as diagnostic link interruptions, data distortion, and the inability to accurately locate module-level faults. Fault judgment logic can be directly applied to the individual photovoltaic module level. Therefore, the efficiency of photovoltaic module fault diagnosis is improved, enabling rapid identification and handling of module-level faults.

[0007] Optionally, the operating parameters can be analyzed via a cloud server to obtain a fault weight score, including: Each photovoltaic smart module is assigned corresponding position coordinates based on its physical layout. Establish a data box for the current moment for photovoltaic smart modules based on their location coordinates; Write the location coordinates and operating parameters into the data box as data to be processed; Based on the data to be processed, perform neighborhood data statistical processing to obtain the structural data to be analyzed; The structural data is sent to the weighted fault judgment system for fault weighting calculation to obtain the fault weight score.

[0008] By adopting the above technical solution, the cloud server can automatically generate location coordinates based on the physical layout of the photovoltaic smart modules, and construct a data box containing time-series operational data, giving each photovoltaic module a clear spatial identifier and data carrier within the system. By performing neighborhood data statistics based on the data box, the operational deviation characteristics of the target module relative to its neighborhood can be extracted. Furthermore, the structured data is input into a weighted fault judgment system for quantitative calculation, obtaining a precise fault weight score. This achieves fine-grained evaluation of the module-level operational status, thereby improving the accuracy and efficiency of fault diagnosis and troubleshooting for photovoltaic smart modules.

[0009] Optionally, neighborhood data statistical processing is performed based on the data to be processed to obtain the structured data to be analyzed, including: Based on the target location coordinates recorded in the data box of the target photovoltaic smart module, obtain the photovoltaic smart modules within a preset range and obtain a set of neighboring candidate modules; Remove neighboring candidate components with negative coordinate indices from the set of neighboring candidate components to obtain the set of neighboring components; Extract the running parameters from the data boxes corresponding to the neighboring components in the neighboring component set to obtain the neighboring running data; The neighborhood mean and neighborhood variance of the neighborhood operation data are obtained according to the preset neighborhood data processing method. The neighborhood mean, neighborhood variance, and current acquisition time are written back to the data box corresponding to the target photovoltaic smart module to obtain the structural data to be analyzed.

[0010] By adopting the above technical solution, based on obtaining the location coordinates of the target photovoltaic smart module, photovoltaic modules within a preset range around it can be screened, and invalid modules with negative coordinate indices can be eliminated. Furthermore, operating parameters can be uniformly extracted from the neighboring module set, and the neighborhood mean and variance can be calculated to obtain the neighborhood operating status. Finally, the above statistical information, along with the current acquisition time, is written back to the target module's data box, forming structured data that can be directly used for subsequent judgment processes. This improves the accuracy and stability of neighborhood data aggregation and enhances the accuracy of input data for subsequent fault judgment modules.

[0011] Optionally, the neighborhood mean and neighborhood variance of the neighborhood operation data are obtained according to a preset neighborhood data processing method, including: Obtain the set of operating voltage, operating current, and operating temperature from the neighborhood operating data; The mean values ​​of the operating voltage set, operating current set, and operating temperature set are calculated separately to obtain the neighborhood mean values ​​of operating voltage, operating current, and operating temperature. The neighborhood mean values ​​of operating voltage, operating current, and operating temperature are integrated to obtain the neighborhood mean value. The variance of the operating voltage neighborhood is calculated based on the mean of the operating voltage neighborhood. The variance of the operating current neighborhood is calculated based on the neighborhood mean of the operating current. The variance of the operating temperature neighborhood is calculated based on the neighborhood mean of the operating temperature. The neighborhood variance is obtained by integrating the neighborhood variances of operating voltage, operating current, and operating temperature.

[0012] By adopting the above technical solution, the mean and variance of operating voltage, operating current, and operating temperature within a neighborhood can be independently and accurately statistically analyzed, forming representative neighborhood mean and variance. This statistical result provides a more reliable local feature basis for subsequent fault diagnosis, making the analysis process more stable and the judgment results more accurate.

[0013] Optionally, the structured data is sent to a weighted fault assessment system for fault weighting calculation to obtain fault weight scores, including: Based on the location coordinates and time information in the data boxes, the structural data are classified to obtain a set of data of the same type; Perform time-series differencing on similar data in a similar dataset to obtain the target change slope of each similar data item; Based on the neighborhood component set, the slope of change that is consistent with the current acquisition time of the target photovoltaic smart component is obtained to obtain the neighborhood slope dataset; Perform statistical calculations on the domain slope dataset to obtain the mean slope of the neighboring domains; Calculate the variance of the neighborhood slope dataset based on the mean of the neighborhood slope; The weight scores are determined based on the slope of the target change, the mean of the neighborhood slope, and the variance of the neighborhood slope. The fault weight score is obtained by weighted summation of the weight scores.

[0014] By employing the above technical solution, the difference between the slope of the target photovoltaic module and the mean and variance of the slopes in the neighborhood can be used to make a more discriminative judgment on the module's status. When the target photovoltaic module exhibits abnormal changes while the neighborhood modules remain generally stable, the deviation between the target module's slope and the neighborhood slope characteristics increases significantly, allowing for timely identification of independent faults in the target module. Conversely, when the neighborhood modules all show similar changing trends, even if the target module's changes are large, the difference between its slope characteristics and those of the neighborhood modules is relatively small, avoiding misjudging changes in environmental consistency as module faults. This method effectively distinguishes between "individual anomalies" and "overall changes," thereby improving the accuracy and reliability of fault identification.

[0015] Optionally, perform time-series differencing on similar data within the same dataset to obtain the target change slope for each similar data item, including: Obtain the slope of operating voltage change, operating current change, operating temperature change, neighborhood mean change, and neighborhood variance change in the neighborhood slope dataset; The average values ​​of the slopes of the above five types of changes are calculated to obtain the average slope of operating voltage, the average slope of operating current, the average slope of operating temperature, the average slope of neighborhood mean, and the average slope of neighborhood variance. The mean slope values ​​of operating voltage, operating current, operating temperature, neighborhood mean slope, and neighborhood variance slope are integrated to obtain the mean slope value of the neighborhood. Variance calculations were performed on the slopes of the same type of change based on their respective means to obtain the variances of the slopes of each type of change. By integrating the variances of various slope variations, the neighborhood slope variance is obtained.

[0016] By adopting the above technical solution, a unified mean and variance calculation can be performed on the slopes of various parameters such as voltage, current, temperature, and neighborhood statistics, enabling a comprehensive quantification of the overall trend and fluctuation level of the neighborhood. The resulting neighborhood slope mean and variance are more comprehensive, providing a more reliable reference for calculating weighted scores.

[0017] Optionally, weight scores are determined based on the target change slope, the mean of the neighborhood slope, and the variance of the neighborhood slope, including: The first evaluation parameter is obtained based on the difference between the target change slope and the mean slope of the neighborhood. The second evaluation parameter is obtained based on the correspondence between the target change slope and the variance of the neighborhood slope; The first evaluation parameter and the second evaluation parameter are linearly combined according to the preset weighting coefficients to obtain the initial weight. The initial weights are normalized to obtain the weight scores.

[0018] By adopting the above technical solution, the generation of fault weight scores can simultaneously reflect the degree of deviation of the target component from the normal trend of the neighborhood and the overall fluctuation level of the neighborhood. By weighting, fusing and normalizing the two types of evaluation parameters, a more stable and comparable weight score can be obtained, thereby improving the accuracy of fault identification.

[0019] Secondly, this application provides a monitoring system based on direct communication between photovoltaic modules, employing the following technical solution: A monitoring system based on direct communication between photovoltaic modules includes: The acquisition module is used to acquire runtime parameters; A memory for storing the program of the monitoring method based on direct communication between photovoltaic modules; The processor and the program in the memory can be loaded and executed by the processor to implement the monitoring method based on direct communication between photovoltaic modules.

[0020] By adopting the above technical solution, the direct collection and uploading of operating parameters of photovoltaic smart modules is achieved, and the operating parameters are analyzed on the cloud server side. Fine-grained quantitative analysis of the operating deviations between the module itself and its neighbors is possible. This eliminates reliance on string-level or third-party intermediate devices for indirect diagnosis, reducing problems such as diagnostic link interruptions, data distortion, and the inability to accurately locate module-level faults. Fault judgment logic can be directly applied to the individual photovoltaic module level. Therefore, the efficiency of photovoltaic module fault diagnosis is improved, enabling rapid identification and handling of module-level faults.

[0021] Thirdly, this application provides a smart terminal, which adopts the following technical solution: A smart terminal includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of the above methods.

[0022] Fourthly, this application provides a computer storage medium capable of storing corresponding programs, which facilitates improved efficiency in troubleshooting photovoltaic modules, and adopts the following technical solution: A computer-readable storage medium storing a computer program that can be loaded by a processor and executed any of the above-described monitoring methods based on direct communication with photovoltaic modules.

[0023] In summary, this application includes at least one of the following beneficial technical effects: 1. Direct acquisition and uploading of operating parameters of photovoltaic smart modules are achieved, and the operating parameters are analyzed on the cloud server side. Fine-grained quantitative analysis of the operating deviation between the module itself and its neighbors is possible. This eliminates reliance on string-level or third-party intermediate devices for indirect diagnosis, reducing problems such as diagnostic link interruptions, data distortion, and the inability to accurately locate module-level faults. Fault diagnosis logic can be directly applied to the individual photovoltaic module level. This improves the efficiency of photovoltaic module fault diagnosis and enables rapid identification and handling of module-level faults. 2. The cloud server can automatically generate location coordinates based on the physical layout of photovoltaic smart modules, and construct a data box containing time-series operational data, giving each photovoltaic module a clear spatial identifier and data carrier within the system. By performing neighborhood data statistics based on the data box, the operational deviation characteristics of the target module relative to its neighborhood can be extracted. Furthermore, the structured data is input into a weighted fault judgment system for quantitative calculation, yielding a precise fault weight score. This enables fine-grained evaluation of the module-level operational status, thereby improving the accuracy and efficiency of fault diagnosis and troubleshooting for photovoltaic smart modules. 3. Based on obtaining the location coordinates of the target photovoltaic smart module, it can filter photovoltaic modules within a preset range around it, eliminating invalid modules with negative coordinate indices. Furthermore, it can uniformly extract operating parameters from the neighboring module set and calculate the neighborhood mean and variance to obtain the neighborhood operating status. Finally, the above statistical information, along with the current acquisition time, is written back to the target module's data box, forming structured data that can be directly used for subsequent judgment processes. This improves the accuracy and stability of neighborhood data aggregation and enhances the accuracy of input data for subsequent fault judgment modules. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of a monitoring system based on direct communication between photovoltaic modules in an embodiment of this application.

[0025] Figure 2This is an off-site schematic diagram of a monitoring method based on direct communication between photovoltaic modules in an embodiment of this application.

[0026] Figure 3 This is a flowchart illustrating a method for analyzing operating parameters in an embodiment of this application.

[0027] Figure 4 This is a flowchart illustrating a structural data processing method in an embodiment of this application.

[0028] Figure 5 This is a flowchart illustrating a method for obtaining neighborhood mean and neighborhood variance in an embodiment of this application.

[0029] Figure 6 This is a flowchart illustrating a method for obtaining fault weight scores in an embodiment of this application.

[0030] Figure 7 This is a flowchart illustrating a method for obtaining the slope of target change in an embodiment of this application.

[0031] Figure 8 This is a flowchart illustrating a weighted score confirmation method in an embodiment of this application. Detailed Implementation

[0032] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figure 1 - Appendix Figure 8 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0033] This application discloses a monitoring method based on direct communication between photovoltaic modules, referring to... Figure 1 This method is executed by a monitoring system based on direct communication between photovoltaic modules. The monitoring system based on direct communication between photovoltaic modules includes intelligent photovoltaic modules, a data concentrator, and a cloud server.

[0034] Reference Figure 2 Monitoring methods based on direct communication between photovoltaic modules: Step S101: Obtain the operating parameters of the photovoltaic smart module through the data detection module. The operating parameters include operating voltage, operating current and operating temperature.

[0035] A photovoltaic smart module is a photovoltaic power generation module with data acquisition, communication, and self-diagnostic capabilities. It incorporates data detection and communication modules to collect key parameters such as operating voltage, operating current, and operating temperature.

[0036] The data detection module is used to sample the operating status of the photovoltaic smart module and generate corresponding data records. This module is integrated into the photovoltaic smart module itself and can work in conjunction with the communication module to ensure that the aforementioned operating parameter data is obtained in real time at each sampling point.

[0037] Operating parameters refer to the data collected by the photovoltaic smart module, including operating voltage, operating current, and operating temperature.

[0038] Step S102: Transmit the running parameters to the data concentrator via the communication module.

[0039] The communication module is a communication unit integrated inside the photovoltaic smart module, used to realize direct communication connection between the module and the data concentrator. Based on the actual layout of the module array, the communication module can autonomously select the reachable data concentrator and report the operating parameters using a preset communication method (such as a direct communication link).

[0040] The data concentrator is an edge aggregation device installed at the photovoltaic array site. It is used to uniformly receive operating parameters from multiple photovoltaic smart components and perform preliminary caching or preprocessing of these parameters so that they can be uploaded to the cloud platform for in-depth analysis later.

[0041] After collecting the operating parameters, the data detection module transmits them to the communication module. The communication module encapsulates the operating parameters into communication data frames, which are then sent by the component's built-in communication module. The communication module can automatically select an accessible data concentrator based on the component's physical layout and report the encapsulated operating parameters to the corresponding data concentrator according to a preset communication method. During transmission, the communication module maintains a direct communication link with the data concentrator to ensure that the operating parameters are stably received in each acquisition cycle.

[0042] Step S103: Transmit the running parameters to the cloud server through the data concentrator.

[0043] The cloud server is used to centrally manage and analyze the operating data sent by each photovoltaic smart module, and supports functions such as large-scale data storage, structured analysis, neighborhood statistics, and fault weighted model operation.

[0044] Step S104: Analyze the operating parameters through the cloud server to obtain the fault weight score.

[0045] Fault weight score is a quantitative indicator generated by the cloud server based on the structural data of photovoltaic smart components. It is used to characterize the degree of anomaly that the target component may have at the current acquisition time.

[0046] The specific steps for analyzing operating parameters through a cloud server to obtain the fault weight score can be found in [reference needed]. Figure 3 The content of the examples.

[0047] Step S105: Determine whether the fault weight score meets the preset conditions.

[0048] Preset conditions refer to the thresholds used to determine whether the photovoltaic smart module is operating normally. Preset conditions are set in advance during the system design phase based on module characteristics, scenario experience, or historical big data training results.

[0049] Step S106: If not, obtain control commands based on the fault weight score.

[0050] In another invention, if the analysis results do not meet the preset conditions, normal monitoring of the photovoltaic smart module is maintained.

[0051] The cloud server analyzes the results to determine the scope and type of the anomaly, and generates corresponding control commands based on preset strategies to enable the photovoltaic smart modules to enter a safe mode or perform corrective actions, thereby preventing the anomaly from escalating further or causing a reduction in power generation efficiency.

[0052] Step S107: Send control commands to the photovoltaic smart module via the cloud server.

[0053] The cloud server first determines the path to the corresponding data concentrator based on the photovoltaic smart component number contained in the control command, and then encapsulates the command into a recognizable data frame format. Subsequently, the cloud server sends the data frame to the data concentrator through the communication module; after receiving the command, the data concentrator transmits the command to the corresponding photovoltaic smart component through a direct communication link, thus realizing the command distribution.

[0054] Reference Figure 3 By analyzing the operating parameters through a cloud server, a fault weight score is obtained, including: Step S201: Set the corresponding position coordinates for each photovoltaic smart module according to the physical layout of the photovoltaic smart module.

[0055] Physical layout refers to the spatial arrangement of various photovoltaic smart components in a photovoltaic power station, such as using a row and column arrangement structure, forming a matrix coordinate system based on row and column numbers.

[0056] Location coordinates are used to identify the coordinate information of the physical location of photovoltaic smart modules in a photovoltaic array.

[0057] In one feasible embodiment, the physical layout parameters of the photovoltaic array are first read, including the total number of rows and columns of the components, the physical installation order, and string information. A component coordinate system is established based on the physical layout parameters. The cloud server numbers the components according to a "top-down, left-to-right" rule and generates corresponding row and column coordinates for each component. For example, the coordinates of the first component in the first row are (0,0), the coordinates of the second component in the first row are (0,1), and so on, setting corresponding position coordinates for each photovoltaic smart component.

[0058] Step S202: Establish a data box for the photovoltaic smart module at the current moment based on the location coordinates.

[0059] The data box is a data container for the photovoltaic smart module at the current acquisition time, used to store the module's operating parameters, location coordinates, and related derived data.

[0060] Each data box corresponds to a photovoltaic smart module and a point in time, and has a unique number. It can be regarded as "a complete data snapshot of the photovoltaic smart module at this moment".

[0061] Step S203: Write the position coordinates and running parameters into the data box as data to be processed.

[0062] Data to be processed refers to data written into the data box for subsequent analysis. This includes, but is not limited to: Location coordinates of the photovoltaic smart module; operating voltage at the current acquisition time; operating current at the current acquisition time; operating temperature at the current acquisition time; current acquisition time.

[0063] Step S204: Perform neighborhood data statistical processing based on the data to be processed to obtain the structural data to be analyzed.

[0064] Neighborhood data statistical processing refers to the process of extracting operating parameters and performing statistical analysis on neighboring modules within a preset range, based on the position of the target photovoltaic smart module in the module matrix. Specific steps for neighborhood data statistical processing based on the data to be processed can be found in [reference needed]. Figure 4 The content of the examples.

[0065] The structured data to be analyzed refers to data boxes that have been augmented with statistical information such as neighborhood mean, neighborhood variance, and current collection time on top of the basic operating parameters; it is considered structured data.

[0066] Step S205: Send the structural data to the weighted fault judgment system for fault weighting calculation to obtain the fault weight score.

[0067] The weighted fault assessment system is an analysis module deployed on a cloud server. It is used to classify the structural data of photovoltaic smart modules, calculate change trends, calculate neighborhood fluctuations, and generate weighted scores. The system comprehensively evaluates the degree of abnormal operation of the modules through indicators such as slope changes and the magnitude of neighborhood fluctuations, and finally outputs a fault weight score.

[0068] Reference Figure 4 Based on the data to be processed, neighborhood data statistical processing is performed to obtain the structured data to be analyzed, including: Step S301: Obtain photovoltaic smart modules within a preset range based on the target location coordinates recorded in the data box of the target photovoltaic smart module, and obtain a set of neighboring candidate modules.

[0069] The preset range is a preset constant that can be adjusted according to actual conditions.

[0070] The target photovoltaic smart module refers to the photovoltaic smart module that needs to be tested and analyzed. The target position coordinates are the position coordinates of the target photovoltaic module in the module coordinate system.

[0071] Using the target location coordinates of the target photovoltaic smart module as the center point, obtain the location coordinates of all photovoltaic smart modules within a preset range of the center point. Combine these location coordinates with the target location coordinates to obtain a neighborhood candidate module set. The neighborhood range is determined by a preset total number of columns m and total number of rows n, where m represents the total number of columns in the neighborhood candidate module set along the first direction, and n represents the total number of rows in the neighborhood candidate module set along the second direction. The first direction is the column direction, corresponding to the x-axis of the coordinate system; the second direction is the row direction, corresponding to the y-axis of the coordinate system. Both m and n are odd numbers greater than or equal to 3 and can be configured according to actual needs.

[0072] Using the target location coordinates (x, y) as the center point, obtain the location coordinates of photovoltaic smart modules that meet the preset range, and combine them with the target location coordinates to form a neighborhood candidate module set PF(x, y):

[0073] Where i and j are integers, and the neighborhood candidate component set PF(x,y) does not contain the coordinates of out-of-bounds or non-corresponding photovoltaic smart components.

[0074] For example, when m=3 and n=3, the neighborhood candidate component set PF(x,y) is: PF(x,y)={(x-1,y-1),(x,y-1),(x+1,y-1),(x-1,y),(x,y),(x+1,y),(x-1,y+1),(x,y+1),(x+1,y+1)}.

[0075] Step S302: Remove neighboring candidate components with negative coordinate indices from the set of neighboring candidate components to obtain the set of neighboring components.

[0076] Neighborhood candidate components with negative coordinate indices refer to photovoltaic smart modules in the neighborhood candidate set whose row or column index is less than 0. These coordinates do not correspond to actual existing photovoltaic smart modules and therefore need to be eliminated.

[0077] In one feasible implementation: Using (x,y) as the target coordinates, generate a set of neighborhood candidate components; For each coordinate (i,j) in the neighborhood candidate component set, perform the following judgment: If i < 0 or j < 0, then remove the component; If i≥1 and j≥1, then it is retained as a valid component; The result of filtering the neighborhood candidate component set is the neighborhood component set.

[0078] Step S303: Extract the running parameters from the data boxes corresponding to the neighbor components in the neighbor component set to obtain the neighbor running data.

[0079] Neighborhood components are sub-items in the neighborhood component set. Neighborhood runtime data is the set of runtime parameters for all neighborhood components.

[0080] Each photovoltaic smart module generates a data box during each data collection cycle. Therefore, the operating parameters in the data boxes are extracted from the neighboring modules (photovoltaic smart modules) in the neighboring module set and aggregated to obtain the neighborhood operating data.

[0081] Step S304: Obtain the neighborhood mean and neighborhood variance of the neighborhood running data according to the preset neighborhood data processing method.

[0082] The neighborhood mean refers to the average level of several neighboring components in terms of operating voltage, operating current, and operating temperature, and is used to reflect the central trend of the neighborhood's operating status.

[0083] Neighborhood variance refers to the degree of dispersion of several neighborhood operating parameters relative to the neighborhood mean. It is used to assess the degree of fluctuation in the neighborhood operating state. The larger the variance, the greater the difference within the neighborhood.

[0084] The specific steps for obtaining the neighborhood mean and neighborhood variance of the neighborhood operation data according to the preset neighborhood data processing method can be found in [reference needed]. Figure 5 The steps in the embodiments.

[0085] Step S305: Write the neighborhood mean, neighborhood variance, and current acquisition time back to the data box corresponding to the target photovoltaic smart module to obtain the structural data to be analyzed.

[0086] The data box corresponding to the target photovoltaic smart module includes location coordinates, operating voltage, operating current, and operating temperature. After the neighborhood mean, neighborhood variance, and current acquisition time are written into the data box, the data box includes location coordinates, operating voltage, operating current, operating temperature, neighborhood mean, neighborhood variance, and current acquisition time. The structure formed by the above data is the structural data to be analyzed.

[0087] Reference Figure 5 The neighborhood mean and neighborhood variance of the neighborhood operation data are obtained according to a preset neighborhood data processing method, including: Step S401: Obtain the set of operating voltage, set of operating current, and set of operating temperature from the neighborhood operating data.

[0088] Extract the operating voltage from the operating parameters of all neighboring components from the neighboring operating data and aggregate them to obtain the operating voltage set.

[0089] Extract the operating current from the operating parameters of all neighboring components from the neighboring operating data and aggregate them to obtain the operating current set.

[0090] Extract the operating temperature from the operating parameters of all neighboring components from the neighboring operating data and aggregate them to obtain the operating temperature set.

[0091] Step S402: Calculate the mean values ​​of the operating voltage set, operating current set, and operating temperature set respectively to obtain the neighborhood mean value of operating voltage, the neighborhood mean value of operating current, and the neighborhood mean value of operating temperature.

[0092] The neighborhood operation data is obtained by selecting the operating parameters within an m×n region centered on the coordinates of the target photovoltaic smart module. For any type of operating parameter PFC (such as operating voltage, operating current, or operating temperature), its neighborhood mean PFS(x,y) can be obtained using the following mathematical expression:

[0093] in: PFC(x,y) represents the running parameter value corresponding to the position coordinates (x,y) of the candidate component in the neighborhood candidate component set PF(x,y); This indicates that there are a total of m×n data points in the m×n neighborhood; PFS(x,y) represents the neighborhood mean of the target photovoltaic smart module on this operating parameter.

[0094] The average value of several operating voltages in the set of operating voltages is calculated to obtain the neighborhood average value of the operating voltage.

[0095] The average value of several operating currents in the set of operating currents is calculated to obtain the neighborhood mean of the operating current.

[0096] The average value of several operating temperatures in the set of operating temperatures is calculated to obtain the neighborhood mean of the operating temperature.

[0097] Step S403: Integrate the neighborhood average of operating voltage, the neighborhood average of operating current, and the neighborhood average of operating temperature to obtain the neighborhood average.

[0098] The neighborhood averages of operating voltage, operating current, and operating temperature are consolidated into a single data structure to form a set of "comprehensive averages" describing the operating conditions of the neighborhood, thus obtaining the neighborhood average.

[0099] Step S404: Calculate the variance of the operating voltage set based on the mean of the operating voltage neighborhood to obtain the operating voltage neighborhood variance.

[0100] Operating voltage neighborhood variance refers to the variance calculated by taking the operating voltage mean as the center and examining the dispersion of the operating voltage of each photovoltaic smart module within its neighborhood at the same data acquisition time. This variance reflects the fluctuation of the neighborhood voltage.

[0101] After obtaining the running parameters PFC(x,y) of the neighborhood candidate components in the neighborhood candidate component set PF(x,y), the neighborhood variance PFV(x0,y0) of the parameter can be calculated using a preset formula.

[0102] formula:

[0103] PFS(x0,y0) is the mean of the neighborhood centered at the target position coordinates (x0,y0).

[0104] Step S405: Calculate the variance of the operating current set based on the neighborhood mean of the operating current to obtain the neighborhood variance of the operating current.

[0105] The operating current neighborhood variance refers to the variance calculated by taking the operating current mean of the operating current neighborhood as the center and examining the dispersion of the operating current of each photovoltaic smart module within the neighborhood of the photovoltaic smart module at the same data acquisition time. This variance reflects the fluctuation of the neighborhood current.

[0106] After obtaining the running parameters PFC(x,y) of the neighborhood candidate components in the neighborhood candidate component set PF(x,y), the neighborhood variance PFV(x0,y0) of the parameter can be calculated using a preset formula.

[0107] formula:

[0108] PFS(x0,y0) is the mean of the neighborhood centered at the target position coordinates (x0,y0).

[0109] Step S406: Calculate the variance of the operating temperature set based on the neighborhood mean of the operating temperature to obtain the neighborhood variance of the operating temperature.

[0110] Operating temperature neighborhood variance refers to the variance calculated by taking the operating temperature of each photovoltaic smart module within its neighborhood as the mean of the operating temperature neighborhood as the center, at the same data acquisition time. This variance reflects the fluctuation of the neighborhood temperature.

[0111] After obtaining the running parameters PFC(x,y) of the neighborhood candidate components in the neighborhood candidate component set PF(x,y), the neighborhood variance PFV(x0,y0) corresponding to the parameters can be calculated using a preset formula.

[0112] formula:

[0113] PFS(x0,y0) is the mean of the neighborhood centered at the target position coordinates (x0,y0).

[0114] Step S407: Integrate the neighborhood variance of operating voltage, the neighborhood variance of operating current, and the neighborhood variance of operating temperature to obtain the neighborhood variance.

[0115] The neighborhood variances of line voltage, operating current, and operating temperature are unified into a single data structure, forming a set of "comprehensive variances" that describe the operating conditions of the neighborhood, thus obtaining the neighborhood variance.

[0116] Reference Figure 6 The structural data is sent to the weighted fault assessment system for fault weighting calculation to obtain fault weight scores, including:

[0117] Step S501: Based on the location coordinates and time information in the data boxes, classify the structural data to obtain a set of data of the same type.

[0118] A similar data set refers to a data set formed by classifying structural data according to attributes based on the location coordinates and acquisition time information recorded by photovoltaic smart modules in the data box.

[0119] When classifying similar data, they can be categorized according to structured data such as voltage, current, temperature, neighborhood mean, and neighborhood variance.

[0120] In one feasible embodiment, the location coordinates recorded in the structural data are first read, and data from all photovoltaic smart modules corresponding to the same location coordinates are selected. Data belonging to the same location coordinates are then grouped according to their corresponding acquisition timestamps, so that each time corresponds to a set of structural data. Furthermore, the structural data under the same timestamp are split by category, for example: operating voltage structural data set, operating current structural data set, operating temperature structural data set, neighborhood mean structural data set, and neighborhood variance structural data set. After the above grouping, multiple "same-type data sets" are obtained for subsequent steps to perform time-series difference processing on each sub-item.

[0121] Step S502: Perform time-series difference operation on the same type of data in the same data set to obtain the target change slope of each sub-item of the same type of data.

[0122] Temporal differential operation refers to performing temporal difference calculations on two structural data sets of the same category but acquired at different times to obtain information on the rate of change of that category of data over time.

[0123] A sub-item of the same type of data is a sub-item within a set of data of the same type.

[0124] In this embodiment, the timing difference operation can be performed using the following formula:

[0125] Wherein, RTPVK(x,y) n This represents the target change slope of the nth type of operational data (e.g., voltage, current, temperature, neighborhood mean, neighborhood variance, etc.) located at position coordinates (x, y). R(xt) represents the value of this type of running data at the current acquisition time t; R(xt-1) represents the value of this type of running data at the previous acquisition time t-1; t-(t-1) is the time interval between two acquisition times, which can be determined according to the sampling period of the acquisition device.

[0126] Based on the above formula, differential processing can be performed on different types of structural data to obtain the rate of change of each type of data at adjacent acquisition times. For example, when R represents the operating voltage, the obtained RTPVK(x,y)... n This represents the slope of the operating voltage change; when R represents the operating current, RTPVK(x,y) n This is the slope of the operating current change; when R represents temperature data or the mean or variance of the neighborhood, the corresponding slope can be calculated in the same way.

[0127] In actual calculations, the above differential calculation can be performed on two sets of structural data sorted by time from the same dataset. The resulting slopes of operating voltage, operating current, operating temperature, neighborhood mean, and neighborhood variance are used as the basis for subsequent statistical calculations. Similarly, the same differential calculation can be performed on all structural data to obtain the target slope for fault diagnosis.

[0128] The specific steps for performing time-series differencing on similar data within the same dataset to obtain the target change slope for each similar data item can be found in [reference needed]. Figure 6 The steps in the embodiments.

[0129] Step S503: Based on the neighborhood component set, obtain the slope of change that is consistent with the current acquisition time of the target photovoltaic smart component, and obtain the neighborhood slope dataset.

[0130] Based on the established set of neighboring components, extract the slope value that is the same as the target photovoltaic smart module's current acquisition time from the data box corresponding to each neighboring component. Since each photovoltaic smart module's data box records slope information at different acquisition times, it is necessary to filter out the synchronized slope value according to the target photovoltaic smart module's current timestamp.

[0131] These synchronized slope changes are categorized into different sets (operating voltage change slope, operating current change slope, operating temperature change slope, operating neighborhood mean change slope, and operating neighborhood variance change slope) and added to the corresponding sets to form a complete neighborhood slope dataset.

[0132] For example, assuming there are 8 neighborhood components, we can obtain 8 operating voltage change slopes, 8 operating current change slopes, 8 operating temperature change slopes, 8 operating neighborhood mean change slopes, and 8 operating neighborhood variance change slopes.

[0133] These data, after being categorized by type, constitute a neighborhood slope dataset, for example: The set of slopes for the operating voltage change is {0.12, 0.10, 0.15, ...}; The set of slopes for the operating temperature change is {0.30, 0.28, 0.32, ...}.

[0134] Step S504: Perform statistical calculations on the domain slope dataset to obtain the mean slope of the neighborhood.

[0135] The mean of the slope for each class of change in the neighborhood slope dataset is calculated. During the calculation, multiple slope values ​​are read according to the slope class, and the mean slope for each class is obtained using the arithmetic mean formula. Then, the mean slopes of each class are integrated to obtain the mean slope of the neighborhood.

[0136] The mean slope of the neighborhood can be calculated using the following statistical formula:

[0137] Wherein, DPF(t,y) represents the mean slope of the change of the neighborhood component set at the current acquisition time.

[0138] RTPVK(x,y) j This represents the slope of the change at position (x, y) for the j-th neighboring component within the set of neighboring components; w represents the number of neighboring components within the set of neighboring components; N(i) is the set of neighborhood indices corresponding to the target neighborhood component.

[0139] Step S505: Calculate the variance of the neighborhood slope dataset based on the mean of the neighborhood slope.

[0140] After obtaining the mean slope of the neighborhood component set, the deviation of each slope value in the neighborhood component set from the mean is further calculated, and the neighborhood slope variance is obtained accordingly.

[0141] The variance of the neighborhood slope can be calculated using the following formula:

[0142] DPFV(t,y) represents the neighborhood slope variance, which reflects the degree of fluctuation in the slope of change within the neighborhood component set.

[0143] This variance is used to quantify the stability of changes within the neighborhood: The larger the variance, the more significant the difference in the slope of change within the set of neighboring components, which may indicate the presence of local anomalies. The smaller the variance, the more stable the changes in the neighborhood.

[0144] Step S506: Determine the weight score based on the target change slope, the mean of the neighborhood slope, and the variance of the neighborhood slope.

[0145] By comparing the target change slope of the target photovoltaic smart module with the mean slope of its neighboring module set, and combining the variance of the neighboring slope, a weighted evaluation is performed on the operational changes of the target photovoltaic smart module at the current acquisition time.

[0146] For example, when the variance of the neighborhood slope is small, it indicates that the changing trend of the neighborhood component set is consistent and the operating state is relatively stable. At this time, if the changing slope of the target photovoltaic smart component deviates significantly from the mean of the neighborhood slope, the deviation is more indicative of anomalies, and the corresponding weight score is increased accordingly.

[0147] When the variance of the neighborhood slope is large, it indicates that the overall change and fluctuation of the neighborhood component set is relatively drastic, and there are significant differences in change within the neighborhood itself. In this case, even if the target change slope deviates from the mean of the neighborhood slope, it should not be directly judged as an abnormal state. However, this deviation is still included in the weight calculation process as a change behavior that needs attention. Its impact on the weight score will be adjusted in conjunction with the variance of the neighborhood slope to avoid misjudgment due to overall fluctuations in the neighborhood.

[0148] It should be noted that the weight adjustment method based on neighborhood slope variance in this embodiment is a local judgment approach in the anomaly assessment process, used to improve the reliability and robustness of anomaly identification, and does not limit the specific implementation form of anomaly judgment.

[0149] The specific steps for determining the weight scores based on the target change slope, the mean of the neighborhood slope, and the variance of the neighborhood slope can be found in [reference needed]. Figure 8 The content of the embodiment. A weighted score that reflects the degree of anomaly of the target photovoltaic smart module is obtained.

[0150] Step S507: Sum the weighted scores to obtain the fault weight score.

[0151] Among them, the weighted score refers to the sub-scores calculated for the operating voltage, operating current, operating temperature, and neighborhood-related parameters (neighborhood slope mean, neighborhood slope variance, etc.) of the target photovoltaic smart module, which are used to reflect the degree of contribution of each parameter to the fault trend.

[0152] The fault weight score is a comprehensive score obtained by summing all the individual weight scores according to a set weighting coefficient. It is used to reflect the risk level of the target photovoltaic smart module deviating from normal behavior. The weighting coefficients used in the weighted summation are preset values.

[0153] Reference Figure 7 Perform time-series differencing on similar data within the same dataset to obtain the target change slope for each similar data item, including:

[0154] Step S601: Obtain the slope of the operating voltage change, the slope of the operating current change, the slope of the operating temperature change, the slope of the neighborhood mean change, and the slope of the neighborhood variance change in the neighborhood slope dataset.

[0155] This step is used to separate five classes of slope values ​​from the neighborhood slope dataset, including:

[0156] The slope of the change in operating voltage of neighboring components; the slope of the change in operating current of neighboring components; the slope of the change in operating temperature of neighboring components; the slope of the change in the mean of the neighborhood; the slope of the change in the variance of the neighborhood.

[0157] These slope data are used for subsequent "slope mean calculation" and "slope variance calculation", thus forming the entire "neighborhood statistical structure of the slope dimension".

[0158] In one feasible embodiment, the operating voltage, operating current, operating temperature, and corresponding neighborhood mean and neighborhood variance of the neighboring components are first read from the data box at the current acquisition time and the data box at the previous acquisition time, respectively.

[0159] Subsequently, according to By performing time-series differencing on the above parameters, five types of change slopes are obtained.

[0160] Step S602: Calculate the mean values ​​of the above five types of slope changes to obtain the mean slope of operating voltage, the mean slope of operating current, the mean slope of operating temperature, the mean slope of neighborhood mean, and the mean slope of neighborhood variance.

[0161] Since the same dataset contains different types of slope data, it is necessary to calculate the average for the slopes of operating voltage, operating current, operating temperature, neighborhood mean, and neighborhood variance to obtain the overall trend of each type of slope within the neighborhood components. By calculating the average, a trend benchmark value for each type of slope can be obtained, providing a central reference point for subsequent slope variance calculations. This allows the system to further determine whether there are abnormal fluctuations in this type of slope within the neighborhood.

[0162] Step S603: Integrate the mean slope of operating voltage, the mean slope of operating current, the mean slope of operating temperature, the mean slope of neighborhood average, and the mean slope of neighborhood variance to obtain the mean slope of neighborhood.

[0163] The average slope values ​​of operating voltage, operating current, operating temperature, neighborhood mean slope, and neighborhood variance slope are all organized into a single data structure to form a set of "comprehensive slope averages" that describe the operating conditions of the neighborhood, thus obtaining the neighborhood slope average.

[0164] Step S604: Calculate the variance of the slope of the same type of change based on the corresponding mean to obtain the variance of the slope of each type of change.

[0165] Here, the corresponding mean refers to the mean value of a certain type of slope obtained by calculating the mean of the slopes for a specific category; the same type of slope refers to the set of data slopes belonging to the same category. For example, for the slope of operating voltage, firstly, the mean of all its slopes is calculated to obtain the mean of the operating voltage slope; then, using the mean of the operating voltage slope as the statistical center, the variance of all the slopes of the operating voltage category is calculated to obtain the variance of the operating voltage slope.

[0166] Similarly, the variance of the slope of the operating current change is calculated with the mean of the operating current slope as the center; the variance of the slope of the operating temperature change is calculated with the mean of the operating temperature slope as the center; the variance of the slope of the neighborhood mean change is calculated with the mean of the neighborhood mean slope as the center; and the variance of the neighborhood variance change is calculated with the mean of the neighborhood variance slope as the center.

[0167] By calculating the variances of the above five types of slope changes, we can obtain the variances of the operating voltage slope change, the operating current slope change, the operating temperature slope change, the neighborhood mean slope change, and the neighborhood variance slope change, respectively. These variances are used to characterize the degree of fluctuation of each type of slope change within the neighborhood range, providing a basis for generating a comprehensive neighborhood slope variance in the future.

[0168] Step S605: Integrate the variances of various slope variations to obtain the variances of the neighborhood slope.

[0169] The variances of operating voltage change slope, operating current change slope, operating temperature change slope, neighborhood mean change slope, and neighborhood variance change slope are all organized into a single data structure, forming a set of "comprehensive change slope variances" that describe the operating conditions of the neighborhood, thus obtaining the neighborhood slope variance.

[0170] Reference Figure 8 The weighted scores are determined based on the slope of the target change, the mean of the neighborhood slopes, and the variance of the neighborhood slopes, including:

[0171] Step S701: Obtain the first evaluation parameter based on the difference between the target change slope and the mean slope of the neighborhood.

[0172] The first evaluation parameter refers to the evaluation value obtained by looking up the difference between the target slope and the mean slope of the neighborhood according to a preset rule table. It is used to reflect the degree of deviation of the target photovoltaic smart module from the overall trend of the neighborhood. The preset rule table is a preset comparison table between the slope difference and the evaluation parameter.

[0173] Step S702: Obtain the second evaluation parameter based on the correspondence between the target change slope and the variance of the neighborhood slope.

[0174] A dynamic sensitivity coefficient is constructed based on the neighborhood slope variance. By performing an inverse scaling process on the neighborhood slope variance, a dynamic sensitivity coefficient is obtained to characterize the neighborhood fluctuation sensitivity.

[0175] The smaller the neighborhood slope variance, the larger the sensitivity coefficient; conversely, the larger the neighborhood slope variance, the smaller the sensitivity coefficient. A feasible expression is: Dynamic Sensitivity Coefficient = Suppression Coefficient × Neighborhood Slope Variance. The suppression coefficient can be obtained from a pre-defined suppression coefficient-slope variance lookup table.

[0176] The target change slope is combined with a dynamic sensitivity coefficient to assign different offset weights to the target change slope under different fluctuation backgrounds. Specifically, when the neighborhood fluctuation is small, the sensitivity coefficient is large, which can amplify the anomaly of the target change slope; when the neighborhood fluctuation is large, the sensitivity coefficient is small, which can suppress false judgments. A feasible mapping method is: Second evaluation parameter = |Target change slope| × Dynamic sensitivity coefficient. Ultimately, the second evaluation parameter is used to characterize the relative anomaly of the target photovoltaic smart module's change trend under the background of neighborhood fluctuations.

[0177] Step S703: Linearly combine the first evaluation parameter and the second evaluation parameter according to the preset weighting coefficient to obtain the initial weight.

[0178] The preset weighting coefficients are pre-set coefficients used to control the relative contribution of the first evaluation parameter and the second evaluation parameter. They usually include the first weighting coefficient and the second weighting coefficient, which can satisfy the requirement that the sum of the two is 1, or can be configured according to the actual application.

[0179] The initial weight refers to the comprehensive evaluation index that has not yet been normalized. Its value is determined by the first evaluation parameter and the second evaluation parameter, and is used to quantify the degree of abnormality or potential failure trend of the target photovoltaic smart module.

[0180] In one feasible implementation:

[0181] The system has a first weighting coefficient α and a second weighting coefficient β, which are used to adjust the contribution of the first evaluation parameter and the second evaluation parameter to the initial weight.

[0182] By combining the first evaluation parameter and the second evaluation parameter in the following linear manner:

[0183] W init =α×P1+β×P2

[0184] in: W init The initial weights; P1 is the first evaluation parameter; P2 is the second evaluation parameter; α and β are weighting coefficients that satisfy α + β = 1.

[0185] In practical calculations, α and β can be adjusted based on the power plant type, component consistency, and the complexity of the operating environment. For example, in scenarios with significant environmental fluctuations, α can be reduced and β increased to mitigate misjudgments caused by abnormal fluctuations.

[0186] Step S704: Normalize the initial weights to obtain the weight scores.

[0187] In one feasible implementation: the initial weights can be linearly normalized based on a preset normalization method. The normalization method can be range normalization, i.e.: Score = (W init -W min ) / (W max -W min )

[0188] in:

[0189] Score is the normalized weighted score; W min With W max These represent the minimum and maximum initial weights of all photovoltaic smart modules in the same batch within the current statistical period.

[0190] This normalization method ensures that all weight scores are within the range of 0 to 1, which facilitates subsequent fault level judgment, threshold comparison and alarm strategy execution.

[0191] Based on the same inventive concept, embodiments of this application provide a monitoring system based on direct communication between photovoltaic modules, comprising:

[0192] The acquisition module is used to acquire runtime parameters;

[0193] A memory for storing the program of the monitoring method based on direct communication between photovoltaic modules described above;

[0194] The processor and memory programs can be loaded and executed by the processor to implement the aforementioned monitoring method based on direct communication between photovoltaic modules.

[0195] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0196] This application provides a computer-readable storage medium storing a computer program that can be loaded and executed by a processor, which is a monitoring method based on direct communication between photovoltaic modules.

[0197] Computer storage media include, for example, USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media that can store program code.

[0198] Based on the same inventive concept, embodiments of this application provide a smart terminal, including a memory and a processor, wherein the memory stores a computer program that can be loaded and executed by the processor for a monitoring method based on direct communication with photovoltaic modules.

[0199] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0200] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A monitoring method based on direct communication between photovoltaic modules, characterized in that, include: The operating parameters of the photovoltaic smart module are obtained through the data detection module. These operating parameters include operating voltage, operating current, and operating temperature. The operating parameters are transmitted to the data concentrator via the communication module; The runtime parameters are transmitted to the cloud server via a data concentrator; By analyzing the operating parameters through a cloud server, a fault weight score is obtained. Determine whether the fault weight score meets the preset conditions; If not, then control commands are obtained based on the fault weight score; Control commands are sent to the photovoltaic smart modules via a cloud server.

2. The monitoring method based on direct communication between photovoltaic modules according to claim 1, characterized in that, The process of analyzing operating parameters through a cloud server to obtain fault weight scores includes: Each photovoltaic smart module is assigned corresponding position coordinates based on its physical layout. Establish a data box for the current moment for photovoltaic smart modules based on their location coordinates; Write the location coordinates and operating parameters into the data box as data to be processed; Based on the data to be processed, perform neighborhood data statistical processing to obtain the structural data to be analyzed; The structural data is sent to the weighted fault judgment system for fault weighting calculation to obtain the fault weight score.

3. The monitoring method based on direct communication between photovoltaic modules according to claim 2, characterized in that, The process of performing neighborhood data statistical processing based on the data to be processed to obtain the structural data to be analyzed includes: Based on the target location coordinates recorded in the data box of the target photovoltaic smart module, obtain the photovoltaic smart modules within a preset range and obtain a set of neighboring candidate modules; Remove neighboring candidate components with negative coordinate indices from the set of neighboring candidate components to obtain the set of neighboring components; Extract the running parameters from the data boxes corresponding to the neighboring components in the neighboring component set to obtain the neighboring running data; The neighborhood mean and neighborhood variance of the neighborhood operation data are obtained according to the preset neighborhood data processing method. The neighborhood mean, neighborhood variance, and current acquisition time are written back to the data box corresponding to the target photovoltaic smart module to obtain the structural data to be analyzed.

4. The monitoring method based on direct communication between photovoltaic modules according to claim 3, characterized in that, The step of obtaining the neighborhood mean and neighborhood variance of the neighborhood operation data according to the preset neighborhood data processing method includes: Obtain the set of operating voltage, operating current, and operating temperature from the neighborhood operating data; The mean values ​​of the operating voltage set, operating current set, and operating temperature set are calculated separately to obtain the neighborhood mean values ​​of operating voltage, operating current, and operating temperature. The neighborhood mean values ​​of operating voltage, operating current, and operating temperature are integrated to obtain the neighborhood mean value. The variance of the operating voltage neighborhood is calculated based on the mean of the operating voltage neighborhood. The variance of the operating current neighborhood is calculated based on the neighborhood mean of the operating current. The variance of the operating temperature neighborhood is calculated based on the neighborhood mean of the operating temperature. The neighborhood variance is obtained by integrating the neighborhood variances of operating voltage, operating current, and operating temperature.

5. The monitoring method based on direct communication between photovoltaic modules according to claim 4, characterized in that, The step of sending structural data to a weighted fault assessment system for fault weighting calculation to obtain fault weight scores includes: Based on the location coordinates and time information in the data boxes, the structural data are classified to obtain a set of data of the same type; Perform time-series differencing on similar data in a similar dataset to obtain the target change slope of each similar data item; Obtain the slope of change that is consistent with the current acquisition time of the target photovoltaic smart module from the set of neighboring components to obtain the neighborhood slope dataset; Perform statistical calculations on the domain slope dataset to obtain the mean slope of the neighboring domains; Calculate the variance of the neighborhood slope dataset based on the mean of the neighborhood slope; The weight scores are determined based on the slope of the target change, the mean of the neighborhood slope, and the variance of the neighborhood slope. The fault weight score is obtained by weighted summation of the weight scores.

6. The monitoring method based on direct communication between photovoltaic modules according to claim 5, characterized in that, The step of performing time-series differencing on similar data within a similar dataset to obtain the target change slope for each similar data item includes: Obtain the slope of operating voltage change, operating current change, operating temperature change, neighborhood mean change, and neighborhood variance change in the neighborhood slope dataset; The average values ​​of the slopes of the above five types of changes are calculated to obtain the average slope of operating voltage, the average slope of operating current, the average slope of operating temperature, the average slope of neighborhood mean, and the average slope of neighborhood variance. The mean slope values ​​of operating voltage, operating current, operating temperature, neighborhood mean slope, and neighborhood variance slope are integrated to obtain the mean slope value of the neighborhood. Variance calculations were performed on the slopes of the same type of change based on their respective means to obtain the variances of the slopes of each type of change. By integrating the variances of various slope variations, the neighborhood slope variance is obtained.

7. The monitoring method based on direct communication between photovoltaic modules according to claim 5, characterized in that, The determination of weighted scores based on the target change slope, the mean of the neighborhood slope, and the variance of the neighborhood slope includes: The first evaluation parameter is obtained based on the difference between the target change slope and the mean slope of the neighborhood. The second evaluation parameter is obtained based on the correspondence between the target change slope and the variance of the neighborhood slope; The first evaluation parameter and the second evaluation parameter are linearly combined according to the preset weighting coefficients to obtain the initial weight. The initial weights are normalized to obtain the weight scores.

8. A monitoring system based on direct communication between photovoltaic modules, characterized in that, The system is used to execute the monitoring method based on direct communication between photovoltaic modules as described in any one of claims 1 to 7, including: The acquisition module is used to acquire runtime parameters; A memory for storing the program of the monitoring method based on direct communication between photovoltaic modules; The processor and the program in the memory can be loaded and executed by the processor to implement the monitoring method based on direct communication between photovoltaic modules.

9. A smart terminal, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and execute the method as described in any one of claims 1 to 7.

Images