Component-level shadow detection method and device for photovoltaic system

Abstract

The invention relates to the technical field of data processing, and discloses a component-level shadow detection method and device for a photovoltaic system. The method comprises the following steps: carrying out synchronous sampling and data storage processing on real-time voltage data of each photovoltaic module in a group string to obtain a sampling data packet containing a module voltage value, a timestamp and a slave station ID (Identity); performing characteristic analysis processing based on a voltage deviation ratio on component voltage distribution in the sampling data packet to obtain component-level abnormal voltage identification information; performing multi-dimensional analysis processing on the abnormal voltage identification information based on time domain features to obtain a shadow feature index including a shielding degree and duration; and carrying out grading judgment processing on the shadow characteristic index based on a preset threshold value to obtain a shadow detection report containing an alarm grade and an optimization strategy. According to the invention, the efficiency and accuracy of component-level shadow detection of the photovoltaic system are improved.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular, to a method and device for component-level shadow detection of a photovoltaic system. Background Art

[0002] With the wide application of photovoltaic power generation systems, photovoltaic modules are affected by environmental factors, especially shadow occlusion, which has become one of the main reasons affecting the power generation efficiency of the system. Component-level shadow detection technology has been proposed to monitor the voltage changes of photovoltaic modules in real time. Through technical means such as distributed acquisition networks and clock synchronization, voltage data of photovoltaic modules are obtained and analyzed to identify shadow occlusion and perform warning and optimization. Current photovoltaic shadow detection systems are usually deployed at the component level of each photovoltaic string. By collecting data such as voltage values, timestamps, and component IDs, voltage feature analysis is performed, and combined with parameters such as reference voltage and deviation rate, a shadow feature index is constructed. Further, the system evaluates the occlusion degree and duration of abnormal data through multi-dimensional analysis, providing real-time optimization strategies for the photovoltaic system.

[0003] Although existing component-level shadow detection methods can improve the accuracy of shadow detection to a certain extent, there are still some key problems. First, the distributed deployment of multi-channel voltage acquisition devices and the implementation of high-precision clock synchronization are difficult. Especially in large-scale photovoltaic arrays, the clock synchronization of slave devices is vulnerable to transmission delay and environmental factors, which may lead to inaccurate sampling. Second, the accuracy of the existing detection methods for analyzing abnormal voltage features is low, and it is difficult to effectively distinguish the voltage changes caused by shadow occlusion from the effects of other factors (such as temperature, component aging, etc.) under complex environmental conditions. These deficiencies may cause false alarms or missed alarms in the system's identification of shadow occlusion, affecting the overall power generation efficiency. Summary of the Invention

[0004] This application provides a method and device for component-level shadow detection of a photovoltaic system, which is used to improve the efficiency and accuracy of component-level shadow detection of a photovoltaic system.

[0005] In a first aspect, this application provides a method for component-level shadow detection of a photovoltaic system, and the method for component-level shadow detection of the photovoltaic system includes:

[0006] Deploy multi-channel voltage acquisition devices for photovoltaic modules in a photovoltaic string to obtain a distributed acquisition network including a master device and multiple slave devices;

[0007] Perform clock synchronization clearing processing on the counters of each slave device based on the IEEE1588 protocol to obtain a microsecond-level synchronous sampling time base;

[0008] Synchronously sample and process the real-time voltage data of each photovoltaic module in the string to obtain a sampling data packet containing the module voltage value, timestamp, and slave station ID;

[0009] Perform feature analysis processing based on the voltage deviation rate on the component voltage distribution in the sampling data packet to obtain component-level abnormal voltage identification information;

[0010] Perform multi-dimensional analysis processing based on time-domain characteristics on the abnormal voltage identification information to obtain a shadow feature index containing the degree of occlusion and duration;

[0011] Perform hierarchical judgment processing based on a preset threshold on the shadow feature index to obtain a shadow detection report containing the alarm level and optimization strategy.

[0012] In a second aspect, the present application provides a component-level shadow detection device for a photovoltaic system. The component-level shadow detection device for the photovoltaic system includes:

[0013] A deployment module for deploying multi-channel voltage acquisition devices for the photovoltaic modules in the photovoltaic string to obtain a distributed acquisition network including a master station device and multiple slave station devices;

[0014] A clearing module for performing clock synchronization and clearing processing on the counters of each slave station device based on the IEEE1588 protocol to obtain a microsecond-level synchronous sampling time base;

[0015] A storage module for synchronously sampling and data storage processing of the real-time voltage data of each photovoltaic module in the string to obtain a sampling data packet containing the module voltage value, timestamp, and slave station ID;

[0016] An analysis module for performing feature analysis processing based on the voltage deviation rate on the component voltage distribution in the sampling data packet to obtain component-level abnormal voltage identification information;

[0017] A processing module for performing multi-dimensional analysis processing based on time-domain characteristics on the abnormal voltage identification information to obtain a shadow feature index containing the degree of occlusion and duration;

[0018] A judgment module for performing hierarchical judgment processing based on a preset threshold on the shadow feature index to obtain a shadow detection report containing the alarm level and optimization strategy.

[0019] In the technical solution provided by this application, through the deployment of multi-channel voltage acquisition devices, a distributed acquisition network including a master station device and multiple slave station devices is constructed, enabling the system to achieve refined voltage acquisition at the component level and ensuring the comprehensiveness and accuracy of data acquisition. Secondly, based on the clock synchronization and clearing process of the IEEE 1588 protocol, a microsecond-level synchronous sampling time base is provided for each slave station device. This high-precision clock synchronization technology enables the voltage acquisition of each component to be carried out under the same time reference, thus avoiding data errors caused by time asynchronization and providing a reliable basis for subsequent data analysis. In addition, by synchronously sampling and storing the real-time voltage data of each photovoltaic component, and including the voltage value, timestamp, and slave station ID in the sampling data packet, the system can effectively trace the voltage change process of each component, helping to identify the shading situation and its duration suffered by a specific component. In terms of data processing, through the feature analysis based on the voltage deviation rate of the component voltage distribution in the sampling data packet, the system can accurately identify abnormal voltage values and generate component-level abnormal voltage identification information. Based on this feature analysis process, the system can accurately judge the voltage changes caused by shadow shading and separately mark such abnormal voltage information for subsequent analysis. At the same time, through the multi-dimensional analysis process of the abnormal voltage identification information, the system can obtain the shadow feature index including the shading degree and duration. Such a multi-dimensional analysis method enables the system not only to identify the occurrence of shadow shading but also to further quantify the degree of shading and its specific impact on the performance of photovoltaic components. Finally, through the hierarchical judgment process based on a preset threshold for the shadow feature index to generate a shadow detection report including the alarm level and optimization strategy, the system realizes the integrated function from shadow detection to alarm generation and strategy optimization. The advantage of this method is that the system not only stays at the basic function of shadow detection but further provides hierarchical alarm information to help users take corresponding optimization measures under different levels of shadow influence to ensure the maximum power generation efficiency of the photovoltaic system. Generally speaking, through the combination of technical features such as clock synchronization, voltage feature analysis, and multi-dimensional judgment, this method not only greatly improves the accuracy and real-time performance of shadow detection but also effectively reduces the possibility of false alarms and missed alarms, providing a solid technical guarantee for the efficient operation of the photovoltaic system. Brief Description of the Drawings

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0021] Figure 1Schematic diagram of an embodiment of the component-level shadow detection method for a photovoltaic system in an embodiment of the present application;

[0022] Figure 2 Schematic diagram of an embodiment of the component-level shadow detection device for a photovoltaic system in an embodiment of the present application. Detailed implementation manners

[0023] Embodiments of the present application provide a component-level shadow detection method and device for a photovoltaic system. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims, and above-mentioned drawings of the present application are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments described here can be implemented in an order other than those illustrated or described here. In addition, the term "including" or "having" and any deformation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily need to be limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0024] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 An embodiment of the component-level shadow detection method for a photovoltaic system in an embodiment of the present application includes:

[0025] Step S101: Deploy multi-channel voltage acquisition devices for photovoltaic modules in a photovoltaic string to obtain a distributed acquisition network including a master station device and multiple slave station devices;

[0026] Step S102: Perform clock synchronization and clearing processing on the counters of each slave station device based on the IEEE1588 protocol to obtain a microsecond-level synchronous sampling time base;

[0027] Step S103: Perform synchronous sampling and data storage processing on the real-time voltage data of each photovoltaic module in the string to obtain a sampling data packet including the component voltage value, time stamp, and slave station ID;

[0028] Step S104: Perform feature analysis processing on the component voltage distribution in the sampling data packet based on the voltage deviation rate to obtain component-level abnormal voltage identification information;

[0029] Step S105: Perform multi-dimensional analysis processing on the abnormal voltage identification information based on time-domain features to obtain a shadow feature index including the occlusion degree and duration;

[0030] Step S106: Perform hierarchical judgment processing on the shadow feature index based on a preset threshold to obtain a shadow detection report including an alarm level and an optimization strategy.

[0031] It can be understood that the execution entity of this application can be a component-level shadow detection device of a photovoltaic system, or a terminal or a server. Specifically, it is not limited here. In the embodiments of this application, the server is used as the execution entity for illustration.

[0032] Specifically, through the arrangement of multi-channel voltage acquisition devices, a distributed acquisition network including a master device and multiple slave devices is built. Here, the master device serves as the central management node, connecting multiple slave devices and forming a chain-like communication structure to ensure that the voltage data of each photovoltaic module can be collected by the master station in real time. To achieve component-level monitoring, the slave devices are given independent address assignments and communication protocol configurations, enabling precise identification and acquisition of the status information of each photovoltaic module in a large-scale photovoltaic array, ensuring comprehensive and accurate voltage acquisition. The IEEE1588 protocol is used to perform clock synchronization processing on the counters of the slave devices, and a zero-clearing synchronization operation is performed with a microsecond-level precision. This high-precision synchronization ensures that each slave device can collect data under the same time reference. The IEEE1588 protocol is time-synchronized by the master station, uses a delay request and response mechanism to measure the transmission delay, and then corrects the local counters of each device, effectively eliminating the time differences between different devices, enabling the voltage sampling data to be highly consistent and avoiding data deviations caused by clock asynchronization.

[0033] The collected voltage data is subjected to real-time acquisition and storage processing through a synchronous sampling time base, and finally a complete sampling data packet including voltage values, timestamps, and slave IDs is generated. In this step, each slave device eliminates noise through anti-aliasing filtering and then performs digital conversion to obtain accurate voltage values. After associating the data with the timestamp and marking it with the slave ID, accurate storage of the data is achieved, laying a stable data foundation for the subsequent voltage feature analysis. Abnormal voltage identification is performed on the sampling data packet through feature analysis of the voltage deviation rate. First, the system calculates the voltage mean value of each component to establish a reference voltage. Then, the voltage value of each component is compared with the reference voltage one by one to determine the degree of deviation and obtain abnormal voltage information. The greater the degree of deviation, the higher the possibility that the component is affected by abnormalities. Through the abnormal voltage identification information obtained by the analysis, it is possible to identify which components are affected during the current time period.

[0034] Perform multi-dimensional analysis on the time-domain characteristics of abnormal voltage information to generate a shadow feature index containing the degree of occlusion and the duration. Specifically, the abnormal voltage information will be subjected to time-series analysis to evaluate the duration of the abnormal state and perform feature classification based on a set time window. Through a multi-level threshold division system, the voltage deviation rate can be graded to determine the degree of occlusion, and the final shadow feature index is comprehensively generated by combining the duration of the occlusion. Through hierarchical judgment to process the shadow feature index, a shadow detection report containing the alarm level and optimization strategy is generated. The system will set different thresholds according to different levels of occlusion to ensure accurate judgment of different degrees of occlusion. For moderate or severe occlusion situations, the system generates corresponding alarms and recommends adjusting the operating parameters of the component to reduce the power loss caused by occlusion, thereby helping the photovoltaic system adjust strategies under different conditions to ensure the optimal power generation efficiency.

[0035] For example, in practical applications, assume that the reference voltage mean of a certain string is 60V, and the collected data shows that the voltage of a certain component drops to 54V. The system identifies this deviation as abnormal voltage information. In the subsequent analysis, it is found that the voltage of this component has been abnormal for 30 minutes. This persistence information, combined with the preset occlusion degree threshold, rates its occlusion as moderate occlusion. According to the generated report, the system recommends adjusting the operating parameters of this component to reduce the power loss caused by shadow occlusion and optimize the overall power generation efficiency.

[0036] In the embodiments of the present application, through the deployment of multi-channel voltage acquisition devices, a distributed acquisition network including a master station device and multiple slave station devices is constructed, enabling the system to achieve refined voltage acquisition at the component level and ensuring the comprehensiveness and accuracy of data acquisition. Secondly, based on the clock synchronization and clearing process of the IEEE 1588 protocol, a microsecond-level synchronous sampling time base is provided for each slave station device. This high-precision clock synchronization technology enables the voltage acquisition of each component to be carried out under the same time reference, thus avoiding data errors caused by time asynchronization and providing a reliable basis for subsequent data analysis. In addition, by synchronously sampling and storing the real-time voltage data of each photovoltaic component, and including the voltage value, timestamp, and slave station ID in the sampling data packet, the system can effectively trace the voltage change process of each component, helping to identify the shading situation and its duration of a specific component. In terms of data processing, through the feature analysis based on the voltage deviation rate of the component voltage distribution in the sampling data packet, the system can accurately identify abnormal voltage values and generate component-level abnormal voltage identification information. Based on this feature analysis process, the system can accurately judge the voltage changes caused by shadow shading and separately mark such abnormal voltage information for subsequent analysis. At the same time, through the multi-dimensional analysis process of the abnormal voltage identification information, the system can obtain a shadow feature index including the shading degree and duration. Such a multi-dimensional analysis method enables the system not only to identify the occurrence of shadow shading but also to further quantify the degree of shading and its specific impact on the performance of photovoltaic components. Finally, through the hierarchical judgment process based on a preset threshold for the shadow feature index, a shadow detection report including an alarm level and an optimization strategy is generated, and the system realizes the integrated function from shadow detection to alarm generation and strategy optimization. The advantage of this method is that the system not only stays at the basic function of shadow detection but further provides hierarchical alarm information to help users take corresponding optimization measures under different levels of shadow effects to ensure the maximum power generation efficiency of the photovoltaic system. Generally speaking, through the combination of technical features such as clock synchronization, voltage feature analysis, and multi-dimensional judgment, this method not only greatly improves the accuracy and real-time performance of shadow detection but also effectively reduces the possibility of false alarms and missed alarms, providing a solid technical guarantee for the efficient operation of the photovoltaic system.

[0037] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0038] (1) Perform connection analysis on the topological structure of the photovoltaic string to obtain the positional relationship of the components in the string, and perform slave station device addressing on the positional relationship to obtain a slave station ID allocation table;

[0039] (2) Perform communication interface configuration processing on the slave station ID allocation table based on the POWERBUS bus to obtain a communication framework supporting 255 nodes, and perform MODBUS protocol adaptation processing on the communication framework to obtain the master-slave communication protocol set;

[0040] (3) Perform data acquisition module configuration processing on the master-slave communication protocol set to obtain an acquisition channel group with a range of 0 - 60V, and perform ADC configuration processing on the acquisition channel group to obtain a digital acquisition unit with an accuracy of 0.5%;

[0041] (4) Perform opto-isolated electrical isolation processing on the digital acquisition unit to obtain an isolated acquisition module with an isolation strength of 2.5kV, and perform IP65 protection configuration processing on the isolated acquisition module to obtain a protected acquisition device;

[0042] (5) Perform physical wiring processing on the protected acquisition device to obtain a connection line with a communication distance of 3000 meters, and perform non-polar connection processing on the connection line to obtain a secure connection network;

[0043] (6) Perform working environment adaptation processing on the secure connection network to obtain a distributed acquisition network including a master station device and multiple slave station devices.

[0044] Specifically, perform connection analysis processing on the topological structure of the photovoltaic string to identify the positional relationship of each component. By analyzing the relative positions of the components within the string, the arrangement order and topological positions of each component can be accurately determined, and each component can be associated with a specific slave station device. Based on the positional relationship, the system generates a slave station ID allocation table, and each component is assigned an independent ID, thus establishing the address allocation of the slave station devices. This allocation table is the basis for subsequent communication and acquisition, ensuring that each slave station device can be accurately located and participate in data acquisition. Through the communication interface configuration of the POWERBUS bus, a communication framework supporting 255 nodes is obtained. Under this communication framework, the system adapts the MODBUS protocol to achieve standardized communication between the master station and the slave stations. The MODBUS protocol provides a simple and reliable transmission mechanism in master-slave communication, enabling the master station to effectively manage and schedule all slave station devices, while ensuring that the slave station devices can transmit the acquired voltage data back to the master station. Such a communication framework not only supports large-scale node connections but also ensures the stability and anti-interference ability of data transmission.

[0045] The master-slave communication protocol set is used to configure the data acquisition module, and a set of acquisition channels with a range of 0 - 60V is constructed. This configuration ensures that the voltage acquisition range can cover the normal operating voltage of the photovoltaic modules. The acquisition channel set is configured with an ADC (Analog-to-Digital Converter) to obtain a digital acquisition unit with an accuracy of 0.5%, ensuring extremely low acquisition error of voltage data and laying a data foundation for high-precision shadow detection. To avoid the influence of electrical interference on the acquired data, electrical isolation processing based on optocouplers is implemented on the digital acquisition unit to obtain an isolated acquisition module with an isolation strength of 2.5kV. Electrical isolation effectively eliminates interference between components and improves the safety of the system by blocking the electrical path. The isolated acquisition module is configured with IP65 protection to meet the requirements of outdoor equipment for dust and water protection, enabling it to work stably for a long time in the harsh environment of the photovoltaic power station.

[0046] To achieve stable long-distance communication, physical wiring is carried out for the protected acquisition device, and a connection line that can cover a communication distance of 3000 meters is designed and formed. The wiring process is simplified by means of non-polar connection. In non-polar connection, the installation and maintenance do not need to consider the polarity direction, greatly reducing the complexity of on-site construction, reducing the risk of miscontact, and at the same time improving the connection efficiency and safety of the system. The working environment of the secure connection network is adapted to form a distributed acquisition network including a master station device and multiple slave station devices. This network operates stably in the complex environment of the photovoltaic system and realizes real-time acquisition and transmission of voltage data for each component. The entire distributed network architecture not only covers a large range of photovoltaic arrays but also ensures the precise positioning of each slave station device and the efficient transmission of data, enabling the shadow detection system to achieve real-time monitoring and accurate analysis at the component level.

[0047] Through the above configuration process, the system can fully adapt to complex environmental changes in the actual photovoltaic array. For example, during the deployment of a large photovoltaic array, the voltage acquisition network of 100 components was successfully built using the above configuration method. The master station device manages each slave station device in real time, and through the slave station ID allocation table, communication interface configuration, and electrical isolation module, a stable data acquisition link with efficient communication and electrical safety is formed. During actual operation, even if a single component is short-term blocked, the acquisition system can record the voltage deviation in real time and quickly transmit it back to the master station for accurate analysis and application of optimization strategies in subsequent shadow detection.

[0048] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0049] (1) Perform frequency stability detection processing on the local clock sources of each slave station device to obtain the frequency deviation value of the clock source, and perform clock drift analysis processing based on Allan variance on the frequency deviation value to obtain the clock stability parameter;

[0050] (2) Perform GPS-based timing calibration processing on the clock signal of the master station device to obtain a UTC time reference signal, and perform timestamp generation processing based on the PTP protocol on the UTC time reference signal to obtain a master station reference timestamp;

[0051] (3) Perform delay measurement processing on the counters of each slave station device based on the delay request-response mechanism to obtain the transmission delay between the master and slave devices, and perform symmetry compensation processing on the transmission delay to obtain a delay compensation parameter;

[0052] (4) Perform synchronous clearing processing on the slave station counter based on the master station reference timestamp to obtain a unified counting start point, and perform clock taming processing on the counting start point to obtain a microsecond-level synchronous sampling time base.

[0053] Specifically, the system detects the frequency stability of the local clock source of each slave station device and obtains the frequency deviation value of the clock source of each slave station device. Frequency deviation refers to the deviation amount of the clock source relative to the ideal frequency. After obtaining the frequency deviation value, the system further performs clock drift analysis on the frequency deviation based on the Allan variance. The Allan variance is a commonly used statistic for analyzing the short-term stability of clock signals and helps to evaluate the fluctuations of the clock source in a short period. By calculating the Allan variance, the system can quantify the stability parameter of the clock to ensure that the clocks of the slave station devices do not cause data deviation due to frequency drift during the acquisition process. This stability parameter provides a calibration basis for subsequent clock synchronization.

[0054] The system performs GPS timing calibration on the clock signal of the master station device to obtain an accurate UTC (Coordinated Universal Time) time reference. GPS timing is a high-precision timing method that provides a standard UTC time through satellite signals, eliminating the drift error in the time of the master station device. After obtaining the UTC reference, the master station device generates a high-precision timestamp based on the PTP (Precision Time Protocol) to form a reference timestamp for the master station. The PTP protocol is a network time synchronization protocol that can achieve sub-microsecond synchronization accuracy within a local area network, providing a reliable time reference for the master station device and making the master station device the time reference in the entire acquisition system. To further accurately synchronize the time of the slave stations with the master station device, the system performs delay measurement on the counters of each slave station device based on the delay request-response mechanism. The delay request-response mechanism is a common measurement method. By having the slave station device send a request and receive a response from the master station, the system can calculate the transmission delay between the master and slave devices. After obtaining the delay data, the system performs symmetry compensation processing, that is, eliminates the asymmetry in the network delay to obtain an accurate delay compensation parameter. The delay compensation parameter is used to correct the timing error of the slave station so that the slave station device can better align with the master station time and avoid the impact of transmission delay errors on the accuracy of data acquisition.

[0055] Based on the reference timestamp of the master station, the system synchronously clears the counters of each slave device to obtain a unified counting starting point. This clearing operation enables the counters of each slave device to start counting from the same time reference, ensuring that the sampling of each slave device is completed on the same time reference. In addition, the system performs clock taming on the cleared counting starting point. Clock taming is a dynamic adjustment technique used to fine-tune the clock source of the slave station to better track the master station time and achieve a microsecond-level synchronous sampling time base. This operation fine-tunes the frequency and phase of the slave station clock, enabling the counters of each slave device to remain synchronized within the accuracy of microseconds.

[0056] For example, in practical applications, if the reference time of the master station device is 10:00:00.000000 UTC, the slave station device measures the transmission delay with the master station as 10 microseconds through the delay request-response mechanism. After the system performs symmetry compensation, this delay is accurately corrected to zero, and the clearing and synchronization operation is executed, so that the sampling time of each slave device is exactly the same as that of the master station device. When the system starts sampling, the data acquisition of each slave device is completed with microsecond-level accuracy based on the time reference of the master station. This high-precision synchronous sampling time base not only eliminates the time error in data acquisition but also ensures component-level real-time monitoring in large-scale photovoltaic systems, making shadow detection more accurate and reliable.

[0057] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0058] (1) Perform sampling period division processing on the synchronous sampling time base to obtain uniformly distributed sampling moments, and generate broadcast trigger signals for the sampling moments to obtain a global synchronous sampling instruction;

[0059] (2) Perform anti-aliasing filtering processing on the acquisition channels of each slave device to obtain a band-limited voltage signal, and perform digital conversion processing on the voltage signal based on a 12-bit ADC to obtain a voltage sampling value;

[0060] (3) Perform timestamp association processing on the voltage sampling values to obtain sampled data with timestamps, and perform slave station ID marking processing on the sampled data with timestamps to obtain a complete sampling data packet.

[0061] Specifically, the sampling period of the synchronous sampling time base is divided to obtain uniformly distributed sampling moments. This means that the acquisition system distributes sampling points according to the set time period, ensuring that the intervals between each sampling moment are consistent, thereby guaranteeing a constant sampling frequency for the system. After the division of the synchronous sampling time base is completed, the system generates a broadcast trigger signal to make all slave devices start sampling at the same time point. This operation is called the global synchronous sampling instruction. While broadcasting the global synchronous sampling instruction, it ensures the synchronization of the acquisition actions of all devices, avoids the problem of time deviation between different devices, and lays a foundation for unified data acquisition.

[0062] The acquisition channels of each slave device are subjected to anti-aliasing filtering to obtain a band-limited voltage signal. Anti-aliasing filtering is a signal processing technique that restricts the frequency range of the voltage signal to avoid interference from high-frequency noise on the acquisition result. During the acquisition process in the photovoltaic system, the anti-aliasing filter will remove high-frequency signals exceeding the sampling frequency, ensuring that the acquired voltage data accurately reflects the actual voltage situation. Subsequently, the system performs 12-bit ADC (analog-to-digital converter) processing on the filtered voltage signal to convert the analog voltage signal into a digital signal. 12-bit ADC means dividing the analog signal into 4096 different digital magnitudes, enabling extremely high voltage acquisition accuracy, ensuring that the system can capture the minute changes in the voltage of photovoltaic modules, and enhancing the resolution and accuracy of the sampling data. The system performs timestamp association processing on the obtained voltage sampling values, that is, matching each sampling value with the corresponding timestamp. This timestamp is generated based on the time reference after synchronization between the master station and the slave stations, ensuring that the moment of each sampling data is unified and accurate. After the timestamp association processing is completed, the system marks the voltage data with timestamps with the slave station ID to form a complete sampling data packet. Each sampling data packet contains information such as voltage values, timestamps, and slave station IDs. These information enable the system to accurately trace the specific components and sampling times of each voltage data, facilitating subsequent analysis and processing.

[0063] For example, during the actual sampling process, assume that the sampling period set by the system is 10 milliseconds. Through the sampling period division in step (1), all slave devices synchronously collect voltage data every 10 milliseconds. After anti-aliasing filtering, the voltage signal is converted into 12-bit digital data. At a certain sampling moment, the voltage value collected by a slave device is 45.6V, which is converted into the corresponding digital quantity through ADC, and this data has an accurate timestamp, such as "10:00:01.010" and the ID of the slave device "ID102". These information are integrated into a complete data packet, such as "ID102,10:00:01.010,45.6V", and transmitted to the master device for storage. The complete sampling data packet not only ensures the unique correspondence between time and device, but also enables the system to accurately locate specific components and their voltage changes in subsequent shadow detection and analysis, providing accurate data support for the efficient operation of the photovoltaic system.

[0064] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0065] (1) Perform intra-string statistical analysis on the voltage values in the sampling data packet to obtain the intra-string voltage mean value, and perform reference voltage calculation on the intra-string voltage mean value to obtain the normal operating reference voltage;

[0066] (2) Perform deviation calculation on each component voltage value with the reference voltage to obtain the voltage deviation rate, and perform threshold comparison on the voltage deviation rate to obtain the component-level abnormal voltage identification information.

[0067] Specifically, the system performs intra-string statistical analysis on the voltage values in the sampling data packet. This step mainly summarizes and averages the voltage sampling data of all components within the entire string, so as to obtain an average value representing the overall voltage level of the string. This average value reflects the voltage distribution of the components within the string under the condition of no obvious shading, providing a basis for judging the normal operating state of the system. Then, through the calculation of the reference voltage for the intra-string voltage mean value, a standard operating voltage value, that is, the "normal operating reference voltage", is obtained. The normal operating reference voltage is obtained through long-term observation and statistics of the historical operation data of the string, representing the voltage reference of the string under ideal conditions.

[0068] The system further calculates the deviation between the voltage values of each component and the normal operating reference voltage to obtain the voltage deviation rate. The calculation of the voltage deviation rate reveals the relative deviation degree of each component's voltage from the reference voltage, providing a basis for identifying abnormal voltages. If the voltage deviation rate of a component is low, it indicates that the component's voltage is basically within the normal range; while a high deviation rate means that the component may be shaded or there are other faults. The system compares the voltage deviation rate with a preset threshold value and determines the degree of voltage abnormality according to the deviation degree. After the threshold comparison process, the system generates component-level abnormal voltage identification information, laying a foundation for identifying shaded components.

[0069] For example, in a photovoltaic string, the system statistically analyzes the voltage values of multiple components and obtains the average voltage of the string as 60V. Based on historical data, the system confirms that the normal operating reference voltage of this string is 59.5V. Subsequently, the system calculates the deviation rate of each component's voltage value from the reference voltage one by one. For example, if the sampled voltage of a certain component is 55V, the voltage deviation rate of this component reflects the obvious difference between its working state and the reference voltage. In the threshold comparison, if the voltage deviation rate exceeds the preset tolerance range, the system generates abnormal voltage identification information for this component, indicating that it may be shaded. This information is further transmitted to the monitoring system for corresponding optimization strategies or further diagnosis, providing a reliable basis for ensuring the normal operation of the photovoltaic system.

[0070] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0071] (1) Perform a temporal continuity analysis process on the abnormal voltage identification information to obtain the duration of voltage abnormality, and perform a time window partitioning process on the duration to obtain time characteristic parameters;

[0072] (2) Perform a multi-level threshold classification process on the voltage deviation rate to obtain the shading degree level, and perform a feature fusion process on the shading degree level and the time characteristic parameters to obtain a shadow feature index.

[0073] Specifically, the system performs a temporal continuity analysis on the abnormal voltage identification information to determine the duration of voltage abnormality. The purpose of the temporal continuity analysis is to determine whether the component voltage deviation is a short-term fluctuation or a continuous abnormality by analyzing the continuity of voltage abnormality data over time. The system determines whether the abnormality is a temporary instantaneous shading or a long-term shadow shading by detecting the duration of voltage abnormality. Subsequently, the system performs a time window partitioning process on the duration to obtain time characteristic parameters. Time window partitioning is to divide the time period into multiple time characteristics to describe the voltage change situation in different time periods. This process provides support in the time dimension for further analysis of shadow features.

[0074] The system performs multi - level threshold classification processing on the voltage deviation rate to determine the level of occlusion. The multi - level threshold classification divides the voltage deviation rate into different levels such as mild occlusion, moderate occlusion, and severe occlusion by setting multiple deviation rate thresholds. This can visually quantify the degree of occlusion of the components. For example, a deviation rate within the range below the mild threshold is considered mild occlusion, within the moderate range is moderate occlusion, and so on. This multi - level classification method can help the system accurately evaluate the intensity of occlusion of the photovoltaic modules. Subsequently, the level of occlusion is fused with the time - characteristic parameters to form a comprehensive shadow - characteristic index. Feature fusion refers to integrating different types of information (i.e., the degree of occlusion and time persistence) so that the system can judge the impact of the shadow in a more comprehensive way.

[0075] For example, in practical applications, the voltage deviation of a certain photovoltaic module has been in an abnormal state for up to 15 minutes, and the deviation rate conforms to the threshold range of moderate occlusion. Through time - series continuity analysis, the system records the 15 - minute abnormal duration and divides it into multiple 5 - minute time windows. The voltage deviation within each window remains consistent, so the time - characteristic parameter "15 - minute continuous abnormality" is obtained. Combining with the multi - level threshold classification, this module is rated as "moderate occlusion". Finally, the system fuses the occlusion level with the time characteristics to generate a shadow - characteristic index, which is used to visually show the degree and duration of occlusion of this module, thus providing a basis for subsequent optimization strategies.

[0076] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0077] (1) Perform multi - level threshold judgment processing on the shadow - characteristic index to obtain the alarm - level judgment result, and perform alarm - information generation processing on the alarm - level judgment result to obtain classified alarm information;

[0078] (2) Perform persistence analysis processing on the classified alarm information to obtain the shadow - type judgment result, and perform optimization - strategy matching processing on the shadow - type judgment result to obtain a shadow - detection report including the alarm level and the optimization strategy.

[0079] Specifically, the system performs multi-level threshold judgment on the shadow feature index to determine the alarm level. The multi-level threshold judgment is based on multiple threshold levels of the preset occlusion severity, such as mild, moderate, and severe occlusion, etc. By comparing the shadow feature index with these thresholds, the system classifies the occlusion situation of the component into different alarm levels. The higher the alarm level, the more serious the impact of the occlusion. After completing the judgment of the alarm level, the system generates corresponding alarm information, which includes the alarm level of the component, the description of the occlusion situation, etc. The classification of the alarm information facilitates users to quickly understand the operating status of the photovoltaic system and take different levels of inspection or maintenance according to different levels. The system conducts continuous analysis on the generated classified alarm information to further determine the shadow type. Continuous analysis means observing the continuity of the alarm information over a period of time to judge whether the occlusion situation is instantaneous (such as passing clouds) or long-term (such as tree occlusion). By statistically analyzing the duration of the occlusion event, the system determines the type of the shadow. According to the results of the continuous analysis, the system classifies the shadow type into categories such as temporary occlusion and permanent occlusion. Different shadow types have different impacts on photovoltaic components, so the system selects a matching optimization strategy based on the shadow type. For temporary occlusion, the system may recommend increasing the load-bearing capacity of the component in the short term, while for long-term occlusion, the system may recommend adjusting the installation position of the component or optimizing the overall operating parameters of the system.

[0080] After determining the alarm level and the shadow type, the system integrates this information to form a shadow detection report. The report includes the alarm level, shadow type, and recommended optimization strategy of each component. This report can be used by the photovoltaic system management personnel to help them timely understand the operating status of the components and take targeted maintenance and optimization measures. For example, in a certain photovoltaic system, the system analyzes the shadow feature index of component A and finds that its occlusion degree reaches the moderate threshold range, so it determines that the component has a moderate alarm and generates the information of "moderate alarm". Subsequently, through continuous analysis, it is found that this occlusion phenomenon occurs every day in the past two weeks, and the system determines that the occlusion of this component is "long-term occlusion". Based on this shadow type, the system recommends in the report that the management personnel check and trim the occluder, or consider adjusting the position of the component to reduce the long-term impact of the occlusion on the power generation efficiency. The finally generated shadow detection report details the alarm level, shadow type, and recommended optimization strategy of component A, thus providing clear data support for management decisions.

[0081] The above describes the component-level shadow detection method of the photovoltaic system in the embodiment of the present application. Next, the component-level shadow detection device of the photovoltaic system in the embodiment of the present application will be described. Please refer to Figure 2 , an embodiment of the component-level shadow detection device of the photovoltaic system in the embodiment of the present application includes:

[0082] The deployment module 201 is used to deploy multi-channel voltage acquisition devices for photovoltaic modules in a photovoltaic string, obtaining a distributed acquisition network including a master station device and multiple slave station devices;

[0083] The clearing module 202 is used to perform clock synchronization clearing processing based on the IEEE 1588 protocol on the counters of each slave station device, obtaining a microsecond-level synchronous sampling time base;

[0084] The storage module 203 is used to perform synchronous sampling and data storage processing on the real-time voltage data of each photovoltaic module in the string, obtaining a sampling data packet including the component voltage value, timestamp, and slave station ID;

[0085] The analysis module 204 is used to perform feature analysis processing based on the voltage deviation rate on the component voltage distribution in the sampling data packet, obtaining component-level abnormal voltage identification information;

[0086] The processing module 205 is used to perform multi-dimensional analysis processing based on time-domain features on the abnormal voltage identification information, obtaining a shadow feature index including the occlusion degree and duration;

[0087] The judgment module 206 is used to perform hierarchical judgment processing based on a preset threshold on the shadow feature index, obtaining a shadow detection report including the alarm level and optimization strategy.

[0088] Through the collaborative cooperation of the above-mentioned various components and the deployment of multi-channel voltage acquisition devices, a distributed acquisition network including a master station device and multiple slave station devices is constructed, enabling the system to achieve refined voltage acquisition at the component level and ensuring the comprehensiveness and accuracy of data acquisition. Secondly, based on the clock synchronization zero-clearing process of the IEEE1588 protocol, a microsecond-level synchronous sampling time base is provided for each slave station device. This high-precision clock synchronization technology enables the voltage acquisition of each component to be carried out under the same time reference, thus avoiding data errors caused by time asynchronization and providing a reliable basis for subsequent data analysis. In addition, by synchronously sampling and storing the real-time voltage data of each photovoltaic component, and including the voltage value, timestamp, and slave station ID in the sampling data packet, the system can effectively trace the voltage change process of each component and help identify the shading situation and its duration of a specific component. In terms of data processing, through the feature analysis based on the voltage deviation rate of the component voltage distribution in the sampling data packet, the system can accurately identify abnormal voltage values and generate component-level abnormal voltage identification information. Based on this feature analysis process, the system can accurately judge the voltage changes caused by shadow shading and separately mark such abnormal voltage information for subsequent analysis. At the same time, through the multi-dimensional analysis process of the abnormal voltage identification information, the system can obtain the shadow feature index including the shading degree and duration. Such a multi-dimensional analysis method enables the system to not only identify the occurrence of shadow shading but also further quantify the degree of shading and its specific impact on the performance of photovoltaic components. Finally, through the hierarchical judgment process based on a preset threshold for the shadow feature index to generate a shadow detection report including the alarm level and optimization strategy, the system realizes the integrated function from shadow detection to alarm generation and strategy optimization. The advantage of this method is that the system not only stays at the basic function of shadow detection but further provides hierarchical alarm information to help users take corresponding optimization measures under different levels of shadow impact to ensure the maximum power generation efficiency of the photovoltaic system. Generally speaking, through the combination of technical features such as clock synchronization, voltage feature analysis, and multi-dimensional judgment, this method not only greatly improves the accuracy and real-time performance of shadow detection but also effectively reduces the possibility of false alarms and missed alarms, providing a solid technical guarantee for the efficient operation of the photovoltaic system.

[0089] As described above, the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method for component-level shadow detection in a photovoltaic system, characterized in that, The component-level shadow detection method for the photovoltaic system includes: Deploy multi-channel voltage acquisition devices for the photovoltaic components in the photovoltaic string to obtain a distributed acquisition network including a master station device and multiple slave station devices; Perform clock synchronization and clearing processing based on the IEEE1588 protocol on the counters of each slave station device to obtain a microsecond-level synchronous sampling time base; Perform synchronous sampling and data storage processing on the real-time voltage data of each photovoltaic component in the string to obtain a sampling data packet including the component voltage value, timestamp, and slave station ID; Perform feature analysis processing based on the voltage deviation rate on the component voltage distribution in the sampling data packet to obtain component-level abnormal voltage identification information; Perform multi-dimensional analysis processing based on time-domain features on the abnormal voltage identification information to obtain a shadow feature index including the occlusion degree and duration; Perform hierarchical judgment processing based on a preset threshold on the shadow feature index to obtain a shadow detection report including the alarm level and optimization strategy.

2. The component-level shadow detection method for a photovoltaic system according to claim 1, wherein, The deployment of multi-channel voltage acquisition devices for the photovoltaic components in the photovoltaic string to obtain a distributed acquisition network including a master station device and multiple slave station devices includes: Perform connection analysis processing on the topological structure of the photovoltaic string to obtain the positional relationship of the components in the string, and perform slave station device addressing processing on the positional relationship to obtain a slave station ID allocation table; Perform communication interface configuration processing based on the POWERBUS bus on the slave station ID allocation table to obtain a communication framework supporting 255 nodes, and perform MODBUS protocol adaptation processing on the communication framework to obtain a master-slave communication protocol set; Perform data acquisition module configuration processing on the master-slave communication protocol set to obtain an acquisition channel group with a range of 0-60V, and perform ADC configuration processing on the acquisition channel group to obtain a digital acquisition unit with an accuracy of 0.5%; Perform electrical isolation processing based on optocouplers on the digital acquisition unit to obtain an isolation acquisition module with an isolation strength of 2.5kV, and perform IP65 protection configuration processing on the isolation acquisition module to obtain a protected acquisition device; Perform physical wiring processing on the protected acquisition device to obtain a connection line with a communication distance of 3000 meters, and perform non-polar connection processing on the connection line to obtain a safe connection network; Perform working environment adaptation processing on the safe connection network to obtain a distributed acquisition network including a master station device and multiple slave station devices.

3. The component-level shadow detection method for a photovoltaic system according to claim 1, characterized in that The clock synchronization and clearing processing based on the IEEE1588 protocol on the counters of each slave station device to obtain a microsecond-level synchronous sampling time base includes: Perform frequency stability detection processing on the local clock source of each slave station device to obtain the frequency deviation value of the clock source, and perform clock drift analysis processing based on Allan variance on the frequency deviation value to obtain the clock stability parameter; Perform timing calibration processing based on GPS on the clock signal of the master station device to obtain a UTC time reference signal, and perform timestamp generation processing based on the PTP protocol on the UTC time reference signal to obtain a master station reference timestamp; Perform delay measurement processing on the counters of each slave device based on the delayed request-response mechanism to obtain the transmission delay between the master and slave devices, and perform symmetry compensation processing on the transmission delay to obtain a delay compensation parameter; Perform synchronous clearing processing on the slave counters based on the master station reference timestamp to obtain a unified counting starting point, and perform clock taming processing on the counting starting point to obtain a microsecond-level synchronous sampling time base.

4. The component-level shadow detection method for a photovoltaic system according to claim 1, wherein The synchronous sampling and data storage processing of the real-time voltage data of each photovoltaic module in the string to obtain a sampling data packet including the module voltage value, timestamp, and slave ID includes: Perform sampling period division processing on the synchronous sampling time base to obtain uniformly distributed sampling moments, and perform broadcast trigger signal generation processing on the sampling moments to obtain a global synchronous sampling instruction; Perform anti-aliasing filtering processing on the acquisition channels of each slave device to obtain a band-limited voltage signal, and perform digital conversion processing on the voltage signal based on a 12-bit ADC to obtain a voltage sampling value; Perform timestamp association processing on the voltage sampling values to obtain timestamped sampling data, and perform slave ID marking processing on the timestamped sampling data to obtain a complete sampling data packet.

5. The component-level shadow detection method for a photovoltaic system according to claim 1, characterized in that, The feature analysis processing of the module voltage distribution in the sampling data packet based on the voltage deviation rate to obtain module-level abnormal voltage identification information includes: Perform statistical analysis processing on the voltage values in the sampling data packet within the string to obtain the string voltage average value, and perform reference voltage calculation processing on the string voltage average value to obtain the normal operating reference voltage; Perform deviation calculation processing on each module voltage value from the reference voltage to obtain the voltage deviation rate, and perform threshold comparison processing on the voltage deviation rate to obtain module-level abnormal voltage identification information.

6. The component-level shadow detection method for a photovoltaic system according to claim 1, wherein, The multi-dimensional analysis processing of the abnormal voltage identification information based on the time domain characteristics to obtain a shadow feature index including the occlusion degree and duration includes: Perform sequential continuity analysis processing on the abnormal voltage identification information to obtain the duration of the voltage abnormality, and perform time window division processing on the duration to obtain time feature parameters; Perform multi-level threshold classification processing on the voltage deviation rate to obtain the occlusion degree level, and perform feature fusion processing on the occlusion degree level and the time feature parameters to obtain the shadow feature index.

7. The method for component-level shadow detection of a photovoltaic system according to claim 1, characterized in that, The hierarchical judgment processing of the shadow feature index based on a preset threshold to obtain a shadow detection report including the alarm level and optimization strategy includes: Perform multi-level threshold judgment processing on the shadow feature index to obtain the alarm level determination result, and perform alarm information generation processing on the alarm level determination result to obtain hierarchical alarm information; Perform persistence analysis processing on the hierarchical alarm information to obtain the shadow type determination result, and perform optimization strategy matching processing on the shadow type determination result to obtain a shadow detection report including the alarm level and optimization strategy.

8. A component-level shadow detection device for a photovoltaic system, which is used to implement the component-level shadow detection method of the photovoltaic system described in any one of claims 1-7, characterized in that, The module-level shadow detection device of the photovoltaic system includes: A deployment module for deploying multi-channel voltage acquisition devices for the photovoltaic modules in the photovoltaic string to obtain a distributed acquisition network including a master station device and multiple slave devices; A clearing module, which is used to perform clock synchronization clearing processing on the counters of each slave device based on the IEEE 1588 protocol to obtain a microsecond-level synchronous sampling time base; A storage module, which is used to perform synchronous sampling and data storage processing on the real-time voltage data of each photovoltaic module in the string to obtain a sampling data packet containing the component voltage value, timestamp, and slave ID; An analysis module, which is used to perform feature analysis processing on the component voltage distribution in the sampling data packet based on the voltage deviation rate to obtain component-level abnormal voltage identification information; A processing module, which is used to perform multi-dimensional analysis processing on the abnormal voltage identification information based on time-domain features to obtain a shadow feature index containing the occlusion degree and duration; A judgment module, which is used to perform hierarchical judgment processing on the shadow feature index based on a preset threshold to obtain a shadow detection report containing the warning level and optimization strategy.

Images