A zero-current photovoltaic string branch automatic alarm method and system

Abstract

The application discloses a zero-current photovoltaic string branch automatic alarm method and system, the method comprises the following steps: collecting the direct-current data of each photovoltaic string branch under each busbar box, and synchronously collecting the corresponding inverter output power data, busbar identification data, branch number data and branch maintenance state data; uploading the collected data to the power monitoring background and establishing a string branch real-time monitoring database; determining the effective power generation working condition based on the comparison result of the inverter output power and the preset power generation power threshold; performing zero-current screening on the branch in the effective power generation working condition, and performing carding shielding determination in combination with the branch maintenance state; performing fault positioning on the zero-current fault candidate branch; triggering alarm output according to the positioning result, and generating zero-current alarm information of the corresponding branch. The application can realize automatic identification, accurate positioning and hierarchical alarm of the zero-current fault branch, and is suitable for large photovoltaic power station operation monitoring scenes.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power plant operation monitoring and fault diagnosis technology, and in particular to an automatic alarm method and system for zero-current photovoltaic string branches. Background Technology

[0002] As the installed capacity of photovoltaic power plants continues to increase, the number of photovoltaic strings within these plants is also growing. A single plant area often contains a large number of combiner boxes, inverters, and string branches. As the basic unit for power collection and output, the operating status of the photovoltaic strings directly affects the power generation efficiency and operational safety of the entire power plant. If a string branch experiences a disconnection, poor contact, a damaged fuse, component malfunction, or other fault, the current in that branch may drop to zero. If such faults are not detected in time, it will not only result in power generation loss but also increase the difficulty of subsequent troubleshooting and maintenance.

[0003] Currently, in photovoltaic power plants, the common practice is to collect the DC current of each string branch on the DC side of the combiner box or the box-inverter integrated unit, and upload the collection results to the monitoring backend so that operators can view the real-time current data of each branch. This method can reflect the operating values ​​of each string, has basic data monitoring capabilities, and plays a certain role in daily operation.

[0004] However, current technologies largely focus on data acquisition and display, emphasizing the display of current values ​​rather than automatically identifying and proactively alarming for abnormal conditions. For operators, when the power plant is large and has a large number of strings, the backend interface often displays a large amount of monitoring data simultaneously. Manually checking the current status of each string branch is not only labor-intensive but also difficult to maintain consistently high efficiency and accuracy. In particular, when a branch experiences zero current, without a clear automatic alert, operators often fail to detect the problem immediately, easily leading to the fault remaining silent for an extended period.

[0005] Furthermore, whether the current in a photovoltaic string branch is zero cannot be determined in isolation from actual operating conditions. When the inverter is operating under low load, when sunlight is weak in the early morning or late evening, when the weather changes suddenly, or when the branch is under manual maintenance, the current may be low or even close to zero. If a simple judgment is made based on a single current value, it is easy to misjudge normal operating conditions as faults, resulting in false alarms. Too many false alarms will reduce the operators' trust in alarm information and thus affect the efficiency of handling real faults.

[0006] Therefore, although existing technologies can collect and display photovoltaic string branch current data, they still have the following shortcomings: First, they lack the ability to actively identify branches with zero current faults; second, they lack an intelligent judgment mechanism that combines power generation conditions for comprehensive judgment; and third, they lack methods to prevent false alarms by differentiating and processing maintenance status. As a result, they cannot meet the actual needs of large-scale photovoltaic power plants for rapid fault detection, accurate fault location, and automatic alarm.

[0007] Based on this, the core problem that the existing technology urgently needs to solve is how to establish an automatic alarm method and system for zero-current photovoltaic string branches using existing photovoltaic string current acquisition data without significantly increasing system complexity. This system can automatically identify zero-current fault branches based on operating conditions, accurately distinguish between normal low current and abnormal zero-current states, and actively alarm and accurately locate faulty branches. Summary of the Invention

[0008] The purpose of this invention is to address the problems in existing technologies where zero-current faults in photovoltaic string branches mainly rely on manual monitoring data, resulting in delayed fault detection, difficulty in accurately judging faults based on actual power generation conditions, and susceptibility to false alarms during maintenance. This invention provides an automatic alarm method and system for zero-current photovoltaic string branches. Without significantly increasing system complexity, it utilizes existing photovoltaic string branch current acquisition data, inverter output power data, and branch maintenance status data to establish an automatic identification, fault location, and hierarchical alarm processing mechanism for string-level zero-current faults. This enables proactive identification, accurate differentiation, rapid location, and automatic alarm for zero-current fault branches.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides an automatic alarm method for zero-current photovoltaic string branches, comprising the following steps: First, DC current data of each photovoltaic string branch under each combiner box in the photovoltaic power station is collected, and inverter output power data, combiner box identification data, branch number data and branch maintenance status data corresponding to each photovoltaic string branch are collected simultaneously to generate the original monitoring dataset of the string branches.

[0010] Specifically, the DC current signal of each photovoltaic string branch in each combiner box is acquired branch by branch, and the inverter output power signal is acquired through the corresponding inverter operation monitoring interface; at the same time, the combiner box address information, branch circuit number information and inverter association information corresponding to each photovoltaic string branch are read, and the above basic identification information is bound one-to-one with the collected branch electrical operation data; furthermore, the branch maintenance label status mark pre-set in the power monitoring background is read, and the branch maintenance label status mark is matched to the corresponding photovoltaic string branch.

[0011] Based on this, the generated data records are processed for time synchronization, data integrity verification, and abnormal null value filtering. The data that passes the verification is then encapsulated according to a unified field structure to generate the original monitoring dataset of the string branch that can be directly called for subsequent judgment and processing.

[0012] Then, the original monitoring dataset of the string branches is transmitted to the measurement and control acquisition device via the communication link, and then uploaded to the power monitoring backend by the measurement and control acquisition device. The DC current data, inverter output power data, combiner box identification data, branch number data and branch maintenance status data of each photovoltaic string branch are associated and stored to form a real-time monitoring database of the string branches.

[0013] Preferably, the DC current data, inverter output power data, combiner box identification data, branch number data, and branch maintenance status data in the original monitoring dataset of the string branches are encapsulated into data transmission messages according to a preset communication message format, and timestamp information and equipment source identification information are written in. The measurement and control acquisition device performs message parsing, source verification, and field splitting on the received data transmission messages. Then, based on the equipment source identification information, combiner box identification data, branch number data, and inverter correspondence, the parsed data are merged and reordered to generate background upload data records that correspond one-to-one with each photovoltaic string branch. Afterward, the power monitoring backend establishes a data index according to the combiner box identification, branch number, and inverter identification, and writes the uploaded data into the branch monitoring records in the string branch real-time monitoring database.

[0014] Based on this, the power monitoring backend uses the real-time monitoring database of the string branches to determine the operating conditions of the inverter output power data corresponding to each photovoltaic string branch, in order to distinguish whether the current branch is in an effective power generation state that can participate in the zero-current fault determination or an ineffective power generation state that does not meet the determination conditions.

[0015] Specifically, the real-time output power value, rated power value, and current sampling period identifier of the inverter corresponding to each photovoltaic string branch are read. The real-time output power value and rated power value are matched according to the inverter identifier to obtain the basic data for inverter power determination. Based on the basic data for inverter power determination, the output power ratio of each inverter in the current sampling period is calculated. The output power ratio is then compared with a preset power generation threshold. When the output power ratio is greater than or equal to the preset power generation threshold, the corresponding inverter is determined to be in an effective power generation state, and each photovoltaic string branch associated with the inverter is marked as a branch to be screened under the effective power generation determination condition.

[0016] Furthermore, the preset power generation threshold can be determined using a fixed threshold method, a time-segmented threshold method, or a weather-adaptive threshold method. The weather-adaptive threshold method determines the power proportion threshold for the current sampling period by reading light intensity information, weather status information, or environmental irradiance information, and combining it with the preset threshold mapping relationship, so as to improve the accuracy of power generation status determination under different operating conditions.

[0017] Next, zero-current screening is performed on each photovoltaic string branch under the effective power generation determination condition. Specifically, the effective monitoring data of the string branch corresponding to the branch to be screened under the effective power generation determination condition is read, and the combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value and branch maintenance status corresponding to each branch to be screened are extracted. The zero-current threshold parameter preset in the power monitoring background is called, and the branch DC current value corresponding to each branch to be screened is compared with the zero-current threshold parameter. When the branch DC current value is less than the zero-current threshold parameter, it is determined that the corresponding branch to be screened meets the zero-current initial screening condition. Furthermore, the zero current threshold parameter can be determined according to a fixed zero current threshold method, a branch type zero current threshold method, or a zero current threshold method based on different operating conditions, in order to enhance the adaptability to different branch characteristics and different operating conditions.

[0018] After meeting the initial screening conditions for zero current, the branch maintenance status is further combined with the tagging and shielding judgment to eliminate false alarms caused by non-fault shutdowns such as manual maintenance and tagging isolation. Specifically, the maintenance status, maintenance tagging start time, maintenance tagging end time and tagging operation identification information of the branch to be screened that meets the initial screening conditions for zero current are read, and it is determined whether the current sampling time falls within the valid tagging time interval. If the branch under maintenance status indicates that the branch to be screened is under maintenance and the current sampling time falls within the valid time interval of the maintenance status, the branch to be screened will be identified as a maintenance shielded branch and prohibited from entering the fault alarm process. If the branch is not under maintenance or the current sampling time does not fall within the valid time interval of the maintenance status, the branch to be screened will be allowed to continue to enter the zero-current fault candidate branch determination process. The zero-current abnormal branch mark will be generated by combining its corresponding combiner box identifier, branch number and inverter identifier, and the branch to be screened will be identified as a zero-current fault candidate branch.

[0019] After identifying candidate branches for zero-current faults, fault location processing is performed on these branches. Specifically, valid monitoring data and status records of the string branches corresponding to the candidate branches for zero-current faults are read, and the inverter identifier, combiner box identifier, branch number, branch DC current value, inverter output power value, branch maintenance status, and corresponding timestamp information corresponding to each candidate branch for zero-current faults are extracted. Using the inverter identifier, combiner box identifier, and branch number as the main index for fault location, a unique location identifier is encoded for each candidate branch for zero-current faults. Next, the unique location identifier code is matched with the device topology relationship data pre-stored in the power monitoring backend to determine the inverter location, combiner box location, and specific branch location of each zero-current fault candidate branch in the photovoltaic power station. Finally, the matched inverter location, combiner box location, and specific branch location are associated and combined with the corresponding branch DC current value, inverter output power value, and corresponding timestamp information to generate the zero-current fault location result of the corresponding zero-current fault candidate branch.

[0020] Finally, based on the zero-current fault location results, the power monitoring backend is triggered to output a fault alarm, generating and displaying zero-current alarm information corresponding to the target photovoltaic string branch. Specifically, the zero-current fault location results are read, and the unique location identifier code, inverter identifier, combiner box identifier, branch number, branch DC current value, inverter output power value, branch maintenance status, corresponding timestamp information, and topology location information of the corresponding zero-current fault candidate branch are extracted. Based on the zero-current fault location results, the corresponding zero-current alarm information is generated. Then, the zero-current alarm information is compared with the alarm level determination rules pre-configured in the power monitoring backend to determine the alarm level of the corresponding zero-current alarm information, and an alarm output control command corresponding to the alarm level is generated. The power monitoring backend triggers the corresponding alarm output mode according to the alarm output control command. The alarm output mode includes at least the monitoring interface pop-up prompt, alarm list display, audible and visual alarm prompt, and fault record writing. The zero current alarm information that has triggered the alarm output is written into the alarm event log database for operation personnel to perform fault viewing, fault confirmation and subsequent maintenance.

[0021] The present invention also provides an automatic alarm system for zero-current photovoltaic string branches, including a data acquisition unit, a data transmission and database construction unit, an operating condition determination unit, a zero-current screening unit, a fault location unit, and an alarm output unit.

[0022] The data acquisition unit collects DC current data of each photovoltaic string branch under each combiner box in the photovoltaic power station, and simultaneously collects inverter output power data, combiner box identification data, branch number data, and branch maintenance status data corresponding to the photovoltaic string branch, generating the original monitoring dataset of the string branch; the data transmission and database construction unit transmits the original monitoring dataset of the string branch to the measurement and control acquisition device via the communication link, and the measurement and control acquisition device uploads it to the power monitoring background, and performs association and storage of DC current data, inverter output power data, combiner box identification data, branch number data, and branch maintenance status data of each photovoltaic string branch, forming a real-time monitoring database of the string branch; the operating condition determination unit is used to determine the operating condition of the inverter output power data corresponding to each photovoltaic string branch based on the real-time monitoring database of the string branch, and determines that the corresponding photovoltaic string branch is in an effective power generation determination condition when the inverter output power meets the preset power generation threshold. The zero-current screening unit is used to determine the zero-current status of the DC current data of each photovoltaic string branch under valid power generation conditions. When the DC current of the corresponding photovoltaic string branch is less than the preset zero-current threshold, it combines the branch maintenance status data of the photovoltaic string branch to determine the alarm shielding. If the branch maintenance status data indicates that the photovoltaic string branch is not under maintenance, the photovoltaic string branch is identified as a candidate branch for zero-current fault. The fault location unit is used to perform fault location processing on the candidate branch for zero-current fault, extract the inverter identifier, combiner box identifier and branch number corresponding to the candidate branch for zero-current fault, and generate a zero-current fault location result. The alarm output unit is used to trigger the fault alarm output of the power monitoring background according to the zero-current fault location result, and generate and display the zero-current alarm information corresponding to the target photovoltaic string branch.

[0023] Furthermore, the data acquisition unit may include a current acquisition subunit, an identifier reading subunit, a maintenance status matching subunit, and a dataset generation subunit; the data transmission and database construction unit may include a message encapsulation subunit, a message parsing subunit, a branch object merging subunit, and a database writing subunit; the operating condition determination unit may include a power ratio calculation subunit, a threshold setting subunit, and an operating condition output subunit; the zero current screening unit may include a zero current threshold calling subunit, a zero current initial screening subunit, a tag shielding determination subunit, and a candidate branch output subunit; the fault location unit may include a location code generation subunit, a topology matching subunit, and a location result generation subunit; the alarm output unit may include an alarm information generation subunit, an alarm level determination subunit, an alarm triggering subunit, and an event retention subunit. Each subunit works collaboratively with the data flow and processing logic corresponding to the above method steps to complete the automatic alarm process for zero current faults.

[0024] Compared with the prior art, the present invention has at least the following beneficial effects: 1. It achieves active identification and automatic alarm for zero-current fault branches, solving the problems of existing technologies that mainly rely on manual viewing of monitoring data and have delayed fault detection, as detailed below: Existing technologies typically only collect and display the DC current of each photovoltaic string branch, remaining largely at the data display level. Operators need to manually search for abnormal branches from a large amount of monitoring data, which is prone to omissions and makes it difficult to promptly detect zero-current fault branches that have been in a silent state for a long time. This invention unifies the collection, association, storage, and rule-based determination of DC current data, inverter output power data, combiner box identification data, branch number data, and branch maintenance status data of each photovoltaic string branch. This enables the system to automatically complete zero-current fault screening in the background and directly trigger zero-current alarm information output when conditions are met. This upgrades the traditional passive monitoring method to an active identification and active alarm method, significantly improving the timeliness of fault detection.

[0025] 2. A composite judgment mechanism combining power generation conditions and maintenance status has been established, which can accurately distinguish between normal low current conditions and abnormal zero current conditions, reducing false alarms and missed alarms, as detailed below: In existing technologies, judging faults based solely on a single current value can easily lead to misjudging normal operating conditions as faults, such as when the inverter is operating under low load, in the early morning or evening, during weather fluctuations, or when a branch is under maintenance and out of service. This invention first reads the inverter's real-time output power and rated power, calculates the inverter's output power percentage within the current sampling period, and compares it with a preset power generation threshold to determine whether the corresponding photovoltaic string branch is in a valid power generation condition. Then, it performs zero-current screening only on branches in a valid power generation condition and further combines this with the branch's maintenance status and the effective tagging time interval for tagging and shielding determination. By organically combining power generation condition determination, zero-current threshold determination, and tagging and shielding determination, this invention can more accurately identify truly faulty branches, reduce false alarms caused by changes in operating conditions or manual maintenance, and lower the risk of missed alarms due to overly simplistic judgment conditions.

[0026] 3. It enables precise location and tiered handling of zero-current fault branches, shortening the fault diagnosis path and improving operation and maintenance efficiency, as detailed below: Existing technologies, even when detecting anomalies, often only provide current value changes, lacking direct indication of the specific fault branch location. Operators still need to conduct further step-by-step investigations, resulting in low processing efficiency. This invention, after identifying candidate branches with zero-current faults, extracts the corresponding inverter identifier, combiner box identifier, and branch number to uniquely locate and encode the candidate branch. It then combines this with equipment topology data to determine the location of the associated inverter, combiner box, and specific branch, ultimately generating a zero-current fault location result. Simultaneously, the system can determine the alarm level based on the fault duration, the branch's current state, the equipment's hierarchical location, and relevant operating condition information, triggering monitoring interface pop-up prompts, alarm list displays, audible and visual alarm prompts, and alarm event recording. Therefore, this invention not only outputs the specific location of the faulty branch but also generates an alarm handling method matching the fault severity, thereby shortening the on-site investigation path and improving the efficiency of maintenance personnel in locating and handling faulty branches.

[0027] 4. The intelligent upgrade, achieved by fully utilizing existing data acquisition links and monitoring backends, demonstrates good engineering feasibility and application value, as detailed below: This invention does not require a large-scale addition of hardware equipment. Instead, it builds upon existing photovoltaic string branch current acquisition, measurement and control acquisition devices, communication links, and power monitoring backends. By optimizing the design of data acquisition, data verification, communication message organization, branch object merging, operating condition judgment, threshold setting, fault location, and alarm output logic, a complete automatic alarm method and system for zero-current photovoltaic string branches is constructed. This solution is applicable to operating scenarios with a large number of combiner boxes and strings in large-scale photovoltaic power plants, and it can also be well integrated with existing monitoring systems and operation and maintenance management processes. Therefore, it has the advantages of clear implementation path, low deployment cost, high reliability, and convenient promotion and application. Attached Figure Description

[0028] Figure 1 This is a flowchart of an automatic alarm method for zero-current photovoltaic string branches proposed in this invention; Figure 2 This is a block diagram of an automatic alarm system for zero-current photovoltaic string branches proposed in this invention. Detailed Implementation

[0029] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0030] The technical solution of the present invention will be further described in detail below with reference to the embodiments of the present invention. It should be noted that the following embodiments are only used to illustrate the present invention and are not intended to limit the scope of protection of the present invention. Equivalent substitutions, conventional improvements or modifications made by those skilled in the art based on the disclosure of the present invention without departing from the concept of the present invention should all fall within the scope of protection of the present invention.

[0031] This invention provides an automatic alarm method and system for zero-current photovoltaic (PV) string branches, applicable to scenarios in large-scale PV power plants where operational monitoring, zero-current fault identification, fault location, and background alarm output are performed on each PV string branch under each combiner box. This method can be implemented using existing PV power plant string current acquisition links, measurement and control acquisition devices, communication links, and power monitoring backends. It does not require large-scale hardware modifications to the existing monitoring system. Instead, it introduces string-level data association, operating condition judgment, maintenance shielding, fault location, and alarm output logic based on existing acquisition data, thereby achieving automatic identification and accurate alarming of zero-current fault branches. Implementation

[0032] Reference Figure 1 In this embodiment, an automatic alarm method for zero-current photovoltaic string branches includes six stages: data acquisition and preprocessing, background transmission and database construction, operating condition determination, zero-current screening and tagging shielding, fault location, and alarm output.

[0033] I. Data Acquisition and Preprocessing Stage First, data is collected from each photovoltaic string branch under each combiner box within the photovoltaic power station. Specifically, current acquisition elements, preferably Hall effect sensors, are installed on each photovoltaic string branch within each combiner box to detect the current of each branch's output conductor, obtaining the DC current signal of each branch at the current sampling time. Simultaneously, the output power signal of the corresponding inverter within the same sampling period is read through the inverter operation monitoring interface corresponding to the electrical circuit of each photovoltaic string branch. This generates string branch electrical operation data that characterizes the current electrical operating status of the target photovoltaic string branch.

[0034] After acquiring current and power signals, the system further reads the basic identification information corresponding to each photovoltaic string branch. The basic identification information preferably includes combiner box address information, branch loop number information, and inverter association information. Specifically, the combiner box address information identifies the location of the combiner box to which the target photovoltaic string branch belongs in the monitoring network; the branch loop number information distinguishes different photovoltaic string branches under the same combiner box; and the inverter association information indicates which inverter the photovoltaic string branch is connected to via the current combiner box. The system matches and binds the aforementioned basic identification information with the acquired string branch electrical operation data one by one to obtain string branch identification association data, ensuring that each data record accurately corresponds to a unique combiner box, a unique branch loop, and a unique inverter association object.

[0035] Subsequently, the system reads the pre-set branch maintenance status tags from the power monitoring backend. Ideally, these branch maintenance status tags are pre-written by operators in the monitoring backend before planned maintenance, manual isolation, outage troubleshooting, or fault handling. The system matches the read branch maintenance status tags to the corresponding photovoltaic string branch data records according to the combiner box address, branch circuit number, inverter association, or unique branch identifier. This ensures that each string branch identifier association data has a corresponding maintenance status attribute, forming string branch status data. This method allows for the differentiation between actually faulty branches and manually repaired branches during the subsequent fault diagnosis stage.

[0036] After completing the above data acquisition and status matching, the system further performs preprocessing on the data. Preprocessing preferably includes time synchronization, data integrity verification, and filtering of abnormal null values. Specifically, the system reads the current sampling timestamp and power sampling timestamp from the string branch identifier associated data, and the status update timestamp from the string branch status data, and establishes a time window with a preset sampling period. The DC current value, inverter output power value, and branch maintenance status corresponding to the same combiner box identifier, the same branch number, and the same inverter identifier are mapped to the same time window. When the current sampling timestamp, power sampling timestamp, and status update timestamp all fall within the same time window, the system determines that the data set has completed time synchronization.

[0037] After time synchronization is complete, the system continues to perform data integrity verification. This verification checks whether each data record simultaneously contains the combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value, branch maintenance status, and corresponding timestamp information. If any necessary field is missing, the field format is abnormal, or the identifier correspondence is inconsistent, the data record is determined to be incomplete. For data not determined to be incomplete, the system further performs abnormal null value filtering, that is, it judges whether the branch DC current value, inverter output power value, and branch maintenance status are null values, invalid default values, non-numeric values, abnormal values ​​exceeding the preset collection range, or illegal values ​​that cannot participate in subsequent alarm judgment; when any of the above conditions are met, the data record is determined to be invalid. Data records that simultaneously meet the requirements of time synchronization, integrity verification, and abnormal null value filtering are retained to form valid monitoring data for string branches.

[0038] In a preferred embodiment, for data records determined to be invalid and with anomalies such as null values ​​for branch DC current, inverter output power, or branch maintenance status, the system can also initiate a supplementary sampling request to the corresponding data source according to the corresponding combiner box identifier, branch number, and inverter identifier to reacquire the missing data items in the current sampling period. After the supplemented data is written back to the corresponding invalid data record, time synchronization processing, data integrity verification, and anomaly null value filtering are performed again. If the supplemented data record passes the aforementioned verification and filtering, it is restored as a valid data record. If the supplemented data record still contains null values, abnormal values, illegal values, or exceeds the preset supplementary sampling limit, it is determined to be a discarded data record and is prohibited from entering the subsequent zero-current alarm determination process.

[0039] After preprocessing, the system structures and encapsulates the valid monitoring data of the string branches according to unified data organization rules. Preferably, data items belonging to the same photovoltaic string branch and that have passed the validity processing are combined into a complete data record according to fixed fields such as combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value, and branch maintenance status, generating the original monitoring dataset of the string branch for subsequent backend transmission and logical judgment.

[0040] II. Backend Transmission and Database Building Stage After generating the original monitoring dataset for the string branch, the system sends the dataset to the measurement and control acquisition device via the communication link and uploads it to the power monitoring backend to form a callable database.

[0041] Specifically, the system first encapsulates the branch DC current data, inverter output power data, combiner box identification data, branch number data, and branch maintenance status data from the original monitoring dataset of the string branches into a data transmission message according to a preset communication message format, and simultaneously writes corresponding timestamp information and equipment source identification information. Preferably, the data transmission message may include a message header field, an equipment source identification field, a combiner box identification field, a branch number field, an inverter identification field, a branch DC current value field, an inverter output power value field, a branch maintenance status field, a timestamp field, a checksum field, and a message tail field. After encapsulation, the data transmission message is sent to the measurement and control acquisition device via the combiner box side communication link.

[0042] After receiving the data transmission message, the monitoring and control acquisition device performs message parsing, source verification, and field splitting to obtain branch DC current data, inverter output power data, combiner box identification data, branch number data, branch maintenance status data, timestamp information, and equipment source identification information corresponding to each photovoltaic string branch. Then, based on the equipment source identification information, combiner box identification data, branch number data, and inverter correspondence, the parsed data is merged and reordered to generate a background upload data record corresponding one-to-one with each photovoltaic string branch.

[0043] In a preferred embodiment, the branch object merging process may include the following operations: using the combiner box identifier, branch number, and inverter identifier as branch object identification keys, grouping and classifying the parsed data items; merging data items with the same combiner box identifier, branch number, and inverter identifier into a candidate data set corresponding to the same photovoltaic string branch; sorting each data item in the same candidate data set according to the order of timestamps, and then reorganizing the positions according to the preset field arrangement order to form a sequential data record that uniquely corresponds to the target photovoltaic string branch.

[0044] Furthermore, to improve the availability of uploaded data, the system can also perform data anomaly correction processing on sequential data records. Specifically, this includes: performing consistency comparison on combiner box identifiers, branch numbers, and inverter identifiers in sequential data records; if two or more inconsistent identifiers exist, they are determined to be identifier conflict records; performing field counting verification on branch DC current data, inverter output power data, branch maintenance status data, and timestamp information; if two or more field values ​​of the same type exist at the same field position, the field value with the timestamp closest to the current sampling period and whose field verification is correct is retained as the target field value, and the remaining field values ​​of the same type are discarded; if necessary fields are missing, a missing field replenishment retrieval is initiated from the original data transmission message or the cached record of the measurement and control acquisition device according to the device source identifier; when timestamp information falls within the same period time window that does not correspond to the current sampling period, time relocation processing is performed on the same branch data record within the adjacent sampling period. After completing the aforementioned corrections, consistency verification, integrity verification, and time window consistency verification are re-executed. If the verification passes, the corrected background uploaded data record is output; if the verification still fails, an anomaly isolation record is output and prohibited from entering the background upload process.

[0045] Finally, the system uploads the data records from the backend to the power monitoring backend via the backend communication link. The power monitoring backend then creates a data index based on the combiner box identifier, branch number, and inverter identifier, and associates and writes the branch DC current data, inverter output power data, branch maintenance status data, and timestamp information from the uploaded data to form the branch monitoring records in the string branch real-time monitoring database. This database serves as the basic data source for subsequent operating condition determination, zero-current screening, fault location, and alarm output.

[0046] III. Operating Condition Determination Phase In this stage, the system performs an operating condition determination on the inverter output power data corresponding to each photovoltaic string branch to identify whether the current branch is in an effective power generation state that can participate in zero-current fault determination.

[0047] Specifically, the system reads the real-time output power value, rated power value, and current sampling period identifier of the inverter corresponding to each photovoltaic string branch. It then matches the real-time output power value with the rated power value according to the inverter identifier to obtain basic data for inverter power determination. Based on this basic data, the system calculates the output power percentage of each inverter within the current sampling period, where the output power percentage is the ratio of the real-time output power value to the rated power value of the inverter. Next, the output power percentage is compared with a preset power generation threshold. When the output power percentage is greater than or equal to the preset power generation threshold, the corresponding inverter is determined to be in an effective power generation state within the current sampling period, and each photovoltaic string branch associated with that inverter is marked as a branch to be screened under the effective power generation determination condition. When the output power percentage is less than the preset power generation threshold, the corresponding inverter is determined to not meet the zero-current alarm determination condition within the current sampling period, and each photovoltaic string branch associated with that inverter is prohibited from entering the zero-current fault screening process of the current sampling period.

[0048] In a preferred embodiment, the preset power generation threshold can be flexibly set according to different scenarios. Preferably, the system reads the threshold configuration parameters pre-stored in the power monitoring backend. The threshold configuration parameters include at least fixed threshold parameters, time-segmented threshold parameters, and weather-condition adaptive threshold parameters. The method for calling the preset power generation threshold is determined based on the time information and environmental condition information corresponding to the current sampling period. If the fixed threshold method is used, the pre-set fixed power percentage value is directly called as the preset power generation threshold. If the time-segmented threshold method is used, the power percentage value corresponding to the time segment of the current sampling period is read from the corresponding time-segmented threshold parameters. If the weather-condition adaptive threshold method is used, at least one of the environmental condition information (light intensity information, weather status information, or environmental irradiance information) corresponding to the current sampling period is read, and the power percentage value corresponding to the current environmental condition is determined as the preset power generation threshold for the current sampling period based on the preset threshold mapping relationship.

[0049] IV. Zero-current screening and tagging / shielding stage In this stage, the system only performs zero-current screening on the branches to be screened under the condition that has been determined to be a valid power generation condition, thereby avoiding unnecessary misjudgments under low power generation conditions.

[0050] Specifically, the system reads the valid monitoring data of the string branches corresponding to the branches to be screened under the valid power generation judgment conditions marked in step three, and extracts the combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value, and branch maintenance status for each branch to be screened. Then, it calls the pre-set zero-current threshold parameter in the power monitoring backend and compares the branch DC current value corresponding to each branch to be screened with the zero-current threshold parameter; when the branch DC current value is less than the zero-current threshold parameter, it determines that the corresponding branch to be screened meets the initial zero-current screening condition.

[0051] In a preferred embodiment, the zero-current threshold parameter can be determined using a fixed zero-current threshold method, a branch-type zero-current threshold method, or a zero-current threshold method based on different operating conditions. Specifically, the system reads the zero-current threshold configuration parameters pre-stored in the power monitoring backend. These configuration parameters include at least fixed zero-current threshold parameters, branch-type zero-current threshold parameters, and zero-current threshold parameters based on different operating conditions. Based on the branch type information corresponding to the branch to be screened and the operating condition information corresponding to the current sampling period, the system determines the method for calling the zero-current threshold parameters. When using the fixed zero-current threshold method, a pre-set fixed current value is directly called. When using the branch-type zero-current threshold method, the corresponding current threshold is read according to the branch type to which the branch to be screened belongs. When using the zero-current threshold method based on different operating conditions, the current threshold corresponding to the current operating condition is read according to at least one of the following operating condition information: the inverter output power ratio corresponding to the current sampling period, environmental operating condition information, or historical branch operating characteristics.

[0052] After a branch meets the initial zero-current screening condition, the system further performs a tagging and shielding determination. Specifically, it reads the branch maintenance status, tagging start time, tagging end time, and tagging operation identifier information corresponding to the branch to be screened that meets the initial zero-current screening condition, and uses the combiner box identifier, branch number, and inverter identifier as the basis for matching the maintenance status. It determines whether the branch maintenance status is a tagged maintenance status, and further determines whether the current sampling time falls within the effective tagging time interval defined by the tagged start time and end time. If the branch maintenance status is a tagged maintenance status and the current sampling time falls within the effective tagging time interval, the corresponding branch to be screened is determined to be a tagged and shielded branch, and a corresponding tagging and shielding mark is generated, prohibiting it from entering the subsequent fault alarm process. If the branch maintenance status is not a tagged maintenance status, or the current sampling time does not fall within the effective tagging time interval, the corresponding branch to be screened does not meet the tagging and shielding condition, and the branch to be screened is allowed to continue entering the zero-current fault candidate branch determination process.

[0053] Finally, for branches that were not identified as shielded branches requiring maintenance but meet the initial zero-current screening criteria, a zero-current abnormal branch marker is generated by combining the corresponding combiner box identifier, branch number, and inverter identifier. The corresponding branch to be screened is then identified as a candidate branch for zero-current fault. This completes the zero-current screening phase.

[0054] V. Fault Location Phase In this stage, the system performs fault location processing on zero-current fault candidate branches to locate the specific faulty branch from the string-level fault screening results.

[0055] Specifically, the system reads the valid monitoring data and status records of the string branches corresponding to the zero-current fault candidate branches, and extracts the inverter identifier, combiner box identifier, branch number, branch DC current value, inverter output power value, branch maintenance status, and corresponding timestamp information for each zero-current fault candidate branch. Then, using the inverter identifier, combiner box identifier, and branch number as the primary index for fault location, a unique location identifier is encoded for each zero-current fault candidate branch, generating a branch location index record that corresponds one-to-one with each zero-current fault candidate branch.

[0056] In a preferred embodiment, the process of generating a unique location identifier code includes: reading the inverter identifier, combiner box identifier, and branch number corresponding to each zero-current fault candidate branch, and establishing multi-level location identifier fields according to the inverter level, combiner box level, and branch level; performing field normalization processing on the inverter identifier, combiner box identifier, and branch number respectively, so that each field meets the preset character length, field encoding format, and field value range requirements; sequentially concatenating the normalized inverter identifier, combiner box identifier, and branch number according to the preset encoding order, and writing a level separator between adjacent fields to generate a basic location code; then reading the current sampling period identifier or corresponding timestamp information and appending it to the basic location code to generate a time-series location code corresponding to the current alarm event; subsequently, performing a uniqueness check on the time-series location code, and determining it as the unique location identifier code for the corresponding zero-current fault candidate branch when there is no duplicate code, and continuing to add preset supplementary distinguishing fields until the uniqueness requirement is met when there is a duplicate code.

[0057] After generating the unique location identifier code, the system matches the location index records of each branch with the pre-stored equipment topology relationship data in the power monitoring backend to determine the inverter location, combiner box location, and specific branch location of each zero-current fault candidate branch in the photovoltaic power station. Preferably, the equipment topology relationship data includes at least an inverter equipment table, a combiner box equipment table, a string branch connection relationship table, and an equipment hierarchy mapping table. The system uses the inverter identifier field, combiner box identifier field, and branch number field in the unique location identifier code as topology matching keys to retrieve the equipment record corresponding to the target zero-current fault candidate branch in the aforementioned equipment table and relationship table. The system performs a hierarchy consistency check on the retrieved equipment record to determine whether the target inverter equipment record, the target combiner box equipment record, and the target branch connection relationship record satisfy a preset connection relationship. The preset connection relationship includes at least the branch belonging to the corresponding combiner box and the combiner box belonging to the corresponding inverter.

[0058] When the hierarchical consistency check passes, the inverter installation location, combiner box installation location, and branch access location corresponding to the target zero-current fault candidate branch are extracted and determined as the topology location information of the zero-current fault candidate branch. When the hierarchical consistency check fails, the corresponding device record is judged as a topology inconsistency record, and the historical connection relationship record or the most recent valid topology mapping record in the power monitoring background is called for secondary matching and correction. Finally, the inverter location, combiner box location, and specific branch location obtained through the initial matching or secondary matching and correction are associated and combined with the corresponding branch DC current value, inverter output power value, and corresponding timestamp information to generate the zero-current fault location result of the corresponding zero-current fault candidate branch.

[0059] VI. Alarm Output Stage In this stage, the system triggers a fault alarm output from the power monitoring backend based on the zero-current fault location result, generating zero-current alarm information that can be directly viewed and handled by operators.

[0060] Specifically, the system reads the zero-current fault location result output in step five and extracts the unique location identifier code, inverter identifier, combiner box identifier, branch number, branch DC current value, inverter output power value, branch maintenance status, corresponding timestamp information, and topology location information for the corresponding zero-current fault candidate branch. Based on the aforementioned information, the system generates corresponding zero-current alarm information, which includes at least the fault branch identifier information, fault occurrence time information, fault electrical status information, and fault location indication information.

[0061] In a preferred embodiment, the system compares the zero-current alarm information with the alarm level determination rules pre-configured in the power monitoring backend to determine the alarm level of the corresponding zero-current alarm information. Specifically, the system reads the branch DC current value, inverter output power value, fault occurrence time information, unique location identifier code, topology location information, and branch maintenance status from the corresponding zero-current alarm information, and calls the alarm level configuration parameters pre-stored in the power monitoring backend. Based on the deviation of the branch DC current value from the zero-current threshold parameter, the output power ratio corresponding to the inverter output power value, the fault duration, the equipment level location of the faulty branch, and whether it is under maintenance, the system assesses the severity of the corresponding zero-current alarm information to obtain the corresponding alarm judgment result. When the alarm judgment result indicates that the branch DC current value is continuously lower than the zero-current threshold parameter, the corresponding inverter is in an effective power generation state, the fault duration reaches the preset duration threshold, and the branch is not under maintenance, the system determines that the zero-current alarm information meets the valid alarm conditions. Then, according to the alarm level configuration parameters, the zero-current alarm information that meets the valid alarm conditions is divided into at least two alarm levels, and alarm output control commands corresponding to the alarm levels are output.

[0062] Subsequently, the power monitoring backend triggers the corresponding alarm output mode according to the alarm output control command. Preferably, the alarm output mode includes at least a monitoring interface pop-up prompt, alarm list display, audible and visual alarm prompt, and fault record writing. Specifically, the system reads the alarm output control command and the alarm level, unique location identifier code, inverter identifier, combiner box identifier, branch number, fault occurrence time information, and topology location information from the corresponding zero current alarm information; determines the corresponding alarm output mode combination according to the alarm level, wherein a lower alarm level at least triggers a monitoring interface pop-up prompt and alarm list display, and a higher alarm level further triggers an audible and visual alarm prompt in addition to triggering a monitoring interface pop-up prompt and alarm list display; generates corresponding alarm display content in the power monitoring backend according to the alarm output mode combination, and sends the alarm display content to the monitoring display terminal; when the audible and visual alarm prompt is triggered, the power monitoring backend sends an audible and visual trigger command to the corresponding alarm device, and keeps the audible and visual alarm prompt in an active state before the operator performs the alarm confirmation operation, or continuously outputs according to the preset alarm duration.

[0063] Finally, after completing at least one alarm output, the system writes the corresponding alarm output result into the alarm event log database, and associates the zero current alarm information with the corresponding unique location identifier code, the current alarm status, and the processing status for operation personnel to perform fault viewing, fault confirmation, and subsequent maintenance. Implementation

[0064] Reference Figure 2In this embodiment, an automatic alarm system for zero-current photovoltaic string branches includes a data acquisition unit, a data transmission and database construction unit, an operating condition determination unit, a zero-current screening unit, a fault location unit, and an alarm output unit. Each unit can be deployed in the same power monitoring backend, or it can be deployed separately in the combiner box-side acquisition equipment, the measurement and control acquisition device, and the monitoring backend server according to the field architecture, and collaboratively complete the zero-current fault identification, location, and alarm output process through a communication link.

[0065] I. Data Acquisition Unit The data acquisition unit is used to acquire and preprocess photovoltaic string branch operation data, basic identification data, and maintenance status data to form a raw monitoring dataset of the string branches that can be directly used for background judgment. Preferably, the data acquisition unit includes a current acquisition subunit, an identification reading subunit, a maintenance status matching subunit, and a dataset generation subunit.

[0066] The current acquisition subunit samples the output circuit of each photovoltaic string branch within each combiner box branch by branch. Specifically, for each photovoltaic string branch, the current acquisition subunit reads the DC current signal of the corresponding branch from the current acquisition element installed on that branch, and reads the output power signal of the inverter belonging to that branch within the same sampling period through the operation monitoring interface of the corresponding inverter. The output of the current acquisition subunit is a set of branch DC current values ​​and inverter output power values ​​generated based on the sampling period, used to characterize the electrical operating status of the target photovoltaic string branch within the current sampling period.

[0067] The identification reading subunit is connected to the current acquisition subunit and is used to read the combiner box identifier, branch number, and inverter association relationship corresponding to each photovoltaic string branch. Specifically, the identification reading subunit can extract the corresponding fields from the device communication configuration table, branch wiring relationship table, or monitoring background device mapping table, and bind the combiner box identifier, branch number, and inverter association relationship to the branch DC current signal and inverter output power signal output by the current acquisition subunit, so that each sampled data record has a unique branch object identity. Through this binding process, subsequent processing units no longer need to re-identify the branch source, but can directly perform judgment and calculation based on the unified branch object identifier.

[0068] The maintenance status matching subunit is connected to the identifier reading subunit. It reads pre-set branch maintenance tag status markers from the power monitoring backend and matches these markers to the corresponding photovoltaic string branches according to the combiner box identifier and branch number. Preferably, the maintenance status matching subunit can not only read the current branch maintenance status but also simultaneously read the corresponding maintenance tag start time, maintenance tag end time, and tag operation identifier information. This information is then written into the corresponding branch data record for subsequent zero-current screening unit to perform tag shielding determination.

[0069] The dataset generation subunit is connected to the aforementioned three subunits. It performs time synchronization processing, data integrity verification, and abnormal null value filtering on the bound and matched data. The verified data is then structured and encapsulated according to combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value, and branch maintenance status to generate the string branch original monitoring dataset. Specifically, the dataset generation subunit may include a time synchronization processing module, an integrity verification module, and an abnormal null value filtering module. The time synchronization processing module reads the current sampling timestamp, power sampling timestamp, and status update timestamp, and establishes a time window with a preset sampling period, mapping data of the same branch object to the same time window. The integrity verification module checks whether the necessary fields in each record are complete and whether the identification relationships are consistent. The abnormal null value filtering module determines whether each field is null, a default invalid value, a non-numeric value, or an out-of-bounds abnormal value. Only data records that have passed the above processing are retained and written into the dataset recording unit according to a unified field structure, outputting the string branch original monitoring dataset.

[0070] In a preferred embodiment, the dataset generation subunit can also be connected to the supplementary acquisition and discard processing module. When the abnormal null value filtering module determines that the anomaly type of a record is a null value anomaly, the supplementary acquisition and discard processing module can initiate a supplementary acquisition request to the corresponding acquisition source according to the corresponding combiner box identifier, branch number, and inverter identifier; if the supplementary acquisition is successful, the supplemented data is re-sent to the time synchronization processing module, integrity verification module, and abnormal null value filtering module; if the supplementary acquisition fails or exceeds the preset number of supplementary acquisitions, the record is marked as a discard record, and a data anomaly tag associated with the corresponding branch object is generated.

[0071] II. Data Transmission and Database Construction Unit The data transmission and database construction unit is used to convert the raw monitoring dataset of string branches output by the data acquisition unit into structured data suitable for communication transmission and backend writing, and to form a real-time monitoring database of string branches in the power monitoring backend that can be queried, compared, and invoked according to rules. Preferably, the data transmission and database construction unit includes a message encapsulation subunit, a message parsing subunit, a branch object merging subunit, and a database writing subunit.

[0072] The message encapsulation subunit is connected to the data acquisition unit and is used to encapsulate the branch DC current data, inverter output power data, combiner box identification data, branch number data, and branch maintenance status data in the original monitoring dataset of the string branches into a data transmission message according to a preset communication message format. Preferably, the message encapsulation subunit constructs a data transmission message structure including a message header field, a device source identifier field, a combiner box identifier field, a branch number field, an inverter identifier field, a branch DC current value field, an inverter output power value field, a branch maintenance status field, a timestamp field, a check field, and a message tail field. The corresponding fields are formatted according to preset byte length, field encoding rules, and field separation rules. Then, a check code is calculated based on the formatted data payload content and written into the check field to generate a standard message.

[0073] The message parsing subunit is located on the measurement and control acquisition device side. Upon receiving a data transmission message, it performs message parsing, source verification, and field splitting. Specifically, the message parsing subunit first determines the message boundaries using the message header and message tail fields. Then, it verifies whether any content anomalies occurred during transmission based on the verification fields. Subsequently, it splits each field to obtain the branch DC current data, inverter output power data, combiner box identification data, branch number data, branch maintenance status data, timestamp information, and equipment source identification information corresponding to each photovoltaic string branch.

[0074] The branch object merging subunit is connected to the message parsing subunit. It performs branch object merging and sequential reorganization on the parsed data based on equipment source identification information, combiner box identification data, branch number data, and inverter correspondence. Specifically, the branch object merging subunit uses the combiner box identifier, branch number, and inverter identifier as branch object identification keys. It groups and categorizes the parsed fields, merging data items with the same combiner box identifier, branch number, and inverter identifier into candidate data sets corresponding to the same photovoltaic string branch. Then, it sorts the data items in the same candidate data set according to their timestamps and reorganizes their positions according to a preset field arrangement order, forming a sequential data record uniquely corresponding to the target photovoltaic string branch.

[0075] In a preferred embodiment, the branch object merging subunit may further include an anomaly correction module. This anomaly correction module performs consistency comparison on the combiner box identifier, branch number, and inverter identifier in the sequential data records. When two or more inconsistent identifiers are found, the corresponding record is determined to be an identifier conflict record. It also performs field counting verification on branch DC current data, inverter output power data, branch maintenance status data, and timestamp information. When two or more field values ​​of the same type exist at the same field position, it retains the field value with the timestamp closest to the current sampling period and whose field verification is correct, and discards the remaining field values. Furthermore, when necessary fields are missing, it initiates a missing field replenishment retrieval from the original data transmission message or the cached records of the measurement and control acquisition device. Finally, when timestamp information falls within a time window corresponding to a period other than the current sampling period, it calls the data records of the same branch in adjacent sampling periods to perform time relocation processing. After correction, consistency verification, integrity verification, and time window consistency verification are performed again. If the verification passes, the corrected data record uploaded to the backend is output; if the verification fails, an anomaly isolation record is output and prohibited from entering the database writing subunit.

[0076] The database writing subunit is connected to the branch object merging subunit to upload data records from the backend to the power monitoring backend. A data index is established based on the combiner box identifier, branch number, and inverter identifier. The uploaded data, including branch DC current data, inverter output power data, branch maintenance status data, and timestamp information, is correlated and written to form branch monitoring records in the string branch real-time monitoring database. These branch monitoring records can serve as a unified data source for the operating condition determination unit, zero-current screening unit, and fault location unit.

[0077] III. Operating Condition Determination Unit The operating condition determination unit is used to determine whether the inverter to which each photovoltaic string branch belongs is in an effective power generation state before performing zero-current screening, thereby avoiding incorrect judgment of the branch current under low power generation or non-power generation conditions. Preferably, the operating condition determination unit includes a power ratio calculation subunit, a threshold setting subunit, and an operating condition output subunit.

[0078] The power percentage calculation subunit reads the real-time output power value and rated power value of the inverter corresponding to each photovoltaic string branch from the real-time monitoring database of the string branches, and calculates the output power percentage of each inverter in the current sampling period. The output power percentage is the ratio of the real-time output power value to the rated power value of the inverter. The power percentage calculation result serves as the basic calculation quantity for determining whether the inverter is currently in an effective power generation state.

[0079] The threshold setting subunit is connected to the power proportion calculation subunit and is used to determine the preset power generation threshold corresponding to the current sampling period based on the preset power generation threshold configuration parameters. Preferably, the threshold setting subunit may include a fixed threshold calling module, a time-segmented threshold calling module, and a weather-condition adaptive threshold calling module. The fixed threshold calling module directly calls the preset fixed power proportion value; the time-segmented threshold calling module reads the corresponding power proportion value from the time-segmented threshold parameters according to the time interval to which the current sampling period belongs; the weather-condition adaptive threshold calling module reads at least one of the environmental condition information, such as light intensity information, weather state information, or environmental irradiance information corresponding to the current sampling period, and determines the power proportion value corresponding to the current environmental condition based on the preset threshold mapping relationship.

[0080] In a preferred embodiment, the weather-condition adaptive threshold calling module can also be connected to a threshold mapping relationship establishment module. This module reads the inverter output power value, inverter rated power value, light intensity information, weather status information, and environmental irradiance information from historical operating cycles, and aligns them according to the sampling period to obtain historical operating condition sample data. Based on the historical operating condition sample data, it calculates the historical output power ratio corresponding to each historical operating cycle. Then, according to preset environmental operating condition classification rules, it classifies the light intensity information, weather status information, and environmental irradiance information into multiple environmental operating condition intervals. It statistically analyzes the historical output power ratio data within each environmental operating condition interval and calculates the reference power ratio value for the corresponding environmental operating condition interval. Finally, it establishes a correspondence between each environmental operating condition interval and its corresponding reference power ratio value, generates a preset threshold mapping relationship, and stores it in the power monitoring backend. Preferably, the reference power ratio value can be determined through the sample mean, sample median, preset quantile value, or a weighted combination result.

[0081] The operating condition output subunit compares the output power ratio with the preset power generation threshold. When the output power ratio is greater than or equal to the preset power generation threshold, it determines that the corresponding inverter is in an effective power generation state and marks each photovoltaic string branch associated with the inverter as a branch to be screened under the effective power generation judgment operating condition. When the output power ratio is less than the preset power generation threshold, it determines that the corresponding inverter does not meet the operating condition conditions for zero current alarm judgment and prohibits each photovoltaic string branch associated with the inverter from entering the zero current fault screening process of the current sampling period.

[0082] IV. Zero Current Screening Unit The zero-current screening unit is used to perform initial zero-current screening on branches under valid power generation conditions, and to shield false alarms in conjunction with the maintenance status. Preferably, the zero-current screening unit includes a zero-current threshold call subunit, a zero-current initial screening subunit, a maintenance shielding determination subunit, and a candidate branch output subunit.

[0083] The zero-current threshold calling subunit reads the pre-stored zero-current threshold configuration parameters in the power monitoring backend, and determines the zero-current threshold parameters corresponding to the branch to be screened based on the branch type information and the operating condition information corresponding to the current sampling period. Preferably, this subunit can call the corresponding threshold in a fixed zero-current threshold mode, a branch type zero-current threshold mode, or an operating condition-based zero-current threshold mode.

[0084] The zero-current initial screening unit compares the DC current value of the branch to be screened with the zero-current threshold parameter. If the DC current value is less than the zero-current threshold parameter, the corresponding branch to be screened is determined to meet the zero-current initial screening condition. The comparison result can serve as a prerequisite for subsequent tagging and shielding determination and the generation of zero-current candidate branches.

[0085] The tagging and shielding determination subunit reads the branch maintenance status, maintenance tagging start time, maintenance tagging end time, and tagging operation identification information corresponding to the branch to be screened that meets the zero-current initial screening condition, and determines whether the current sampling time falls within the valid tagging time interval. If the branch maintenance status is under tagging and the current sampling time falls within the valid tagging time interval defined by the maintenance tagging start time and maintenance tagging end time, then the corresponding branch to be screened is determined to be a maintenance-shielded branch and a tagging and shielding mark is generated; if the branch maintenance status is not under tagging or the current sampling time does not fall within the valid tagging time interval, then the corresponding branch to be screened does not meet the tagging and shielding conditions.

[0086] When the branch to be screened is not under maintenance or the current sampling time does not fall within the valid time interval of the maintenance, the candidate branch output subunit generates a zero-current abnormal branch mark by combining the corresponding combiner box mark, branch number and inverter mark, and outputs the branch to be screened as a zero-current fault candidate branch for the fault location unit to call.

[0087] V. Fault Location Unit The fault location unit is used to further refine the zero-current fault candidate branch from the "judgment result" into a "specific locatable branch object". Preferably, the fault location unit includes a location code generation subunit, a topology matching subunit, and a location result generation subunit.

[0088] The location coding generation subunit reads the inverter identifier, combiner box identifier, and branch number corresponding to the zero-current fault candidate branch. It establishes multi-layered location identifier fields according to inverter, combiner box, and branch levels. After normalizing each field, it concatenates them in a preset coding order and adds the current sampling period identifier or timestamp information to generate a unique location identifier code for the corresponding zero-current fault candidate branch. In this way, the faulty branch no longer exists with only a single number, but forms a location code that simultaneously includes equipment hierarchy and time attributes.

[0089] The topology matching subunit reads the pre-stored device topology relationship data in the power monitoring backend and uses the inverter identifier, combiner box identifier, and branch number fields from the unique location identifier code as topology matching keys. It then searches the inverter device table, combiner box device table, string branch connection relationship table, and device hierarchy mapping table for the device record corresponding to the target zero-current fault candidate branch. If the hierarchy consistency check passes, the successfully matched device hierarchy information is output; if the hierarchy consistency check fails, the corresponding device record is determined to be a topology inconsistency record, and a second matching correction is performed by calling historical connection relationship records or the most recent valid topology mapping record; if the second matching correction still fails, a topology anomaly flag is output, and the corresponding branch is sent to the manual review queue.

[0090] When the hierarchical consistency check passes, the location result generation subunit extracts the corresponding inverter installation location, combiner box installation location, and branch access location. It then associates and combines these information with the corresponding timestamps to generate a zero-current fault location result. This result retains both the faulty branch's identification information and the electrical state and equipment location indication information at the time of the fault.

[0091] VI. Alarm Output Unit The alarm output unit is used to generate zero-current alarm information based on the fault location result, perform alarm level determination, and trigger the corresponding alarm mode. Preferably, the alarm output unit includes an alarm information generation subunit, an alarm level determination subunit, an alarm triggering subunit, and an event retention subunit.

[0092] The alarm information generation subunit generates corresponding zero-current alarm information based on the zero-current fault location result. The zero-current alarm information includes at least the fault branch identification information, fault occurrence time information, fault electrical status information, and fault location indication information.

[0093] The alarm level determination subunit assesses the severity of zero-current alarm information based on the deviation of the branch DC current value from the zero-current threshold parameter, the output power ratio corresponding to the inverter output power value, the duration of the fault, the equipment level location of the faulty branch, and whether it is under maintenance. It then outputs the corresponding alarm level and alarm output control command. Preferably, zero-current alarm information with a longer fault duration, a more critical affected location, or a higher corresponding inverter output power ratio can be classified as a higher alarm level.

[0094] The alarm triggering subunit triggers the corresponding alarm output mode according to the alarm output control command. Specifically, the alarm triggering subunit can determine the corresponding alarm output mode combination according to the alarm level. Among them, the lower alarm level will at least trigger the monitoring interface pop-up prompt and alarm list display, while the higher alarm level will trigger the audible and visual alarm prompt in addition to triggering the monitoring interface pop-up prompt and alarm list display. It can also generate alarm display content including fault branch identification information, fault occurrence time information, branch current status information and equipment location indication information, and send it to the monitoring display terminal. When the audible and visual alarm prompt is triggered, an audible and visual trigger command is sent to the corresponding alarm device, and the audible and visual alarm prompt is kept in an active state before the operator performs the alarm confirmation operation, or it can continuously output according to the preset alarm hold duration.

[0095] The event retention subunit is used to write the zero-current alarm information that has been triggered into the alarm event log library, and to associate and store the zero-current alarm information with the corresponding unique location identifier code, the current alarm status and the processing status, so that operators can perform fault viewing, fault confirmation and subsequent maintenance. It can also serve as a data source for subsequent alarm level optimization, threshold correction and historical false alarm rate analysis.

[0096] In summary, the zero-current photovoltaic string branch automatic alarm system in this embodiment, through the coordinated operation of the data acquisition unit, data transmission and database construction unit, operating condition judgment unit, zero-current screening unit, fault location unit, and alarm output unit, can automatically identify, accurately distinguish, precisely locate, and classify zero-current fault branches on the basis of the existing photovoltaic monitoring link, thereby improving the detection and handling efficiency of zero-current faults in photovoltaic power plants.

[0097] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An automatic alarm method for zero-current photovoltaic string branches, characterized in that, Includes the following steps: S1. Collect DC current data of each photovoltaic string branch under each combiner box in the photovoltaic power station, and simultaneously collect inverter output power data, combiner box identification data, branch number data and branch maintenance status data corresponding to each photovoltaic string branch, and generate the original monitoring dataset of string branches. S2. The original monitoring dataset of the string branch is transmitted to the measurement and control acquisition device via the communication link, and then uploaded to the power monitoring background by the measurement and control acquisition device. The DC current data, inverter output power data, combiner box identification data, branch number data and branch maintenance status data of each photovoltaic string branch are associated and stored to form a real-time monitoring database of the string branch. S3. The power monitoring backend reads the output power data of the inverter corresponding to each photovoltaic string branch based on the real-time monitoring database of the string branch, and compares the output power of the corresponding inverter with the preset power generation threshold. When the output power of the corresponding inverter meets the preset power generation threshold, it is determined that the photovoltaic string branch associated with the inverter is in the effective power generation judgment condition. S4. Perform zero-current screening on each photovoltaic string branch that is in the effective power generation judgment condition. Compare the DC current of the corresponding photovoltaic string branch with the preset zero-current threshold. When the DC current of the corresponding photovoltaic string branch is less than the preset zero-current threshold, determine the labeling and shielding judgment based on the branch maintenance status of the photovoltaic string branch. When the corresponding photovoltaic string branch is not in the labeling and maintenance status, determine the photovoltaic string branch as a candidate branch for zero-current fault. S5. Perform fault location processing on the candidate branch of zero current fault, extract the inverter identifier, combiner box identifier and branch number of the corresponding candidate branch of zero current fault, and generate the zero current fault location result. S6. Based on the zero-current fault location result, trigger the fault alarm output of the power monitoring backend, generate and display the zero-current alarm information corresponding to the target photovoltaic string branch.

2. The automatic alarm method for zero-current photovoltaic string branches according to claim 1, characterized in that, Step S1 includes: The DC current signal of each photovoltaic string branch is obtained by the current acquisition element set in each combiner box, and the output power signal of the inverter is obtained through the operation monitoring interface of the corresponding inverter to obtain the electrical operation data of the string branch. Read the combiner box address information, branch circuit number information and inverter association information corresponding to each photovoltaic string branch, and bind the combiner box address information, branch circuit number information and inverter association information to the string branch electrical operation data one by one to obtain string branch identification association data; Read the pre-set branch maintenance tag status flag in the power monitoring backend, and match the branch maintenance tag status flag to the corresponding photovoltaic string branch to obtain the string branch status data. Time synchronization processing, data integrity verification, and abnormal null value filtering are performed on the string branch identifier association data and string branch status data to obtain effective monitoring data for string branches. The effective monitoring data of the string branch is structured and encapsulated according to the combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value and branch maintenance status to generate the original monitoring dataset of the string branch.

3. The automatic alarm method for zero-current photovoltaic string branches according to claim 2, characterized in that, Time synchronization processing, data integrity verification, and abnormal null value filtering include: Read the current sampling timestamp, power sampling timestamp and status update timestamp in the string branch identifier association data, and establish a time window with the same period based on the preset sampling period. The DC current value, inverter output power value and branch maintenance status corresponding to the same combiner box identifier, the same branch number and the same inverter identifier are mapped to the same period time window. When the current sampling timestamp, power sampling timestamp and status update timestamp all fall into the same period time window, the corresponding data is determined to complete the time synchronization process. After the time synchronization process is completed, perform data integrity verification on the data to determine whether each data record contains the combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value, branch maintenance status and corresponding timestamp information. If any field is missing, the field format is abnormal or the identifier correspondence is inconsistent, the data record is determined to be an incomplete data record. For data records that are not determined to be incomplete, perform abnormal null value filtering, determine whether the branch DC current value, inverter output power value and branch maintenance status are null values, invalid default values, abnormal values ​​that exceed the preset collection range or illegal values ​​that cannot participate in subsequent alarm determination, and determine data records that meet any of the above conditions as invalid data records; Retain data records that simultaneously meet the requirements of time synchronization, data integrity verification, and abnormal null value filtering, and generate effective monitoring data for the string branch.

4. The automatic alarm method for zero-current photovoltaic string branches according to claim 3, characterized in that, Abnormal null value filtering also includes: For data records that are determined to be invalid and whose anomaly type is null DC current value, null output power value, or null maintenance status of branch, a supplementary sampling request is initiated to the corresponding data acquisition source according to the corresponding combiner box identifier, branch number, and inverter identifier, so as to reacquire the missing data items in the current sampling period. Write the DC current value of the branch, the output power value of the inverter, or the maintenance status of the branch obtained from the supplementary sampling into the corresponding invalid data record, and re-execute time synchronization processing, data integrity verification, and abnormal null value filtering on the supplementary data record; When the supplemented data record undergoes time synchronization processing, data integrity verification, and abnormal null value filtering, the supplemented data record is restored to a valid data record. If the data record after supplementary sampling still contains null values, abnormal values, illegal values, or exceeds the preset limit for the number of supplementary samplings, the supplementary data record will be judged as a discarded data record and will be prohibited from entering the subsequent zero-current alarm judgment process.

5. The automatic alarm method for zero-current photovoltaic string branches according to claim 1, characterized in that, Step S2 includes: The DC current data, inverter output power data, combiner box identification data, branch number data, and branch maintenance status data in the original monitoring dataset of the string branch are encapsulated into data transmission messages according to the preset communication message format, and the corresponding timestamp information and equipment source identification information are written in. The data transmission message is sent to the measurement and control acquisition device through the communication link on the combiner box side. The measurement and control acquisition device performs message parsing, source verification and field splitting on the received data transmission message. The measurement and control acquisition device merges and reorganizes the branch objects based on the equipment source identification information, combiner box identification data, branch number data and inverter correspondence, and generates background upload data records that correspond one-to-one with each photovoltaic string branch. The data uploaded from the backend is transmitted to the power monitoring backend via the backend communication link. The power monitoring backend then establishes a data index based on the combiner box identifier, branch number, and inverter identifier, and associates and writes the branch DC current data, inverter output power data, branch maintenance status data, and timestamp information in the uploaded data. The data records that have been associated and written are stored as branch monitoring records in the real-time monitoring database of the group branch.

6. The automatic alarm method for zero-current photovoltaic string branches according to claim 1, characterized in that, Step S3 includes: Read the real-time output power value, rated power value, and current sampling period identifier of the inverter corresponding to each photovoltaic string branch, and match the real-time output power value with the rated power value of the inverter according to the inverter identifier to obtain the basic data for inverter power determination; Based on the basic data for inverter power determination, the output power ratio of each inverter in the current sampling period is calculated. The output power ratio is the ratio of the real-time output power value to the rated power value of the inverter. The output power ratio is compared with the preset power generation threshold. When the output power ratio is greater than or equal to the preset power generation threshold, it is determined that the corresponding inverter is in an effective power generation state in the current sampling period, and each photovoltaic string branch associated with the inverter is marked as a branch to be screened under the effective power generation judgment condition. When the output power ratio is less than the preset power generation threshold, it is determined that the corresponding inverter does not meet the operating conditions for zero current alarm judgment in the current sampling period, and each photovoltaic string branch associated with the inverter is prohibited from entering the zero current fault screening process in the current sampling period.

7. The automatic alarm method for a zero-current photovoltaic string branch according to claim 6, characterized in that, The preset power generation threshold settings include: Read the threshold configuration parameters pre-stored in the power monitoring backend. The threshold configuration parameters include at least fixed threshold parameters, time-based threshold parameters, and weather-adaptive threshold parameters. Based on the time information and environmental conditions corresponding to the current sampling period, the method for calling the preset power generation threshold is determined. When using the fixed threshold method, the preset fixed power percentage value is directly called as the preset power generation threshold. When using the time-segmented threshold method, the power percentage value corresponding to the time segment to which the current sampling period belongs is read from the corresponding time-segmented threshold parameter as the preset power generation threshold. When using the weather-adaptive threshold method, at least one of the following environmental conditions information is read from the current sampling period: light intensity information, weather status information, or environmental irradiance information. Based on the environmental condition information, the power ratio value corresponding to the current environmental condition is determined in the preset threshold mapping relationship as the preset power generation threshold.

8. The automatic alarm method for zero-current photovoltaic string branches according to claim 1, characterized in that, Step S4 includes: Read the effective monitoring data of the string branch corresponding to the branch to be screened under the effective power generation judgment condition marked in step S3, and extract the combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value and branch maintenance status corresponding to each branch to be screened. The system calls the pre-set zero current threshold parameter in the power monitoring backend and compares the DC current value of each branch to be screened with the zero current threshold parameter. When the DC current value of the branch is less than the zero current threshold parameter, it is determined that the corresponding branch to be screened meets the zero current initial screening condition. For branches that meet the zero-current initial screening conditions, read their corresponding branch maintenance status and determine whether the branch maintenance status indicates that the branch to be screened is under maintenance. If the branch maintenance status indicates that the branch to be screened is under maintenance, determine the branch to be screened as a maintenance shielded branch and prohibit it from entering the fault alarm process. For branches that are not identified as shielded branches under maintenance and meet the initial screening conditions for zero current, a zero current abnormal branch mark is generated by combining the corresponding combiner box mark, branch number and inverter mark, and the corresponding branch to be screened is identified as a candidate branch for zero current fault.

9. An automatic alarm system for a zero-current photovoltaic string branch, characterized in that, include: The data acquisition unit is configured to collect DC current data of each photovoltaic string branch under each combiner box in the photovoltaic power station, and simultaneously collect inverter output power data, combiner box identification data, branch number data and branch maintenance status data corresponding to the photovoltaic string branch, and generate the original monitoring dataset of the string branch. The data transmission and database building unit is configured to transmit the original monitoring dataset of the string branches to the measurement and control acquisition device via the communication link, and then upload it to the power monitoring backend by the measurement and control acquisition device. It associates and stores the DC current data, inverter output power data, combiner box identification data, branch number data and branch maintenance status data of each photovoltaic string branch to form a real-time monitoring database of the string branches. The operating condition determination unit is configured to determine the operating condition based on the real-time monitoring database of the string branches and the inverter output power data corresponding to each photovoltaic string branch. When the inverter output power meets the preset power generation threshold, the corresponding photovoltaic string branch is determined to be in an effective power generation determination condition. The zero-current screening unit is configured to perform zero-current determination on the DC current data of each photovoltaic string branch that is in the effective power generation determination condition, and when the DC current of the corresponding photovoltaic string branch is less than the preset zero-current threshold, it combines the branch maintenance status data of the photovoltaic string branch to perform alarm shielding determination. When the branch maintenance status data indicates that the photovoltaic string branch is not in the listed maintenance state, the photovoltaic string branch is determined to be a candidate branch for zero-current fault. The fault location unit is configured to perform fault location processing on the zero-current fault candidate branch, extract the inverter identifier, combiner box identifier and branch number corresponding to the zero-current fault candidate branch, and generate the zero-current fault location result. The alarm output unit is configured to trigger the fault alarm output of the power monitoring backend based on the zero-current fault location result, and generate and display the zero-current alarm information corresponding to the target photovoltaic string branch.

10. An automatic alarm system for a zero-current photovoltaic string branch according to claim 9, characterized in that: The data acquisition unit includes a current acquisition subunit, an identifier reading subunit, a maintenance status matching subunit, and a dataset generation subunit. The current acquisition subunit samples the output circuit of each photovoltaic string branch within each combiner box branch, acquiring the corresponding branch DC current signal and reading the inverter output power signal of the corresponding inverter within the same sampling period. The identifier reading subunit reads the combiner box identifier, branch number, and inverter association relationship corresponding to each photovoltaic string branch, and then correlates the combiner box identifier, branch number, and inverter association relationship with the branch DC current signal and inverter... The output power signals are bound one by one; the maintenance status matching subunit reads the pre-set branch maintenance status tags from the power monitoring backend and matches the branch maintenance status tags to the corresponding photovoltaic string branches according to the combiner box identifier and branch number; the dataset generation subunit performs time synchronization processing, data integrity verification and abnormal null value filtering on the bound and matched data, and encapsulates the verified data in a structured manner according to the combiner box identifier, branch number, inverter identifier, branch DC current value, inverter output power value and branch maintenance status to generate the original monitoring dataset of the string branches; The data transmission and database construction unit includes a message encapsulation subunit, a message parsing subunit, a branch object merging subunit, and a database writing subunit. The message encapsulation subunit encapsulates the branch DC current data, inverter output power data, combiner box identification data, branch number data, and branch maintenance status data from the original monitoring dataset of the string branches into data transmission messages according to a preset communication message format, and writes the corresponding timestamp information and device source identification information. The message parsing subunit executes the message after the measurement and control acquisition device receives the data transmission message. The process involves parsing, source verification, and field splitting. The branch object merging subunit performs branch object merging and sequential reorganization on the parsed data based on the equipment source identification information, combiner box identification data, branch number data, and inverter correspondence, generating a background upload data record that corresponds one-to-one with each photovoltaic string branch. The database writing subunit uploads the background upload data record to the power monitoring backend and establishes a data index according to the combiner box identifier, branch number, and inverter identifier. It then writes the uploaded data in association to form the branch monitoring record in the string branch real-time monitoring database. The operating condition determination unit includes a power percentage calculation subunit, a threshold setting subunit, and an operating condition output subunit. The power percentage calculation subunit reads the real-time output power value and rated power value of the inverter corresponding to each photovoltaic string branch, and calculates the output power percentage of each inverter in the current sampling period. The threshold setting subunit determines the preset power generation threshold corresponding to the current sampling period based on the preset power generation threshold configuration parameters. The operating condition output subunit compares the output power percentage with the preset power generation threshold. When the output power percentage is greater than or equal to the preset power generation threshold, it determines that the corresponding inverter is in an effective power generation state, and marks each photovoltaic string branch associated with the inverter as a branch to be screened under the effective power generation determination operating condition. The zero-current screening unit includes a zero-current threshold retrieval subunit, a zero-current initial screening subunit, a tagging and shielding determination subunit, and a candidate branch output subunit. The zero-current threshold retrieval subunit reads the pre-stored zero-current threshold configuration parameters in the power monitoring backend and determines the zero-current threshold parameter for the branch to be screened based on the branch type information and the operating condition information corresponding to the current sampling period. The zero-current initial screening subunit compares the branch DC current value corresponding to the branch to be screened with the zero-current threshold parameter. If the branch DC current value is less than the zero-current threshold parameter, it determines that the branch is not selected. The branch to be screened should meet the initial screening condition of zero current; the tagging and shielding determination subunit reads the maintenance status, maintenance tagging start time, maintenance tagging end time and tagging operation identification information of the branch to be screened that meets the initial screening condition of zero current, and determines whether the current sampling time falls within the effective tagging time interval; when the branch to be screened is not in the tagging maintenance state or the current sampling time does not fall within the effective tagging time interval, the candidate branch output subunit generates a zero current abnormal branch mark by combining the corresponding combiner box mark, branch number and inverter mark, and outputs the branch to be screened as a zero current fault candidate branch; The fault location unit includes a location code generation subunit, a topology matching subunit, and a location result generation subunit. The location code generation subunit reads the inverter identifier, combiner box identifier, and branch number corresponding to the zero-current fault candidate branch. It establishes multi-level location identifier fields according to the inverter level, combiner box level, and branch level. After normalizing each field, it concatenates them according to a preset coding order and adds the current sampling period identifier or timestamp information to generate a unique location identifier code for the corresponding zero-current fault candidate branch. The topology matching subunit reads the pre-stored equipment topology relationship data in the power monitoring backend and uses a unique identifier... The inverter identifier field, combiner box identifier field, and branch number field in the bit identifier code are used as topology matching keys. The device record corresponding to the candidate branch of the target zero current fault is retrieved in the inverter device table, combiner box device table, string branch connection relationship table, and device hierarchy mapping table. When the hierarchy consistency check passes, the location result generation subunit extracts the corresponding inverter installation location, combiner box installation location, and branch access location. It then associates and combines the inverter installation location, combiner box installation location, branch access location, branch DC current value, inverter output power value, and corresponding timestamp information to generate the zero current fault location result. The alarm output unit includes an alarm information generation subunit, an alarm level determination subunit, an alarm triggering subunit, and an event retention subunit. The alarm information generation subunit generates corresponding zero-current alarm information based on the zero-current fault location result. The alarm level determination subunit assesses the severity of the zero-current alarm information based on the deviation of the branch DC current value from the zero-current threshold parameter, the output power ratio corresponding to the inverter output power value, the fault duration, the equipment level of the faulty branch, and whether it is under maintenance. It then outputs the corresponding alarm level and alarm output control command. The alarm triggering subunit triggers the corresponding alarm output mode according to the alarm output control command. The event retention subunit writes the zero-current alarm information that has already triggered the alarm output into the alarm event log database and associates the zero-current alarm information with the corresponding unique location identifier code, the current alarm status, and the processing status.

Images