Operation object construction system, method and device based on KKS coding and medium

Abstract

The invention discloses an operation object construction system, method and device based on KKS coding and a medium, and the method comprises the following steps: obtaining and expanding a KKS coding hierarchy, constructing an operation object containing an element level based on the expanded KKS coding hierarchy, and forming an operation object relation tree; constructing a spatial relation tree used for representing spatial position relations in the preset area; constructing functional attributes of the operation object; and constructing an operation object library based on the operation object relation tree, the spatial relation tree and the functional attributes of the operation objects. According to the method, a refined operation object relation tree, a spatial relation tree and rich functional attributes are constructed, and are associated and integrated, so that a digital operation object library with unified standard and comprehensive information is formed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of the hydropower industry, KKS coding, power plant equipment identification management, and in particular relates to a KKS coding-based operation object construction system, method, device and medium. BACKGROUND

[0002] KKS coding (Kraftwerk-Kennzeichen-System), commonly translated as power plant identification system, is a standardized coding system for uniquely identifying and classifying all equipment, components, buildings / structures and systems in a power plant or other industrial facilities. The KKS coding of existing hydropower stations generally adopts a structured four-level hierarchical coding method according to the "Hydropower Plant Identification System Coding Guide" (GB / T 35707-2017), and the coding levels include the whole plant level, system level, equipment level and component level, as shown in Figure 1

[0003] This KKS coding-based equipment object construction method provides standardized support for the equipment operation and maintenance management of hydropower plants, but there are still the following problems or deficiencies in actual work: 1. The existing coding hierarchy structure is not refined enough, and there is no operation construction of equipment objects, and no accurate coding to the equipment element level, which makes these equipment unable to be accurately referred to in operation, and the equipment description is easy to vary from person to person.

[0004] 2. The existing equipment object construction method lacks the definition of device operation-related attributes, such as device spatial attributes, device state, interval attributes, and industrial television association, which limits the digital application scenarios of the existing equipment management method, and the scalability and functionality are poor. SUMMARY

[0005] The purpose of the present application is to provide a KKS coding-based operation object construction system, method, device and medium, in order to solve the technical problems existing in the background art.

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions: One or more embodiments of the present application provide a KKS coding-based operation object construction method, which comprises: acquiring and extending the KKS coding hierarchy, constructing operation objects containing element levels based on the extended KKS coding hierarchy, and forming an operation object relationship tree; constructing a spatial relationship tree for representing the positional relationship of each space in a preset area; constructing the functional attributes of the operation objects; and constructing an operation object library based on the operation object relationship tree, the spatial relationship tree and the functional attributes of the operation objects.

[0007] ​The method forms a standard and unified digital operation object library with comprehensive information by constructing a refined operation object relationship tree, a spatial relationship tree, and rich functional attributes, and associating and integrating them. This solves the problem of unclear description of terminal operation objects and missing attribute information in existing KKS coding applications, and provides precise and reliable underlying data support for digital operation and maintenance of hydropower stations (such as electronic operation tickets, intelligent inspection, and equipment full life cycle management), greatly improving the fine level of equipment management and the efficiency and safety of business operations.

[0008] As a preferred solution, the extended KKS coding hierarchy includes adding element-level coding based on KKS coding using a structured hierarchical coding method to identify the terminal operation object. By adding element-level coding, the identification granularity of KKS coding is extended from the device / component level to the most basic and independently operable elements (such as air switches, buttons, and valves) in a hydropower station, achieving unique and accurate identification of the terminal operation object. This overcomes the problem of existing technologies that cannot accurately refer to and operate specific elements due to insufficient coding granularity, laying a solid foundation for subsequent fine management and automated control.

[0009] As a preferred solution, the construction of the spatial relationship tree includes determining the spatial hierarchy relationship based on the positional relationship of each space in the preset area, and constructing the spatial relationship tree based on the spatial hierarchy relationship. By constructing a structured spatial relationship tree, abstract physical space positional relationships are converted into hierarchical data models that can be recognized and processed by computers. This enables each operation object to be accurately associated with a specific spatial location, enabling device positioning, regional management, and location-based inspection path planning, and enriching the dimensions of digital management.

[0010] As a preferred solution, the construction of the operation object library includes storing the operation object relationship tree and the spatial relationship tree in structured form data, and mapping the functional attributes of the operation objects to the structured form data. The structured form data with mapped functional attributes is imported into the digital management system through an ETL tool to form a digital operation object library.

[0011] Structured form storage facilitates batch processing and verification of data, and batch import through an ETL tool or database operation enables efficient and automated deployment of offline constructed data models into the digital management system in the production environment. This method ensures the efficiency and accuracy of data migration, greatly shortening the period from construction to use of the operation object library.

[0012] As a preferred solution, the step of constructing the operation object library further comprises: utilizing a machine learning model to intelligently analyze the unstructured text data, automatically extract the potential functional attributes of the operation objects, and verify and complete the constructed operation object library. By utilizing the machine learning model to intelligently analyze the massive unstructured text, the device attribute information that may be missed by human can be automatically discovered and extracted, and the existing library can be automatically verified and completed. This greatly reduces the work burden of manual sorting, overcomes the problems of low efficiency and easy errors of manual methods, and significantly improves the automation degree, data integrity and accuracy of the operation object library construction.

[0013] As a preferred solution, the step of constructing the operation object library of the hydropower station further comprises: based on historical operation data, utilizing association rule mining or a graph neural network model to automatically learn and recommend the association relationships between the operation objects and tools, between the operation objects and other operation objects, and updating the corresponding functional attributes. By analyzing the historical data to automatically mine the implicit association relationships between the operation objects and between the operation objects and tools, and intelligently recommending to the system, the functional attributes of the operation object library not only depend on the preset rules, but also can continuously learn and evolve from the actual operation experience. This enhances the intelligence of the system, and provides data-driven decision support for optimizing the operation process, recommending necessary tools, and preventing misoperation.

[0014] As a preferred solution, the method further comprises: accessing real-time sensor data of the operation objects; utilizing a time series prediction or anomaly detection model to calculate the predicted state information of the devices; and associating the predicted state information as dynamic update attributes to the corresponding operation objects in the operation object library.

[0015] One or more embodiments of the present specification provide a KKS coding-based operation object construction system, which comprises: an operation object construction module configured to extend the KKS coding hierarchy, construct operation objects containing element level, and form an operation object relationship tree; a space construction module configured to construct a space relationship tree representing the positional relationship of each space in a preset area; an attribute definition module configured to construct the functional attributes of the operation objects; and an object library generation module configured to construct a hydropower station operation object library based on the operation object relationship tree, the space relationship tree and the functional attributes.

[0016] As a preferred solution, the operation object construction module is further configured to increase element level coding based on KKS coding in a structured hierarchical coding manner, for identifying the endmost operation objects. This module ensures the uniqueness and standardization of the element level coding in the entire station range by executing standardized coding extension rules, eliminates coding conflicts and chaos from the system level, and ensures the quality of the basic data.

[0017] As a preferred embodiment, the spatial construction module is further configured as follows: the construction of the operation object library includes: storing the operation object relationship tree and spatial relationship tree as structured form data; mapping the functional attributes of the operation objects to the structured form data; and importing the structured form data mapped with functional attributes into the digital management system through an ETL tool to form a digital operation object library. This module realizes the spatial data from model to storage to final application, ensuring that spatial information can be efficiently identified and utilized by the digital management system, and improving the integration efficiency and data processing capabilities of the entire system.

[0018] This specification provides one or more embodiments of an operation object construction apparatus based on KKS encoding. The apparatus includes at least one processor and at least one memory; the at least one memory is used to store computer instructions; the at least one processor is used to execute at least a portion of the computer instructions to implement the method described above.

[0019] This specification provides one or more embodiments of a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, can implement the method described in any of the above-described embodiments.

[0020] This device and storage medium provide a concrete physical implementation, enabling the advanced method of building an operation object library based on KKS encoding to be executed and deployed by any computing-capable device. This greatly enhances the practicality and scalability of the method, allowing it to move beyond theoretical solutions and be transformed into practical products and applications, generating economic benefits.

[0021] The beneficial effects that the KKS-encoded object construction system, method, apparatus, and medium disclosed in this application may bring include, but are not limited to: (1) Expand the application scope of KKS encoding The hydropower station operation object library constructed according to this invention can further expand the application scope of KKS coding. Compared with the conventional KKS coding method for constructing equipment objects, the construction of operation objects emphasizes equipment components as operation objects, expanding its attributes covered in hydropower station operation business, and digitizing the operation actions of operators on equipment components. Furthermore, based on assigning a unique identification code to each operation object in the hydropower station, related functional attributes are expanded to enable it to record various status information, making equipment operation management more refined and standardized.

[0022] (2) Improve equipment information recording Compared with traditional KKS coding, which only records equipment names and codes, the hydropower station operation object database constructed by this invention can, on the one hand, record the business relationship between the equipment and its related production auxiliary equipment (such as tools and industrial televisions) and the location relationship between the equipment and the spatial environment; on the other hand, it can record various status information generated by the equipment in production activities, which can provide traceable means and basis for each production business link.

[0023] (3) Provide data support for digital applications The hydropower station operation object database constructed by this invention provides underlying data support for the digital management of hydropower station equipment and production operations, greatly improving the efficiency and accuracy of engineering design, construction, operation, maintenance, management, and information exchange. Examples include electronic operation tickets, electronic work tickets, equipment inspections, equipment lifecycle management, and procurement / material / project management. Attached Figure Description

[0024] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein: Figure 1 This is an exemplary flowchart illustrating a method for constructing hydropower station operation objects based on KKS encoding, according to some embodiments of this specification. Figure 2 This is an exemplary block diagram of a hydropower station operation object construction system based on KKS encoding, as shown in some embodiments of this specification. Figure 3 This is a schematic diagram of the operation object relationship tree according to some embodiments of this specification; Figure 4 This is a schematic diagram of a spatial relationship tree according to some embodiments of this specification; Figure 5 This is a schematic diagram of a space form according to some embodiments of this specification; Figure 6 This is a schematic diagram of a single operational object—a spatial form—illustrated according to some embodiments of this specification. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0026] Conversely, this application covers any substitutions, modifications, equivalent methods, and schemes made within the spirit and scope of this application as defined in the claims. Furthermore, to provide the public with a better understanding of this application, certain specific details are described in detail below. However, this application can be fully understood by those skilled in the art even without these detailed descriptions.

[0027] The method and system for constructing operation objects based on KKS encoding provided in this application are applicable to hydropower stations, power plants, large production workshops, etc.

[0028] Figure 1 This is an exemplary block diagram of an operation object construction system based on KKS encoding, according to some embodiments of this specification. The operation object construction system 100 based on KKS encoding includes an operation object construction module 110, a space construction module 120, an attribute definition module 130, and an object library generation module 140.

[0029] The operation object construction module 110 is used to obtain the standard KKS coding library of the hydropower station. KKS coding is an international standard power plant identification that conforms to the "Coding Guidelines for Hydropower Plant Identification System" GB / T 35707-2017, and expands the coding level, that is, adds component-level coding under the standard four-level structure. In some embodiments, the operation object construction module is configured to expand the KKS coding level, construct operation objects containing component levels, and form an operation object relationship tree.

[0030] In some embodiments, an operational object refers to an entity in a hydropower station that needs to be operated, monitored, or managed. In this scheme, it specifically refers to equipment objects identified by extended KKS coding, down to the component level. Each operational object has a unique KKS code, which serves as the basic data unit for building digital management.

[0031] The spatial construction module 120 is used to construct a spatial relationship tree representing the positional relationships of various spaces within a preset area; wherein, the preset area can be a hydropower station, a thermal power station, a large production workshop, etc.

[0032] The attribute definition module 130 is used to construct the functional attributes of the operation objects; the attribute definition module is configured to extend the functional attributes of each operation object and structure its attributes so that each operation object has relevant attribute tags and can establish a connection with each space.

[0033] The object library generation module 140 is used to construct a hydropower station operation object library based on the operation object relationship tree, spatial relationship tree and the functional attributes of the operation objects.

[0034] In some embodiments, constructing the operation object library includes: storing the operation object relationship tree and spatial relationship tree as structured form data; mapping the functional attributes of the operation objects to the structured form data; and importing the structured form data mapped with functional attributes into the digital management system using an ETL tool to form a digital operation object library. This module realizes the transformation of spatial data from model to storage to final application, ensuring that spatial information can be efficiently identified and utilized by the digital management system, and improving the integration efficiency and data processing capabilities of the entire system.

[0035] In some embodiments, the digital management system can be a software platform system used for operation and maintenance management in a hydropower station, such as a production management system (MIS), equipment asset management system (EAM), supervisory control and monitoring system (SCADA), or intelligent operation and maintenance platform. The operation object library is ultimately imported into the system to provide standardized data services for its various application functions, such as issuing operation tickets, inspections, and asset management.

[0036] The operation object construction system based on KKS coding provided in this embodiment adopts a modular design, in which the functions of operation object construction, space construction, attribute definition and object library generation are implemented by dedicated modules. This system can transform traditional, scattered and non-standard management information into centralized and standardized digital assets, providing a core data engine for the construction of a digital operation and maintenance platform for hydropower stations.

[0037] Figure 2 This is an exemplary flowchart of an exemplary process 200 for constructing an operation object based on KKS encoding, as shown in some embodiments of this specification.

[0038] In some embodiments, process 200 is executed by a processor. All or part of the modules of the KKS-based object construction system 100 can be integrated into the processor. For example... Figure 2 As shown, process 200 includes the following steps.

[0039] Step 210: Obtain and expand the KKS encoding level. Based on the expanded KKS encoding level, construct operation objects containing component levels and form an operation object relationship tree. By standardizing the description and structured encoding of all operation objects in the operation object relationship tree, each device has a unique and standard name and code, forming a structured form - the operation object form.

[0040] Structured forms refer to data files or database tables that store the relational tree, spatial relational tree, and functional attributes of operation objects in tabular form. Each column represents an independent object, an operation object, or a space, and each row represents an attribute of the table object, such as KKS code, name, parent node ID, status, and space code. This format facilitates manual viewing, batch processing, and machine reading, and is the standard data source format for ETL processes. Common formats include Excel spreadsheets, CSV files, or database tables.

[0041] KKS coding is an international standard power plant identification system that conforms to the "Coding Guidelines for Hydropower Plant Identification Systems" (GB / T 35707-2017). It is a structured, hierarchical coding system used to uniquely identify and classify all equipment, systems, and structures within a hydropower station. Its standard levels include plant-wide, system-wide, equipment-wide, and component-wide.

[0042] The extended KKS coding hierarchy refers to the construction of a component-level operational object library for hydropower stations, based on the standard KKS coding library established by the hydropower station enterprise and in accordance with KKS coding rules. A structured hierarchical coding approach is adopted, adding a finer-grained coding level—the component level—below the four-level structure of the standard KKS coding: plant-wide, system-wide, equipment-wide, and component-wide. This level identifies the most basic, independently operable physical entities in the hydropower station, representing the smallest unit of operation. Examples include an independent circuit breaker (KK001), pushbutton (AN001), fuse (RD001), or valve (FM001).

[0043] Specifically, based on the company's existing KKS coding library, the equipment included in each level of equipment will be further reviewed. Following the methods in the technical solution, uncoded equipment will be expanded to improve the equipment tree structure. For example, if the equipment "Transformer Protection C Cabinet" actually contains switching element equipment, but this equipment is not in the existing coding library, then a one-level coding extension needs to be made at the "Transformer Protection C Cabinet" level to code the switching element equipment within that equipment.

[0044] The main levels of object encoding: Level 0: Plant-wide code, representing unit or shared system equipment. For example: Unit 1F, compressed air system. Encoding example: 01.

[0045] Level 1: System code, representing the unit or a similar system or subsystem within the Level 0 code range. Examples include: power output and plant auxiliary power systems, generator circuit breakers and disconnectors. Encoding example: 0BAC01.

[0046] Level 2: Equipment code, indicating the classification of mechanical and electrical equipment within the Level 1 system. For example: electrical equipment (switchgear), mechanical equipment (valves). Example code: GS001.

[0047] Level 3: Component code, indicating the classification of components and signals within a device unit. For example: electrical components (power switchgear), mechanical components (gate valves, ball valves). Encoding examples: -Q, KA.

[0048] Level 4: Component Code. This represents the classification of basic, independent components within the equipment, and is the final operational object in a hydropower station. The coding method is based on the equipment type, using the first letter of the type's pinyin followed by a number. For example: circuit breaker, fuse, coding example: KK001.

[0049] Example of specific encoding rules: Taking code 010BAC01GS001-QZS001 as an example, 01 represents Unit 1F, belonging to Level 0 operation object; 0BAC01 represents generator circuit breaker and switching, belonging to Level 1 operation object; GS001 represents switchgear, belonging to Level 2 operation object; -Q represents circuit breaker, belonging to Level 3 operation object; and ZS001 represents status indication, belonging to Level 4 operation object. The complete names and codes of each level of operation object are shown in Table 1 below: Table 1

[0050] See Figure 3 In some embodiments, the operation object relationship tree is a data structure based on a parent-child hierarchy, used to represent the physical or logical subordination relationships between all operation objects in a hydropower station. The root node of the tree represents Unit 1F, the intermediate nodes represent generator circuit breakers and disconnectors, switching equipment, and circuit breakers, and the leaf nodes represent the status indications of the circuit breakers at the very end of the component-level operation objects. This tree structure intuitively reflects the hierarchical relationship of the various functional components of a unit.

[0051] Step 220: Construct a spatial relationship tree to represent the positional relationships of each space within the preset area; In some embodiments, a spatial relationship tree is a data structure based on inclusive relationships, which can represent the hierarchical relationship of physical spatial locations within a preset area. The spatial coding of each space within the preset area still adopts a structured multi-level hierarchical coding method. Depending on the complexity of the spatial structure, the coding levels can be flexibly expanded. The spatial hierarchy is determined based on the positional relationship of each space or region, constructing a spatial relationship tree. The spatial code consists of a combination of letters and numbers, and the length of the coded characters depends on the coding level and the order of the spaces. By standardizing the description and structured coding of all spaces or regions in the spatial relationship tree, each space or region on-site has a unique and standard name and code, forming a structured form—a spatial form.

[0052] For example, based on existing construction drawings of the hydropower station, identify the location, name, dimensions, elevation, and other information of each space or area within the hydropower station. Then, code each space according to the methods outlined in the technical plan.

[0053] Main levels of spatial coding: Level 0: Full plant code, representing the plant / station. For example: left bank power station of a certain hydropower station, right bank power station of a certain hydropower station. Encoding examples: 01, 02.

[0054] Level 1: Building code, indicating the plant or elevation. For example: generator floor main plant, switchyard 510m floor. Encoding example: 1UAB01.

[0055] Level 2: Room or area code, indicating a specific equipment room or area division. For example: 1F unit control room, left bank intermediate pressure unit room. Encoding example: R001.

[0056] Example of specific encoding rules: Taking code 011UAB01R001 as an example, 01 represents the left bank power station of a hydropower station, belonging to level 0 space; 1UAB01 represents the first layer of the GIS switch station, belonging to level 1 space; R001 represents the GIS equipment room, belonging to level 2 space. The complete names and codes of each level of space are shown in Table 2 below: Table 2

[0057] Using data processing tools such as Excel, the coded spaces are formatted as structured lists to create a structured form – a space form, such as... Figure 5 As shown.

[0058] For example, in a hydroelectric power station, this tree represents the hierarchical relationship between all physical spaces within the station, such as powerhouses, floors, rooms, and areas. The root node represents the entire left bank power station, intermediate nodes represent the first level of the GIS switchyard, and leaf nodes represent specific rooms or areas, such as the GIS equipment room. This tree defines the spatial hierarchy; for example, the power station includes the first level of the GIS switchyard, and the first level of the GIS switchyard includes the GIS equipment room, and so on. Figure 4 As shown.

[0059] Step 230: Construct the functional attributes of the operation object; the functional attributes are a set of structured data fields attached to each operation object, used to describe the various characteristics and relationships of the object excluding identification information. They include at least: Equipment Number: For the lowest level of equipment, a number generated according to the enterprise identification coding standard is used for the dual naming description of the equipment in the operation ticket or work order. For example: switch number 1-4K1, valve number 1DF1.

[0060] b. Equipment Code: The code is formed according to the enterprise identification coding standard, using a simple and easy-to-understand coding method to facilitate quick equipment identification by users. It is generally used for F3 level equipment. For example, the code ABC-BH-ZB-01BC represents a hydropower station - relay protection system - transformer protection system - transformer protection cabinet C of the No. 1 transformer.

[0061] c. Status attributes: used to reflect various states of the device, including normal state and actual state.

[0062] d. Spatial attributes: used to establish the relationship between devices and spaces, reflecting the space name and code where the device is located.

[0063] e. Identification Attributes: Reflects the name and code of the sign hanging on the equipment.

[0064] f. Grounding wire identification attribute: reflects the name and code of the grounding wire suspended on the equipment.

[0065] g. Tool and equipment attributes: Reflects what tools or keys are needed to operate the equipment.

[0066] h. Industrial TV Attributes: The industrial TV camera channel that reflects the space where the device is located.

[0067] i. Application attributes: including but not limited to "whether it belongs to the operation interval", "number of times the operation ticket is used", "whether it belongs to the signal linkage point of the monitoring system", "whether it belongs to the equipment status verification point during the flood season", "whether it belongs to the inspection point", etc.

[0068] Step 240: Based on the operation object relationship tree, spatial relationship tree, and functional attributes of the operation objects, construct an operation object library. Store the operation object relationship tree and spatial relationship tree as structured form data, namely operation object forms and spatial forms, such as tables in Excel, CSV, or relational databases.

[0069] The construction of the operation object library includes: storing the operation object relationship tree and spatial relationship tree as structured form data; mapping the functional attributes of the operation objects to the structured form data; and importing the structured form data mapped with functional attributes into the digital management system using an ETL tool to form a digital operation object library. After the data is successfully imported into the library, the operation object library is no longer a collection of scattered files, but a set of interconnected tables in a relational database.

[0070] In some embodiments, mapping the functional attributes of an operation object to the structured form data specifically includes: after obtaining the functional attribute data of the operation object, adding it as an additional row to the operation object form. Each functional attribute corresponds to one or more rows in the form. The specific attribute values ​​of each operation object are then filled into the corresponding row of the form.

[0071] In the functional attributes of each operation object, a special spatial attribute field can be defined. The value of this field is directly filled with the unique spatial code corresponding to its physical space in the spatial relationship tree. For example, 011UAB01R001 represents "GIS equipment room".

[0072] When building the object library, the system uses this shared "spatial code" as a key to link the object to a specific spatial node in the spatial relationship tree.

[0073] Once the data is imported into the database, this association manifests as relationships between tables. By executing SQL join queries, such as JOIN operations, it's easy to find the attributes of the space where a specific device resides, or all devices within a given space. Ultimately, the structured form becomes a single operational object—a space form—containing all identifying information, hierarchical relationships, and functional attributes.

[0074] See Table 3 below for details. This constitutes a single operation object – a spatial form. Figure 6 .

[0075] Table 3

[0076] After a successful ETL job execution, the operation object library is formally established in the database of the digital management system, forming a digital operation object library. It is no longer a discrete file, but becomes an organic component of the system. Through database relationships and foreign key constraints, it maintains data integrity and consistency, and can be called by upper-level business applications (such as operation tickets, inspections, and mobile applications) at any time through SQL queries or application programming interfaces (APIs).

[0077] Using structured form storage facilitates batch data processing and validation, while batch import via ETL tools or database operations enables the efficient and automated deployment of offline-built data models into the production environment's digital management system. This approach ensures the efficiency and accuracy of data migration, significantly shortening the cycle from building the operational object library to its deployment.

[0078] In a preferred embodiment, to enhance the intelligence and automation of the operation object library construction, the processor can also integrate machine learning models (such as Neural Network (NN) models or Deep Neural Network (DNN) models). The step of constructing the operation object library further includes: using machine learning models to intelligently analyze unstructured text data, automatically extracting the potential functional attributes of operation objects, and verifying and completing the constructed operation object library. Utilizing machine learning models to intelligently analyze massive amounts of unstructured text can automatically discover and extract device attribute information that may be missed manually, and automatically verify and complete the existing library. This greatly reduces the workload of manual sorting, overcomes the inefficiency and error-prone nature of manual methods, and significantly improves the automation level, data integrity, and accuracy of the operation object library construction.

[0079] In some embodiments, the machine learning model can be trained based on multiple labeled training samples.

[0080] In some embodiments, each training sample may include unstructured text data. The processor can obtain the unstructured text data by parsing historical sample requests, etc., from the user's end.

[0081] In some embodiments, tags can be sample functional attributes corresponding to sample unstructured text data and a verified and completed library of operation objects. Tags can be obtained through manual labeling. For example, those skilled in the art can determine the potential functional attributes of operation objects based on historical sample requests, and verify and complete the operation object library to obtain tags.

[0082] In some embodiments, the processor can acquire one or more training samples and the label corresponding to each training sample; perform multiple iterations, and terminate the iteration when the iteration termination condition is met, thus obtaining a trained machine learning model. Each iteration includes: selecting a training sample from the training samples, inputting the training sample into the initial machine learning model, and obtaining the predicted output of the initial machine learning model corresponding to the training sample; calculating the value of the loss function by substituting the predicted output and the label of the training sample into a predefined loss function formula; and updating the model parameters in the initial machine learning model in reverse based on the value of the loss function. The model parameters in the reverse machine learning model can be updated using various methods. For example, the model parameters in the initial machine learning model can be updated in reverse based on gradient descent. The iteration termination condition may include reaching a threshold number of iterations, etc.

[0083] Unstructured text data refers to information that lacks a predefined data model or is not organized in a fixed format. In this embodiment, it specifically refers to textual materials generated during the operation and management of hydropower stations, including equipment manuals, maintenance reports, historical work orders, operation logs, safety regulations, etc. This data is an important data source for intelligent analysis by machine learning models.

[0084] Manually sorting and inputting massive amounts of equipment components and their attributes is prone to errors and omissions, and is also very labor-intensive. Using Natural Language Processing (NLP) models, such as BERT and Transformer, to analyze unstructured text data such as equipment manuals, operation manuals, and historical work orders, automatically extracts information such as equipment name, model, possible status, and associated tools. This information is then compared, verified, and automatically completed with a pre-built operation object database, greatly improving efficiency and accuracy.

[0085] In some embodiments, the system accesses existing unstructured text data sources of the hydropower station, including but not limited to: equipment manuals, maintenance reports, historical electronic work tickets / operation tickets (database records), safety regulations, etc.

[0086] First, raw text is extracted from these documents using text parsing tools such as Apache PDFBox. Then, the text is preprocessed, including word segmentation, stop word removal, part-of-speech tagging, and initial cleaning using Named Entity Recognition (NER).

[0087] A pre-trained model based on the Transformer architecture, such as BERT or RoBERTa, is used, and domain adaptation is performed on top of it. The training data is labeled by relevant technical personnel to construct a sequence label dataset containing labels such as "equipment name", "part type", "operation action", "required tools", and "hazard source". The learning objective of the model is to be able to accurately identify "201FK" as the operation object (circuit breaker), "disconnect" as the operation action, and "do not close, people are working" as the label attribute from a text such as "Disconnect the high-voltage side circuit breaker 201FK of the main transformer of Unit 1F and hang the sign 'Do Not Close, People Are Working'".

[0088] For example, based on drawings and experience, the operator initially constructs the relationship tree and core attributes of the operation objects, forming the operation object library V1.0. The initially constructed operation object library, such as equipment names and codes, is then matched with entities extracted from text using similarity calculations, such as cosine similarity.

[0089] NLP models perform batch analysis on massive amounts of text. For example, if the model finds that multiple historical reports mention "a special crank is required to operate 201FK," and the "tool attribute" of the current 201FK object is empty, the system will issue a verification alert or autocomplete suggestion to the administrator. Similarly, the model may find from the procedures that "after operating the 202DK grounding switch, a three-phase grounding wire must be suspended at a designated location," and thus suggest adding a "grounding wire identification attribute" to the 202DK object.

[0090] The system presents all potential attribute changes identified by intelligent analysis to the administrator in a highlighted manner. After the administrator confirms, the changes are updated in batches to the operation object library, generating a richer and more accurate operation object library version 2.0.

[0091] This embodiment frees manual labor from tedious information mining and verification, greatly improves the efficiency of building the object database and the quality of data, ensures the comprehensiveness and accuracy of attribute information, and provides a high-quality data foundation for digital applications from the source.

[0092] In some embodiments, the step of constructing the hydropower station operation object library further includes: based on historical operation data, using association rule mining or graph neural network models to automatically learn and recommend the association relationships between operation objects and tools, and between operation objects and other operation objects, and updating the corresponding functional attributes. By analyzing historical data, implicit association relationships between operation objects and between operation objects and tools are automatically mined and intelligently recommended to the system, so that the functional attributes of the operation object library not only depend on preset rules, but can also continuously learn and evolve from actual operating experience. This enhances the intelligence of the system and provides data-driven decision support for optimizing operation processes, recommending necessary tools, and preventing misoperations. The graph neural network model is a machine learning model. The training samples of the graph neural network model can be historical operation data, and the labels can be the preferred operation objects and tools, and the association relationships between operation objects and other operation objects corresponding to the historical operation data. These labels can be obtained based on manual annotation. The specific training method for the graph neural network model can be implemented using the same training method as the aforementioned machine learning model; see the preceding text for details.

[0093] In some embodiments, historical operation data refers to the collection of data on various operational tasks that have been completed in the past, recorded in the hydropower station's digital management system. It typically exists in the form of structured database tables, recording information such as the task type, operation object, tools used, operators, and timestamps for each operation. This data forms the basis for training association rule mining and graph neural network models.

[0094] The relationships between equipment attributes, such as "tool attributes" and "operation objects," may initially require manual definition, which can be complex and potentially incomplete. By using association rule mining and learning, such as the Apriori algorithm or graph neural networks (GNNs), historical operation data can be analyzed to automatically discover frequently occurring "equipment-tool" combinations or sequential patterns where "operating equipment A usually requires operating equipment B." This allows for the automatic recommendation or improvement of the "tool attributes" and "linkage relationship attributes" of equipment in the operation object database.

[0095] In some embodiments, the method further includes: accessing real-time sensor data of the operation object; calculating the predicted state information of the device using a time-series prediction or anomaly detection model; and associating the predicted state information as a dynamically updated attribute with the corresponding operation object in the operation object library.

[0096] In response to connecting real-time sensor data to the operational object library, such as temperature, vibration, and current, time-series prediction models (e.g., LSTM, Prophet) or anomaly detection models (e.g., isolated forest, autoencoder) are used to monitor and predict equipment status in real time. The prediction results (e.g., "high risk of overheating in the next 24 hours") can be fed back into the operational object library as a new, dynamic "predicted status attribute," providing decision support for preventative maintenance.

[0097] For example, the system can obtain real-time sensor data from key equipment such as transformers, circuit breakers, and water pumps through the API interface of the plant's monitoring system, including time-series data such as temperature, vibration, current, voltage, and oil level.

[0098] Using a Long Short-Term Memory (LSTM) network model, taking the main transformer winding temperature as an example, a multivariate time-series prediction model is trained using historical temperature data from the past year, including features such as ambient temperature and load current. The model's task is to learn the temperature variation patterns under normal operating conditions and predict the temperature value for a future period, such as 24 hours.

[0099] The specific steps are as follows: Real-time data streams are connected to a big data platform for data cleaning and standardization. The processed data is then input into a pre-trained LSTM model for real-time inference to obtain predicted values ​​of the device's core state parameters. The system compares the predicted values ​​with the actual measured values; if the deviation continuously exceeds a threshold, an anomaly alarm is triggered. Simultaneously, the trend predicted by the model (such as a continuous rise in temperature) is transformed into a high-level predicted state attribute, such as "Healthy," "Caution," "Warning," or "Danger."

[0100] The generated "predicted status attributes" will be automatically written to the "status attribute" field of the operation object through a database update operation. This means that the device status in the operation object library is no longer static data recorded in the previous shift, but dynamic data that changes over time.

[0101] This embodiment endows the operation object library with "predictive" capabilities. Before issuing an operation ticket involving a certain piece of equipment, operators can first view the real-time "predictive status attributes" of that operation object. If the system displays a "warning" status, it may indicate that maintenance or additional safety measures are needed first, thereby transforming the operation and maintenance mode from "reactive maintenance" to "predictive maintenance," greatly improving the safety and reliability of hydropower station operation.

[0102] This specification provides one or more embodiments of an operation object construction apparatus based on KKS encoding. The apparatus includes at least one processor and at least one memory; the at least one memory is used to store computer instructions; the at least one processor is used to execute at least a portion of the computer instructions to implement the method described above.

[0103] This specification provides one or more embodiments of a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, can implement the method described in any of the above-described embodiments.

[0104] This device and storage medium provide a concrete physical implementation, enabling the advanced method of building an operation object library based on KKS encoding to be executed and deployed by any computing-capable device. This greatly enhances the practicality and scalability of the method, allowing it to move beyond theoretical solutions and be transformed into practical products and applications, generating economic benefits.

[0105] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for constructing operation objects based on KKS encoding, characterized in that, The method includes: Obtain and extend the KKS encoding level, construct operation objects containing element levels based on the extended KKS encoding level, and form an operation object relationship tree; Construct a spatial relationship tree to represent the spatial relationships between various locations within a preset area; Construct the functional properties of the operation object; Based on the operation object relationship tree, spatial relationship tree, and functional attributes of the operation objects, an operation object library is constructed.

2. The method for constructing an operation object based on KKS encoding as described in claim 1, characterized in that, The extended KKS encoding hierarchy includes: adding element-level encoding on the basis of KKS encoding using a structured hierarchical encoding method to identify the last operational object.

3. The method for constructing an operation object based on KKS encoding as described in claim 1, characterized in that, The construction of the spatial relationship tree includes: determining the spatial hierarchy relationship based on the positional relationship of each space within a preset area, and constructing the spatial relationship tree based on the spatial hierarchy relationship.

4. The method for constructing an operation object based on KKS encoding as described in claim 1, characterized in that, The construction of the operation object library includes: storing the operation object relationship tree and spatial relationship tree in a structured form; mapping the functional attributes of the operation objects to the structured form; and importing the structured form data mapped with functional attributes into the digital management system through an ETL tool to form a digital operation object library. The mapping of the functional attributes of the operation object to the structured form data specifically includes: adding the acquired functional attribute data of the operation object to the operation object form as additional rows; each functional attribute corresponds to one or more rows in the form; and filling the corresponding row of the form with the specific attribute value of each operation object. In the functional attributes of each operation object, a special spatial attribute field is defined. The value of this field is directly filled with the unique spatial code corresponding to its physical space in the spatial relationship tree. When constructing the operation object library, the spatial code is used as a key to link the operation object with a specific spatial node in the spatial relationship tree.

5. The method for constructing an operation object based on KKS encoding as described in claim 4, characterized in that, The steps of constructing the operation object library also include: using a machine learning model to analyze unstructured text data, automatically extracting the potential functional attributes of the operation objects, and verifying and completing the constructed operation object library.

6. The method for constructing an operation object based on KKS encoding as described in claim 4, characterized in that, The step of constructing the operation object library further includes: based on historical operation data, using association rule mining or graph neural network models, automatically learning and recommending the association relationships between operation objects and tools, and between operation objects and other operation objects, and updating the corresponding functional attributes.

7. The method for constructing an operation object based on KKS encoding as described in claim 4, characterized in that, The method further includes: accessing real-time sensor data of the operation object; calculating the predicted state information of the device using a time-series prediction or anomaly detection model; and associating the predicted state information as a dynamically updated attribute with the corresponding operation object in the operation object library.

8. A system for constructing operation objects based on KKS encoding, characterized in that, include: The operation object construction module is used to extend the KKS coding hierarchy, construct operation objects containing element levels, and form an operation object relationship tree; The spatial construction module is used to build a spatial relationship tree that represents the positional relationships of various spaces within a preset area; The attribute definition module is used to construct the functional attributes of the manipulated object; The object library generation module is used to construct an operation object library based on the operation object relationship tree, spatial relationship tree, and functional attributes of the operation objects.

9. A device for constructing operation objects based on KKS encoding, characterized in that, The device includes at least one processor and at least one memory; The at least one memory is used to store computer instructions; The at least one processor is configured to execute at least a portion of the computer instructions to implement the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed by a processor, enable the implementation of the method according to any one of claims 1 to 7.

Images