Optical storage and charging integrated device account digital management method and system

Abstract

The application discloses a light storage and charging integrated equipment account digital management method and system, and relates to the technical field of new energy station management.The system comprises an edge gateway and a cloud platform.The application identifies communication protocols and analyzes data through the edge gateway multi-protocol analysis unit, and automatically matches account field templates in cooperation with an equipment category identification classifier;extracts maintenance work order entities and relationships by using a natural language processing model, updates account maintenance and component records through semantic vector matching, forms a work order closed loop, and caches account modification logs in the form of a vector clock when communication is interrupted, synchronizes and detects conflicts and merges after recovery, and guarantees the final consistency of the edge cloud.The cloud platform gathers data, provides full life cycle history tracking, hierarchical permission control and reports, automatically synchronizes the scrap state, meets the requirements of asset fine management and compliance.

Description

Technical Field

[0001] This invention relates to the field of new energy power station management technology, specifically to a digital management method and system for integrated photovoltaic, energy storage and charging equipment ledgers. Background Technology

[0002] Integrated photovoltaic-energy storage-charging power stations are composite energy stations that deeply integrate photovoltaic power generation, energy storage systems, and charging facilities. These stations contain a wide variety of equipment, including photovoltaic inverters, energy storage converters, battery management systems, charging pile controllers, and power distribution monitoring units. The communication protocols used by these various devices are inconsistent, their data formats differ, and their operating parameters and status information are stored in the proprietary closed systems of the equipment manufacturers.

[0003] Currently, the establishment of equipment ledgers and routine maintenance are still mainly done manually. Maintenance personnel need to check the nameplates of each piece of equipment, manually fill in basic information such as equipment code, specifications, serial number, and installation location, and then enter it into spreadsheets or independent management software. Afterwards, they periodically update the paper maintenance records and parts replacement records into the ledger database. When equipment faces changes in status such as scrapping or dismantling, manual adjustments are also required in the system.

[0004] This model has significant shortcomings: manual registration is time-consuming and intensive, often resulting in a vacuum period after new equipment is put into operation where there are no records to check; the completeness and accuracy of the record information rely excessively on the subjective factors of the operators, making it difficult to eliminate errors and omissions; key information in maintenance work orders—such as which parts were replaced and what kind of repairs were performed—cannot be promptly and accurately transmitted back to the equipment record, creating an information gap between the work order closure and record updates; even when equipment has been scrapped or dismantled, the record remains in an operational state for a long time, resulting in a serious discrepancy between the record and the actual situation.

[0005] As the number of photovoltaic, energy storage, and charging stations increases and their geographical distribution becomes more dispersed, the aforementioned contradictions become increasingly prominent, making it difficult for operation and maintenance personnel to accurately grasp the real-time status of assets, thereby affecting the timeliness of maintenance decisions and the accuracy of compliance reports. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and system for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment, thus solving the problems mentioned in the background section.

[0007] To achieve the above objectives, the present invention provides a method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment, applied to a digital management system for ledgers of integrated photovoltaic, energy storage, and charging equipment that includes an edge gateway and a cloud platform, comprising the following steps: S1. Multi-protocol parsing: A multi-protocol parsing unit is configured in the edge gateway. The multi-protocol parsing unit calls the protocol parsing template library to identify the communication protocol and parse the data of each device in the optical storage and charging station, and obtains the original operating data items of each device. S2. Equipment Category Identification: When the edge gateway identifies and parses data based on the communication protocol, it matches the protocol type and the device identifier extracted from the original operating data items. It then calls a pre-built device category identification classifier to determine the device category to which the device belongs. Based on the device category, it retrieves the general attribute field generation template and the device category-specific attribute field generation template from the field generation rule base. The parameter values ​​obtained directly or indirectly from the original operating data items are automatically filled into the general attribute field generation template and the device category-specific attribute field generation template to generate the initial ledger record for the device. S3, Assigning Electronic Identifier: Each device is assigned a unique electronic identifier, and a mapping relationship is established between the electronic identifier and the corresponding ledger record. The edge gateway sends the assigned electronic identifier to the tag generation device connected to it to generate physical tags with the electronic identifier embedded or printed on them.

[0008] S4, Work Order Analysis: The edge gateway receives maintenance work order data generated by the site management system or mobile terminal, uses a natural language processing model to perform entity recognition and relationship extraction on the maintenance work order data, obtains the equipment clues, work types and replacement parts information associated with the maintenance work order data, performs semantic similarity matching between the obtained equipment clues and the stored ledger records, determines the target equipment index, and automatically updates the ledger inspection record and parts ledger fields of the equipment according to the work type and replacement parts information. S5. Ledger record update: When the edge gateway receives a notification that the device scrapping or dismantling process is completed, it automatically updates the corresponding device's ledger record status to scrapped or offline. S6. Synchronize ledger records: Synchronize the ledger records generated by the edge gateway to the cloud platform.

[0009] As a further aspect of the present invention, the protocol parsing template library contains protocol parsing templates that correspond one-to-one with various industrial communication protocols. The protocol parsing templates include protocol frame start identifier, data field offset, data type, byte order, and parsing function index. The method by which the edge gateway performs protocol self-adaptation identification on the received raw data frame is as follows: the byte sequence of the data frame is matched with the protocol frame start identifier of each protocol parsing template. If the match is successful, the corresponding protocol parsing template is loaded, and the raw running data item is parsed out according to the data field offset and data type in the protocol parsing template.

[0010] As a further aspect of the present invention, the equipment category identification classifier is a classification model trained by taking the equipment protocol type, statistical characteristics of equipment operating parameters and communication message characteristics as inputs and the equipment category of each device as the classification label. The statistical characteristics of the equipment operating parameters include the mean and variance of voltage, current, and power, while the characteristics of the communication messages include the message period, heartbeat signal interval, and message-specific register address range.

[0011] As a further aspect of the present invention, the equipment category identification classifier takes the statistical features and communication message features corresponding to each equipment as input, outputs the probability of belonging to each equipment category, and takes the equipment category corresponding to the maximum probability as the identification result.

[0012] As a further aspect of the present invention, the process of using a natural language processing model to perform entity recognition and relation extraction on maintenance work order data is as follows: the maintenance work order text is input into a pre-trained named entity recognition and relation extraction model, and the equipment code field, component name field, operation type field, and replacement material list field are extracted. The extracted device code field and the device number field in the ledger record are compared using a natural language processing model to generate semantic encoding vectors for the device code field and the device number field in the ledger record. Semantic similarity matching based on semantic encoding is then performed, and the similarity of the matching is obtained. If the similarity of the matching exceeds a preset threshold, the matched ledger record is identified as the target device. For each item in the replacement bill of materials, compare it with the existing parts ledger field of the target equipment, and add new parts or mark replacement parts.

[0013] As a further aspect of the present invention, the specific method of semantic similarity matching is as follows: the semantic encoding vector of the device encoding field generated by the natural language processing model and the semantic encoding vector of the device number field in the ledger record are calculated with cosine similarity, and a pre-set similarity threshold is extracted. When the similarity is greater than the similarity threshold, the match is considered successful.

[0014] As a further aspect of the present invention, the method for generating the notification of completion of the equipment scrapping or dismantling process is as follows: after the edge gateway detects that the heartbeat signal of the target device is interrupted and the duration exceeds the preset time, it sends an on-site confirmation task to the site management system and receives a manual confirmation completion notification returned by the site management system.

[0015] As a further aspect of the present invention, the method further includes: when communication between the edge gateway and the cloud platform is interrupted, the edge gateway caches log entries of the ledger modification operation locally, and each log entry includes an operation timestamp, operation type, target device identifier, and a snapshot of the field values ​​before and after the modification. When communication is restored, the edge gateway and the cloud perform bidirectional log synchronization, and use a vector clock-based conflict detection and merging algorithm to handle inconsistencies between the two sets of data.

[0016] As a further aspect of the present invention, the conflict detection and merging algorithm based on vector clocks includes: each edge gateway and the cloud maintains a vector clock V1 and V2, and each log entry carries the vector clock at the time of its generation; During synchronization, the vector clock constraint relationship of log entries is compared. If V1 < V2, the cloud record is directly overwritten by the edge gateway log. If V1 and V2 are concurrently conflicting, the version generated by the operation with the higher privilege level will be retained according to the predefined arbitration rules.

[0017] A digital management system for integrated photovoltaic, energy storage, and charging equipment ledgers. This system is used to implement digital management methods for integrated photovoltaic, energy storage, and charging equipment ledgers. The system includes: The edge gateway is configured with a multi-protocol parsing unit, a device identification and documentation unit, a work order semantic parsing unit, and a local cache synchronization unit. The multi-protocol parsing unit is used to identify the communication protocols and parse the data of each device in the optical storage and charging station by calling the protocol parsing template library, and to obtain the original operating data items of each device. The equipment identification and filing unit is used to determine the equipment category by calling a pre-built equipment category identification classifier based on the protocol type matched during parsing and the equipment identifier extracted from the original operation data items. Based on the equipment category, it retrieves the general attribute field generation template and the equipment category-specific attribute field generation template from the field generation rule base, automatically fills in the initial ledger record, assigns a unique electronic identifier to each device, and establishes a mapping relationship between the electronic identifier and the corresponding ledger record. The work order semantic parsing unit is used to receive maintenance work order data, use natural language processing models to perform entity recognition and relation extraction, obtain equipment clues, work types and replacement parts information, determine target equipment through semantic matching and automatically update the ledger maintenance records and parts ledger fields, and update the ledger status when a scrapping or dismantling notice is received. The local cache synchronization unit is used to cache the ledger modification operation log locally when the network is down, and to perform bidirectional log synchronization with the cloud platform when the network is restored, using conflict detection and merging based on vector clocks; The cloud platform communicates with the edge gateway to receive ledger records.

[0018] As a further aspect of the present invention, the protocol parsing template library stores protocol parsing templates corresponding to ModbusTCP, CAN and MQTT protocols; the equipment category identification classifier is a multi-classification model trained based on gradient boosting decision trees; the natural language processing model is a pre-trained language model based on the Transformer architecture, which is fine-tuned on manually annotated work order corpora.

[0019] This invention provides a method and system for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment. Compared with existing technologies, it has the following advantages: The cloud platform aggregates all site log data, providing full lifecycle traceability from equipment installation to scrapping, as well as standardized report output. Simultaneously, upon receiving notification of equipment scrapping or dismantling completion, it automatically updates the log status to scrapped or offline, ensuring consistency between records and actual inventory and meeting the needs for refined asset management and compliance supervision. The cloud platform can also further provide role-based hierarchical access control.

[0020] By utilizing a pre-trained natural language processing model to perform entity recognition and relation extraction on maintenance work orders, the system obtains equipment codes, component names, operation types, and replacement material lists. Semantic encoding vector similarity calculations are then used to match target equipment, automatically updating the maintenance log and component log fields. This creates a closed-loop synchronization between work orders and the logbook, reducing errors from manual recording and update delays.

[0021] When communication between the edge gateway and the cloud platform is interrupted, ledger modification operations are cached locally as log entries, with each log entry carrying a vector clock. After the network is restored, bidirectional log synchronization is performed, the constraints of the vector clocks are compared to detect concurrent conflicts, and they are merged according to arbitration rules to ensure eventual consistency of ledger records between the edge and the cloud, thereby improving the reliability of the system under network outage conditions.

[0022] The cloud platform aggregates all site log data, providing full lifecycle tracking from equipment installation to scrapping, role-based hierarchical access control, and standardized report output. Simultaneously, upon receiving notification of equipment scrapping or dismantling completion, the log status is automatically updated to scrapped or offline, ensuring consistency between records and actual inventory and meeting the needs of refined asset management and compliant supervision. Attached Figure Description

[0023] Figure 1 This is a system block diagram of the digital management system for the integrated photovoltaic, energy storage, and charging equipment ledger of the present invention.

[0024] Figure 2 This is a flowchart illustrating the digital management method for the ledger of the integrated photovoltaic, energy storage, and charging equipment of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Please see Figure 1 and Figure 2 As shown, the embodiments of the present invention provide the following technical solutions: As an embodiment of the present invention: This invention relates to a method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment. This method is executed within a digital management system for integrated photovoltaic, energy storage, and charging equipment comprised of an edge gateway and a cloud platform, and specifically includes the following steps: S1. Configure a multi-protocol parsing unit in the edge gateway. This multi-protocol parsing unit integrates a protocol parsing template library, which contains corresponding protocol parsing templates for the industrial communication protocols in the optical storage and charging field station.

[0027] The edge gateway connects to devices such as photovoltaic inverters, energy storage converters, charging pile controllers, and power distribution monitoring units within the photovoltaic-energy storage-charging station, and receives raw data frames through the physical communication links established with these devices.

[0028] Among them, raw data frames refer to binary data packets directly generated by each device according to its industrial communication protocol without protocol parsing or processing.

[0029] The multi-protocol parsing unit identifies and parses the communication protocols of the raw data frames from each device by calling the protocol parsing template library, thereby obtaining the raw operational data items of each device. The raw operating data items are single field values ​​parsed from the message. The raw operating data items include instantaneous voltage values, instantaneous current values, power output values, cumulative power values, and device status values.

[0030] S2. The edge gateway identifies the protocol type matched during communication protocol identification and data parsing, as well as the device identifier extracted from the original running data items, and calls the device category identification classifier pre-built and deployed on the edge gateway to determine the device category to which the device belongs.

[0031] The equipment categories are selected from corresponding equipment categories such as photovoltaic modules, energy storage equipment, charging piles, and power distribution equipment.

[0032] The equipment category identification classifier is a classification model pre-trained based on various communication and electrical features.

[0033] The protocol parsing template library includes templates for generating general attribute fields and templates for generating equipment category-specific attribute fields. The general attribute field generation templates define common ledger fields shared by all equipment categories, i.e., general attribute fields; the equipment category-specific attribute field generation templates define ledger fields that differ depending on the equipment category, i.e., equipment category-specific attribute fields.

[0034] After determining the device category, the edge gateway retrieves the general attribute field generation template and the device category-specific attribute field generation template from the field generation rule base based on the device category.

[0035] The field generation rule base is a key-value pair database, where the key is the category identifier output by the device category identification classifier, and the value is a set of predefined template pointers. The edge gateway uses the category identifier as an index to perform a hash lookup in the field generation rule base to quickly locate and retrieve the corresponding general attribute field generation template and device category-specific attribute field generation template. The general attribute fields include at least the device number, serial number, purchase date, supplier, rated parameters, and installation location. Different device categories correspond to different device category-specific attribute fields.

[0036] The edge gateway will automatically fill the corresponding field positions in the template with parameter values ​​obtained directly or indirectly from the original running data items, thereby generating an initial ledger record for the device.

[0037] Among them, the parameter values ​​obtained directly refer to the parameters that are already included in the original operating data and whose format is consistent with the requirements of the ledger fields, such as the device serial number, firmware version, instantaneous current at the moment of power-on, etc., which can be directly read and assigned values; the original operating data is the message reported by the device.

[0038] Indirectly calculated parameter values ​​refer to parameters obtained by performing mathematical operations on one or more data items in the original operating data. For example, the continuous operating time can be obtained by subtracting the start and end times of the operating time, the average operating temperature can be obtained by averaging multiple temperature sampling values, or the average flow rate can be obtained by the ratio of the cumulative flow rate to the cumulative time.

[0039] In this embodiment, the protocol parsing template built into the edge gateway marks the value retrieval method for each ledger field, including direct mapping and mathematical operation formulas. When the value retrieval method is a mathematical operation formula, the protocol parsing template stores the corresponding calculation expression. After parsing, the edge gateway uses the calculation expression to calculate the original running data item to obtain the indirectly calculated parameter value.

[0040] S3. While generating the initial ledger record, the edge gateway assigns a unique electronic identifier to each device and establishes a mapping relationship between the electronic identifier and the corresponding ledger record.

[0041] The edge gateway sends the assigned electronic identifier to a tag generation device connected to it to generate a physical tag embedded with or printed with the electronic identifier. In this embodiment, the electronic identifier is encoded as a QR code or RFID tag information. Staff affix the electronic identifier to the device casing, and the corresponding ledger record can be retrieved by scanning the code.

[0042] In this context, the label generation device refers to a label printer or RFID writer. The edge gateway sends the assigned electronic identifier to the label printer or RFID writer connected to it to generate a QR code label or RFID tag containing the electronic identifier. This establishes a mapping relationship between the electronic identifier and the corresponding ledger record.

[0043] S4. During daily operation, the station management system or on-site mobile terminal generates maintenance work order data based on engineering activities such as repair and maintenance.

[0044] The work order semantic parsing unit in the edge gateway receives maintenance work order data and uses a pre-trained natural language processing model to perform entity recognition and relation extraction on the data. This yields equipment clues, including equipment identifiers, equipment names, or equipment locations, as well as job type and replacement part information associated with the work order data. The obtained equipment clues are then semantically matched with stored ledger records to determine the target equipment index. Based on the job type and replacement part information, the ledger maintenance record and parts ledger fields for that equipment are automatically updated.

[0045] If a maintenance work order records the replacement of a battery module of an energy storage device, the corresponding module information in the component ledger field of that device will be updated, and a maintenance record will be added.

[0046] In this embodiment, the natural language processing model is existing technology, so it will not be described in detail.

[0047] S5. When equipment reaches the end of its life cycle and enters the scrapping process or is dismantled, the site management system issues a notification that the equipment scrapping or dismantling process is completed.

[0048] After receiving notification that the scrapping or dismantling process of the device is completed, the edge gateway updates the corresponding device's ledger record status to scrapped or offline and records the operation time, so that the ledger record is synchronized with the actual physical status of the device.

[0049] In this embodiment, the equipment scrapping or dismantling process completion notification is generated as follows: when the edge gateway detects that the heartbeat signal or communication connection of the target device is interrupted, and the duration of the interruption exceeds a preset duration threshold, and after the on-site confirmation task dispatched by the site management system has been manually verified and returned, the site management system generates the corresponding equipment scrapping or dismantling process completion notification.

[0050] S6 and the edge gateway synchronize the generated ledger records to the cloud platform via wired or wireless networks.

[0051] When the network connection between the edge gateway and the cloud platform is normal, real-time synchronization of ledger records is performed directly. When the network connection is interrupted, a local caching mechanism is activated to cache the ledger modification operation logs to be synchronized in the edge gateway. Two-way log synchronization is performed after communication is restored. The cloud platform aggregates ledger record information from all devices in the entire site and provides full lifecycle history query, hierarchical access control, and standardized report generation services.

[0052] Through full lifecycle history tracking, managers can trace information from the initial installation of any device, through all maintenance and component replacements, to its eventual scrapping. Hierarchical access control ensures that different roles can only access and manipulate authorized data. Standardized report generation supports one-click output of verification reports according to regulatory requirements.

[0053] As a second embodiment of the present invention: In specific implementation, compared with Embodiment 1, the only difference between the technical solution of this embodiment and Embodiment 1 is that: The protocol parsing template library contains protocol parsing templates that correspond one-to-one with various industrial communication protocols. Specifically, the library stores protocol parsing templates for ModbusTCP (Modbus Transmission Control Protocol, an Ethernet-based Modbus protocol), CAN (Controller Area Network), and MQTT (Message Queuing Telemetry Transport) protocols. Each protocol parsing template defines at least the protocol frame start identifier, data field offset, data type, byte order, and parsing function index.

[0054] The protocol frame start identifier is used to distinguish data frames of different protocols. The data field offset indicates the starting position and length of the bytes carrying the device operating parameters in the data frame. The data type indicates the numerical format of the field. The byte order indicates whether the bytes are big-endian or little-endian. The parsing function index points to the calculation function that converts binary data into actual physical quantities.

[0055] The edge gateway performs communication protocol identification and data parsing on the received raw data frames as follows: The edge gateway receives raw data frames sent by the device. The raw data frame is a byte sequence consisting of multiple bytes. The starting portion of the byte sequence is matched against the protocol frame start identifier of each protocol parsing template. This involves comparing the starting byte sequence of the original data frame with the protocol frame start identifier byte by byte. A match is considered successful if and only if all bytes are identical; otherwise, the next template is tried. If a match is successful, the corresponding protocol parsing template is loaded. Based on the data field offset and data type defined in the loaded protocol parsing template, the original runtime data item is parsed from the corresponding position in the byte sequence.

[0056] The method for matching the beginning of the byte sequence with the protocol frame start identifier of each protocol parsing template is as follows: Let M be the number of protocol parsing templates in the protocol parsing template library, and let H be the protocol frame start identifier of the m-th protocol parsing template. m m = 1, 2, ..., M, where m is the sequence index of the protocol parsing template. For the current original data frame byte sequence B, the first L bytes are calculated sequentially. m byte and H m The matching degree is determined, and when they are completely consistent, the template is adopted. Then, byte segments are extracted from the specified position and converted into actual physical quantities to realize protocol parsing.

[0057] Meanwhile, the edge gateway identifies and extracts the device identifier by recognizing the device address field or the device identifier in the message during the data parsing process through the communication protocol.

[0058] The equipment category identification classifier takes the statistical characteristics of equipment protocol type, equipment operating parameters and communication message characteristics as input. In this embodiment, the classification model is trained with four equipment categories: photovoltaic modules, energy storage equipment, charging piles and power distribution equipment as classification labels.

[0059] Among them, the device protocol type is the aforementioned identified protocol category; the statistical characteristics of the device operating parameters include the mean and variance of voltage, current, and power; the communication message characteristics include message period, heartbeat signal interval, and message-specific register address range.

[0060] The mean and variance in the statistical features are calculated as follows: the edge gateway collects the target parameter values ​​from multiple consecutive data frames of the device using a sliding time window.

[0061] The sliding time window is initially empty. After receiving the first W sample values, the window is filled and the mean and variance are calculated. Thereafter, each time a new sample value is received, the window slides, the oldest sample value is removed, and the mean and variance are recalculated based on the new W sample values.

[0062] Taking voltage parameters as an example, let the i1th sample value within the window be... If i1 = 1, 2, ..., W, where W is the preset window length, then... and Calculate the mean value of this voltage parameter respectively. and variance : Similarly, the mean and variance of the parameters corresponding to current and power can be calculated and combined into a vector of statistical characteristics.

[0063] The equipment category identification classifier takes the statistical characteristics and communication message characteristics of each equipment as input, and outputs the probability p1, p2, ... p of each equipment category. u Satisfying p1+p2+......+p u =1, and the equipment category corresponding to the maximum probability is taken as the identification result. u refers to the number of equipment categories. In this embodiment, u is 4. p1, p2, p3, and p4 refer to the probabilities of the corresponding equipment categories of photovoltaic modules, energy storage devices, charging piles, and power distribution equipment, respectively.

[0064] For the heartbeat signal interval in the communication message characteristics, the edge gateway obtains it by recording the arrival timestamps of two consecutive messages from the same device with the same function code and calculating their time difference. For the specific register address range of the message, the edge gateway parses the original data frames of the identified protocol and extracts the numerical range formed by the start and end addresses of the registers used to carry device operating parameters. The start and end addresses of this range are encoded and used as one of the input features of the classification model.

[0065] In this embodiment, the equipment category identification classifier employs a gradient boosting decision tree algorithm during the training phase, using historical data of labeled equipment categories as the sample set for learning. During inference, it is directly deployed on the edge gateway. The gradient boosting decision tree algorithm is an existing algorithm and will not be elaborated upon. The training process of this equipment category identification classifier is completed on a cloud platform, with training samples derived from historically accessed labeled equipment data from optical storage and charging stations. After the equipment category identification classifier is trained in the cloud and passes accuracy verification, it is deployed and run on the edge gateways of each station via OTA firmware upgrade.

[0066] After determining the device category, the edge gateway retrieves the general attribute field generation template and the device category-specific attribute field generation template from the field generation rule library based on the device category, and automatically fills them in using the collected parameters to form the initial ledger record.

[0067] As an embodiment of the present invention: In its specific implementation, compared to Embodiment 1 and Embodiment 2, the technical solution of this embodiment is to combine the solutions of Embodiment 1 and Embodiment 2. The only difference between the technical solution of this embodiment and Embodiment 1 and Embodiment 2 is that: After the on-site maintenance personnel fill out the maintenance work order, the maintenance work order data is transmitted to the edge gateway in text form.

[0068] For maintenance work order data, the work order semantic parsing unit is configured with a natural language processing model to automatically extract equipment code, component name, operation type, and replacement material list fields from the maintenance work order data. The construction of the natural language processing model and the use of the natural language processing model to perform entity recognition and relation extraction on the maintenance work order data are as follows: In this embodiment, the natural language processing model is a pre-trained language model based on the Transformer architecture (such as BERT (Bidirectional Encoder Representations from Transformers), RoBERTa (Robustly optimized BERT approach), or ALBERT (A Lite BERT)), which is fine-tuned on manually annotated work order corpus.

[0069] Specifically, an entity recognition output layer and a relation extraction output layer are added side-by-side on top of the encoding layer of the pre-trained language model; The entity recognition output layer uses sequence labeling to output the entity label corresponding to each character / word. The relation extraction output layer outputs predefined relation categories using a classification method.

[0070] During fine-tuning, the model is jointly trained using a weighted sum of entity labeling loss and relation classification loss as the total loss, enabling the model to simultaneously perform entity recognition and relation extraction from the same input work order text.

[0071] For example, a work order corpus reads: "On-site inspection revealed an over-temperature alarm on the B-phase IGBT module of the #2 energy storage converter. The module needs to be replaced and the temperature sensor calibrated." The corresponding entity recognition annotation results are: Equipment code field: "#2 energy storage converter"; Component name fields: "B-phase IGBT module", "Temperature sensor"; Fault phenomenon field: "Over-temperature alarm"; Operation type fields: "Replacement", "Calibration"; Replacement material list field includes one item: "B-phase IGBT module". The relation extraction annotation results are: (#2 energy storage converter, includes, B-phase IGBT module), (#2 energy storage converter, includes, temperature sensor), (Replacement, target, B-phase IGBT module), (Calibration, target, temperature sensor), (Over-temperature alarm, originates from, B-phase IGBT module). At least 2000 work order corpora annotated in this way were used to fine-tune the natural language processing model.

[0072] The maintenance work order data is input into a natural language processing model, and the equipment code field, component name field, operation type field, and replacement material list field are extracted.

[0073] After extracting the device code field, the extracted device code field is matched with the device number field in the ledger record based on semantic similarity. The semantic encoding vector of the device encoding field generated by the natural language processing model is compared with the semantic encoding vector of the device number field in the ledger record by calculating the cosine similarity. Specifically, the device encoding string With the i-th equipment number field in the ledger record Convert to semantic encoding vector =(x1,x2,......,x n ), =(y1,y2,......,y n ), where n is the dimension of the semantic encoding vector, through Calculate the device code string With the i-th equipment number field in the ledger record similarity between In the formula, k = 1, 2, ..., n; Extract the pre-set similarity threshold θ sim In this embodiment, the similarity threshold is determined based on the performance of the natural language processing model on the validation set, and its value is between 0.7 and 0.95. Greater than θ sim If the match is successful, the device with the highest similarity will be selected as the target device.

[0074] Once the target device is matched, the ledger is automatically updated based on the extracted operation type and the changed bill of materials field.

[0075] For each item in the replacement bill of materials, compare it with the existing parts ledger field of the target equipment, and add new parts or mark replacement parts.

[0076] At the same time, add an entry to the maintenance log containing the work order number, time, work type and conclusion to complete the closed loop between the work order and the log.

[0077] As an embodiment of the present invention: In specific implementation, compared with Embodiments 1, 2, and 3, the only difference between this embodiment and Embodiments 1, 2, and 3 is that: This embodiment addresses the scenario where communication between the edge gateway and the cloud platform is interrupted. This embodiment implements a network outage cache synchronization mechanism in the edge gateway.

[0078] When communication between the edge gateway and the cloud platform is interrupted, the edge gateway caches log entries for ledger modification operations locally. Each log entry includes an operation timestamp, operation type, target device identifier, and snapshots of field values ​​before and after the modification.

[0079] When communication is restored, the edge gateway and the cloud platform perform bidirectional log synchronization, employing a vector clock-based conflict detection and merging algorithm to handle data inconsistencies. The edge gateway uploads locally cached log entries in batches, while the cloud provides feedback on log entries generated during this period. The final version is determined by comparing vector clocks. The specific method is as follows: Each edge gateway contains a local first vector clock V1, the cloud platform contains a second vector clock V2, and each log entry carries a snapshot of the corresponding node's vector clock at that time when it is generated; During bidirectional log synchronization, for any log entry, the constraint relationship between the vector clock snapshot it carries and the current vector clock of the target node is compared: If the vector clock V1 carried by a log entry from the edge gateway satisfies V1 > V2 with the current vector clock V2 in the cloud, it indicates that the log entry is an update after the cloud has a known state. Then, the log entry is applied to the cloud for recording and V2 is updated. If V1 and V2 are concurrent, it is determined that the edge gateway and the cloud platform have concurrently modified the same field. In this case, the version generated by the operation with the higher permission level will be retained according to the predefined arbitration rules.

[0080] Specifically, a permission level score is assigned to each type of operation. Cloud-based automated approval operations have a higher score than on-site manual operations. In case of conflict, the version with the higher score is retained, and a conflict log is generated for auditing.

[0081] In this embodiment, the predefined rules for the permission level score of operation type are as follows: The cloud platform's automated approval process assigns a tiered permission score of 3. The operation triggered after manual review on the cloud platform is assigned a permission level score of 2. The edge gateway assigns a permission level score of 1 for local manual operation or on-site mobile terminal operation.

[0082] When the vector clock determines that two log entries are in concurrent conflict, it compares the permission level scores they carry, retains the field value corresponding to the log entry with the higher score as the final version, and records the version with the lower score in the conflict log for manual auditing and backtracking.

[0083] If both parties have the same permission level score, the version with the later operation timestamp will prevail.

[0084] The present invention also provides a digital management system for the ledger of integrated photovoltaic, energy storage and charging equipment. The system is used to implement the digital management method for the ledger of integrated photovoltaic, energy storage and charging equipment. The system includes an edge gateway deployed at the site and a cloud platform located in the cloud.

[0085] The edge gateway is configured with a multi-protocol parsing unit, a device identification and documentation unit, a work order semantic parsing unit, and a local cache synchronization unit.

[0086] The multi-protocol parsing unit is used to identify the communication protocols and parse the data of the photovoltaic inverters, energy storage converters, charging pile controllers and power distribution monitoring units in the photovoltaic-storage-charging station by calling the protocol parsing template library, and to obtain the original operating data items of each device.

[0087] The equipment identification and filing unit is used to determine the equipment category by calling a pre-built equipment category identification classifier based on the protocol type matched during parsing and the equipment identifier extracted from the original operation data items. Based on the equipment category, it retrieves the general attribute field generation template and the equipment category-specific attribute field generation template from the field generation rule base, automatically fills in the initial ledger record, assigns a unique electronic identifier to each device, and establishes a mapping relationship between the electronic identifier and the corresponding ledger record.

[0088] The work order semantic parsing unit is used to receive maintenance work order data, use natural language processing models to perform entity recognition and relation extraction, obtain equipment clues, work types and replacement parts information, determine target equipment through semantic matching and automatically update the ledger maintenance records and parts ledger fields, and update the ledger status when a scrapping or dismantling notice is received.

[0089] The local cache synchronization unit is used to cache the ledger modification operation log locally when the network is down, and to perform bidirectional log synchronization with the cloud platform when the network is restored, using a conflict detection and merging based on vector clocks.

[0090] The cloud platform communicates with the edge gateway to receive ledger records and provide full lifecycle history query, hierarchical access control, and standardized report generation services.

[0091] It should be stated that all data collected in this application was collected with the user's consent and authorization, and the use of the data is legal and compliant, and the use and processing of the data comply with the relevant laws, regulations and standards of the relevant regions.

[0092] Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0093] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0094] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

[0095] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A digital management method for ledgers of integrated photovoltaic, energy storage, and charging equipment, characterized in that: Includes the following steps: S1. Configure a multi-protocol parsing unit in the edge gateway. The multi-protocol parsing unit calls the protocol parsing template library to identify the communication protocol and parse the data of each device in the optical storage and charging station, and obtain the original operating data items of each device. S2. When the edge gateway identifies and parses data according to the communication protocol, it matches the protocol type and the device identifier extracted from the original running data item, calls the pre-built device category identification classifier to determine the device category to which the device belongs, and retrieves the general attribute field generation template and the device category-specific attribute field generation template from the field generation rule base based on the device category. The parameter values ​​obtained from the original operation data items will be automatically filled into the general attribute field generation template and the equipment category-specific attribute field generation template to generate the initial ledger record for the equipment. S3. Assign a unique electronic identifier to each device and establish a mapping relationship between the electronic identifier and the corresponding ledger record; S4. The edge gateway receives maintenance work order data generated by the site management system or mobile terminal, uses a natural language processing model to perform entity recognition and relationship extraction on the maintenance work order data, obtains the equipment clues, work types and replacement parts information associated with the maintenance work order data, performs semantic similarity matching between the obtained equipment clues and the stored ledger records, determines the target equipment index, and automatically updates the ledger inspection record and parts ledger fields of the equipment according to the work type and replacement parts information. S5. When the edge gateway receives a notification that the device scrapping or dismantling process is completed, it automatically updates the corresponding device's ledger record status to scrapped or offline. S6. Synchronize the ledger records generated by the edge gateway to the cloud platform.

2. The method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment according to claim 1, characterized in that, The protocol parsing template library contains protocol parsing templates that correspond one-to-one with various industrial communication protocols. The protocol parsing template includes the protocol frame start identifier, data field offset, data type, byte order, and parsing function index. The method by which the edge gateway performs protocol self-adaptation identification on the received raw data frame is as follows: the byte sequence of the data frame is matched with the protocol frame start identifier of each protocol parsing template. If the match is successful, the corresponding protocol parsing template is loaded, and the raw running data item is parsed out according to the data field offset and data type in the protocol parsing template.

3. The method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment according to claim 1, characterized in that, The equipment category identification classifier is a classification model trained by taking the statistical characteristics of equipment protocol type, equipment operating parameters and communication message characteristics as input and the equipment category of each device as the classification label. The statistical characteristics of the equipment operating parameters include the mean and variance of voltage, current, and power, while the characteristics of the communication messages include the message period, heartbeat signal interval, and message-specific register address range.

4. The method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment according to claim 3, characterized in that, The equipment category identification classifier takes the statistical features and communication message features corresponding to each equipment as input, outputs the probability of belonging to each equipment category, and takes the equipment category corresponding to the maximum probability as the identification result.

5. The method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment according to claim 1, characterized in that, The process of using a natural language processing model to perform entity recognition and relation extraction on maintenance work order data is as follows: input the maintenance work order text into a pre-trained named entity recognition and relation extraction model, and extract the equipment code field, component name field, operation type field, and replacement material list field. The extracted device code field and the device number field in the ledger record are compared using a natural language processing model to generate semantic encoding vectors for the device code field and the device number field in the ledger record. Semantic similarity matching based on semantic encoding is then performed, and the similarity of the matching is obtained. If the similarity of the matching exceeds a preset threshold, the matched ledger record is identified as the target device. For each item in the replacement bill of materials, compare it with the existing parts ledger field of the target equipment, and add new parts or mark replacement parts.

6. The method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment according to claim 5, characterized in that, The specific method of semantic similarity matching is as follows: the semantic encoding vector of the device encoding field generated by the natural language processing model is used to calculate the cosine similarity with the semantic encoding vector of the device number field in the ledger record, and a pre-set similarity threshold is extracted. When the similarity is greater than the similarity threshold, the match is considered successful.

7. The method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment according to claim 1, characterized in that, Also includes: When communication between the edge gateway and the cloud platform is interrupted, the edge gateway caches log entries of ledger modification operations locally. Each log entry includes an operation timestamp, operation type, target device identifier, and snapshots of field values ​​before and after modification. When communication is restored, the edge gateway and the cloud perform bidirectional log synchronization, and use a vector clock-based conflict detection and merging algorithm to handle inconsistencies between the two sets of data.

8. The method for digital management of ledgers for integrated photovoltaic, energy storage, and charging equipment according to claim 7, characterized in that, The collision detection and merging algorithm based on vector clocks includes: each edge gateway and the cloud maintains a vector clock V1 and V2, and each log entry carries the vector clock at the time of its generation; During synchronization, the vector clock constraint relationship of log entries is compared. If V1 < V2, the cloud record is directly overwritten by the edge gateway log. If V1 and V2 are concurrently conflicting, the version generated by the operation with the higher privilege level will be retained according to the predefined arbitration rules.

9. A digital management system for integrated photovoltaic, energy storage, and charging equipment ledgers, the system being used to execute the digital management method for integrated photovoltaic, energy storage, and charging equipment ledgers as described in any one of claims 1-8, characterized in that, The system includes: The edge gateway is configured with a multi-protocol parsing unit, a device identification and documentation unit, a work order semantic parsing unit, and a local cache synchronization unit. The multi-protocol parsing unit is used to identify the communication protocols and parse the data of each device in the optical storage and charging station by calling the protocol parsing template library, and to obtain the original operating data items of each device. The equipment identification and filing unit is used to determine the equipment category by calling a pre-built equipment category identification classifier based on the protocol type matched during parsing and the equipment identifier extracted from the original operation data items. Based on the equipment category, it retrieves the general attribute field generation template and the equipment category-specific attribute field generation template from the field generation rule base, automatically fills in the initial ledger record, assigns a unique electronic identifier to each device, and establishes a mapping relationship between the electronic identifier and the corresponding ledger record. The work order semantic parsing unit is used to receive maintenance work order data, use natural language processing models to perform entity recognition and relation extraction, obtain equipment clues, work types and replacement parts information, determine target equipment through semantic matching and automatically update the ledger maintenance records and parts ledger fields, and update the ledger status when a scrapping or dismantling notice is received. The local cache synchronization unit is used to cache the ledger modification operation log locally when the network is down, and to perform bidirectional log synchronization with the cloud platform when the network is restored, using conflict detection and merging based on vector clocks; The cloud platform communicates with the edge gateway to receive ledger records.

10. The digital management system for integrated photovoltaic, energy storage, and charging equipment ledgers according to claim 9, characterized in that, The protocol parsing template library stores protocol parsing templates corresponding to ModbusTCP, CAN, and MQTT protocols; the equipment category identification classifier is a multi-classification model trained based on gradient boosting decision trees; the natural language processing model is a pre-trained language model based on the Transformer architecture, fine-tuned on manually annotated work order corpora.

Images