A photovoltaic power station monitoring system and method based on data analysis

Abstract

The application belongs to the technical field of photovoltaic power station monitoring, and discloses a photovoltaic power station monitoring system and method based on data analysis, which comprises a space-time operation mapping module, which is used for collecting photovoltaic component operation parameters and environmental parameters, and performing time synchronization and space mapping processing to form a space-time operation data set; a power deviation calculation module, which is used for normalizing the output power of the photovoltaic component according to the space-time operation data set to construct a component performance benchmark; the same photovoltaic components are divided into comparison groups, the power deviation of each photovoltaic component relative to the reference value in the group is calculated, and the direction state of the power deviation is extracted; and a component state tracking module, which is used for aggregating the direction state of the power deviation of each photovoltaic component, judging whether the reference value in the group deviates, and automatically reconstructing the reference value in the group when it is judged that the reference value in the group deviates; which is helpful to improve the overall power generation efficiency and operation and maintenance management level of the photovoltaic power station.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power plant monitoring technology, and more specifically, to a photovoltaic power plant monitoring system and method based on data analysis. Background Technology

[0002] Existing photovoltaic power plant monitoring methods rely on instantaneous output power, daily power generation, or equipment ratings for performance evaluation, comparing measured data with fixed thresholds to determine if modules are abnormal. While this method is simple to implement, it has limitations in practical applications.

[0003] First, the output power of photovoltaic modules fluctuates with solar irradiance. Power data collected under different weather conditions and at different times vary greatly. If irradiance conditions are not processed uniformly, environmental changes may be misjudged as module abnormalities or the true performance degradation may be masked.

[0004] Second, existing technologies use fixed reference values ​​or component nameplate parameters as comparison benchmarks. However, photovoltaic modules age over long-term operation, and their normal output levels change. Fixed reference values ​​cannot reflect long-term evolution, leading to discrepancies between the assessment results and the actual state, making it difficult to determine the degradation trend.

[0005] Third, when comparing components, the differences in component model, rated power, installation tilt angle and orientation were not fully considered. Different components have natural differences in output, and direct comparison can easily misjudge normal differences as abnormalities, reducing the accuracy of monitoring.

[0006] Fourth, existing methods only focus on the magnitude of power deviation and do not utilize the information on the positive and negative directions of the deviation. This information is of great significance for identifying the overall offset trend within a group and judging whether the reference benchmark is distorted. Ignoring this information will limit the ability to perform anomaly analysis and trend judgment.

[0007] Fifth, existing technologies often only issue simple alarms after detecting component anomalies, lacking quantitative analysis of the scope of the anomaly and power generation loss, and cannot automatically determine the priority of handling. Maintenance personnel need to make manual judgments, which is a large workload and makes it difficult to handle anomalies that have a significant impact on power generation revenue in a timely manner.

[0008] In view of this, the present invention proposes a photovoltaic power plant monitoring system and method based on data analysis to solve the above problems. Summary of the Invention

[0009] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a photovoltaic power plant monitoring system based on data analysis, comprising: The spatiotemporal operation mapping module is used to collect photovoltaic module operation parameters and environmental parameters, and perform time synchronization and spatial mapping processing to form a spatiotemporal operation dataset; The power deviation calculation module is used to normalize the output power of photovoltaic modules based on the spatiotemporal operation dataset and construct a module performance benchmark; it divides similar photovoltaic modules into comparison groups, calculates the power deviation of each photovoltaic module relative to the reference value within the group, and extracts the directional state of the power deviation. The module status tracking module is used to aggregate the directional status of the power deviation of each photovoltaic module, determine whether the reference value within the group has shifted, automatically reconstruct the reference value within the group when it is determined that the reference value within the group has shifted, and generate a module attenuation positioning mark based on the module performance benchmark and the reconstructed reference value within the group. The power generation loss assessment module is used to determine the anomaly type and impact range based on the component attenuation location markers and the preset hierarchical relationship between the component strings, combiner boxes, and inverters; and to calculate the power generation loss and loss contribution value corresponding to the anomaly based on the anomaly type and impact range. The monitoring decision output module is used to generate a maintenance priority list based on the anomaly type, impact range, power generation loss, and loss contribution value, and output the corresponding monitoring alarm results.

[0010] Preferably, the method for collecting photovoltaic module operating parameters and environmental parameters includes: The output voltage and output current are collected at the string output terminals of each photovoltaic module, and the output power of each photovoltaic module is calculated based on the output voltage and output current; temperature sensors are installed on the back panel of the photovoltaic modules to collect the module temperature. Irradiance sensors are deployed in the area where the photovoltaic array formed by the photovoltaic modules is located to collect the solar irradiance on the receiving surface of the modules; the output power, module temperature and solar irradiance of the photovoltaic modules are combined to form the operating parameters of the photovoltaic modules; In the meteorological monitoring unit of the photovoltaic power station, ambient temperature sensors, wind speed sensors and humidity sensors are deployed to collect ambient temperature, wind speed and humidity to form environmental parameters; the operating parameters of the photovoltaic modules and environmental parameters are obtained according to the preset sampling period, and the collection timestamp and equipment number are added to each parameter.

[0011] Preferably, the method for forming a spatiotemporal runtime dataset includes: The photovoltaic module operating parameters and environmental parameters with additional collection timestamps are resampled according to a preset time granularity, and the preset time granularity is used as a unified time reference to align the parameters within the same time interval, thus obtaining the set of synchronization parameters corresponding to each photovoltaic module at each unified moment. Read the equipment configuration file of the photovoltaic power station. The equipment configuration file records the equipment number of each photovoltaic module and its position coordinates in the photovoltaic array. Based on the equipment number, the synchronization parameter set is associated and integrated with the corresponding position coordinates to form a spatiotemporal operation dataset.

[0012] Preferably, the method for constructing a component performance benchmark includes: The output power and solar irradiance of each photovoltaic module at each unified moment are extracted from the spatiotemporal operation dataset. The output power is normalized according to the solar irradiance to obtain the normalized power of each photovoltaic module. The normalized power average of each photovoltaic module that meets the preset operating conditions within a preset number of consecutive days of operation is used as the initial performance benchmark; the normalized power that meets the preset operating conditions is extracted according to the preset update cycle, and the corresponding periodic performance benchmark value is calculated; the initial performance benchmark and the periodic performance benchmark values ​​are combined to form a module performance benchmark sequence, which serves as the module performance benchmark characterizing the performance changes of each photovoltaic module.

[0013] Preferably, the method for extracting the directional state of the power deviation includes: All photovoltaic modules are grouped according to their module model, rated power, installation tilt angle, and installation orientation. PV modules with the same parameters or that meet the preset consistency conditions are divided into the same comparison group; the periodic performance benchmark value of each PV module in the current update cycle is extracted. Calculate the average value of the periodic performance benchmark value of all photovoltaic modules in each comparison group in the current update cycle, and use it as the reference value within the comparison group; determine the power deviation of each photovoltaic module based on the difference between the periodic performance benchmark value of each photovoltaic module in the current update cycle and the reference value within the group. Based on the sign of the power deviation corresponding to each photovoltaic module, the direction state of the power deviation corresponding to each photovoltaic module is extracted. When the sign of the power deviation is positive, the direction state of the corresponding photovoltaic module is determined to be positive, indicating that the periodic performance benchmark value of the photovoltaic module is higher than the reference value within the group. When the power deviation is negative, the directional state of the corresponding photovoltaic module is determined to be negative, indicating that the periodic performance benchmark value of the photovoltaic module is lower than the reference value within the group; when the power deviation is zero, the directional state of the corresponding photovoltaic module is determined to be neutral, indicating that the periodic performance benchmark value of the photovoltaic module is consistent with the reference value within the group.

[0014] Preferably, the method for automatically reconstructing the intra-group reference value when an intra-group reference value shift occurs includes: The directional state of all photovoltaic modules in each comparison group is accumulated in the current update cycle to obtain the directional state aggregation quantity, which represents the overall deviation trend in the group. Based on the ratio of the absolute value of the directional state aggregation quantity to the number of photovoltaic modules in the comparison group, it is determined whether there is a systematic unidirectional offset in the comparison group. When the ratio exceeds the preset offset threshold, it is determined that the reference value within the group has shifted, and an offset mark for the reference value within the group is generated. After generating the offset mark for the reference value within the group, the photovoltaic modules are sorted according to the absolute value of the power deviation corresponding to each photovoltaic module, and photovoltaic modules whose absolute value of power deviation is in the top preset proportion are removed. The reference value within the group is recalculated based on the periodic performance benchmark value corresponding to the remaining photovoltaic modules in the current update cycle, and the recalculated reference value within the group is used as the reconstructed reference value within the group.

[0015] Preferably, the method for generating component attenuation positioning identifiers includes: Extract the periodic performance benchmark value of each photovoltaic module from the module performance benchmark corresponding to the current update cycle, and obtain the reconstructed intra-group reference value of the comparison group to which each photovoltaic module belongs in the current update cycle; calculate the reconstructed power deviation of each photovoltaic module relative to the reconstructed intra-group reference value based on the difference between the periodic performance benchmark value corresponding to each photovoltaic module and the reconstructed intra-group reference value. The relative performance loss ratio of each photovoltaic module is calculated based on the ratio of the difference between the reconstructed intra-group reference value and the corresponding periodic performance benchmark value to the reconstructed intra-group reference value. When the relative performance loss ratio exceeds the preset loss ratio threshold and the corresponding periodic performance benchmark value is lower than the reconstructed intra-group reference value, the corresponding photovoltaic module is determined to have abnormal degradation. Based on the position coordinates of the photovoltaic module with abnormal degradation in the photovoltaic array, a corresponding module degradation location identifier is generated.

[0016] Preferably, the method for calculating the power generation loss and loss contribution value corresponding to the anomaly includes: Based on the location coordinates in the component attenuation location markers, the corresponding component string number, combiner box number, and inverter number are queried from the pre-established equipment hierarchy table; the component attenuation location markers are collected and statistically analyzed according to the three levels of component string, combiner box, and inverter. The anomaly type is determined based on the distribution of abnormal components at each level. Specifically, when an anomaly involves only a single photovoltaic module, it is determined to be a module-level anomaly. When there are several abnormal photovoltaic modules in the same module string, it is determined to be a module string-level anomaly. When there are several abnormal module strings under the same combiner box, it is determined to be a combiner box-level anomaly. When there are several abnormal combiner boxes under the same inverter, it is determined to be an inverter-level anomaly. The set of devices covered by the corresponding anomaly type is taken as the scope of influence; the power generation loss corresponding to each abnormal photovoltaic module is calculated based on the absolute value of the reconfiguration power deviation of each abnormal photovoltaic module within the scope of influence and the duration of the current update cycle. The power generation losses of all abnormal photovoltaic modules within the same scope of the abnormal event are summed to obtain the total power generation loss corresponding to the abnormal event. The loss contribution value of each abnormal event is calculated based on the proportion of the total power generation loss corresponding to each abnormal event to the total power generation loss of all abnormal events in the current statistical period.

[0017] Preferably, the method for outputting the corresponding monitoring alarm results includes: An operation and maintenance assessment record is constructed for each abnormal event. The operation and maintenance assessment record includes the abnormality type, scope of impact, amount of power generation loss, loss contribution value and corresponding equipment number; and all operation and maintenance assessment records are sorted. The sorting is based on the loss contribution value as the first sorting criterion, the power generation loss as the second sorting criterion, and the preset hierarchical weight corresponding to the anomaly type as the third sorting criterion; after sorting, an operation and maintenance priority list is generated based on the sorting results. According to the preset alarm output format, the abnormal type, impact range, power generation loss, loss contribution value and equipment number corresponding to each abnormal event in the operation and maintenance priority list are encapsulated to generate the corresponding monitoring alarm results and output to the photovoltaic power station monitoring terminal.

[0018] A data analysis-based method for monitoring photovoltaic power plants includes: S1. Collect photovoltaic module operating parameters and environmental parameters, and perform time synchronization and spatial mapping processing to form a spatiotemporal operating dataset; S2. Normalize the output power of photovoltaic modules based on the spatiotemporal operation dataset to construct a module performance benchmark; divide similar photovoltaic modules into comparison groups, calculate the power deviation of each photovoltaic module relative to the reference value within the group, and extract the direction state of the power deviation. S3. Aggregate the directional state of the power deviation of each photovoltaic module, determine whether the reference value within the group has shifted, automatically reconstruct the reference value within the group when it is determined that the reference value within the group has shifted, and generate a module attenuation positioning mark based on the module performance benchmark and the reconstructed reference value within the group. S4. Based on the component attenuation location markers and the preset hierarchical relationship between the component strings, combiner boxes, and inverters, determine the anomaly type and its impact range; based on the anomaly type and its impact range, calculate the power generation loss and loss contribution value corresponding to the anomaly. S5. Generate an operation and maintenance priority list based on the anomaly type, impact range, power generation loss, and loss contribution value, and output the corresponding monitoring and alarm results.

[0019] Compared with the prior art, the present invention has the following beneficial effects: By extracting output power and solar irradiance from spatiotemporal operational datasets and normalizing the output power using reference irradiance under standard test conditions, the measured output under different irradiance conditions is converted to the same reference condition, solving the problem of indirect comparison of power data at different sampling times and improving the consistency and comparability of performance evaluation. An initial performance benchmark is constructed using the normalized average power value that meets preset operating conditions within a consecutive preset number of days during the initial operation of photovoltaic modules. Periodic performance benchmark values ​​are continuously calculated according to a preset update cycle, forming a module performance benchmark sequence and establishing a dynamically updated performance reference system over time. This sequence retains the initial performance level of the module, records the long-term performance change trajectory, weakens the impact of short-term disturbances, reveals the true degradation trend, and improves the accuracy, stability, and long-term tracking capability of photovoltaic module performance evaluation, providing a reliable data foundation for subsequent anomaly location and other tasks.

[0020] Photovoltaic modules are grouped according to their model, rated power, installation tilt angle, and orientation. Horizontal comparisons are only made between modules with identical or similar operating characteristics, addressing the issue of direct comparison between different types of modules and improving the rationality of comparison results. The average value of the periodic performance benchmark within each comparison group is calculated as a reference value within the group. The difference between each module's periodic performance benchmark and the group reference value is used as the power deviation, quantifying the relative performance level of each module. Then, the power deviation is discretized using a sign function, mapping continuous power deviations to positive, negative, and neutral states, representing module performance higher than, lower than, and approximately equal to the group's average level, respectively. This method preserves the directional information of the power deviation, converting continuous quantities into directional states, providing a unified data expression format, improving the accuracy of horizontal comparisons of similar modules, the ability to express directional information, and the stability of statistical analysis, providing a reliable foundation for the identification and precise location of abnormal attenuation. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of a photovoltaic power plant monitoring system based on data analysis according to the present invention; Figure 2 This is a schematic diagram of a photovoltaic power plant monitoring method based on data analysis according to the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0023] Please see Figure 1 As shown, this embodiment provides a photovoltaic power plant monitoring system based on data analysis, specifically including the following steps: The spatiotemporal operation mapping module is used to collect photovoltaic module operation parameters and environmental parameters, and perform time synchronization and spatial mapping processing to form a spatiotemporal operation dataset; The power deviation calculation module is used to normalize the output power of photovoltaic modules based on the spatiotemporal operation dataset and construct a module performance benchmark; it divides similar photovoltaic modules into comparison groups, calculates the power deviation of each photovoltaic module relative to the reference value within the group, and extracts the directional state of the power deviation. The module status tracking module is used to aggregate the directional status of the power deviation of each photovoltaic module, determine whether the reference value within the group has shifted, automatically reconstruct the reference value within the group when it is determined that the reference value within the group has shifted, and generate a module attenuation positioning mark based on the module performance benchmark and the reconstructed reference value within the group. The power generation loss assessment module is used to determine the anomaly type and impact range based on the component attenuation location markers and the preset hierarchical relationship between the component strings, combiner boxes, and inverters; and to calculate the power generation loss and loss contribution value corresponding to the anomaly based on the anomaly type and impact range. The monitoring decision output module generates a maintenance priority list based on anomaly type, impact range, power generation loss, and loss contribution, and outputs corresponding monitoring alarm results. All modules are connected via wired and / or wireless means to enable data transmission between them.

[0024] Methods for collecting photovoltaic module operating parameters and environmental parameters include: The output voltage and output current are collected at the string output terminals of each photovoltaic module, and the output power of each photovoltaic module is calculated based on the output voltage and output current; temperature sensors are installed on the back panel of the photovoltaic modules to collect the module temperature. Irradiance sensors are deployed in the area where the photovoltaic array formed by the photovoltaic modules is located to collect the solar irradiance on the receiving surface of the modules; the output power, module temperature and solar irradiance of the photovoltaic modules are combined to form the operating parameters of the photovoltaic modules; In the meteorological monitoring unit of the photovoltaic power station, ambient temperature sensors, wind speed sensors and humidity sensors are deployed to collect ambient temperature, wind speed and humidity to form environmental parameters; the operating parameters of the photovoltaic modules and environmental parameters are obtained according to the preset sampling period, and the collection timestamp and equipment number are added to each parameter.

[0025] In this embodiment, it should be noted that the data acquisition device in the photovoltaic power station monitoring system periodically triggers the parameter acquisition process according to a preset sampling period. The preset sampling period can be set according to the monitoring accuracy requirements, for example, 1 second, 10 seconds, 30 seconds, 1 minute, or 5 minutes. At the end of each sampling period, the data acquisition device sends an acquisition command to the monitoring unit corresponding to each photovoltaic module and the meteorological monitoring unit to obtain the current photovoltaic module operating parameters and environmental parameters.

[0026] The meteorological monitoring unit for a photovoltaic power station is a meteorological data acquisition device installed at the photovoltaic power station site. It is used to centrally collect environmental information surrounding the power station, including ambient temperature sensors, wind speed sensors, and humidity sensors. This meteorological monitoring unit can be deployed near the photovoltaic array in the form of an integrated small meteorological station. Its output data reflects the actual environmental conditions of the photovoltaic modules. During each data acquisition process, after receiving various parameters, the data acquisition device reads the system clock to generate the current acquisition time and writes the acquisition time into the corresponding data record as an acquisition timestamp. The acquisition timestamp can be in the format of "year-month-day hour:minute:second".

[0027] Simultaneously, the data acquisition device reads the device number corresponding to the currently acquired object from the device configuration table and writes the device number into the corresponding data record. The device number is used to uniquely identify the data source object; for example, it can be a "component number," "string number," or "sensor number." Therefore, after each sampling cycle of data acquisition is completed, each data record generated by the system includes the device number, the acquisition timestamp, and the corresponding parameter value.

[0028] For example, during a certain sampling period, the system collected data showing that the photovoltaic module numbered PV-A03-015 had an output power of 428.6W, a module temperature of 46.2℃, and a solar irradiance of 8122W / m². 2 Simultaneously, the ambient temperature was 31.5℃, the wind speed was 2.8m / s, and the humidity was 67%. The system read the current time "2026-04-16 10:30:00" as the acquisition timestamp and appended the device number "PV-A03-015" to the record, forming a data record containing the device number, acquisition timestamp, output power, component temperature, solar irradiance, and environmental parameters for subsequent time synchronization and spatial mapping processing.

[0029] Methods for generating spatiotemporal runtime datasets include: The photovoltaic module operating parameters and environmental parameters with additional collection timestamps are resampled according to a preset time granularity, and the preset time granularity is used as a unified time reference to align the parameters within the same time interval, thus obtaining the set of synchronization parameters corresponding to each photovoltaic module at each unified moment. The system sets a uniform time granularity, which can be set to 10 seconds, 30 seconds, 1 minute, or 5 minutes according to monitoring requirements. Then, continuous time intervals are divided using this time granularity, and each parameter is mapped to its corresponding time interval according to its collection timestamp. For multiple parameter values ​​falling within the same time interval, the valid value within that time interval is directly selected. If no corresponding parameter value exists within a certain time interval, a forward hold method can be used, taking the most recent valid value from the previous time interval as the parameter value for the current time interval; alternatively, a linear interpolation method can be used to calculate the parameter value for the current time interval based on two adjacent valid values.

[0030] After the above processing, at each unified moment, the system can obtain the output power, module temperature, solar irradiance, and ambient temperature, wind speed, and humidity corresponding to the same photovoltaic module, thus forming a set of synchronization parameters for the photovoltaic module at the unified moment.

[0031] Read the equipment configuration file of the photovoltaic power station. The equipment configuration file records the equipment number of each photovoltaic module and its position coordinates in the photovoltaic array. Based on the equipment number, the synchronization parameter set is associated and integrated with the corresponding position coordinates to form a spatiotemporal operation dataset.

[0032] After completing the time synchronization process, the system reads the equipment configuration file of the photovoltaic power station. The equipment configuration file records the equipment number of each photovoltaic module and its position coordinates in the photovoltaic array. The position coordinates can be represented in two-dimensional row and column coordinate form, such as the r-th row and c-th column, or in three-dimensional coordinate form.

[0033] Subsequently, the system uses the device number as the association key to search for the corresponding device number in the synchronization parameter set, and reads the location coordinates corresponding to that device number from the device configuration file, appending the location coordinates to the synchronization parameter set. After association, each synchronization parameter set contains not only the operating parameters and environmental parameters at a unified time, but also the spatial position of the corresponding photovoltaic module in the array. The synchronization parameter sets and their location coordinates corresponding to all photovoltaic modules at each unified time are integrated to form a spatiotemporal operation dataset.

[0034] For example, the system maps the output power data collected at 10:00:03, the component temperature data collected at 10:00:20, and the irradiance data collected at 10:00:45 to the time interval 10:00:00~10:01:00, corresponding to the unified time 10:01:00. After alignment, a synchronization parameter set is obtained for device number PV-A03-015. The system finds the location coordinates (A03, 15) corresponding to this device number in the device configuration file and appends these coordinates to the synchronization parameter set, ultimately forming the following record: Device number PV-A03-015, unified time 10:01:00, output power 428.6 W, component temperature 46.2℃, solar irradiance 812W / m². 2 The ambient temperature is 31.5℃, the wind speed is 2.8m / s, the humidity is 67%, and the location coordinates are (A03, 15). This record constitutes one data point in the spatiotemporal operation dataset.

[0035] Methods for building component performance benchmarks include: The output power and solar irradiance of each photovoltaic module at each unified moment are extracted from the spatiotemporal operation dataset. The output power is normalized according to the solar irradiance to obtain the normalized power of each photovoltaic module. Normalized power output for each photovoltaic module: ;in, Indicates that it is located at the th Line 1 The photovoltaic modules in the row at the same time The corresponding normalized power represents the equivalent output power obtained by converting the measured output power to standard irradiation conditions; Indicates that it is located at the th Line 1 The photovoltaic modules in the row at the same time The measured output power; Indicates that it is located at the th Line 1 The photovoltaic modules in the row at the same time Corresponding solar irradiance; Indicates the reference irradiance under standard test conditions; This indicates the row number of the photovoltaic module in the photovoltaic array; Indicates the column number of the photovoltaic module in the photovoltaic array; An index identifier representing a unified time; The normalized power average of each photovoltaic module that meets the preset operating conditions within a preset number of consecutive days of operation is used as the initial performance benchmark. Initial performance baseline: ;in, Indicates that it is located at the th Line 1 Initial performance benchmarks for photovoltaic modules in the series; This indicates the number of valid sampling points that meet the preset operating conditions within a preset number of consecutive days of operation. Indicates the valid sampling point number; Indicates the first The sampling time corresponding to each valid sampling point; Indicates the first Each valid sampling point at time... The corresponding normalized power; Extract the normalized power that meets the preset operating conditions according to the preset update cycle, and calculate the corresponding periodic performance benchmark value. Cyclic performance benchmark: ;in, Indicates the first The corresponding performance baseline value within each update cycle; Indicates the update cycle number; Indicates the first The number of valid sampling points that meet the preset operating conditions within each update cycle; The initial performance benchmark and the performance benchmark values ​​of each period are combined to form a module performance benchmark sequence, which serves as the module performance benchmark characterizing the performance changes of each photovoltaic module.

[0036] It should be noted that the preset operating conditions include that the solar irradiance at the corresponding sampling time is not lower than a preset solar irradiance threshold, and the normalized power corresponding to that sampling time is used as a valid sampling point in the performance benchmark calculation; when the solar irradiance is lower than the preset solar irradiance threshold, the corresponding sampling point is removed and not included in the performance benchmark calculation. For example, when the preset solar irradiance threshold is set to 700W / m 2 If the solar irradiance at a certain sampling moment is 812 W / m 2 If the sampling time meets the preset operating conditions; if the solar irradiance is 430W / m 2 If the sampling time does not meet the preset operating conditions, it will not participate in the calculation of the initial performance benchmark and periodic performance benchmark values.

[0037] Methods for extracting the directional state of power deviation include: All photovoltaic modules are grouped according to their module model, rated power, installation tilt angle, and installation orientation. PV modules with the same parameters or that meet the preset consistency conditions are divided into the same comparison group; the periodic performance benchmark value of each PV module in the current update cycle is extracted. Calculate the average value of the periodic performance benchmark value of all photovoltaic modules in each comparison group in the current update cycle, and use it as the reference value within the comparison group. Within-group reference values: ;in, Indicates the first Within each update cycle, the comparison group's intra-group reference value; Indicates the number of photovoltaic modules within the comparison group; Indicates the comparison within the group. The photovoltaic module is in the first The periodic performance baseline value for each update cycle; This indicates the index of the photovoltaic modules within the comparison group; The power deviation of each photovoltaic module is determined based on the difference between the periodic performance benchmark value and the reference value within the group corresponding to the current update cycle. Power deviations for each photovoltaic module: ;in, Indicates the first Within the [number] update cycle, located at [number]th [location] Line 1 Power deviation of the photovoltaic modules in the column; Based on the sign of the power deviation corresponding to each photovoltaic module, the direction state of the power deviation corresponding to each photovoltaic module is extracted. When the sign of the power deviation is positive, the direction state of the corresponding photovoltaic module is determined to be positive, indicating that the periodic performance benchmark value of the photovoltaic module is higher than the reference value within the group. Directional state of power deviation: ;in, Indicates the directional state of the power deviation of the photovoltaic module; Symbolic functions are used to map continuous numerical values ​​to discrete symbolic states. When the power deviation is negative, the directional state of the corresponding photovoltaic module is determined to be negative, indicating that the periodic performance benchmark value of the photovoltaic module is lower than the reference value within the group; when the power deviation is zero, the directional state of the corresponding photovoltaic module is determined to be neutral, indicating that the periodic performance benchmark value of the photovoltaic module is consistent with the reference value within the group.

[0038] Methods for automatically reconstructing intra-group reference values ​​when an intra-group reference value shift is detected include: The directional state of all photovoltaic modules in each comparison group is accumulated in the current update cycle to obtain the directional state aggregation quantity, which represents the overall deviation trend in the group. Based on the ratio of the absolute value of the directional state aggregation quantity to the number of photovoltaic modules in the comparison group, it is determined whether there is a systematic unidirectional offset in the comparison group. Directional state aggregation quantity: ;in, Indicates the moment of unification The corresponding directional state aggregation quantity is used to characterize the overall deviation direction of all photovoltaic modules in the comparison group; Indicates the comparison within the group. Each photovoltaic module at the same time The corresponding directional state; When the ratio exceeds the preset offset threshold, it is determined that the reference value within the group has shifted, and an offset mark for the reference value within the group is generated. After generating the offset mark for the reference value within the group, the photovoltaic modules are sorted according to the absolute value of the power deviation corresponding to each photovoltaic module, and photovoltaic modules whose absolute value of power deviation is in the top preset proportion are removed. The reference value within the group is recalculated based on the periodic performance benchmark value corresponding to the remaining photovoltaic modules in the current update cycle, and the recalculated reference value within the group is used as the reconstructed reference value within the group.

[0039] Methods for generating component attenuation location markers include: Extract the periodic performance benchmark value of each photovoltaic module from the module performance benchmark corresponding to the current update cycle, and obtain the reconstructed intra-group reference value of the comparison group to which each photovoltaic module belongs in the current update cycle; calculate the reconstructed power deviation of each photovoltaic module relative to the reconstructed intra-group reference value based on the difference between the periodic performance benchmark value corresponding to each photovoltaic module and the reconstructed intra-group reference value. Reconfiguration power deviation: ;in, Indicates the photovoltaic module in the first The reconstructed deviation value corresponding to each update cycle; Indicates the corresponding comparison group in the th The group reference value after reconstruction in each update cycle; The relative performance loss ratio of each photovoltaic module is calculated based on the ratio of the difference between the reconstructed intra-group reference value and the corresponding periodic performance benchmark value to the reconstructed intra-group reference value. The relative performance loss ratio for each photovoltaic module: ;in, Indicates the photovoltaic module in the first The relative performance loss percentage corresponding to each update cycle; When the relative performance loss ratio exceeds the preset loss ratio threshold and the corresponding periodic performance benchmark value is lower than the reconstructed intra-group reference value, the corresponding photovoltaic module is determined to have abnormal degradation; based on the position coordinates of the photovoltaic module with abnormal degradation in the photovoltaic array, the corresponding module degradation location identifier is generated.

[0040] The system uses location coordinates as the core positioning field and combines this with the abnormal degradation status of the photovoltaic module in the current update cycle to generate a unique module degradation positioning identifier. This identifier is used to quickly locate the abnormal module during subsequent power generation loss assessment and establish a correspondence between the abnormal module and its degradation level.

[0041] In this embodiment, the component attenuation location identifier can be constructed using a structured data format, including the following fields: The location coordinate field is used to record the row and column numbers of the photovoltaic modules in the photovoltaic array; The update cycle field is used to record the current update cycle number; The abnormal status field is used to record the abnormal degradation status of the photovoltaic module; The attenuation level field is used to record the relative performance loss ratio of the photovoltaic module.

[0042] When generating a component attenuation location identifier, the system first writes the location coordinate field into the location identifier, and then writes the current update cycle number, abnormal status and relative performance loss ratio in sequence to form a structured location record that corresponds one-to-one with the abnormal photovoltaic component.

[0043] Since the location coordinates directly correspond to the physical installation position of the photovoltaic module in the photovoltaic array, the generated module attenuation location marker can not only characterize whether an anomaly has occurred, but also accurately identify the location of the anomaly and its degree of attenuation.

[0044] For example, a photovoltaic module located in row 12, column 8 of a photovoltaic array is identified as having abnormal degradation within the 6th update cycle, with a relative performance loss ratio of 7.5%. The system can generate a module degradation location identifier in the following form: ID_decay={row number=12, column number=8, update cycle=6, abnormal state=degradation abnormal, degradation degree=7.5%}. After subsequent modules read this module degradation location identifier, they can directly determine the specific location of the abnormal module and calculate the impact of the module on the corresponding module string and the power generation capacity of the entire photovoltaic power station based on the degradation degree.

[0045] Methods for calculating the power generation loss and loss contribution value corresponding to anomalies include: Based on the location coordinates in the component attenuation location markers, the corresponding component string number, combiner box number, and inverter number are queried from the pre-established equipment hierarchy table; the component attenuation location markers are collected and statistically analyzed according to the three levels of component string, combiner box, and inverter. In this embodiment, it should be noted that a device hierarchy table is pre-established during the system deployment phase. This table records the correspondence between each photovoltaic module within the photovoltaic power station and its associated module string, combiner box, and inverter. Each record in the device hierarchy table includes the following fields: the location coordinates of the photovoltaic module, module number, module string number, combiner box number, and inverter number.

[0046] Among them, the location coordinates are used to uniquely identify the installation position of the photovoltaic module in the photovoltaic array, and can be represented in a two-dimensional coordinate form of row number-column number; the module string number is used to identify the series branch to which the photovoltaic module belongs; the combiner box number is used to identify the combiner box to which the module string is connected; and the inverter number is used to identify the inverter to which the combiner box is connected.

[0047] Upon receiving the component attenuation positioning identifier, the location coordinate field within the identifier is first parsed to extract the corresponding row and column numbers. Then, using these row and column numbers as query keys, a matching search is performed in the device hierarchy table to find records whose location coordinates match the row and column numbers.

[0048] Once a matching record is found, the corresponding component string number, combiner box number, and inverter number are read from that record. The read hierarchical information is then associated with the component attenuation location identifier to generate an anomaly location result containing device hierarchical information. The device hierarchical relationship table can be stored in a relational database, configuration file, or memory-mapped table. The query process can be implemented using primary key indexes, hash mapping, or dictionary lookups to improve query efficiency.

[0049] For example, the following record exists in the equipment hierarchy table: Location coordinates: (12, 8); Component number: PV-12-08; Component string number: STR-03; Combiner box number: CB-02; Inverter part number: INV-01; When the location coordinates in the component attenuation location identifier are (12, 8), the system can determine the component string number of the abnormal photovoltaic component as STR-03, the combiner box number as CB-02, and the inverter number as INV-01 by querying the device hierarchy table. Subsequently, the type of anomaly and its impact range can be determined based on the above hierarchy information.

[0050] The anomaly type is determined based on the distribution of abnormal components at each level. Specifically, when an anomaly involves only a single photovoltaic module, it is determined to be a module-level anomaly. When there are several abnormal photovoltaic modules in the same module string, it is determined to be a module string-level anomaly. When there are several abnormal module strings under the same combiner box, it is determined to be a combiner box-level anomaly. When there are several abnormal combiner boxes under the same inverter, it is determined to be an inverter-level anomaly. The set of devices covered by the corresponding anomaly type is taken as the scope of influence; the power generation loss corresponding to each abnormal photovoltaic module is calculated based on the absolute value of the reconfiguration power deviation of each abnormal photovoltaic module within the scope of influence and the duration of the current update cycle. The power generation loss corresponding to each abnormal photovoltaic module: ;in, Indicates the position of the photovoltaic array. Line 1 The photovoltaic modules in the column are in the first The amount of power generation loss corresponding to each update cycle; This represents the absolute value of the reconfiguration power deviation, used to indicate the magnitude of power loss; Indicates the first The duration corresponding to each update cycle; The power generation losses of all abnormal photovoltaic modules within the same scope of the abnormal event are summed to obtain the total power generation loss corresponding to the abnormal event. The loss contribution value of each abnormal event is calculated based on the proportion of the total power generation loss corresponding to each abnormal event to the total power generation loss of all abnormal events in the current statistical period.

[0051] The loss contribution value corresponding to each abnormal event: ;in, Indicates the first The loss contribution value corresponding to each abnormal event; Indicates the first The total power generation loss corresponding to each abnormal event; This represents the total power generation loss corresponding to all abnormal events within the current statistical period; This indicates the total number of abnormal events within the current statistical period; Indicates the sequence number index of the abnormal event; Indicates the sequence number of the abnormal event; For example, if an abnormal photovoltaic (PV) module corresponding to a certain module degradation location marker is located in the 12th row and 8th column of the PV array, and its module string is identified as S03, its combiner box as C02, and its inverter as INV01, and if four PV modules in the same module string exhibit abnormalities, then the abnormality type is determined to be a module string-level abnormality, and all abnormal PV modules covered by module string S03 constitute the affected area. If the total power generation loss corresponding to this abnormal event is 3.6 kWh, and the total power generation loss of all abnormal events in the current statistical period is 18 kWh, then the loss contribution value corresponding to this abnormal event is 0.2.

[0052] It should be noted that all calculation formulas in this invention follow the principle of dimensional consistency during the construction process. All parameters participating in the same addition and subtraction operation have the same physical unit, and the unit of the result after participating in multiplication and division operations is clear and corresponds to its physical meaning. Therefore, there will be no problem of calculation result distortion caused by inconsistent dimensions.

[0053] Specifically, normalized power, initial performance baseline, periodic performance baseline, intra-group reference value, power deviation, and reconfiguration power deviation are all used to characterize the power level of photovoltaic modules at different stages, and their physical units are all power units. Orientation state and orientation state aggregation are only used to characterize the deviation direction and the overall deviation trend within the group, and are both dimensionless state quantities.

[0054] The relative performance loss ratio is obtained by comparing the difference between two unit power values ​​to a reference power value; the result is a dimensionless ratio. The power generation loss is obtained by multiplying the power difference by the corresponding time length; the result is in energy units. The loss contribution value is obtained by comparing the power generation loss of a single abnormal event to the total power generation loss of all abnormal events; the result is also a dimensionless ratio. Because the units of each parameter are clear and consistent in the formulas, each calculation result has a clear physical meaning.

[0055] Methods for outputting corresponding monitoring and alarm results include: An operation and maintenance assessment record is constructed for each abnormal event. The operation and maintenance assessment record includes the abnormality type, scope of impact, amount of power generation loss, loss contribution value and corresponding equipment number; and all operation and maintenance assessment records are sorted. The sorting is based on the loss contribution value as the first sorting criterion, the power generation loss as the second sorting criterion, and the preset hierarchical weight corresponding to the anomaly type as the third sorting criterion; after sorting, an operation and maintenance priority list is generated based on the sorting results. In this embodiment, it should be noted that the sorting is first based on the loss contribution value, and then sorted from largest to smallest. The larger the loss contribution value, the greater the impact of the corresponding abnormal event on the total power generation loss of the current statistical period, and the higher the sorting priority.

[0056] When two or more abnormal events have the same loss contribution value, the power generation loss is used as the second ranking criterion, and the events are ranked from largest to smallest. The larger the power generation loss, the higher the ranking priority. When both the loss contribution value and the power generation loss are the same, the preset level weight corresponding to the abnormality type is used as the third ranking criterion. The preset level weight is used to characterize the importance of abnormalities at different equipment levels; the higher the equipment level, the greater the corresponding preset level weight, and the higher the ranking priority.

[0057] The preset hierarchical weights can be set as follows: Inverter-level anomalies: 4; Combiner box-level anomalies: 3; Module-level anomalies: 2; Photovoltaic module-level anomalies: 1. After sorting, priority numbers are assigned to each anomaly according to the sorting order, and the anomalies are arranged in order of priority number to generate an operation and maintenance priority list.

[0058] According to the preset alarm output format, the abnormal type, impact range, power generation loss, loss contribution value and equipment number corresponding to each abnormal event in the operation and maintenance priority list are encapsulated to generate the corresponding monitoring alarm results and output to the photovoltaic power station monitoring terminal.

[0059] The information in the maintenance priority list is encapsulated according to a preset alarm output format. The preset alarm output format is a predefined data structure used to uniformly describe monitoring alarm information, including the following fields: Priority number; anomaly type; scope of impact; power generation loss; loss contribution value; equipment number. The system writes the information corresponding to each anomaly in the maintenance priority list into the above fields, forming a structured monitoring and alarm result, and outputs the monitoring and alarm result to the photovoltaic power station monitoring terminal.

[0060] For example, if three anomalous events occur within a certain statistical period: Abnormal Event A: Component cascade anomaly, device number STR-03, power generation loss 3.6kWh, loss contribution value 0.20; Abnormal Event B: Combiner box level abnormality, equipment number CB-02, power generation loss 2.8kWh, loss contribution value 0.15; Abnormal Event C: Photovoltaic module-level abnormality, equipment number PV-12-08, power generation loss 0.9kWh, loss contribution value 0.05.

[0061] The system first sorts the data according to their loss contribution value, resulting in the order A, B, and C, and assigns them priority numbers 1, 2, and 3 respectively, generating an operation and maintenance priority list. Then, it encapsulates the data according to a preset alarm output format, generating the following monitoring alarm results: Priority 1, cascaded component anomaly, affected area STR-03, power generation loss 3.6kWh, loss contribution value 0.20, equipment number STR-03; Priority 2, combiner box level anomaly, affected area CB-02, power generation loss 2.8kWh, loss contribution value 0.15, equipment number CB-02; Priority 3, photovoltaic module-level anomaly, affected area PV-12-08, power generation loss 0.9kWh, loss contribution value 0.05, equipment number PV-12-08.

[0062] The preset loss ratio threshold is set by staff based on historical data analysis results. This historical analysis process includes the system collecting the relative performance loss ratios of multiple photovoltaic modules and calculating their average value as a reference to obtain the preset loss ratio threshold. Similarly, the preset offset threshold is also set by staff based on the system's historical operating data and the specific application scenario requirements, and is adjusted by staff according to the actual situation during system operation.

[0063] This embodiment extracts output power and solar irradiance from a spatiotemporal operational dataset, normalizes the output power using reference irradiance under standard test conditions, and converts the measured output under different irradiance conditions to the same reference condition. This solves the problem of incomparable power data at different sampling times, improving the consistency and comparability of performance evaluation. An initial performance benchmark is constructed using the normalized average power value that meets preset operating conditions within a preset number of consecutive days during the initial operation of the photovoltaic module. Periodic performance benchmark values ​​are continuously calculated according to a preset update cycle, forming a module performance benchmark sequence and establishing a dynamically updated performance reference system. This sequence retains the initial performance level of the module, records the long-term performance change trajectory, weakens the impact of short-term disturbances, reveals the true degradation trend, and improves the accuracy, stability, and long-term tracking capability of photovoltaic module performance evaluation, providing a reliable data foundation for subsequent anomaly location and other tasks.

[0064] Photovoltaic modules are grouped according to their model, rated power, installation tilt angle, and orientation. Horizontal comparisons are only made between modules with identical or similar operating characteristics, addressing the issue of direct comparison between different types of modules and improving the rationality of comparison results. The average value of the periodic performance benchmark within each comparison group is calculated as a reference value within the group. The difference between each module's periodic performance benchmark and the group reference value is used as the power deviation, quantifying the relative performance level of each module. Then, the power deviation is discretized using a sign function, mapping continuous power deviations to positive, negative, and neutral states, representing module performance higher than, lower than, and approximately equal to the group's average level, respectively. This method preserves the directional information of the power deviation, converting continuous quantities into directional states, providing a unified data expression format, improving the accuracy of horizontal comparisons of similar modules, the ability to express directional information, and the stability of statistical analysis, providing a reliable foundation for the identification and precise location of abnormal attenuation. Example

[0065] Please see Figure 2 As shown, parts not described in detail in this embodiment are described in Embodiment 1. A data analysis-based photovoltaic power plant monitoring method is provided, including: S1. Collect photovoltaic module operating parameters and environmental parameters, and perform time synchronization and spatial mapping processing to form a spatiotemporal operating dataset; S2. Normalize the output power of photovoltaic modules based on the spatiotemporal operation dataset to construct a module performance benchmark; divide similar photovoltaic modules into comparison groups, calculate the power deviation of each photovoltaic module relative to the reference value within the group, and extract the direction state of the power deviation. S3. Aggregate the directional state of the power deviation of each photovoltaic module, determine whether the reference value within the group has shifted, automatically reconstruct the reference value within the group when it is determined that the reference value within the group has shifted, and generate a module attenuation positioning mark based on the module performance benchmark and the reconstructed reference value within the group. S4. Based on the component attenuation location markers and the preset hierarchical relationship between the component strings, combiner boxes, and inverters, determine the anomaly type and its impact range; based on the anomaly type and its impact range, calculate the power generation loss and loss contribution value corresponding to the anomaly. S5. Generate an operation and maintenance priority list based on the anomaly type, impact range, power generation loss, and loss contribution value, and output the corresponding monitoring and alarm results. Example

[0066] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the photovoltaic power plant monitoring system based on data analysis described above.

[0067] Since the electronic device described in this embodiment is the electronic device used to implement the photovoltaic power plant monitoring system and method based on data analysis in the embodiments of this application, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the photovoltaic power plant monitoring system and method based on data analysis described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the photovoltaic power plant monitoring system and method based on data analysis in the embodiments of this application falls within the scope of protection of this application.

[0068] It should be noted that all formulas in this manual are calculated by removing dimensions and taking their numerical values. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0069] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A photovoltaic power plant monitoring system based on data analysis, characterized in that, include: The spatiotemporal operation mapping module is used to collect photovoltaic module operation parameters and environmental parameters, and perform time synchronization and spatial mapping processing to form a spatiotemporal operation dataset; The power deviation calculation module is used to normalize the output power of photovoltaic modules based on the spatiotemporal operation dataset and construct a module performance benchmark; it divides similar photovoltaic modules into comparison groups, calculates the power deviation of each photovoltaic module relative to the reference value within the group, and extracts the directional state of the power deviation. The module status tracking module is used to aggregate the directional status of the power deviation of each photovoltaic module, determine whether the reference value within the group has shifted, automatically reconstruct the reference value within the group when it is determined that the reference value within the group has shifted, and generate a module attenuation positioning mark based on the module performance benchmark and the reconstructed reference value within the group. The power generation loss assessment module is used to determine the anomaly type and impact range based on the component attenuation location markers and the preset hierarchical relationship between the component strings, combiner boxes, and inverters; and to calculate the power generation loss and loss contribution value corresponding to the anomaly based on the anomaly type and impact range. The monitoring decision output module is used to generate a maintenance priority list based on the anomaly type, impact range, power generation loss, and loss contribution value, and output the corresponding monitoring alarm results.

2. The photovoltaic power plant monitoring system based on data analysis according to claim 1, characterized in that, The method for collecting photovoltaic module operating parameters and environmental parameters includes: The output voltage and output current are collected at the string output terminals of each photovoltaic module, and the output power of each photovoltaic module is calculated based on the output voltage and output current; temperature sensors are installed on the back panel of the photovoltaic modules to collect the module temperature. Irradiance sensors are deployed in the area where the photovoltaic array formed by the photovoltaic modules is located to collect the solar irradiance on the receiving surface of the modules; the output power, module temperature and solar irradiance of the photovoltaic modules are combined to form the operating parameters of the photovoltaic modules; In the meteorological monitoring unit of the photovoltaic power station, ambient temperature sensors, wind speed sensors and humidity sensors are deployed to collect ambient temperature, wind speed and humidity to form environmental parameters; the operating parameters of the photovoltaic modules and environmental parameters are obtained according to the preset sampling period, and the collection timestamp and equipment number are added to each parameter.

3. The photovoltaic power plant monitoring system based on data analysis according to claim 2, characterized in that, The method for forming a spatiotemporal runtime dataset includes: The photovoltaic module operating parameters and environmental parameters with additional collection timestamps are resampled according to a preset time granularity, and the preset time granularity is used as a unified time reference to align the parameters within the same time interval, thus obtaining the set of synchronization parameters corresponding to each photovoltaic module at each unified moment. Read the equipment configuration file of the photovoltaic power station. The equipment configuration file records the equipment number of each photovoltaic module and its position coordinates in the photovoltaic array. Based on the equipment number, the synchronization parameter set is associated and integrated with the corresponding position coordinates to form a spatiotemporal operation dataset.

4. The photovoltaic power plant monitoring system based on data analysis according to claim 3, characterized in that, The method for constructing component performance benchmarks includes: The output power and solar irradiance of each photovoltaic module at each unified moment are extracted from the spatiotemporal operation dataset. The output power is normalized according to the solar irradiance to obtain the normalized power of each photovoltaic module. The normalized power average of each photovoltaic module that meets the preset operating conditions within a preset number of consecutive days of operation is used as the initial performance benchmark; the normalized power that meets the preset operating conditions is extracted according to the preset update cycle, and the corresponding periodic performance benchmark value is calculated; the initial performance benchmark and the periodic performance benchmark values ​​are combined to form a module performance benchmark sequence, which serves as the module performance benchmark characterizing the performance changes of each photovoltaic module.

5. A photovoltaic power plant monitoring system based on data analysis according to claim 4, characterized in that, The method for extracting the directional state of the power deviation includes: All photovoltaic modules are grouped according to their module model, rated power, installation tilt angle, and installation orientation. PV modules with the same parameters or that meet the preset consistency conditions are divided into the same comparison group; the periodic performance benchmark value of each PV module in the current update cycle is extracted. Calculate the average value of the periodic performance benchmark value of all photovoltaic modules in each comparison group in the current update cycle, and use it as the reference value within the comparison group; determine the power deviation of each photovoltaic module based on the difference between the periodic performance benchmark value of each photovoltaic module in the current update cycle and the reference value within the group. Based on the sign of the power deviation corresponding to each photovoltaic module, the direction state of the power deviation corresponding to each photovoltaic module is extracted. When the sign of the power deviation is positive, the direction state of the corresponding photovoltaic module is determined to be positive, indicating that the periodic performance benchmark value of the photovoltaic module is higher than the reference value within the group. When the power deviation is negative, the directional state of the corresponding photovoltaic module is determined to be negative, indicating that the periodic performance benchmark value of the photovoltaic module is lower than the reference value within the group; when the power deviation is zero, the directional state of the corresponding photovoltaic module is determined to be neutral, indicating that the periodic performance benchmark value of the photovoltaic module is consistent with the reference value within the group.

6. The photovoltaic power plant monitoring system based on data analysis according to claim 5, characterized in that, The method for automatically reconstructing the intra-group reference value when an intra-group reference value shift occurs includes: The directional state of all photovoltaic modules in each comparison group is accumulated in the current update cycle to obtain the directional state aggregation quantity, which represents the overall deviation trend in the group. Based on the ratio of the absolute value of the directional state aggregation quantity to the number of photovoltaic modules in the comparison group, it is determined whether there is a systematic unidirectional offset in the comparison group. When the ratio exceeds the preset offset threshold, it is determined that the reference value within the group has shifted, and an offset mark for the reference value within the group is generated. After generating the offset mark for the reference value within the group, the photovoltaic modules are sorted according to the absolute value of the power deviation corresponding to each photovoltaic module, and photovoltaic modules whose absolute value of power deviation is in the top preset proportion are removed. The reference value within the group is recalculated based on the periodic performance benchmark value corresponding to the remaining photovoltaic modules in the current update cycle, and the recalculated reference value within the group is used as the reconstructed reference value within the group.

7. A photovoltaic power plant monitoring system based on data analysis according to claim 6, characterized in that, The method for generating component attenuation positioning identifiers includes: Extract the periodic performance benchmark value of each photovoltaic module from the module performance benchmark corresponding to the current update cycle, and obtain the reconstructed intra-group reference value of the comparison group to which each photovoltaic module belongs in the current update cycle; calculate the reconstructed power deviation of each photovoltaic module relative to the reconstructed intra-group reference value based on the difference between the periodic performance benchmark value corresponding to each photovoltaic module and the reconstructed intra-group reference value. The relative performance loss ratio of each photovoltaic module is calculated based on the ratio of the difference between the reconstructed intra-group reference value and the corresponding periodic performance benchmark value to the reconstructed intra-group reference value. When the relative performance loss ratio exceeds the preset loss ratio threshold and the corresponding periodic performance benchmark value is lower than the reconstructed intra-group reference value, the corresponding photovoltaic module is determined to have abnormal degradation. Based on the position coordinates of the photovoltaic module with abnormal degradation in the photovoltaic array, a corresponding module degradation location identifier is generated.

8. A photovoltaic power plant monitoring system based on data analysis according to claim 7, characterized in that, The method for calculating the power generation loss and loss contribution value corresponding to the anomaly includes: Based on the location coordinates in the component attenuation location markers, the corresponding component string number, combiner box number, and inverter number are queried from the pre-established equipment hierarchy table; the component attenuation location markers are collected and statistically analyzed according to the three levels of component string, combiner box, and inverter. The anomaly type is determined based on the distribution of abnormal components at each level. Specifically, when an anomaly involves only a single photovoltaic module, it is determined to be a module-level anomaly. When there are several abnormal photovoltaic modules in the same module string, it is determined to be a module string-level anomaly. When there are several abnormal module strings under the same combiner box, it is determined to be a combiner box-level anomaly. When there are several abnormal combiner boxes under the same inverter, it is determined to be an inverter-level anomaly. The set of devices covered by the corresponding anomaly type is taken as the scope of influence; the power generation loss corresponding to each abnormal photovoltaic module is calculated based on the absolute value of the reconfiguration power deviation of each abnormal photovoltaic module within the scope of influence and the duration of the current update cycle. The power generation losses of all abnormal photovoltaic modules within the same scope of the abnormal event are summed to obtain the total power generation loss corresponding to the abnormal event. The loss contribution value of each abnormal event is calculated based on the proportion of the total power generation loss corresponding to each abnormal event to the total power generation loss of all abnormal events in the current statistical period.

9. A photovoltaic power plant monitoring system based on data analysis according to claim 8, characterized in that, The method for outputting the corresponding monitoring and alarm results includes: An operation and maintenance assessment record is constructed for each abnormal event. The operation and maintenance assessment record includes the abnormality type, scope of impact, amount of power generation loss, loss contribution value and corresponding equipment number; and all operation and maintenance assessment records are sorted. The sorting is based on the loss contribution value as the first sorting criterion, the power generation loss as the second sorting criterion, and the preset hierarchical weight corresponding to the anomaly type as the third sorting criterion; after sorting, an operation and maintenance priority list is generated based on the sorting results. According to the preset alarm output format, the abnormal type, impact range, power generation loss, loss contribution value and equipment number corresponding to each abnormal event in the operation and maintenance priority list are encapsulated to generate the corresponding monitoring alarm results and output to the photovoltaic power station monitoring terminal.

10. A photovoltaic power plant monitoring method based on data analysis, implemented by a photovoltaic power plant monitoring system based on data analysis as described in any one of claims 1 to 9, characterized in that, include: S1. Collect photovoltaic module operating parameters and environmental parameters, and perform time synchronization and spatial mapping processing to form a spatiotemporal operating dataset; S2. Normalize the output power of photovoltaic modules based on the spatiotemporal operation dataset to construct a module performance benchmark; divide similar photovoltaic modules into comparison groups, calculate the power deviation of each photovoltaic module relative to the reference value within the group, and extract the direction state of the power deviation. S3. Aggregate the directional state of the power deviation of each photovoltaic module, determine whether the reference value within the group has shifted, automatically reconstruct the reference value within the group when it is determined that the reference value within the group has shifted, and generate a module attenuation positioning mark based on the module performance benchmark and the reconstructed reference value within the group. S4. Based on the component attenuation location markers and the preset hierarchical relationship between the component strings, combiner boxes, and inverters, determine the anomaly type and its impact range; based on the anomaly type and its impact range, calculate the power generation loss and loss contribution value corresponding to the anomaly. S5. Generate an operation and maintenance priority list based on the anomaly type, impact range, power generation loss, and loss contribution value, and output the corresponding monitoring and alarm results.

Images