Data processing methods, devices and equipment applicable to power generation operations

By converting multimodal information into structured information and determining priority levels, the problem of data continuity and consistency in power production management is solved, achieving business continuity and compliance reliability in complex field environments, and avoiding duplicate entry and data loss.

CN123115274APending Publication Date: 2026-11-17MEIZHOU POWER SUPPLY BUREAU OF GUANGDONG POWER GRID CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611482840.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In power production management, existing technologies struggle to achieve data continuity, consistency, and compliance reliability in complex field environments, leading to repetitive data entry across terminals and systems, insufficient information consistency, and low processing efficiency. In particular, business continuity is affected in weak network or network outage environments.

Method used

By converting multimodal information into structured information, determining priority levels, and storing it in a local queue when the network is interrupted, and synchronizing it to the production system according to priority when the network is restored, the standardized expression and hierarchical management of operation instructions are achieved, avoiding duplicate entry and ensuring business continuity.

Benefits of technology

It improved the continuity, consistency, and compliance reliability of power production business processing, ensured the consistency of data connection and processing efficiency, and avoided data loss.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123115274A_ABST
    Figure CN123115274A_ABST
Patent Text Reader

Abstract

This application provides a data processing method, apparatus, and equipment applicable to power production operations. The method includes: acquiring multimodal information in response to user input; wherein the multimodal information represents operational instructions for power production operations; converting the multimodal information into structured information and determining the priority level of the structured information; wherein the structured information is used to drive the production system to execute business processing; when a network interruption is detected, storing the structured information in a local queue according to its priority level; and when network recovery is detected, synchronizing the structured information in the local queue to the production system according to its priority level. This method can improve the continuity, consistency, and compliance reliability of power production business processing in complex field environments.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of power production business process management and information processing, and in particular to a data processing method, apparatus and equipment applicable to power production business. Background Technology

[0002] In power production management, field operations typically rely on a combination of personal computer (PC) production systems, stand-alone mobile applications, and paper-based processes.

[0003] However, the above methods often require repeated data entry across terminals and systems in field operations, resulting in poor data connectivity and difficulty in continuously processing business in weak or offline network environments. At the same time, relying on manual verification and transcription can easily lead to insufficient information consistency, decreased processing efficiency, and limited stability of procedure execution.

[0004] Therefore, how to improve the continuity, consistency, and compliance reliability of power production operations in complex field environments has become a technical problem that needs to be solved. Summary of the Invention

[0005] The data processing method, apparatus, and equipment provided in this application for power production operations can improve the continuity, consistency, and compliance reliability of power production operations in complex field environments.

[0006] In a first aspect, embodiments of this application provide a data processing method applicable to power generation operations, including:

[0007] In response to user input, multimodal information is acquired; whereby the multimodal information represents the operational instructions for power production operations.

[0008] Multimodal information is converted into structured information, and the priority level of the structured information is determined; the structured information is used to drive the production system to perform business processing.

[0009] When a network interruption is detected, the structured information is stored in a local queue according to its priority level.

[0010] When network recovery is detected, the structured information in the local queue is synchronized to the production system according to priority level.

[0011] Optionally, as described above, the multimodal information is converted into structured information, including:

[0012] Based on the modality type of the multimodal information, the multimodal information is converted into intermediate text; wherein the modality type includes one or more of audio type, image type and text type;

[0013] Convert intermediate text into structured information.

[0014] Optionally, as described above, the multimodal information is converted into intermediate text based on its modality type, including:

[0015] When the modal type includes audio, the multimodal information is converted into intermediate text using a preset speech recognition technology;

[0016] When the modal type includes images, multimodal information is converted into intermediate text using preset image recognition technology;

[0017] When the modal type includes text, the multimodal information is determined to be intermediate text.

[0018] Optionally, as described above, the intermediate text is converted into structured information, including:

[0019] Based on a pre-defined terminology and keyword database, determine the business type corresponding to the intermediate text; wherein, the pre-defined terminology and keyword database includes the business type and the set of keywords corresponding to the business type;

[0020] From the preset set of terminology templates, determine the terminology template corresponding to the business type as the target template; wherein, the terminology template includes multiple data units used to carry information elements, and conversion logic used to establish the association between the data units and the intermediate text;

[0021] Based on the target template, extract the target operation parameters from the intermediate text and populate them into the target template to generate structured information.

[0022] Optionally, as described above, the priority level of structured information is determined, including:

[0023] Based on the business type corresponding to the structured information, a score value for the business type is determined; whereby the score value represents the urgency of the impact of the structured information on the operational security of the production system.

[0024] The priority level of structured information is determined based on the score.

[0025] Optionally, as described above, the business type includes at least one of production planning, equipment defect, and work permit categories;

[0026] Among them, the production plan category represents the formulation and modification of power production tasks, the equipment defect category represents the reporting and tracking of abnormalities in power equipment, and the work permit category represents the safety permits for power operations and personnel changes.

[0027] Optionally, as described above, the structured information in the local queue is synchronized to the production system according to priority levels, including:

[0028] Structured information to be synchronized is extracted from the local queue in descending order of priority.

[0029] The extracted structured information is sent to the production system, and the synchronization results returned by the production system are received.

[0030] When the synchronization result indicates a synchronization failure, the corresponding structured information is re-stored into the local queue, and the step of synchronizing the structured information in the local queue to the production system according to priority level is re-executed when network recovery is detected.

[0031] Optionally, the method described above further includes the following steps before sending the extracted structured information to the production system:

[0032] The structured information is validated to obtain the validation result; the validation process includes at least one of operation qualification validation, operation permission validation, and operation environment validation.

[0033] If the verification result fails, an alarm message will be generated.

[0034] Secondly, embodiments of this application provide a data processing apparatus suitable for power generation operations, comprising:

[0035] The response module is used to respond to user input operations and acquire multimodal information; wherein, the multimodal information represents the operation instructions for power production operations;

[0036] The conversion module is used to convert multimodal information into structured information and determine the priority level of the structured information; the structured information is used to drive the production system to perform business processing.

[0037] The caching module is used to store structured information into a local queue according to the priority level of the structured information when a network interruption is detected.

[0038] The recovery module is used to synchronize structured information in the local queue to the production system according to priority level when network recovery is detected.

[0039] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0040] The memory stores instructions that the computer executes;

[0041] The processor executes computer execution instructions stored in memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0042] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0043] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0044] The data processing method, apparatus, and equipment for power production operations provided in this application are applied to an instant messaging terminal. By responding to user input operations, multimodal information is acquired. Further, the multimodal information is converted into structured information, and the priority level of the structured information is determined. Further, when a network interruption is detected, the structured information is stored in a local queue according to the priority level of the structured information. When the network is detected to be restored, the structured information in the local queue is synchronized to the production system according to the priority level. In this application, the multimodal information represents the operation instructions for power production operations, and the structured information is used to drive the production system to perform business processing. The data processing method for power production operations provided in this application converts multimodal information into structured information, achieving standardized expression of operation instructions. This allows instant messaging terminals to directly drive the production system to execute business processes, avoiding repetitive data entry across terminals and systems, and improving data consistency and processing efficiency. By determining the priority levels of structured information, hierarchical management of different business operations is achieved, providing a basis for subsequent priority-based business processing. Furthermore, by storing structured information in a local queue according to priority levels when a network interruption is detected, and synchronizing it to the production system according to priority levels when the network is restored, business continuity is ensured in weak network or network outage environments, preventing data loss. The method in this application improves the continuity, consistency, and compliance reliability of power production business processing in complex field environments. Attached Figure Description

[0045] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0046] Figure 1 A flowchart illustrating a data processing method applicable to power generation operations provided in this application. Figure 1 ;

[0047] Figure 2 A flowchart illustrating a data processing method applicable to power generation operations provided in this application. Figure 2 ;

[0048] Figure 3 A schematic diagram of the structure of a data processing device applicable to power production operations is provided in this application;

[0049] Figure 4 This is a schematic diagram of the structure of an electronic device provided in this application.

[0050] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0051] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0052] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, they do not violate public order and good morals, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0053] It should be noted that the data processing method, apparatus and equipment applicable to power production business of this application can be used in the field of power production business process management and information processing, and can also be used in any field other than the field of power production business process management and information processing. The application field of the data processing method, apparatus and equipment applicable to power production business of this application is not limited.

[0054] In the field of power production management, on-site operations typically occur in complex environments such as substations, switching stations, high-voltage rooms, underground cable tunnels, and maintenance work areas. These operations include equipment defect reporting, plan additions, defect elimination feedback, and the transmission and processing of various production instructions closely related to on-site operations. These scenarios are characterized by dispersed personnel, numerous work locations, significant network fluctuations, and high timeliness requirements. On-site personnel often need to immediately create a record upon discovering equipment anomalies and submit it to the backend production system to drive subsequent handling processes.

[0055] Under the existing application architecture, business operations are typically supported by instant messaging tools, standalone mobile applications, PC-based production systems, and back-end business platforms. Among them, field terminals are responsible for collecting photos, voice descriptions, and text notes, fixed workstations are responsible for formal data entry and review operations, and the back-end production system is responsible for work order flow, status updates, and ledger maintenance.

[0056] Since the field terminals and the back-end production system are often not directly connected, information can easily become discontinuous when flowing between multiple entry points. Especially when the work area is in a weak network, out-of-network, or high-interference environment, it is difficult for the data collected on-site to enter the production system in a timely manner, resulting in a significant time lag between on-site discovery, on-site recording, and system acceptance.

[0057] In such a business environment, how to enable instant messaging terminals to handle more on-site processing capabilities and adapt to the timeliness, continuity and consistency requirements of power production operations has become a key issue in on-site digital collaboration.

[0058] In existing technologies, a common approach involves on-site personnel first recording problems using methods such as photos, voice recording, or manual input, and then submitting the reports and formally entering them into the database through different systems. For example, when maintenance personnel discover equipment overheating, unusual noises, or abnormal indicators during inspections, they often first notify the work team via chat or temporarily enter brief information in a standalone mobile application, and then return to the PC-based production system to supplement the information with the equipment number, anomaly category, location description, and handling suggestions. For business involving plan adjustments or process changes, it is usually necessary to re-enter the corresponding fields in another system, sometimes even supplementing with paper records for subsequent transcription. The working principle of this type of solution is essentially to separate on-site data collection from system processing, relying on manual data transfer between different platforms, and using manual judgment to determine the priority of business and the timing of submission. While this approach can cover daily production needs to a certain extent, its limitations are particularly prominent in the power field environment.

[0059] First, the lack of a unified data transfer mechanism between different systems means that on-site inputs such as voice, images, and text cannot be directly converted into business data that can be recognized by the production system. This leads to the need for the same event to be interpreted, filled out repeatedly, and reconfirmed multiple times. Information is prone to errors in equipment identification, inconsistent descriptions, or omissions during the relaying process. Second, when the on-site network is interrupted, independent applications often cannot continue submitting data, and messages in chat tools are difficult to directly convert into formal business information. Operators can only record these messages locally or on paper and then re-enter them after the network is restored. This not only increases the probability of forgetting, omissions, and duplicate reporting but also mixes truly urgent business with ordinary business, making it impossible to prioritize data for system processing based on risk level. Third, the data received by the back-end system is often detached from the original on-site context, affecting the continuity of business processing and preventing a smooth closed loop between front-end recording, back-end processing, and subsequent handling.

[0060] It is evident that while traditional technical solutions offer multiple tool entry points in form, they fail to address the issue of continuous processing of field operations in complex network environments, nor do they establish a stable data channel between multimodal inputs and production system execution. Consequently, they exhibit significant shortcomings in terms of efficiency, data consistency, and timeliness of business response.

[0061] In view of this, how to improve the continuity, consistency and reliability of power production business processing in complex field environments has become an urgent technical problem to be solved.

[0062] To address the aforementioned issues, this application provides a data processing method applicable to power production operations. By converting multimodal information into structured information, it standardizes the expression of operational instructions, enabling instant messaging terminals to directly drive the production system to execute business processes. This avoids repetitive data entry across terminals and systems, improving data consistency and processing efficiency. By determining the priority levels of structured information, it achieves hierarchical management of different business operations, providing a basis for subsequent priority-based business processing. Furthermore, by storing structured information in a local queue according to priority levels when a network interruption is detected, and synchronizing it to the production system according to priority levels when the network is restored, it ensures business continuity in weak network or network outage environments, preventing data loss. The method of this application improves the continuity, consistency, and compliance reliability of power production business processing in complex field environments.

[0063] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0064] Figure 1A flowchart illustrating a data processing method applicable to power generation operations provided in this application. Figure 1 This method is applied to instant messaging clients and includes:

[0065] S101. Responding to the user's input operation, acquire multimodal information; wherein, the multimodal information represents the operation instructions for power production operations;

[0066] S102. Convert multimodal information into structured information and determine the priority level of the structured information; wherein, the structured information is used to drive the production system to perform business processing;

[0067] S103. When a network interruption is detected, the structured information is stored in a local queue according to its priority level.

[0068] S104. When network recovery is detected, the structured information in the local queue is synchronized to the production system according to priority level.

[0069] The instant messaging client serves as the application platform for this method, acting as the entry point for user input operations and responsible for the subsequent acquisition, conversion, caching, and synchronization of multimodal information. In the power production field environment, the instant messaging client can be deployed on mobile terminals used by maintenance personnel, repair personnel, and on-duty personnel, or on handheld devices, tablets, or portable industrial terminals with communication capabilities, integrating on-site business entry points and instant communication capabilities into the same application platform.

[0070] In step S101, multimodal information can refer to input information used to characterize power production business operation instructions. As the raw data for subsequent structured processing, it can come from various input forms such as audio, images, and text. It can represent business instructions independently by a single input content, or it can represent the complete business intent by combining multiple input contents.

[0071] Responding to user input can be understood as the instant messaging terminal monitoring and recognizing operation events triggered by the user through the human-computer interaction interface, and calling the corresponding input interface to obtain multimodal information based on the operation event.

[0072] In practice, the instant messaging client can provide a unified input interface that integrates at least one interactive element from text input area, voice recording button, and image capture button. Users select the appropriate input method based on their on-site operational needs: when selecting text input, users type text content including business type identifier, device identifier, and operation parameters in the text input area; when selecting voice input, users long-press the voice recording button to record voice content, and the instant messaging client calls the terminal device's microphone to collect audio data and converts the audio data into text information through its built-in speech recognition engine; when selecting image input, users click the image capture button to activate the camera, take a photo of the on-site equipment, or scan the equipment's QR code / barcode. The instant messaging client performs optical character recognition on the captured image to extract device identification information, or parses the device identification information from the QR code / barcode. For combined input scenarios, users can first take a photo of the equipment and then add text descriptions; the instant messaging client associates and stores the image and text information together to represent the complete business operation instructions.

[0073] It should be understood that the input forms of multimodal information are not limited to the examples above, and may also include other forms such as video information, gesture information, and sensor data information, as long as they can represent the operational instructions for power production operations. Multimodal information in different input forms can be used individually or in combination, and this application does not limit this.

[0074] It should also be understood that the process of acquiring multimodal information may include preliminary verification of the user's identity, such as confirming the legitimacy of the current operator's identity through login status verification, biometric recognition, etc., to ensure the traceability of subsequent business operations. This identity verification step can be performed before or simultaneously with the acquisition of multimodal information, and this application does not limit this.

[0075] In step S102, structured information refers to a data structure organized according to a preset data format and containing clearly defined fields, used to carry standardized business operation instruction information. Structured information, through unified field naming and data types, eliminates format differences caused by different input formats, enabling instant messaging terminals and production systems to interact in a standardized data format, avoiding repeated cross-system data entry and information parsing obstacles.

[0076] Priority level can refer to a scheduling priority identifier determined based on factors such as business type, equipment level, and operational urgency. It is used to guide the sorting and synchronization order of structured information in the local queue, ensuring that urgent or important business operations are processed first.

[0077] The production system can refer to the cluster of information systems within a power company that support production operations, including but not limited to a power grid management platform for equipment management, a defect management system for defect tracking, a planning management system for scheduling, a work order management system for work order management, and a human resources management system for personnel qualification management. Each system interacts with instant messaging terminals through standardized interfaces.

[0078] In practice, the instant messaging client can have a built-in terminology template engine that maintains a terminology mapping table and a priority mapping table.

[0079] For example, the terminology mapping table is pre-configured with mapping relationships between business terminology keywords and structured fields. For instance, keywords such as "PLAN_CHG", "Plan Change", and "Change Plan" are mapped to the business type field value "Plan Change"; keywords such as "DEFECT", "Defect", and "Fault" are mapped to the business type field value "Defect Reporting"; and prefix identifiers such as "SW-" and "ID" are mapped to extraction rules for the device identifier field.

[0080] The priority mapping table is pre-configured with the mapping relationship between business types and priority levels. For example, "defect alarm" is mapped to priority 1 (highest), "work order change" is mapped to priority 2 (high), "plan change" is mapped to priority 3 (medium), and "equipment query" is mapped to priority 4 (low).

[0081] In one possible implementation, after acquiring the preprocessed multimodal information, the terminology template engine first performs word segmentation on the multimodal information to identify business terminology keywords. Then, it converts the business terminology keywords into corresponding structured field values ​​according to a terminology mapping table. Next, it queries a priority mapping table based on the identified business type field values ​​to determine the priority level of the structured information. Finally, it assembles the field values ​​and priority levels into structured information according to a preset data format (such as JSON). After the structured information is assembled, the instant messaging client can perform data integrity verification. Verification can include checking for missing required fields, conforming field formats, and ensuring field values ​​are within valid ranges. If the verification passes, the structured information is passed to subsequent processing stages; if the verification fails, an error message is returned to the user, prompting them to re-enter the information. This achieves standardized conversion from multimodal information to structured information, enabling the instant messaging client to directly drive the production system to execute business processes, avoiding information distortion caused by manual transcription and duplicate entry across systems, and improving data consistency and processing efficiency.

[0082] It should be understood that terminology mapping rules and priority mapping relationships can be configured and adjusted according to actual business needs. Different power companies or different business scenarios can set different mapping rules, and this application does not limit this. The fields included in the structured information can be added or removed according to the business type. For example, structured information for defect reporting can include location information fields, anomaly description fields, and photo attachment fields; structured information for plan changes can include change reason fields and impact scope fields. This application does not limit the specific field composition of the structured information. The terminology template engine can use rule-based segmentation, dictionary-based segmentation, or machine learning-based segmentation methods. This application does not limit the specific segmentation algorithm.

[0083] In step S103, the instant messaging client monitors the network connection status in real time. When a network interruption is detected, the structured information to be processed is stored in the corresponding position of the local queue according to the priority level of the structured information to be processed, so as to ensure that the business operation data is not lost during the network interruption and can continue to be processed in priority order after the network is restored.

[0084] A network interruption can refer to the instant messaging client determining, through active probing or passive listening, that the network connection between the current terminal device and the production system is unavailable or data transmission is blocked.

[0085] For example, the active probing method may include: the instant messaging client periodically sending network probing requests (such as Transmission Control Protocol (TCP) connection requests or Hypertext Transfer Protocol (HTTP) heartbeat requests) to a preset network detection address; if no response is received within a preset timeout period, the network is deemed to be interrupted. The preset network detection address may include at least one of the following: the server address of the production system, the server address of the instant messaging client, or a public network address; the preset timeout period can be configured according to the network environment, for example, set to 3 seconds, 5 seconds, or 10 seconds.

[0086] Passive monitoring methods may include: the instant messaging client listening for network status change events of the terminal device (such as operating system network status notifications, network interface status change events), and determining that the network is interrupted when the network status changes from connected to disconnected.

[0087] A local queue can refer to a data structure deployed in the local storage space of an instant messaging client, ordered according to priority, used to temporarily store structured information to be synchronized during network outages.

[0088] In one possible implementation, the local queue is implemented using a queue data structure that supports priority-based insertion and dequeue operations. Elements in the queue are arranged from highest to lowest priority, and elements of the same priority level are arranged in chronological order of their timestamps. Local storage space can include the terminal device's memory or persistent storage media (such as a local database or file system) to ensure that structured information is not lost due to application closure or device restart during network outages. The local queue supports concurrent access control; when multiple structured information items need to be enqueued simultaneously, a locking mechanism ensures the atomicity and consistency of queue operations.

[0089] It should be understood that network outage detection methods are not limited to periodically sending network probe requests or listening for network status change events. Other methods can also be used, such as monitoring the response status of production system interfaces, monitoring packet transmission failure rates, and monitoring whether network latency exceeds a threshold. When network probe requests occasionally fail but recover quickly, a fault tolerance mechanism can be set to avoid frequent switching of network status determination. For example, a network outage can be determined only after multiple consecutive probe failures, or a cooldown time for network status switching can be set (e.g., after a network outage is determined, a certain number of successful probes are required before network recovery is determined). This application does not limit the specific strategy for determining network outages. In addition, the priority sorting rules of local queues can be adjusted according to business needs. For example, in some scenarios, the priority of business related to plan changes can be increased to high priority, or the priority of the same business type can be further subdivided according to the equipment voltage level. This application does not limit the specific rules for priority sorting.

[0090] In step S104, the instant messaging terminal continuously monitors the network connection status. When the network is determined to be restored, the structured information in the local queue is synchronized to the production system in sequence according to the priority level of the structured information in the queue, so as to complete the follow-up processing of the business operations temporarily stored during the network interruption and ensure the final consistency and integrity of the business data.

[0091] Network recovery detection can refer to the instant messaging client determining, through active probing or passive monitoring, that the network connection between the current terminal device and the production system has been restored and the data transmission path has been re-established.

[0092] For example, the active detection method may include: the instant messaging terminal periodically sends network detection requests to a preset network detection address, and if a response is successfully received within a preset timeout period, the network is determined to have recovered.

[0093] Passive monitoring methods may include: the instant messaging client listening for network status change events of the terminal device, and determining that the network has recovered when the network status changes from disconnected to connected.

[0094] In one possible implementation, to improve the accuracy of the determination, a network recovery confirmation mechanism can be set up. For example, the network recovery can be finally confirmed only after multiple successful probes or after continuously monitoring the connection status for a certain period of time, thus avoiding frequent state switching caused by network fluctuations.

[0095] In one possible implementation, synchronizing structured information in the local queue to the production system according to priority levels can include:

[0096] The instant messaging client retrieves structured information from the head of its local queue in the order it is arranged, and sends it to the corresponding business interface in the production system via the network connection. Upon receiving the structured information, the production system parses the fields, executes the corresponding business processing logic, and returns the processing result to the instant messaging client. Upon receiving the processing result, the instant messaging client determines whether the processing was successful. If successful, it removes the structured information from its local queue; if processing fails or a timeout occurs without a result, it triggers a retransmission mechanism or exception handling procedure. This synchronization process is executed in descending order of priority, ensuring that high-priority business operations are processed and responded to by the production system first.

[0097] In practice, after detecting network recovery, the instant messaging client initiates a synchronization task, which can be executed according to the following process:

[0098] First, check if the local queue is empty. If it is empty, end the synchronization task. If it is not empty, retrieve the first structured information (i.e., the structured information with the highest priority) from the head of the queue. Send this structured information to the corresponding business interface of the production system via a network connection (e.g., a message queue protocol). Set a timeout period and wait for the production system to return the processing result. If a successful processing result is received within the timeout period, delete the structured information from the local queue and record a synchronization success log, then continue processing the next structured information in the queue. If no processing result is received within the timeout period or a failed processing result is received, determine if the number of retransmissions for this structured information has reached the preset maximum number of retransmissions. If not, increase the number of retransmissions and try sending again after a preset retransmission interval. If the preset maximum number of retransmissions has been reached, mark the structured information as abnormal, remove it from the local queue to the abnormal queue, and generate an alarm message to notify relevant personnel. The synchronization task can be configured to be triggered on a timed basis (e.g., check the network status and perform synchronization every 5 minutes), or it can be triggered by a network recovery event (i.e., start synchronization immediately after detecting network recovery), or it can be triggered manually by the user. This application does not limit this.

[0099] In one possible implementation, during the synchronization process, the instant messaging client can compress the structured information to reduce the amount of data transmitted over the network and improve synchronization efficiency; it can also encrypt the transmission process (e.g., using transport layer encryption or application layer encryption) to ensure data transmission security. For synchronization scenarios involving large amounts of structured information, a batch synchronization mechanism can also be used, which packages multiple pieces of structured information into a batch for transmission, reducing the number of network requests and increasing synchronization throughput. This application does not limit the specific compression, encryption, and batch processing methods.

[0100] This embodiment provides a data processing method suitable for power production operations. By converting multimodal information into structured information, it achieves standardized expression of operation instructions, enabling instant messaging terminals to directly drive the production system to execute business processes. This avoids repetitive data entry across terminals and systems, improving data consistency and processing efficiency. By determining the priority level of structured information, it achieves hierarchical management of different business operations, providing a basis for subsequent priority-based business processing. By storing structured information in a local queue according to priority level when a network interruption is detected, and synchronizing it to the production system according to priority level when the network is restored, it ensures the continuity of business in weak network or network outage environments and avoids data loss. The method of this application improves the continuity, consistency, and compliance reliability of power production business processing in complex field environments.

[0101] Figure 2 A flowchart illustrating a data processing method applicable to power generation operations provided in this application. Figure 2 , Figure 2 Implementation examples in Figure 1 Based on the examples, a data processing method applicable to power production operations is described in detail, such as... Figure 2 As shown, the method may include:

[0102] S201. In response to user input, obtain multimodal information; wherein, the multimodal information represents the operation instructions for power production operations.

[0103] S202. Based on the modality type of the multimodal information, convert the multimodal information into intermediate text; wherein the modality type includes one or more of audio type, image type and text type.

[0104] Modality type can refer to the original input form or data format category of multimodal information, used to distinguish the type of information input by the user through different sensory channels or interaction methods. For example, the modality type can be automatically identified and determined by the instant messaging client based on the type of input interface triggered by the user. For instance, input triggered by the user clicking the voice recording button is identified as audio, input triggered by the user clicking the image capture button is identified as image, and input triggered by the user typing content in the text input area is identified as text.

[0105] Intermediate text refers to string information uniformly represented in plain text format after modal transformation, serving as an intermediate data form connecting the original multimodal input and subsequent structured processing. Intermediate text retains the business semantics of the original input, eliminating differences in data format, encoding methods, and storage structures between different modalities, enabling subsequent processing to utilize a unified text parsing method.

[0106] Converting multimodal information into intermediate text aims to establish a standardized conversion channel from heterogeneous input to unified text representation, avoiding parsing complexity and compatibility issues caused by different modal information directly entering subsequent processing stages.

[0107] In one alternative implementation, step S202 may include:

[0108] S2021. When the modal type includes audio, the multimodal information is converted into intermediate text using a preset speech recognition technology.

[0109] The multimodal information of the audio type may include voice data recorded by the user through a microphone, pre-recorded audio files, etc., and its data format may be an audio encoding format.

[0110] In one possible implementation, after acquiring multimodal information of the audio type, the instant messaging client first preprocesses the multimodal information of the audio type, such as noise reduction (removing environmental noise), endpoint detection (identifying valid speech segments), and audio framing (segmenting continuous audio into short frame sequences), in order to improve the accuracy of subsequent speech recognition.

[0111] The pre-defined speech recognition technology can refer to an algorithm or system used to convert speech content in multimodal audio information into a text sequence. For example, the pre-defined speech recognition technology may include an end-to-end speech recognition model based on deep learning (e.g., a model based on the Transformer architecture), a traditional speech recognition engine based on a cascaded acoustic model and language model, or a cloud-based speech recognition application programming interface (API). The speech recognition technology can support the recognition of Mandarin Chinese, dialects, and power industry terminology, and can be domain-adapted to power production scenarios to improve the accuracy of terminology recognition.

[0112] For example, in a noisy substation environment, maintenance personnel input "Postpone Plan 207 to August 25th" via voice. The instant messaging client then uses the microphone of the terminal device to collect the audio data. After collection, the instant messaging client performs noise reduction processing on the audio data to remove background noise generated by the operation of substation equipment, and then extracts the valid speech segments through endpoint detection. The processed audio data is then input into a preset speech recognition engine. Based on acoustic and language models adapted for the power industry, the speech recognition engine recognizes the text sequence "Postpone Plan 207 to August 25th" as the intermediate text output.

[0113] S2022. When the modal type includes an image type, the multimodal information is converted into intermediate text using a preset image recognition technology.

[0114] Among them, multimodal information of image type can include photos of on-site equipment taken by the user through the camera, photos of equipment nameplates, QR code / barcode images, screenshots, etc., and its data format is image encoding format.

[0115] In one possible implementation, after acquiring multimodal information of the image type, the instant messaging client first preprocesses the multimodal information of the image type, such as image enhancement (adjusting brightness and contrast), image rotation correction, image scaling, etc., to improve the accuracy of subsequent image recognition.

[0116] Pre-defined image recognition technologies can refer to algorithms or systems used to extract text information from images or identify the business semantics corresponding to image content. For example, pre-defined image recognition technologies may include Optical Character Recognition (OCR) technology, QR code / barcode parsing technology (e.g., parsing algorithms based on barcode / QR code recognition libraries), and image classification and recognition technology (e.g., image classification models based on Convolutional Neural Networks (CNNs)). OCR technology is used to recognize printed or handwritten text in images, QR code / barcode parsing technology is used to quickly extract information such as device identifiers from encoded images, and image classification and recognition technology is used to identify the business scenario type to which the image content belongs (e.g., device malfunction type, operating environment type).

[0117] For example, an inspector discovers an abnormal temperature in a switchgear cabinet. They take a photo of the equipment nameplate via instant messaging. The photo contains the printed text "Equipment Model: SW-35kV-29", "Rated Voltage: 35kV", and "Installation Location: #2 Substation". The instant messaging client performs image enhancement processing on the photo to improve text clarity. Then, it uses an OCR engine to recognize the text content in the photo, extracting the key information "SW-35kV-29" and "#2 Substation". Combined with the user-added text description "Temperature Anomaly", the intermediate text "Equipment SW-35kV-29 located in #2 Substation, temperature measurement abnormality" is generated.

[0118] S2023. When the modal type includes text type, determine the multimodal information as intermediate text.

[0119] Multimodal text information can include text content typed directly by the user in the text input area, pasted text content, or text content shared through other applications. Since multimodal text information is itself in text format, no modality conversion is required, and it can be used directly as intermediate text. However, instant messaging clients can also preprocess multimodal text information, such as removing leading and trailing whitespace, standardizing character encoding, and normalizing punctuation marks, to ensure consistent formatting of the intermediate text.

[0120] For example, if a user types "PLAN_CHG:ID207_Delayed to 0825" in the text input area, the instant messaging client will directly use this text as the intermediate text "PLAN_CHG:ID207_Delayed to 0825". Alternatively, if a user copies the text "Defect Report: SW-35kV-29 Temperature Measurement Abnormality" from another application and pastes it into the input area, the instant messaging client will remove leading and trailing spaces and use "Defect Report: SW-35kV-29 Temperature Measurement Abnormality" as the intermediate text.

[0121] It is understandable that by identifying different modal types and using corresponding conversion techniques to process them into unified intermediate text, the differences in format, encoding, and expression of multimodal information can be eliminated, thereby providing a consistent input basis for subsequent processing steps and achieving standardized conversion of multimodal input to unified text representation.

[0122] S203. Convert the intermediate text into structured information.

[0123] The process of converting intermediate text into structured information aims to establish a conversion channel from natural language descriptions to standardized data structures that can be parsed by machines and executed by systems.

[0124] For example, the conversion process can be based on a preset terminology keyword library and terminology template set. Through steps such as semantic parsing, keyword matching, parameter extraction, and template filling, the free-format intermediate text can be converted into structured information with fixed field definitions and data formats, enabling the production system to directly identify business intent and execute corresponding processing logic.

[0125] In one alternative implementation, step S203 may include:

[0126] S2031. Determine the business type corresponding to the intermediate text based on the preset terminology and keyword library; wherein, the preset terminology and keyword library includes the business type and the set of keywords corresponding to the business type.

[0127] The pre-set terminology keyword library can refer to the basic data resources that are pre-configured and stored locally on the instant messaging client or on the server side for business type identification.

[0128] For example, the terminology keyword database is organized in the form of key-value pairs or relational tables, where the key is a business type identifier and the value is a set of keywords corresponding to that business type. The keywords in the keyword set can include exact matching keywords (such as the technical term abbreviation "DEFECT") and fuzzy matching keywords (such as synonyms like "change," "modify," and "adjust"), and support adding, deleting, modifying, querying, and dynamically expanding keywords. The terminology keyword database can be regularly updated and maintained according to the development and changes in power production operations.

[0129] In one possible implementation, determining the business type corresponding to the intermediate text based on a pre-defined terminology keyword library may include:

[0130] The intermediate text is segmented into words to obtain the segmentation results; the segmentation results are matched with the keyword sets of each business type in the terminology keyword library to calculate the matching score; the business type with the highest matching score is determined as the business type corresponding to the intermediate text.

[0131] The matching methods can include exact matching (the word segmentation result is completely consistent with the keyword) and fuzzy matching (the word segmentation result is more similar to the keyword than a preset threshold).

[0132] Optionally, the business type includes at least one of production planning, equipment defect, and work permit categories;

[0133] Among them, the production plan category represents the formulation and modification of power production tasks, the equipment defect category represents the reporting and tracking of abnormalities in power equipment, and the work permit category represents the safety permits for power operations and personnel changes.

[0134] For example, the production planning category covers the entire lifecycle management operations of various production tasks in power grid operation, including planning, approval, issuance, execution, and modification. Examples include the formulation of annual maintenance plans, the arrangement of monthly defect elimination plans, the addition of temporary emergency repair tasks, and the time or content adjustment of issued plans. The equipment defect category covers the entire process tracking management operations of power equipment from defect discovery, reporting, classification, dispatching, handling, acceptance to closed-loop archiving. Examples include the reporting of abnormal transformer temperatures, the tracking of circuit breaker operation exceeding limits, and the handling and feedback of line insulator damage. The work permit category covers the application, approval, modification, extension, and termination of safety permits for power field operations. Examples include the application for substation maintenance work tickets, the permission for changes in work crew members, the approval of work time extensions, and the verification of operator qualifications.

[0135] The three business types mentioned above cover the core business scenarios in power production sites. Production planning focuses on the timing of production tasks and resource allocation; equipment defect management focuses on the monitoring and remediation of equipment health; and work permit management focuses on safety control and personnel management for on-site operations. There are business relationships among these three types. For example, handling equipment defects may trigger plan changes, and applying for a work permit may depend on the status of the plan's issuance. The terminology keyword library and terminology template set can support cross-business type related processing.

[0136] Different business types are configured with different keyword sets in the terminology keyword library to achieve differentiated structured processing. The initial priority mapping relationship of each business type can also be different. For example, the initial mapping of equipment defects is high priority, the initial mapping of work permits is high priority, and the initial mapping of production plans is medium priority. However, the specific priority level can be dynamically adjusted according to the business type's score.

[0137] In one possible implementation, the production planning class can further include subtypes such as plan addition (representing the addition of new production task plans), plan change (representing the adjustment of the time, content, or personnel of an existing plan), plan cancellation (representing the termination of a plan in execution), and plan query (representing the retrieval of plan execution status); the equipment defect class can further include subtypes such as defect reporting (representing the initial reporting after the discovery of equipment abnormalities), defect tracking (representing the status update during defect handling), defect acceptance (representing the acceptance confirmation after defect repair), and defect statistics (representing the summary analysis of defect data); the work permit class can further include subtypes such as work order application (representing the application for a new work permit), work order change (representing the change of permit content or personnel), work order extension (representing the extension of the permit validity period), and qualification verification (representing the real-time verification of the qualifications of operators). Each subtype has a more targeted set of keywords in the terminology keyword library to support more refined business type identification. For example, the keyword set for the plan change subtype includes "delay", "early", "rescheduling", "PLAN_CHG", etc., and the keyword set for the defect reporting subtype includes "abnormal temperature measurement", "discharge", "insulation damage", "DEFECT_REPORT", etc.

[0138] The method for determining subtypes includes: after identifying the major business type category, further matching the word segmentation results with the keyword sets of each subtype under that category, and determining the subtype with the highest matching score as the subtype corresponding to the intermediate text. It can be understood that determining the subtype helps to match more accurate terminology templates subsequently, improving the accuracy of structured information generation.

[0139] It is understandable that by using a terminology keyword library and terminology template set that cover the core business types (production planning, equipment defects, work permits) in the power production field, it is possible to support multimodal input processing and standardized conversion of the main business scenarios in power production. This enables the instant messaging terminal to serve as a unified business entry point to handle diverse field operation needs, thereby achieving the centralization and standardization of the power production business processing entry point.

[0140] For example, the terminology keyword library is configured as follows: For business types such as production planning, the keyword set includes "plan", "change", "delay", "add", "cancel", "PLAN", "PLAN_CHG", "PLAN_ADD", etc.; For business types such as equipment defects, the keyword set includes "defect", "fault", "abnormality", "damage", "DEFECT", "DEFECT_REPORT", "temperature measurement abnormality", "discharge", etc.; For business types such as work permits, the keyword set includes "work ticket", "permit", "qualification", "personnel change", "TICKET", "TICKET_CHG", "double review", etc.

[0141] For example, based on the above configuration of the terminology keyword library, for the intermediate text "delaying plan 207 to August 25", the word segmentation results can be "will", "plan", "207", "delay", "to", "August 25". The two keywords "plan" and "delay" are matched with the production planning keyword set, and the matching degree score is the highest, so the business type is determined to be production planning.

[0142] It should be understood that the configuration of the terminology keyword database is not limited to the above examples. Regular expression rules, semantic vector similarity calculations, and machine learning-based text classification models can also be used for business type identification. This application does not limit the specific business type identification algorithm. The terminology keyword database can be deployed locally on the instant messaging client to support offline identification, or it can be deployed on the server to support more complex identification algorithms. This application does not limit the deployment location of the terminology keyword database.

[0143] S2032. From the preset set of terminology templates, determine the terminology template corresponding to the business type as the target template; wherein, the terminology template includes multiple data units used to carry information elements, and conversion logic used to establish the association between the data units and the intermediate text.

[0144] This includes a pre-configured data resource set containing terminology templates for each business type (or subtype), stored locally on the instant messaging client or server. The terminology template set is organized by business type, supporting quick retrieval of corresponding terminology templates by business type. The terminology template set is updated using both full and incremental updates, and the latest template version can be synchronized from the server via the network to ensure synchronization between templates and business rules. Inheritance relationships can exist between terminology templates in the set; for example, a subtype template inherits the common data units of its parent type and extends specific data units on top of this, reducing template redundancy and improving maintenance efficiency.

[0145] A terminology template refers to a predefined data structure template for a specific business type, specifying the field composition, field naming, data type, value range, and logical relationships between fields that should be included in structured information under that business type. For example, a terminology template includes two core parts: first, the data unit definition, which is the metadata description of each field in the structured information; and second, the transformation logic definition, which is the specific rules and methods for extracting field values ​​from intermediate text.

[0146] Data units can be used to carry specific information elements, with each data unit corresponding to a field in structured information. For example, the metadata description of a data unit may include: unit identifier (field name), unit name (field name in Chinese), data type (e.g., string, number, date, enumeration value, etc.), whether it is required, default value, value range, validation rules, etc.

[0147] Transformation logic can be used to establish the association between data units and intermediate text, defining how to parse and extract the values ​​of corresponding data units from the intermediate text. For example, transformation logic may include regular expression matching rules, keyword position extraction rules, delimiter segmentation rules, semantic role labeling rules, etc. Optionally, transformation logic supports the combined use of multiple extraction methods, such as first locating the target segment using regular expressions and then extracting specific values ​​using keywords, to improve the accuracy and robustness of the extraction.

[0148] S2033. Based on the target template, extract the target operation parameters from the intermediate text and fill them into the target template to generate structured information.

[0149] Among them, target operation parameters refer to key parameter information directly related to business operations extracted from intermediate text, and are the core data content driving the production system to execute business processes. For example, target operation parameters include specific business elements such as equipment identification, operation type, operation time, operator, change content, and anomaly description. The accuracy of their extraction directly affects the correctness and effectiveness of subsequent business processing in the production system. The extraction process of target operation parameters follows the transformation logic defined in the target template, accurately extracting them from the intermediate text through methods such as regular expression matching, keyword positioning, and delimiter segmentation.

[0150] In specific implementation, step S2033 may include: traversing each data unit defined in the target template, parsing and extracting the intermediate text according to the conversion logic corresponding to each data unit to obtain the specific value of each data unit (i.e., the target operation parameter); filling the extracted target operation parameters into the corresponding data units of the target template; performing integrity verification on all data units, including whether required fields have been filled, whether the field format conforms to the specifications, whether the field values ​​are within the valid range, and whether the logical relationship between fields meets the requirements; after the verification passes, assembling each data unit and its value into structured information according to the preset data format. If the integrity verification fails, the reason for the verification failure is recorded, and an error message is returned to the user, prompting them to supplement the missing information or re-enter it.

[0151] For example, for the intermediate text "Delay Plan 207 to August 25th", the target template is a production plan template. The target operation parameters are extracted according to the template conversion logic: the regular expression matching rule for the device_id unit is "ID\d+", extracting the target operation parameter "ID207" from "Plan 207"; the keyword extraction rule for the operation_param unit is "delay to" followed by a date, extracting the target operation parameter "delay to 0825" from "delay to August 25th"; the timestamp unit takes the current system time as the target operation parameter "2026-08-01 09:30:00"; and the operator_id unit takes the currently logged-in user identifier as the target operation parameter "USER_001". The above target operation parameters are populated into each data unit of the template, and integrity checks are performed: device_id is required and has been filled with "ID207", and the format conforms to the string type requirements; operation_param is required and has been filled with "delayed to 0825", and the format conforms to the string type requirements; timestamp is required and has been filled with "2026-08-01 09:30:00", and the format conforms to the date and time type requirements; operator_id is required and has been filled with "USER_001", and the format conforms to the string type requirements. All required fields are populated and the format conforms to the requirements; the check passes, and structured information is generated.

[0152] It should be understood that the data format used to generate structured information can include, but is not limited to, lightweight data exchange formats (JSON format), extensible markup language formats (XML format), formats used to express data serialization (YAML format), key-value pair formats, etc., as long as they can be parsed and processed by the production system. After the structured information is generated, it can be compressed or encrypted to reduce the amount of data transmitted or improve data security. This application does not limit the specific compression and encryption methods.

[0153] It is understandable that by using target templates corresponding to business types to extract and fill target operation parameters to generate structured information, it is possible to ensure that the structured information of different business types has a unified field specification and data format. This enables the production system to directly parse the field content and call the corresponding business processing logic without the need for manual secondary transcription or format conversion, thereby achieving automated and standardized conversion of multimodal inputs to standardized business instructions.

[0154] It is understandable that by identifying modal types and converting them into unified intermediate text, and then identifying business types based on a terminology keyword library and matching corresponding terminology templates for structured processing, structured information can be obtained. This allows for the establishment of a complete conversion link from raw multimodal inputs to executable instructions in the production system, thereby eliminating format barriers and parsing obstacles caused by different input formats and achieving standardized, automated, and intelligent processing of power production business operation instructions.

[0155] S204. Determine the score value of the business type based on the business type corresponding to the structured information; wherein, the score value represents the urgency of the impact of the structured information on the operational safety of the production system.

[0156] In one possible implementation, step S204 may include: extracting business type fields (e.g., "business_type" and "sub_type") from structured information; querying a preset business type scoring mapping table to obtain the initial score value corresponding to the business type; and adjusting the initial score value based on additional factors in the structured information (e.g., equipment voltage level, equipment importance level, operation time window, etc.) to obtain the final score value. The score value can be represented numerically, with a larger value indicating a higher urgency of the impact on the safety of the production system's operation, requiring priority handling.

[0157] The business type scoring mapping table can be pre-configured with the mapping relationship between each business type and the initial score value. The initial score value is set based on the potential impact of the business type on power grid security, equipment security, and personnel safety. For example, business types involving equipment failures that directly threaten the stable operation of the power grid are assigned a higher initial score value.

[0158] For example, the adjustment methods for additional factors may include: setting adjustment coefficients based on equipment voltage levels (e.g., an adjustment coefficient of 2.0 for 1000kV equipment, 1.5 for 500kV equipment, and 1.3 for 220kV equipment); setting adjustment coefficients based on equipment importance levels (e.g., a transformer coefficient of 1.5, a line coefficient of 1.3, and a switch coefficient of 1.2); setting time urgency coefficients based on the operation time window (e.g., a coefficient of 1.5 for less than 1 hour before the deadline, 1.3 for less than 4 hours, and 1.1 for less than 24 hours); and setting status coefficients based on equipment operating status (e.g., a coefficient of 1.0 for equipment that is out of service, and 1.4 for equipment operating with defects). Finally, the score for determining the business type can be the product of the initial score, voltage level coefficient, importance level coefficient, time urgency coefficient, and status coefficient.

[0159] S205. Determine the priority level of structured information based on the score.

[0160] In one possible implementation, step S205 may include: comparing the calculated score with a preset priority threshold range to determine the priority range into which the score falls, thereby determining the corresponding priority level. The priority threshold range may be pre-configured with score ranges for each priority level, and the priority level is positively correlated with the score; the higher the score, the higher the priority level.

[0161] In one possible implementation, priority levels are represented using hierarchical coding, such as Priority 1 (highest priority), Priority 2 (highest priority), Priority 3 (medium priority), and Priority 4 (lowest priority). The threshold range for the score value corresponding to each priority level can be configured and adjusted according to actual business needs. For example, the score value range for Priority 1 is ≥100, the score value range for Priority 2 is 70-99, the score value range for Priority 3 is 40-69, and the score value range for Priority 4 is <40.

[0162] It is understandable that by determining the priority level of structured information based on business type, priority allocation can be more aligned with the actual safety risk situation of power production business, thereby ensuring that emergency business is handled first and achieving dynamic matching and precise scheduling between business priority and safety risk.

[0163] S206. When a network interruption is detected, the structured information is stored in a local queue according to its priority level.

[0164] S207. When network recovery is detected, the structured information in the local queue is synchronized to the production system according to priority level.

[0165] In an optional implementation, step S207 may include:

[0166] S2071. Extract the structured information to be synchronized from the local queue in descending order of priority.

[0167] In practice, the instant messaging client can check the current status of the local queue and obtain the priority level of all structured information in the queue. It then sorts the structured information according to priority level from highest to lowest, and for structured information with the same priority level, arranges them according to the order of their enqueue timestamps. Structured information is then retrieved sequentially from the head of the sorted queue as the information to be synchronized. This extraction process ensures that the highest priority structured information is processed first, enabling urgent business involving equipment and personnel safety to be synchronized and processed as soon as the network is restored.

[0168] S2072. Send the extracted structured information to the production system and receive the synchronization results returned by the production system.

[0169] In practice, the instant messaging client can determine the target production system corresponding to the structured information based on the business type field in the structured information (e.g., equipment defect type corresponds to the defect management system, production plan type corresponds to the planning management system, and work permit type corresponds to the work ticket management system). It then sends the structured information to the corresponding business interface of the target production system via a network connection (e.g., message queue protocol). The client waits for the production system to receive and process the structured information, and receives the synchronization result returned by the production system. The synchronization result can include two states: synchronization success and synchronization failure. Synchronization success indicates that the production system has successfully received and correctly processed the structured information, while synchronization failure indicates that the production system failed to process the information, there was a network transmission error, or the interface returned an error.

[0170] For example, the instant messaging client extracts structured information with priority level 1. Based on the "business_type" field being "equipment defect type," it determines the target production system to be the defect management system. This structured information is then sent to the defect creation interface of the defect management system via a message queue protocol. Upon receiving the information, the defect management system parses the field content, creates a defect work order "DEF-20250801-001," and returns a synchronization success result, including the work order number and processing timestamp. The instant messaging client receives and parses this synchronization result, confirming successful synchronization.

[0171] In one alternative implementation, before sending the extracted structured information to the production system, the following may also be included:

[0172] The structured information is validated to obtain a validation result; the validation process includes at least one of operation qualification validation, operation permission validation, and operation environment validation; if the validation result indicates that the validation fails, an alarm message is generated.

[0173] The verification result refers to the judgment conclusion obtained after performing multi-dimensional compliance checks on structured information, which is used to characterize whether the business operation corresponding to the structured information meets the execution conditions.

[0174] For example, the verification result can include two states: pass and fail. If the verification result is pass, it indicates that the current operator has the corresponding qualifications, has the operation permission, and the operation environment meets the requirements, allowing the structured information to be sent to the production system. If the verification result is fail, it indicates that at least one verification requirement is not met, prohibiting the structured information from being sent to the production system and triggering an alarm.

[0175] For example, the verification results can be recorded in the form of Boolean values, status codes, or structured data, and can include detailed judgment information for each verification item and the reason for failure.

[0176] In practice, the verification process can be executed locally on the instant messaging client, either after extracting structured information from the local queue and before sending it to the production system. In one possible implementation, the verification process can synchronously block subsequent sending (i.e., sending only continues after verification passes) or asynchronously interrupt the sending process if verification fails. The verification result can be persistently stored in a local log for subsequent auditing, tracing, and security analysis.

[0177] Operational qualification verification refers to the process of checking whether the current operator possesses the professional qualification certificates required to perform the business operation. This can include at least one of the following: certificate existence verification, certificate validity verification, and certificate scope of business verification. Qualification certificates can include legally required qualifications for power production operations, such as high-voltage electrician operation certificates, live-line working certificates, work permit issuer qualifications, and work supervisor qualifications. Specifically, operational qualification verification can be performed in real-time by calling the relevant human resources management system or qualification management system interface, or by querying locally cached qualification data when the network is interrupted.

[0178] For example, the instant messaging client extracts structured information about a work order change, where the operator_id is "USER_003". The instant messaging client calls the human resources management system interface to query the qualification status of "USER_003": it returns the high-voltage electrician's operating certificate (certificate number: HVD-2020-00123, validity period: 2020-06-01 to 2030-12-31) and the work order issuer's qualification (certificate number: WPI-2021-00456, validity period: 2021-03-01 to 2030-06-30). The current date is 2026-08-01, both qualification certificates are valid, and the operating qualification verification is passed.

[0179] Operation permission verification refers to the process of checking whether the current operator has the system permissions to perform a specific business operation. This can include at least one of the following: business type permission verification, equipment scope permission verification, and data scope permission verification. Specifically, business type permission verification checks whether the operator has the necessary permissions for that business type (e.g., work order modification, defect reporting); equipment scope permission verification checks whether the operator's permissions cover the target equipment (e.g., a specific substation, a specific line); and data scope permission verification checks whether the operator has the right to access or modify target data (e.g., a specific work order, a specific plan). Operation permission verification can be implemented by querying the permission management system or the production system's permission interface.

[0180] Operational environment verification refers to the process of checking environmental factors such as time conditions, spatial conditions, and network conditions of the current operation. For example, it may include at least one of the following: operation time window verification, operation geographical location verification, and network security environment verification. Specifically, operation time window verification checks whether the current time is within the permitted time period for the business operation (e.g., some high-risk operations are only allowed during daytime hours); operation geographical location verification checks whether the current geographical location is within the permitted operating area of ​​the target equipment or work site (e.g., spatial verification using Global Positioning System (GPS), Bluetooth beacon positioning, or base station positioning); and network security environment verification checks whether the current network connection is accessed through a secure channel (e.g., a Virtual Private Network (VPN)) to the production management network.

[0181] In one possible implementation, generating alarm information may include: when a verification result indicates a failure, the instant messaging client generates structured alarm information based on the failed verification item and the specific reason. The alarm information may include alarm type (e.g., qualification anomaly, permission anomaly, environmental anomaly), alarm level (e.g., urgent, important, general), failed verification item, operator information, target business information, suggested handling measures, and a generation timestamp. The alarm information can be sent to preset recipients (e.g., duty officers, security managers, and superiors) via the instant messaging client's push notification function. Simultaneously, an alarm log is recorded locally on the instant messaging client and can be synchronized to the security audit system according to configured policies.

[0182] It should be understood that the specific verification items in the verification process can be flexibly configured and expanded according to business type, enterprise security strategy, and regulatory requirements. For example, for business operations involving dual-person review, verification with both people present simultaneously can be added; for operations involving special equipment, verification of special equipment operator's certificates can be added; for operations involving confined space operations, verification of gas detection qualification can be added. This application does not limit the specific combination of verification items in the verification process. The verification process can be executed independently locally on the instant messaging terminal, or it can be executed collaboratively with external systems such as production systems, human resources systems, and access control systems. This application does not limit the executing entity or the invocation method of the verification process. Alarm information can be sent via instant messaging push, SMS notification, email notification, voice call notification, etc. This application does not limit the specific method of sending alarm information.

[0183] It should also be understood that, in a network outage environment, operation qualification verification and operation permission verification can query qualification and permission data cached locally on the instant messaging client (such as valid qualification information synchronized within the last 30 days and permission configuration snapshots) to support compliance checks in an offline environment. Local cached data can be set with an expiration date, after which a prompt will be displayed to resynchronize the latest data once the network is restored. This application does not limit the update strategy or expiration date settings for locally cached data.

[0184] It is understandable that by performing multi-dimensional verification of operational qualifications, operational permissions, and operational environment on structured information before synchronization, and embedding a compliance check mechanism to intercept non-compliant operations, the compliant execution of power production operations can be forcibly guaranteed from a technical perspective. This can effectively prevent safety risks such as unqualified operations, unauthorized operations, and operations in violation of regulations, thereby reducing the probability of safety accidents caused by human error and violations, and improving the stability, reliability, and security of the execution of procedures.

[0185] S2073. When the synchronization result indicates that synchronization has failed, the corresponding structured information is stored back into the local queue, and the step of synchronizing the structured information in the local queue to the production system according to the priority level is re-executed when network recovery is detected.

[0186] In one possible implementation, when the synchronization result indicates a synchronization failure, the following scenarios may be included: the production system returns an explicit processing failure response (e.g., qualification error, permission error, environment error, etc.); the network transmission times out and no response is received from the production system; the network is interrupted again during the synchronization process, resulting in a connection error; the response returned by the production system fails to be parsed or lacks the necessary success identifier.

[0187] Furthermore, when the synchronization result indicates a synchronization failure, the corresponding structured information is re-stored into the local queue. This can include: the instant messaging client re-inserting the structured information into the corresponding position in the local queue according to its current priority level.

[0188] In one possible implementation, when structured information is re-queued, its priority level can be adjusted by comprehensively considering factors such as the number of retransmissions and the reasons for failure (e.g., structured information that has been retransmitted multiple times but still failed can have its priority appropriately increased, or failures explicitly rejected by the production system can maintain their original priority and wait for manual intervention). After re-queuing, the metadata of the structured information is updated (e.g., adding the number of retransmissions, recording the last failure time, recording the reason for failure, etc.) for reference during subsequent synchronization.

[0189] Furthermore, the synchronization steps are re-executed, forming a cyclical processing mechanism. For example, after processing the current structured information, the instant messaging client checks if the local queue is empty. If the queue is not empty, it retrieves the next structured information to be synchronized in descending priority order, repeating the sending, receiving synchronization results, and judgment processing flow until the local queue is empty. This cyclical processing mechanism ensures that the structured information in the local queue is continuously processed after network recovery, until all synchronization is successful or the process transitions to an exception handling flow.

[0190] It is understandable that by storing structured information back into the local queue when synchronization fails, and retrying synchronization according to priority when the network recovers or in the next synchronization cycle, it can be ensured that business data will not be lost due to synchronization failures caused by temporary network fluctuations, brief unavailability of production systems, or momentary failures. This ensures the final consistency and integrity of business operations, avoids data loss or business omissions, and achieves reliable data transmission and business continuity assurance in unstable network environments.

[0191] It should be understood that when the network connection is normal, the instant messaging client can directly send structured information to the production system without going through the local queue cache, and complete business processing and data synchronization in real time, realizing real-time synchronization in online mode.

[0192] In this embodiment, based on the above embodiments, on the one hand, the problem of multimodal input standardization is solved by converting multimodal information into intermediate text according to the modality type, and then converting it into structured information based on term templates; on the other hand, the priority level is determined based on the business type score, and qualification, permission, and environment verification mechanisms are embedded before synchronization, which further improves the rationality of priority allocation and the compliance of business operations; thereby achieving the effects of continuity, consistency, and compliance reliability in power production business processing under complex field environments, while reducing the risk of human error and illegal operation.

[0193] Figure 3 A schematic diagram of a data processing device applicable to power production operations is provided in this application, as shown below. Figure 3 As shown, the data processing device 30 suitable for power production operations includes: a response module 301, a conversion module 302, a cache module 303, and a recovery module 304.

[0194] The response module 301 is used to respond to user input operations and acquire multimodal information; wherein, the multimodal information represents the operation instructions of power production business;

[0195] The conversion module 302 is used to convert multimodal information into structured information and determine the priority level of the structured information; wherein, the structured information is used to drive the production system to perform business processing;

[0196] The caching module 303 is used to store structured information into a local queue according to the priority level of the structured information when a network interruption is detected.

[0197] Recovery module 304 is used to synchronize structured information in the local queue to the production system according to priority level when network recovery is detected.

[0198] In an optional example, conversion module 302 is also used for:

[0199] Based on the modality type of the multimodal information, the multimodal information is converted into intermediate text; wherein the modality type includes one or more of audio type, image type and text type;

[0200] Convert intermediate text into structured information.

[0201] In an optional example, conversion module 302 is also used for:

[0202] When the modal type includes audio, the multimodal information is converted into intermediate text using a preset speech recognition technology;

[0203] When the modal type includes images, multimodal information is converted into intermediate text using preset image recognition technology;

[0204] When the modal type includes text, the multimodal information is determined to be intermediate text.

[0205] In an optional example, conversion module 302 is also used for:

[0206] Based on a pre-defined terminology and keyword database, determine the business type corresponding to the intermediate text; wherein, the pre-defined terminology and keyword database includes the business type and the set of keywords corresponding to the business type;

[0207] From the preset set of terminology templates, determine the terminology template corresponding to the business type as the target template; wherein, the terminology template includes multiple data units used to carry information elements, and conversion logic used to establish the association between the data units and the intermediate text;

[0208] Based on the target template, extract the target operation parameters from the intermediate text and populate them into the target template to generate structured information.

[0209] In an optional example, conversion module 302 is also used for:

[0210] Based on the business type corresponding to the structured information, a score value for the business type is determined; whereby the score value represents the urgency of the impact of the structured information on the operational security of the production system.

[0211] The priority level of structured information is determined based on the score.

[0212] In one optional example, the business type includes at least one of the following: production planning, equipment defect, and work permit.

[0213] Among them, the production plan category represents the formulation and modification of power production tasks, the equipment defect category represents the reporting and tracking of abnormalities in power equipment, and the work permit category represents the safety permits for power operations and personnel changes.

[0214] In an optional example, recovery module 304 is also used for:

[0215] Structured information to be synchronized is extracted from the local queue in descending order of priority.

[0216] The extracted structured information is sent to the production system, and the synchronization results returned by the production system are received.

[0217] When the synchronization result indicates a synchronization failure, the corresponding structured information is re-stored into the local queue, and the step of synchronizing the structured information in the local queue to the production system according to priority level is re-executed when network recovery is detected.

[0218] In an optional example, recovery module 304 is also used for:

[0219] The structured information is validated to obtain the validation result; the validation process includes at least one of operation qualification validation, operation permission validation, and operation environment validation.

[0220] If the verification result fails, an alarm message will be generated.

[0221] The data processing device applicable to power production operations in this embodiment can execute the technical solutions in the above method. Its specific implementation process and technical principles are the same, and will not be repeated here.

[0222] Figure 4 A schematic diagram of the structure of an electronic device provided in this application, such as... Figure 4 As shown, the electronic device 40 provided in this embodiment includes at least one processor 401 and a memory 402. Optionally, the electronic device 40 further includes a communication component 403. The processor 401, memory 402, and communication component 403 are connected via a bus 404.

[0223] In a specific implementation, at least one processor 401 executes computer execution instructions stored in memory 402, causing at least one processor 401 to perform the above-described method.

[0224] The specific implementation process of processor 401 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0225] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0226] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0227] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0228] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0229] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0230] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as magnetic disks or optical disks.

[0231] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A data processing method applicable to power production operations, characterized in that, Applied to instant messaging clients, the method includes: In response to user input, multimodal information is acquired; wherein the multimodal information represents the operation instructions for power production operations; The multimodal information is converted into structured information, and the priority level of the structured information is determined; wherein, the structured information is used to drive the production system to perform business processing; When a network interruption is detected, the structured information is stored in a local queue according to its priority level. When network recovery is detected, the structured information in the local queue is synchronized to the production system according to the priority level.

2. The method according to claim 1, characterized in that, The step of converting the multimodal information into structured information includes: Based on the modality type of the multimodal information, the multimodal information is converted into intermediate text; wherein, the modality type includes one or more of audio type, image type, and text type; The intermediate text is converted into the structured information.

3. The method according to claim 2, characterized in that, The step of converting the multimodal information into intermediate text based on the modality type of the multimodal information includes: When the modal type includes the audio type, the multimodal information is converted into the intermediate text using a preset speech recognition technology; When the modal type includes the image type, the multimodal information is converted into the intermediate text using a preset image recognition technology; When the modality type includes the text type, the multimodal information is determined to be the intermediate text.

4. The method according to claim 2, characterized in that, The step of converting the intermediate text into the structured information includes: The business type corresponding to the intermediate text is determined based on a preset terminology and keyword database; wherein the preset terminology and keyword database includes business types and a set of keywords corresponding to the business types. From a preset set of terminology templates, a terminology template corresponding to the business type is determined as the target template; wherein, the terminology template includes multiple data units for carrying information elements, and conversion logic for establishing the association between the data units and the intermediate text; Based on the target template, target operation parameters are extracted from the intermediate text and filled into the target template to generate the structured information.

5. The method according to claim 4, characterized in that, Determining the priority level of the structured information includes: Based on the business type corresponding to the structured information, a score value for the business type is determined; wherein, the score value represents the urgency of the impact of the structured information on the operational safety of the production system. The priority level of the structured information is determined based on the score.

6. The method according to claim 4, characterized in that, The business types include at least one of production planning, equipment defect, and work permit types; Among them, the production plan category represents the formulation and modification of power production tasks, the equipment defect category represents the reporting and tracking of abnormalities in power equipment, and the work permit category represents the safety permits for power operations and personnel change operations.

7. The method according to any one of claims 1-6, characterized in that, The step of synchronizing the structured information in the local queue to the production system according to the priority level includes: Structured information to be synchronized is extracted sequentially from the local queue in descending order of priority level. The extracted structured information is sent to the production system, and the synchronization result returned by the production system is received. When the synchronization result indicates a synchronization failure, the corresponding structured information is re-stored into the local queue, and the step of synchronizing the structured information in the local queue to the production system according to the priority level is re-executed when network recovery is detected.

8. The method according to claim 7, characterized in that, Before sending the extracted structured information to the production system, the method further includes: The structured information is validated to obtain a validation result; wherein the validation process includes at least one of operation qualification validation, operation permission validation, and operation environment validation. If the verification result fails, an alarm message is generated.

9. A data processing device suitable for power generation operations, characterized in that, include: The response module is used to respond to user input operations and acquire multimodal information; wherein, the multimodal information represents the operation instructions for power production operations; A conversion module is used to convert the multimodal information into structured information and determine the priority level of the structured information; wherein the structured information is used to drive the production system to perform business processing; A caching module is used to store the structured information into a local queue according to the priority level of the structured information when a network interruption is detected. The recovery module is used to synchronize the structured information in the local queue to the production system according to the priority level when network recovery is detected.

10. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-8.