An industrial field device positioning assistance system and method

Abstract

The application discloses an industrial field device positioning auxiliary system and method, which collects image data and corresponding text description data of an industrial field device, stores an image storage path corresponding to the image data and the text description data in association as a structured record, receives a search term input by a user, performs matching search on the structured record based on the search term, and displays the searched structured record in a form of image-text association to assist the user in confirming a target device. In this way, the user can quickly search for position information of the target device, effectively shortens the time for finding the device, and solves the defects that an operation and maintenance personnel takes a long time to find a remote device and is prone to errors.

Description

Technical Field

[0001] This disclosure relates to the field of inspection and maintenance technology for industrial field equipment, and specifically to an industrial field equipment positioning auxiliary system and method. Background Technology

[0002] During the production and operation of industrial equipment, the equipment piping systems are intricate and complex. For example, thermal power plant equipment is typically equipped with thousands of valves of various types, including but not limited to gate valves, globe valves, butterfly valves, ball valves, regulating valves, and safety valves. Many of these valves are located in remote, high-altitude, or blind-sighted areas. These devices perform crucial functions such as regulating media flow, controlling process flow, and ensuring the safe operation of the equipment.

[0003] During routine inspections, periodic operations, or emergency responses, operators frequently need to quickly locate specific equipment. Currently, this mainly relies on personal experience and memory, consulting outdated and unintuitive paper drawings, or recording equipment information electronically and storing the photos on mobile devices for inspection personnel to review and locate. However, these methods are time-consuming, prone to errors due to similar environments, and difficult to effectively transfer experiential knowledge, failing to meet the modern requirements for efficiency, accuracy, and safety in equipment inspection and maintenance. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of the prior art in locating specific equipment (such as valves) that are too time-consuming and prone to errors, and to provide an auxiliary system and method for locating equipment in industrial field.

[0005] The present invention solves the above-mentioned technical problems through the following technical solution:

[0006] In a first aspect, this disclosure provides an industrial field equipment positioning assistance system, comprising:

[0007] The information acquisition module is used to collect image data and corresponding text description data of industrial field equipment;

[0008] The data storage module is used to associate and store the image storage path corresponding to the image data with the text description data as a structured record. The structured record includes at least a record identifier, an image storage path, and text description content.

[0009] The intelligent retrieval module is used to receive search terms input by the user and perform matching retrieval on the structured records based on the search terms;

[0010] The results display module is used to display the retrieved structured records in a graphic and textual format to help users identify the target device.

[0011] Optionally, the text description data includes at least one of the following: device function description, system information, location feature description, and user-defined tags.

[0012] Optionally, the data storage module further includes a logically derived field, which is generated based on the text description content and the user-defined tags with a preset separator.

[0013] The intelligent search module performs the matching search based on the logically derived fields.

[0014] Optionally, the structured record further includes a creation timestamp, and the data storage module can also be used to build a database index based on the logically derived field and the creation timestamp;

[0015] The intelligent search module is used to perform searches based on the database index and sort the search results in reverse order of creation time.

[0016] Optionally, the intelligent retrieval module can also be used to remove spaces from the search terms entered by the user and perform a retrieval on the text description content through wildcard matching.

[0017] Optionally, the results display module displays the search results in the form of a list of images and text, and each list item includes at least a device thumbnail, key text description and tag information;

[0018] In response to the user's selection of a list item, display the corresponding full image and all text description data.

[0019] Optionally, the system further includes a local area network (LAN) point-to-point data synchronization module for synchronizing structured records among multiple devices in an environment without external network coverage. The data synchronization module includes:

[0020] The device discovery unit discovers peer devices within the local area network based on User Datagram Protocol (UDP) broadcast.

[0021] The connection establishment unit establishes a long-lived connection with the discovered peer devices based on the transmission control protocol;

[0022] An incremental synchronization unit is used to exchange the unique identifier and last update timestamp of structured records in each device to determine an incremental data list, which represents the structured records that need to be updated by the device.

[0023] The conflict handling unit is used to retain the newer version based on the timestamp and issue a prompt to the user when version conflicts exist for the same structured record on different devices.

[0024] Optionally, the incremental synchronization unit can also be used to divide the structured record into data blocks of a preset size for transmission, and to perform a limited number of retries for data blocks that fail to be transmitted.

[0025] And / or,

[0026] The incremental synchronization unit can also be used to retain the transmitted data blocks when the network is interrupted and the data synchronization process is terminated, and to continue transmitting the unfinished data from the point of interruption during the next synchronization.

[0027] Secondly, this disclosure provides an industrial field equipment positioning assistance method, comprising the following steps:

[0028] Collect image data and corresponding text description data of industrial field equipment;

[0029] The image storage path corresponding to the image data is associated with the text description data and stored as a structured record. The structured record includes at least a record identifier, an image storage path, and text description content.

[0030] Receive user-inputted search terms and perform matching and retrieval on the structured records based on the search terms;

[0031] The retrieved structured records are displayed in a graphic and textual format to help users identify the target device.

[0032] Thirdly, this disclosure provides an electronic device including a memory, a processor, and a computer program stored in the memory and for running on the processor, wherein the processor executes the computer program to implement the device positioning assistance method described in any one of the second aspects.

[0033] Fourthly, this disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the device positioning assistance method described in any one of the first aspects.

[0034] Fifthly, this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the device positioning assistance method as described in any one of the first aspects.

[0035] The positive and progressive effects of this invention are as follows: This disclosure collects image data and corresponding text description data of industrial field equipment, and associates the image storage path corresponding to the image data with the text description data to store it as a structured record. It receives user-input search terms, performs matching searches on the structured records based on the search terms, and displays the retrieved structured records in a text-image association format to assist users in identifying target equipment. In this way, by associating and storing images and environmental features of industrial field equipment in a visualized structured record, it enables fast and accurate retrieval through user search requests. Users can quickly retrieve the location information of target equipment, effectively shortening the time spent searching for equipment. This solves the problems of time-consuming and error-prone searches for remote equipment by maintenance personnel. Furthermore, the search results (such as precise descriptions of valves and on-site photos) are displayed to operators for confirmation, preventing misidentification of equipment and improving the safety and accuracy of equipment inspection and maintenance work. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the modules of an industrial field equipment positioning assistance system shown in an example embodiment of the present disclosure;

[0037] Figure 2 This is a schematic diagram of a local area network point-to-point data synchronization module shown in an example embodiment of this disclosure;

[0038] Figure 3 A flowchart illustrating an exemplary embodiment of this disclosure of an industrial field equipment positioning assistance method;

[0039] Figure 4 A schematic diagram of a process for storing structured records of industrial field equipment provided as an exemplary embodiment of this disclosure;

[0040] Figure 5 A schematic diagram of the structure of data entities in a structured record of industrial field equipment provided as an exemplary embodiment of this disclosure;

[0041] Figure 6 A schematic diagram illustrating the process of searching and matching structured records of industrial field equipment in a database, provided as an exemplary embodiment of this disclosure;

[0042] Figure 7 This is a schematic diagram illustrating a process for synchronizing structured records among multiple devices connected to the same local area network, provided as an exemplary embodiment of this disclosure.

[0043] Figure 8 This is a schematic diagram of the structure of an electronic device shown in an example embodiment of the present disclosure. Detailed Implementation

[0044] The present invention will be further illustrated by way of embodiments below, but the present invention is not limited to the scope of the embodiments described herein.

[0045] The prefixes such as "first" and "second" used in this disclosure are merely for distinguishing different descriptive objects and do not limit the position, order, priority, quantity, or content of the described objects. The use of ordinal numbers and other prefixes used to distinguish descriptive objects in this disclosure does not constitute a limitation on the described objects. The description of the described objects is given in the claims or the context of the embodiments, and should not be construed as an unnecessary limitation. Furthermore, in the description of this embodiment, unless otherwise stated, "multiple" means two or more.

[0046] In this embodiment of the disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information comply with relevant laws and regulations and do not violate public order and good morals.

[0047] It is understandable that in large equipment plants (such as thermal power plants), the piping systems are intricate and complex, and there are numerous industrial field devices controlling these systems. For example, some valves are located in remote, high-altitude, deep within pipe racks, or in blind spots ("remote valves"). Maintenance personnel often need to quickly locate valves in specific equipment during routine inspections, periodic operations, or emergency responses. Existing methods for locating industrial field devices primarily rely on:

[0048] (1) Experienced maintenance personnel lead new maintenance personnel to identify industrial field equipment on-site. The location of industrial field equipment depends on the personal memory of experienced maintenance personnel. However, personnel changes in maintenance personnel can easily lead to the loss of equipment location information.

[0049] (2) Review paper-based industrial site equipment ledgers, internal system structure diagrams or equipment lists and other drawings and materials. However, the drawings and materials are not updated in a timely manner, and it is inconvenient to carry the drawings and materials on site. In addition, the drawings and materials are in the form of two-dimensional planar schematic diagrams, which cannot provide maintenance personnel with an intuitive and referable on-site view.

[0050] (3) Use electronic tools such as cameras to take pictures of the location of industrial equipment and record the location of industrial equipment in text using a notebook. However, the photos of the location of industrial equipment taken by the camera and the corresponding text records are stored separately and have poor correlation, which makes it difficult to retrieve the location of industrial equipment later and makes it difficult to quickly and accurately retrieve effective industrial equipment location information when needed.

[0051] Existing methods for locating industrial equipment (such as valves) are too time-consuming, severely impacting operational efficiency and even emergency response speed. Furthermore, due to the complexity and high similarity of industrial environments, relying solely on personal memory or drawings is prone to ambiguity, leading to a high risk of misidentification and misoperation. Therefore, there is an urgent need for an industrial equipment location assistance system and method designed for the complex environments of industrial sites.

[0052] Example 1

[0053] Figure 1 A schematic diagram of a module for an industrial field equipment positioning assistance system provided as an exemplary embodiment of this disclosure, the system comprising:

[0054] Information acquisition module 501: Used to acquire image data and corresponding text description data of industrial field equipment;

[0055] In one example, the information acquisition module 501 includes two sub-units: an image acquisition unit 5011 and a text input unit 5012.

[0056] Image acquisition unit 5011 is used to acquire image data of industrial field equipment (such as valves). Image acquisition unit 5011 can be integrated into a dedicated mobile acquisition device or deployed as a software module on a general-purpose smart terminal. Image acquisition unit 5011 can acquire a standard photo of a single industrial field device in single-shot mode; acquire multiple photos of the same industrial field device from different angles in burst mode; or acquire photos of multiple industrial field devices sequentially in a single operation in batch acquisition mode. No specific limitations are specified here.

[0057] In one example, the image acquisition unit 5011 incorporates an image quality detection algorithm that automatically evaluates the image's sharpness, exposure, and composition after capture. Images that do not meet quality requirements are prompted to be retaken. The image acquisition unit 5011 also supports automatic watermarking, overlaying metadata such as capture time, personnel, and geographical location onto the captured images for easy traceability later.

[0058] The text input unit 5012 is used to receive text description data associated with images of industrial field equipment (such as images of equipment valves). The text input unit 5012 can input text description data through a full keyboard input method, a nine-key input method, a voice input method that converts user speech into text through an integrated speech recognition engine, or a photo recognition method that converts photos into text.

[0059] In one example, the text description data includes at least one of the following: device function description, system information, location feature description, and user-defined tags.

[0060] In one example, the text description data of the industrial field equipment includes the equipment's functional description, system to which it belongs, and location characteristics. For example, the text description of a valve in the industrial field equipment is "Isolation valve for the circulating water return pipeline of the No. 1 turbine low-pressure cylinder, equipment number V-1207. This valve is located on the west wall of the zero-meter floor of the turbine hall, adjacent to the outlet pipeline of the #1 circulating water pump. The valve body is blue, and there is a set of large filters about 2 meters in front of it as a prominent reference." The valve's labels include "#V-1207", "#1 turbine", "isolation valve", and "#circulating water pump outlet pipeline".

[0061] It is understood that the term "industrial field equipment" as used in this disclosure broadly refers to various devices and instruments in industrial production sites that require inspection, operation, or maintenance, including but not limited to valves, pumps, instruments, motors, heat exchangers, compressors, fans, filters, heat exchangers, reactors, storage tanks, pipeline fittings, and various local indicating instruments and actuators. All of the above equipment has a clearly defined physical location and functional attributes, and image and text description data can be collected through this system and stored as structured records for retrieval and location.

[0062] For example, the industrial field equipment is a pump, and the text description of the pump is "Unit 1 circulating water pump, equipment number P-101. This pump is located on the south side of the circulating water pump house, with a rated flow of 3000 m³ / h and a motor power of 800 kW. The pump body is green, and a red 'High Pressure Danger' warning sign is affixed to its base. The operating procedure board for this pump is displayed on the nearby wall." The pump's labels include "#P-101", "#Unit 1", "#Circulating Water Pump", and "#High Pressure Equipment".

[0063] For example, in an industrial field, the equipment is an instrument, and the text description of the motor is "Exhaust fan A motor, equipment number M-301. This motor is located on the east side of the exhaust fan room, rated voltage 10kV, rated power 2000kW. The motor casing is dark gray, its junction box faces east, and there is a yellow warning line on the ground nearby." The motor's label includes "#M-301", "#Exhaust fan", "High voltage motor", and "#10kV system". In one example, the voice input method converts the user's speech into text through an integrated speech recognition engine, supporting Mandarin and dialect recognition.

[0064] In one example, photo recognition extracts text information from photos of equipment in an industrial setting using optical character recognition technology.

[0065] In one example, the information acquisition module 501 also includes a data verification unit for performing integrity and standardization checks on the acquired data. The integrity check ensures that required fields are filled in (such as industrial site equipment numbers and location descriptions); the standardization check ensures that the data format conforms to preset rules (such as correct number format and correct time format). For data that does not meet the requirements, the system prompts the acquisition personnel to supplement or correct it.

[0066] Data storage module 502: used to associate the image storage path corresponding to the image data with the text description data and store it as a structured record, wherein the structured record includes at least a record identifier, an image storage path, and text description content.

[0067] The data storage module can be a hybrid storage architecture, in which a relational database is used to store structured information of structured records of industrial field equipment (such as valves), supporting transaction processing and data integrity constraints; a file system or object storage service stores images of industrial field equipment for efficient access to unstructured data; and a full-text search engine is used to store and index text description data, supporting complex full-text search requirements.

[0068] In one example, the data storage module 502 can also be used for data backup and recovery to prevent data loss. The data storage module supports automatic scheduled backups as well as manual backups; backup data can be stored locally or in the cloud. Data recovery supports full recovery and incremental recovery, allowing users to select the recovery scope and time point as needed.

[0069] In one example, the data storage module further includes a logically derived field, which is generated based on the text description content and the user-defined tags with a preset separator; the intelligent retrieval module performs the matching retrieval based on the logically derived field.

[0070] In one example, the preset separator is the vertical bar symbol "|". For instance, in a structured record of industrial field equipment, the text description is "Isolation valve for the circulating water return pipeline of the No. 1 turbine low-pressure cylinder, equipment number V-1207. This valve is located on the west wall of the turbine hall's zero-meter level, adjacent to the outlet pipeline of the #1 circulating water pump." The user-defined tags are "#V-1207", "#1 turbine", "#isolation valve", and "#circulating water pump outlet pipeline". The data storage module uses the vertical bar "|" as the separator between the text description and the user-defined tags and generates the logical derived field as: "Isolation valve for the circulating water return pipeline of the No. 1 turbine low-pressure cylinder, equipment number V-1207. This valve is located on the west wall of the turbine hall's zero-meter level, adjacent to the outlet pipeline of the #1 circulating water pump. | #V-1207 | #1 turbine | #isolation valve | #circulating water pump outlet pipeline".

[0071] In one example, the preset separator is a semicolon ";". For example, for the same valve record mentioned above, the merged logically derived field is: "Isolation valve for the circulating water return pipeline of the No. 1 turbine low-pressure cylinder, equipment number V-1207. This valve is located on the west wall of the zero-meter floor of the turbine hall, adjacent to the outlet pipeline of the #1 circulating water pump; #V-1207; #1 turbine; #isolation valve; #circulating water pump outlet pipeline".

[0072] In one example, the structured record also includes a creation timestamp, and the data storage module can also be used to create a database index for the logically derived field and the creation timestamp; the intelligent retrieval module is used to perform a retrieval based on the database index and arrange the retrieval results in reverse order of creation time.

[0073] The data storage module supports multiple index types, including a primary key index for quickly locating records based on industrial field equipment identifiers; a full-text index for keyword retrieval based on text content; a tag index for quickly filtering records based on tag categories; and a time index for sorting search results by creation or update time. The data storage module can also be used for dynamic index updates; when structured records of industrial field equipment (such as valves) are added, deleted, or modified, the index can be automatically updated synchronously to maintain data consistency.

[0074] Intelligent retrieval module 503: used to receive search terms input by the user and perform matching retrieval on the structured records based on the search terms.

[0075] In one example, the retrieval module 503 includes a text retrieval unit 5031, a voice retrieval unit 5032, and an image retrieval unit 5033.

[0076] The text retrieval unit 5031 is used to provide retrieval based on text-based keywords. The text retrieval unit 5031 receives search keywords entered by the user, converts them into query conditions, and submits them to the full-text search engine for retrieval. The text retrieval unit 5031 can also support query expansion, automatically identifying synonyms, near-synonyms, and related words of the query terms to broaden the search scope and improve the recall rate. For example, when a user searches for "electric valve," the positioning assistance system simultaneously retrieves structured records of valves containing related expressions such as "electric valve," "electric regulating valve," and "motor-driven valve."

[0077] In one example, the search request is a text description of the corresponding industrial field equipment. For instance, the search request would be "Equipment number V-3105, belonging to the #3 boiler main steam system, located on the high east side of the boiler room." Upon receiving the search request, the retrieval module matches it against the structured records of industrial field equipment in the database, searching for all structured records containing or similar text descriptions, thus obtaining a list of matching industrial field equipment results for user comparison and confirmation.

[0078] In one embodiment, when the search request is a text description of a corresponding industrial field device, the positioning assistance system performs preprocessing on the text description by removing spaces.

[0079] For example, if a user enters "V-3105", the preprocessed result will be "V-3105" to prevent search failures due to space mismatches.

[0080] In one embodiment, when the search request is a text description of the corresponding positioning assistance system, the positioning assistance system performs delay anti-jitter processing on the text description.

[0081] Understandably, to avoid system resource consumption and result list flickering caused by frequent search requests during continuous user input, the positioning assistance system sets a preset delay (e.g., 300 milliseconds) after detecting user input. When the user stops inputting for more than the preset delay, the system matches the structured records of industrial field equipment in the database, searching for all structured records of industrial field equipment that contain or are similar to the above text description, and obtains a list of industrial field equipment matching results for the user to compare and confirm.

[0082] For example, if a user wants to search for "#3 boiler drain valve", the positioning assistance system waits for 300 milliseconds when "#3 boiler" is entered. If the user continues to enter "drain valve" within 300 milliseconds, the timer resets. Only after the user has completely entered "#3 boiler drain valve" and paused for more than 300 milliseconds will the positioning assistance system finally match "#3 boiler drain valve" as the text description with the structured records of the positioning assistance system in the database.

[0083] The voice retrieval unit 5032 provides retrieval based on voice input. It integrates a speech recognition engine that converts user voice input into text before handing it over to the text retrieval unit for execution. The speech recognition engine supports both real-time and streaming modes, outputting recognition results while the user speaks, providing a smooth interactive experience. The voice retrieval unit incorporates noise suppression and echo cancellation algorithms, enabling accurate voice recognition even in noisy power plant environments.

[0084] In one example, the search request is a voice signal obtained through voice input. For instance, maintenance personnel input the voice signal "Find the No. 3 boiler drain valve," and the voice recognition unit of the industrial field equipment positioning assistance system converts the voice signal into the text string "No. 3 boiler drain valve" in real time. The retrieval module receives the converted text string, matches it with the structured records in the database, and returns a list of matching results related to text descriptions such as "No. 3 boiler" and "drain valve" for the user to compare and confirm.

[0085] Image retrieval unit 5033 is used for retrieval based on photographs of the appearance of industrial field equipment. Image retrieval unit 5033 includes an image preprocessing component, a feature extraction component, and a feature matching component. The image preprocessing component performs processing on the user-captured image, such as size normalization, contrast enhancement, and distortion correction. The feature extraction component extracts high-dimensional feature vectors from the preprocessed image using a deep convolutional neural network model, effectively extracting the appearance features of the equipment. The feature matching component 5033 calculates the similarity between the feature vector of the photographed image and the feature vectors of images in the valve structured records stored in the database. Similarity calculation can use vector cosine distance or Euclidean distance as a metric. For each candidate industrial field equipment, the positioning assistance system selects the result with the highest similarity to the photographed image as the matching score for that equipment. Finally, all candidate equipment records with scores exceeding a threshold are returned, sorted in descending order of score.

[0086] In one example, the deep convolutional neural network employs, for instance, the ResNet-50 architecture, comprising 49 convolutional layers and 1 fully connected layer, and is pre-trained on the ImageNet dataset. The network accepts an input image that has been size-normalized (e.g., 224×224 pixels), and extracts hierarchical representations from low-level edge textures to high-level semantic features layer by layer through residual modules, ultimately outputting a 2048-dimensional feature vector. This feature vector is used to characterize the appearance features of industrial field equipment (such as valves) and serves as the basis for subsequent cosine similarity calculations with image feature vectors stored in a database.

[0087] In another example, the deep convolutional neural network employs the MobileNetV3-Large architecture, suitable for lightweight inference on mobile devices. This network optimizes the balance between computational efficiency and accuracy through depthwise separable convolutions and neural architecture search techniques, significantly reducing the number of parameters and floating-point operations while maintaining high recognition accuracy. The network also accepts a 224×224 pixel input image and outputs a 1280-dimensional feature vector. For example, on mobile devices such as smartphones or tablets, the MobileNetV3-Large model can complete feature extraction from a single image within tens of milliseconds, meeting the real-time retrieval needs of industrial sites while reducing device power consumption.

[0088] In one example, the search request is a field image containing the target equipment. For instance, maintenance personnel might take a photo of a difficult-to-identify valve or its distinctive surroundings (such as a unique pipe layout or nearby instrument panel). The image recognition unit of the positioning assistance system extracts and analyzes features from the uploaded image, identifying key objects (such as valve type, number plate, pipe color) or scene features, and automatically generates a set of keywords for text description (such as "red handwheel," "near pressure gauge," "overhead pipeline"). The intelligent retrieval module receives these generated keywords as search criteria, matches them against structured records of industrial field equipment in the database, and finds a list of matching valves with similar visual features for user comparison and confirmation.

[0089] In one example, the fuzzy matching retrieval can be performed based on a string similarity algorithm based on edit distance (such as Levenshtein distance). For instance, by calculating the similarity score between the user-input search terms and the text description content, and comparing the similarity score with a preset similarity score threshold, structured records with scores exceeding the preset threshold are returned to assist the user in identifying the target device.

[0090] In one example, the fuzzy matching retrieval can be performed based on the TF-IDF keyword matching algorithm. Combined with Chinese word segmentation technology (such as jieba word segmentation), the search terms and text description content are segmented and the term frequency-inverse document frequency score is calculated. The record with the highest relevance is returned to help the user confirm the target device.

[0091] For example, if a user inputs the search term "electric valve," the system first uses the jieba word segmentation tool to segment the search term into two terms: "electric" and "valve." Then, the system traverses all structured records of industrial field equipment in the database. For each record, its text description (such as "electric valve," "motor-driven valve," "manual valve," etc.) undergoes the same word segmentation process, and the term frequency (TF) of "electric" and "valve" in that structured record is calculated—that is, the number of times the term appears in that record. Simultaneously, the number of records containing "electric" or "valve" in the entire database is counted to calculate the inverse document frequency (IDF) of each term—that is, the rarity of the term globally. For each structured record, the system sums the TF-IDF values ​​(TF × IDF) of each term in the search term to obtain the overall matching score of that structured record relative to the search term "electric valve." All structured records are then sorted in descending order of their scores, and the top-ranked structured records are returned. For example, if the text of the record "electric valve" in the database contains high frequencies of the words "electric" and "valve", and "electric" is relatively rare in the entire database (high IDF), its overall score will be significantly higher than that of the record containing only "manual valve". The system will prioritize displaying the former, thereby effectively improving the accuracy and recall of the search.

[0092] In another example, fuzzy matching retrieval can be performed based on a deep learning-based semantic matching model (such as a BERT-based semantic similarity model). The search terms and text descriptions are encoded into semantic vectors, the cosine similarity between the vectors is calculated, and the record with the highest semantic relevance is returned to help users identify the target device.

[0093] For example, if a user inputs the search term "red handwheel valve near the pressure gauge," the system inputs this search term into a pre-trained BERT model to generate a high-dimensional semantic vector A. Simultaneously, the system encodes the text description content (e.g., "The valve is a red handwheel, located approximately 0.5 meters below the pressure gauge") in each structured record of industrial field equipment in the database into a semantic vector B using the BERT model. Then, the system calculates the cosine similarity between vector A and vector B to obtain the semantic matching score between the record and the search term. After traversing all records, the system sorts the results with scores exceeding a preset similarity threshold in descending order of score and returns the structured record with the highest semantic relevance. In one example, the image preprocessing component in the image retrieval unit 5033 can also be used to perform motion blur restoration on images of the target industrial field equipment. When the image has linear blur caused by the motion of the capturing equipment, deconvolution processing is performed on the image based on the blur kernel estimation result to restore the image's sharp edges and detailed textures. In another example, the image preprocessing component can also be used to perform defocus blur restoration on images of the target industrial field equipment. For example, when the on-site image of the target industrial equipment has concentric circle blur caused by inaccurate focus, a deconvolution method based on the point spread function can be used to restore the image.

[0094] In one example, the image preprocessing component can also be used to perform noise suppression processing on on-site images of target industrial equipment. For instance, when there is significant noise in the on-site image of a target valve due to low light, a deep learning-based noise reduction algorithm can be used to directly predict a clean image through a trained neural network model, effectively removing noise while preserving details.

[0095] It is understood that search requests are not limited to the forms described above. In possible implementations, the forms of search requests can be diverse, including but not limited to text input, voice input, image input, or any combination thereof listed in the above embodiments. The intelligent retrieval module of the positioning assistance system has a corresponding interface or preprocessing unit to receive, identify, and process these different types of search requests, and uniformly convert them into query instructions that can be matched with structured records of industrial field equipment in the database. Those skilled in the art should understand that any input form that can convert a user's intention to locate industrial field equipment into an input form that can be processed and executed by the system, as long as the technical effect it achieves is "rapid positioning of industrial field equipment based on image and text records," is a reasonable variation of the technical solution of this invention and falls within the protection scope sought by this disclosure.

[0096] Results display module 504: This module displays the retrieved structured records in a graphic and textual format to help users identify the target device.

[0097] The result display module 504 displays the search results in the form of a list of images and text. Each list item contains at least a device thumbnail, key text descriptions, and tag information. In response to the user's selection of a list item, the module displays the corresponding complete image and all text description data.

[0098] In one example, the results display module 504 can include multiple view modes: a list view displays the structured records of candidate industrial field equipment in a table format, with each row showing columns of key information for the industrial field equipment, supporting sorting and pagination; a card view displays the structured records of candidate industrial field equipment in a card format, with each card containing a thumbnail of the industrial field equipment and key information for easy browsing; and a map view marks the candidate industrial field equipment on a power plant floor plan or 3D model, visually showing the location distribution of the industrial field equipment, and users can click on the map markers to view the details of the industrial field equipment.

[0099] In one example, the results display module 504 can also be used to support results comparison. When multiple candidate industrial field devices exist, the user can select multiple candidate records for side-by-side comparison. The comparison view simultaneously displays complete information such as the image, text description, and tags of the selected industrial field device, and highlights the differences between different industrial field devices to facilitate user differentiation and confirmation. For photo search scenarios, the results display module also provides thumbnail comparisons between the image to be searched and the images of candidate industrial field devices to help users intuitively judge the matching results.

[0100] In one example, the results display module 504 supports secondary filtering and refined searching of search results. When the initial search results are too numerous or not precise enough, the user can add filtering conditions based on the current results, such as further limiting the region, industrial field equipment type, operating status, etc. The system continues to filter the current results and provides a more accurate candidate list.

[0101] In one example, the positioning assistance system also includes a data management module 505, used by the user to edit and modify (e.g., update descriptions, add tags) or delete structured records of industrial field equipment. The data editing and modification function allows users to modify and update stored structured records of industrial field equipment. Users can update image information of industrial field equipment, such as adding new angle photos or replacing old photos with poor quality; users can also update text descriptions of industrial field equipment (such as valves), such as correcting errors in location descriptions, adding missing operating instructions, and adding new tag categories.

[0102] For example, the original structured record of the equipment valve was "Equipment number V-3105, system #3 boiler main steam system, located at a high position on the east side of the boiler room". Now, the valve location has been changed from "high position on the east side of the boiler room" to "high position on the west side of the boiler room", and the data management module has updated the digital record of the industrial field equipment to "Equipment number V-3105, system #2 boiler main steam system, located at a high position on the west side of the boiler room".

[0103] The data deletion function allows users to delete unwanted structured records of industrial field equipment. Deletion can be done individually or in batches. To delete a single structured record, the user selects the target record and confirms the deletion. To delete a batch of records, the user can set deletion conditions, such as deleting records created within a specific time period or deleting records from a specific region. The data deletion function supports two modes: soft deletion and hard deletion. Soft deletion marks the record as deleted, preserving the data for recovery; hard deletion physically erases the data record and associated image files, making it unrecoverable.

[0104] For example, the original structured record of the valve was "Equipment number V-3105, system to which it belongs is the #3 boiler main steam system, located on the high ground on the east side of the boiler room". Now, the valve is being removed, and the data management module performs a hard deletion of the valve record "Equipment number V-3105, system to which it belongs is the #2 boiler main steam system, located on the high ground on the west side of the boiler room".

[0105] In one example, the data deletion function also includes cascading deletion. When deleting an entity from the structured record of industrial field equipment, the system automatically detects the data resources associated with that entity, including associated image files, associated equipment relationships, associated operation logs, etc. Cascading deletion can be set to automatic mode or confirmation mode. In automatic mode, all associated resources are deleted automatically without user confirmation; in confirmation mode, all associated resources to be deleted are listed before deletion, and execution is performed after user confirmation.

[0106] In one example, the positioning assistance system also includes a local area network (LAN) point-to-point data synchronization module 506, used to synchronize structured records of industrial field equipment between multiple thermal power plant equipment databases based on the LAN without relying on a central server. Figure 2 As shown, the local area network point-to-point data synchronization module specifically includes:

[0107] The device discovery unit 5061 is specifically used to discover peer devices in the local area network based on the User Datagram Protocol (UDP) broadcast.

[0108] In one example, the device discovery unit is used to initiate data synchronization when a user clicks the "LAN Synchronize" button on the main interface. The power plant device sending a broadcast message to a designated port on a broadcast address (e.g., 255.255.255.255) within the local area network (LAN) contains the device's name, IP address, and system version information. Other devices running the same system on the same LAN automatically reply with a message containing their own information upon receiving the broadcast. The device initiating the synchronization collects these replies and displays all synchronizable industrial field devices in an online list, allowing the user to select which industrial field devices to receive data synchronization from.

[0109] The connection establishment unit 5062 establishes a long connection with the discovered peer device based on the transmission control protocol; specifically, it is used to establish a transmission connection between the industrial field device that initiates data synchronization and the industrial field device that receives data synchronization.

[0110] In one example, when a user selects an industrial field device from the list to receive data synchronization, the two devices establish a reliable long connection via the Transmission Control Protocol (TCP).

[0111] The incremental synchronization unit 5063 is used to exchange the unique identifier and last update timestamp of the structured record in each device to determine the incremental data list, which represents the structured record that needs to be updated by the device.

[0112] For example, the incremental synchronization unit, after obtaining a list of unique IDs and timestamps for all valve structured records on the other device, compares the timestamps of the local records with those of the remote records one by one. Through this comparison, the system identifies valve structured records that exist on the other device but are not present on its own device (i.e., newly added records), as well as valve structured records on its own device that are older than those on the other device (i.e., records that need to be updated). Finally, the positioning assistance system generates an "incremental data list," where each item is data that truly needs to be synchronized, thus avoiding the resource waste and time loss caused by full synchronization.

[0113] In one example, to improve transmission stability and anti-interference capabilities, the incremental synchronization unit can also be used to divide the structured record into data blocks of a preset size for transmission, and perform a limited number of retries for data blocks that fail to be transmitted. The system transmits structured text data and image files. Text data includes information such as record ID, description, tags, and timestamps. After all text data has been transmitted and confirmed to be correct, the system enters the image file transmission stage.

[0114] In one example, the incremental synchronization unit can also be used to retain the transmitted data blocks when the network is interrupted and the data synchronization process is terminated, and to continue transmitting the unfinished data from the point of interruption during the next synchronization.

[0115] In one example, the local area network point-to-point data synchronization module further includes the conflict handling unit 5064, which is used to retain the newer version based on the timestamp and issue a prompt to the user when the same structured record has version conflicts on different devices.

[0116] In one example, the local area network point-to-point data synchronization module also includes a synchronization completion confirmation unit, which automatically verifies the MD5 hash value of the received data after data transmission is completed to ensure data integrity. Upon successful verification, the new structured records of the industrial field devices are merged into the local database, and a synchronization completion notification is sent to the user.

[0117] In one example, the synchronization completion confirmation unit can also be used to display a prompt box after the entire synchronization process is completed, informing the user which conflicts were automatically resolved during this synchronization and which version was ultimately retained. If the user believes that the automatic selection was incorrect, they can manually revert to the old version.

[0118] In practical implementation, the various modules of the positioning assistance system 500 interact with each other and call functions through standardized interfaces. The information acquisition module 501 sends the structured records of industrial field equipment collected to the data storage module 502 through the data interface; the data storage module 502 responds to the query request of the intelligent retrieval module 503 and returns candidate records; the intelligent retrieval module 503 sends the candidate records to the result display module 504 for presentation.

[0119] In this way, through point-to-point synchronization over a local area network, the sharing of structured records of industrial field equipment across multiple devices can be achieved without an external network. This allows for one-click import of accumulated structured records of industrial field equipment, enabling inspection and maintenance personnel to divide the work and record structured records for different areas of industrial field equipment. After synchronization, a unified digital archive of industrial field equipment is formed throughout the plant, facilitating rapid knowledge transfer and collective accumulation. Simultaneously, multi-device synchronization also enables multiple copies of the structured records of industrial field equipment, effectively avoiding the risk of data loss due to single device failure.

[0120] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs.

[0121] Example 2

[0122] Figure 3 A flowchart of an industrial field equipment positioning assistance method provided for an exemplary embodiment of this disclosure.

[0123] S101. Collect image data and corresponding text description data of industrial field equipment.

[0124] In one example, an image acquisition device can be used to photograph the industrial site equipment, obtaining an image that clearly reflects the appearance characteristics of the target industrial site equipment. For example, the image acquisition device can be a smartphone, tablet, digital camera, or other electronic device with a camera function.

[0125] In one example, the text description data includes at least one of the following: device function description, system information, location feature description, and user-defined tags.

[0126] In one example, the text description data for the valve is: "Isolation valve for the circulating water return pipeline of turbine No. 1, equipment number V-1207. This valve is located on the west wall of the zero-meter level of the turbine hall, adjacent to the outlet pipeline of circulating water pump #1. The valve body is blue, and there is a set of large filters about 2 meters in front of it as a prominent reference. The valve is labeled "#V-1207, turbine No. 1, isolation valve, circulating water pump outlet pipeline".

[0127] S102. The image storage path corresponding to the image data is associated with the text description data and stored as a structured record.

[0128] The structured record includes at least a record identifier, an image storage path, and a text description.

[0129] In one example, the text description content and the user-defined tag are matched with a preset separator to generate a logical derived field, and the matching retrieval is performed based on the logical derived field.

[0130] For example, the preset separator uses the vertical bar symbol "|". The unique identifier (ValveID) of a structured record for industrial field equipment is "123456789", and the text description is "Isolation valve for the circulating water return pipeline of the No. 1 turbine low-pressure cylinder, equipment number V-1207. This valve is located on the west wall of the zero-meter floor of the turbine hall, adjacent to the outlet pipeline of the #1 circulating water pump. The valve body is blue, and there is a set of large filter screens about 2 meters in front of it as a prominent reference." The user-defined tags are "#V-1207", "#1 Steam Turbine", "Isolation Valve", and "#Circulating Water Pump Outlet Pipeline". The logically derived field is "Isolation valve for the circulating water return pipeline of the low-pressure cylinder of Steam Turbine No. 1, equipment number V-1207. This valve is located on the west wall of the zero-meter floor of the turbine hall, adjacent to the outlet pipeline of the #1 circulating water pump. The valve body is blue, and there is a large filter screen about 2 meters in front of it as a prominent reference. | #V-1207, #1 Steam Turbine, Isolation Valve, #Circulating Water Pump Outlet Pipeline".

[0131] In one example, a database index is built using the logically derived fields and the creation timestamp; a retrieval is performed based on the database index, and the search results are arranged in descending order of creation time. For example, the image storage path (ImagePath) of a structured record of industrial field equipment is " / storage / emulated / 0 / ValvePhotos / V-1207_20230429_143022.jpg", and the creation time (CreatedDate) is "2023-04-29 14:30:25". A database index is built using the aforementioned logically derived fields and the creation timestamp; a retrieval is performed based on the database index, and the search results are arranged in descending order of creation time.

[0132] In one example, structured records of industrial field equipment can also be edited (such as updating text descriptions, adding labels) or deleted.

[0133] For example, the original structured record for the valve was "Equipment number V-3105, system #3 boiler main steam system, located at a high point on the east side of the boiler room". Now, the valve location has been changed from "high point on the east side of the boiler room" to "high point on the west side of the boiler room", so the structured record should be updated to "Equipment number V-3105, system #2 boiler main steam system, located at a high point on the west side of the boiler room".

[0134] For example, the original structured record of the valve was "Equipment number V-3105, system #3 boiler main steam system, located on the high ground east of the boiler room". Now, if the valve is to be removed, the valve record "Equipment number V-3105, system #2 boiler main steam system, located on the high ground west of the boiler room" will be deleted.

[0135] S103. Receive the search terms input by the user, and perform matching and retrieval of the structured records based on the search terms.

[0136] In one example, a user's search request for the industrial field equipment may be one or more of the following: text description, voice input, or image recognition.

[0137] In one embodiment, when the search request is a text description of a corresponding industrial field device, the text description is preprocessed to remove spaces before matching the search request with the structured records of the industrial field device. For example, if a user enters the text description "V-3105", the preprocessed text description will be "V-3105" to prevent search failures due to space mismatches.

[0138] In one embodiment, when the search request is a text description of the corresponding industrial field equipment, a delay anti-jitter process is performed before matching the search request with the structured record of the industrial field equipment.

[0139] It is understandable that, in order to avoid the consumption of system resources and flickering of the result list caused by users frequently triggering search requests during continuous input, a preset delay (such as 300 milliseconds) is set after user input is detected. When the user stops inputting for more than the preset delay, the system is matched with the structured records of industrial field equipment to find all structured records of industrial field equipment that contain or are similar to the above text description, and a list of industrial field equipment matching results is obtained for the user to compare and confirm.

[0140] For example, if a user wants to search for "#3 furnace drain valve", the positioning assistance system waits for 300 milliseconds when "#3 furnace" is entered. If the user continues to enter "drain valve" within 300 milliseconds, the timer resets until the user completely enters "#3 furnace drain valve" and pauses for more than 300 milliseconds before finally matching "#3 furnace drain valve" as the text description with the structured records of industrial field equipment in the database.

[0141] S104. Display the retrieved structured records in a graphic and textual format to help users identify the target device.

[0142] In one example, the results are displayed as a list of images and text, with each list item containing at least a device thumbnail, key text descriptions, and tag information. In response to a user's selection of a list item, the corresponding complete image and all text description data are displayed. Users can click on a matching result to view a larger image and complete information, enabling comparison between the search results and the on-site image.

[0143] In one example, when identifying the target industrial field equipment, location navigation services are supported, such as displaying the navigation path from the current location to the target industrial field equipment, estimating the walking distance and time, etc., through the inertial sensor and Bluetooth beacon positioning technology of the mobile terminal.

[0144] In one example, when multiple candidate industrial field devices exist, users can select multiple candidate records for side-by-side comparison. The comparison view simultaneously displays complete information such as the image, text description, and tags of the selected industrial field device, and highlights the differences between different industrial field devices to facilitate user differentiation and confirmation. For image search scenarios, the results display module also provides thumbnail comparisons between the image to be searched and the images of candidate industrial field devices to help users intuitively judge the matching results.

[0145] In one example, the matching results list can be further filtered and refined. When the initial search yields too many or insufficiently precise results, the user can add filtering conditions to the current results, such as further specifying the region, industrial field equipment type, or operating status. The system will then continue filtering the current results to provide a more accurate candidate list.

[0146] In this embodiment of the disclosure, the industrial field equipment positioning assistance method includes three core components: a data acquisition process, a data storage process, and a data retrieval process.

[0147] Reference Figure 4 The following describes the process of storing structured records of industrial field equipment:

[0148] S201: Identify the industrial field equipment to be collected.

[0149] When maintenance personnel discover an industrial field device that may require logging, they first search the database using the known device number, location, or functional description. If no structured record matching the industrial field device exists in the database, the device is identified as the "target industrial field device to be logged." Conversely, if a matching structured record for the industrial field device exists in the database and the information is accurate and valid, and after on-site verification, no logging operation is required.

[0150] S202: Acquire images of target industrial equipment.

[0151] In one example, an image acquisition device can be used to photograph the industrial site equipment, obtaining an image that clearly reflects the appearance characteristics of the target industrial site equipment. For example, the image acquisition device can be a smartphone, tablet, digital camera, or other electronic device with a camera function.

[0152] For example, for valves located at high or difficult-to-access locations, the image acquisition device can be a telescopic camera, a drone, or other auxiliary shooting equipment.

[0153] It is understood that when taking pictures of the target industrial field equipment, the target industrial field equipment occupies a sufficiently large proportion in the image to clearly display information such as the valve type, structural features, and nameplate markings.

[0154] S203: Record text description data of the target industrial field equipment image.

[0155] In one example, when capturing images of target industrial equipment, a text description of the target industrial equipment image is recorded. It is understood that the text description data is an important supplement to the industrial equipment image, containing textual content such as the location description, functional description, to which the equipment belongs, its relationship with surrounding equipment, and operating precautions.

[0156] In one example, the text description can be recorded through the text input interface of a mobile terminal, or it can be entered via voice input and then converted into text.

[0157] For example, the text description of the valve is "Isolation valve for the return water pipeline of the low-pressure cylinder of turbine No. 1, equipment number V-1207. The valve is located on the west wall of the zero-meter floor of the turbine hall, adjacent to the outlet pipeline of the #1 circulating water pump. The valve body is blue, and there is a set of large filter screens about 2 meters in front of it as a prominent reference."

[0158] In one example, corresponding tags can also be added to the text description of the target industrial field equipment image. It can be understood that adding tag fields to the text description of industrial field equipment allows for the classification of industrial field equipment according to different dimensions, facilitating batch management and conditional retrieval.

[0159] In one example, the label field can be a pre-defined standardized label, such as "high temperature and high pressure", "hazardous medium", "frequent operation", "maintenance lockout", etc.

[0160] In one example, the label field can also be a custom label derived from the text description of the equipment in the industrial field. For example, the text description of the valve is "Isolation valve for the circulating water return pipeline of the No. 1 turbine low-pressure cylinder, equipment number V-1207. This valve is located on the west wall of the zero-meter floor of the turbine hall, adjacent to the outlet pipeline of the #1 circulating water pump. The valve body is blue, and there is a set of large filters about 2 meters in front of it as a prominent reference." The labels of the valve are "#V-1207", "#1 turbine", "isolation valve", and "#circulating water pump outlet pipeline".

[0161] Reference Figure 5The following describes the data entity of the structured record of the target industrial field equipment. The structured record of the industrial field equipment is a data structure containing complete attribute information of the industrial field equipment. Its core data structure includes the following fields:

[0162] Unique Identifier (ValveID): A string representing a unique identifier for a structured record of industrial field equipment, used to identify and index these records. Each unique identifier for the process field equipment is unique within the database and cannot be duplicated.

[0163] Image storage path (ImagePath): This is a string representing the path or access address of an industrial field equipment image in the database. The path format uses standard file system path notation, such as " / storage / emulated / 0 / ValvePhotos / V-1207_20230429_143022.jpg". The database supports storing multiple industrial field equipment images from different angles or in different states, which can be managed by separating them with commas or by creating an image association table.

[0164] The text description (Description) is a text data type and includes textual content such as the location description, function description, process system to which it belongs, relationship with surrounding equipment, and operating precautions for the industrial site equipment. The text description field supports multi-line text and can include paragraphs and list structures to comprehensively record information about the industrial site equipment. For example, the text description (Description) could be: "Isolation valve for the circulating water return pipeline of the No. 1 turbine low-pressure cylinder, equipment number V-1207. This valve is located on the west wall of the zero-meter level of the turbine hall, adjacent to the outlet pipeline of the #1 circulating water pump. The valve body is blue, and there is a large filter screen approximately 2 meters in front of it as a prominent reference point."

[0165] Tags: This is a string array, representing a collection of tags for categorized management. Each tag is an independent string and can contain pre-defined standardized tags (such as "high temperature and high pressure", "hazardous media", "frequent operation", "maintenance lockout") or custom tags based on the valve's text description (such as "#V-1207", "#1 steam turbine", "isolation valve", "#circulating water pump outlet pipeline"). The tag field supports multiple tags for a single valve.

[0166] Creation Time (CreatedDate): The data type is date and time, and the format is "YYYY-MM-DD HH:mm:ss". The creation time is used to record the creation time of the structured record of the industrial field equipment when it is entered into the database, which facilitates traceability and version management. For example, the creation time (CreatedDate) is "2023-04-29 14:30:25".

[0167] The logically derived field (SearchText) is a text data type. It is used for full-text search indexing, supporting Chinese word segmentation retrieval and fuzzy matching to improve search efficiency and accuracy. The logically derived field is formed by segmenting the text description (Description) field and tags, removing stop words, expanding synonyms, and performing semantic analysis on the text description data to identify key elements such as location information, system information, and device information. Based on preset extraction rules, corresponding information fragments are extracted from the text description and tags to form the logically derived field.

[0168] In one example, when the logically derived field is a location hierarchy description field, this field is used to represent the hierarchical location of industrial field equipment within the power plant's spatial structure. By parsing the text description, information such as the area (e.g., main plant, coal conveyor bridge), floor height (e.g., zero meter, 12.6-meter operating floor), and column position (e.g., column A, column B) of the industrial field equipment can be identified and organized into strings according to hierarchical relationships, such as the logically derived field "main plant 12.6-meter operating floor, column B". This field supports filtering and retrieval by area or floor height.

[0169] In one example, when the logically derived field is an associated device information field, this field is used to represent other devices that have a physical or logical association with the target industrial field device. By performing entity recognition on the text description, information on devices such as pumps, pipes, containers, and instruments associated with the target industrial field device is extracted, and association relationships are established. The associated device information supports bidirectional queries; users can search for associated devices by industrial field device, or search for associated industrial field devices by device. This field is used to analyze device topology relationships and assist in fault diagnosis.

[0170] In one example, when the logically derived field is an operation-related information field, this field is used to represent prompts related to the operation of industrial field equipment. By performing pattern matching on the text description, operation guidance information such as the operating direction (e.g., clockwise to close, counterclockwise to open), operating torque, and precautions can be extracted. This operation-related information, as a logically derived field, can be presented along with the search results, helping users quickly understand how to operate the industrial field equipment and reducing the risk of misoperation.

[0171] In one example, the database also allows users to manually add and edit logically derived fields to suit more personalized management needs. Users can define custom logically derived field types based on the actual management needs of the power plant, such as equipment status assessment and inspection priority evaluation.

[0172] The above fields constitute the basic data structure of structured records for industrial field equipment. When storing structured records of industrial field equipment, a database can be managed using a relational database, or it can be stored using a document-based database or a key-value database. Alternatively, multiple database technologies can be used in combination. For example, a relational database can be used to store structured index information, an object storage service can be used to store image files, and a full-text search engine can be used to provide efficient retrieval capabilities. No specific limitations are made here.

[0173] In one example, structured records of frequently used or important industrial field equipment from the database can be added to a favorites folder for quick access. The favorites folder supports custom grouping, allowing users to categorize and organize their favorite industrial field equipment according to their management needs.

[0174] In one example, the structured records of industrial field equipment can be generated into a shareable link or QR code and shared with other users. The shareable link or QR code contains a functional description of the industrial field equipment, its system, and location characteristics.

[0175] Reference Figure 6 The following describes the process of retrieving structured records of industrial field equipment from the matching database:

[0176] S301: The user initiates a search request.

[0177] A user's search request for the industrial field equipment can be one or more of the following: text description, voice input, or image recognition.

[0178] In one example, the search request is a text description of the corresponding industrial field equipment. In response to the search request for the aforementioned text description, it is matched against structured records of industrial field equipment in the database to find all structured records of industrial field equipment that contain or resemble the aforementioned text description, resulting in a list of matching industrial field equipment for user comparison and confirmation.

[0179] When the search request is a text description of the corresponding industrial field equipment, users can directly enter keywords for basic searching. For example, if a user enters "circulating water pump," it will return all structured records of industrial field equipment containing that keyword. Keywords can match any field, including unique identifiers, text descriptions, tags, and logically derived fields for industrial field equipment.

[0180] When the search request is a text description of the corresponding industrial field equipment, users can use logical operators to combine multiple keywords for combined retrieval. For example, if a user enters "circulating water pump AND electric valve", the system will return structured records of valves that contain both keywords "circulating water pump" and "electric valve"; if a user enters "circulating water pump OR feed water pump", the system will return structured records of valves that contain either the keyword "circulating water pump" or "electric valve".

[0181] When the search request is a text description of the corresponding industrial field equipment, users can set range conditions for numeric fields to perform range searches. For example, they can search for structured valve records created within a specific time period, "October 31, 2023 to December 31, 2023". They can also search for structured valve records within a specific floor height range, where "the valve's distance from the ground is between 5 and 10 meters".

[0182] When the search request is a text description of the corresponding valve, users can use wildcards for fuzzy matching. For example, entering "V-*" will match all structured records of valves whose valve numbers begin with "V"; entering "xun*" will match all structured records of valves whose numbers begin with "xun".

[0183] In one embodiment, when the search request is a text description of the corresponding valve, the text description is subjected to delayed anti-jitter preprocessing before matching the search request with the structured records of industrial field equipment in the database.

[0184] It is understandable that, to avoid system resource consumption and result list flickering caused by frequent search requests during continuous user input, a preset delay (e.g., 300 milliseconds) is set after user input is detected. When the user stops inputting for more than the preset delay, the system is matched against structured records of industrial field equipment in the database to find all structured records of industrial field equipment that contain or are similar to the above text description, resulting in a list of industrial field equipment matching results for the user to compare and confirm.

[0185] For example, if a user wants to search for "#3 boiler steam trap", the system will wait 300 milliseconds when the user enters "#3 boiler". If the user continues to enter "steam trap" within 300 milliseconds, the timer will reset. Only after the user has completely entered "#3 boiler steam trap" and paused for more than 300 milliseconds will "#3 boiler steam trap" be used as the text description and matched with the valve structured record in the database.

[0186] In one embodiment, when the search request is a text description of a corresponding valve, the text description is preprocessed to remove spaces before matching the search request with the structured records of industrial field equipment in the database. For example, if the user enters the text description "V-3105", the preprocessed text description will be "V-3105" to prevent search failures due to space mismatches.

[0187] In one example, the search request is a voice signal obtained through voice input. The user presses the voice input button on a mobile terminal or dedicated device and speaks out the search criteria. The voice signal is then converted into a text description, and the search is performed according to the text description.

[0188] For example, maintenance personnel can input the voice signal "Find the No. 3 boiler drain valve" and the voice signal is converted into the text query string "No. 3 boiler drain valve" in real time. After receiving the converted text string, it is matched with the structured valve records in the database, and a list of valve matching results related to text descriptions such as "No. 3 boiler" and "drain" is returned for the user to compare and confirm.

[0189] In this way, by using voice input to search for items, users' hands are freed up. In scenarios where manual input is inconvenient during inspections, this improves the convenience of human-computer interaction and further shortens the time spent searching for equipment in industrial settings, thereby improving overall operational efficiency and safety.

[0190] In one example, the search request is to take a picture of the target industrial field equipment. The picture of the target industrial field equipment is then used to match the corresponding structured records of industrial field equipment in the database.

[0191] For example, maintenance personnel take a photo of a valve that is difficult to identify. The uploaded image undergoes feature extraction and analysis to identify key objects in the image (such as valve type, number plate, pipe color), and generates a set of textual description keywords (such as "red handwheel," "near pressure gauge," "overhead pipeline"). These keywords are then used as search criteria and matched against structured valve records in the database to find a list of matching valves with similar visual features, allowing users to compare and confirm their selection.

[0192] In one example, when the search request is to capture images of target industrial equipment, the user can use their mobile device's camera to take pictures of the equipment. For situations where it's difficult to capture clear images in a single shot, burst mode and real-time preview are supported to help users obtain the best possible images.

[0193] In one example, when matching a captured image of a target industrial field device as a search request with structured records of industrial field devices in a database, the captured image of the target industrial field device can be preprocessed.

[0194] In one example, the images of the target industrial equipment can be processed by size normalization, contrast adjustment, and color space conversion. Size normalization adjusts the image to a uniform size, facilitating subsequent processing and feature extraction; contrast adjustment enhances image clarity, making the outlines and details of the industrial equipment more prominent; and color space conversion converts the image into a color representation suitable for feature extraction.

[0195] In one example, motion blur restoration can be performed on on-site images of target industrial equipment. When the image contains linear blur caused by the movement of the capturing equipment, deconvolution processing is performed on the image based on the blur result to restore the sharp edges and detailed textures of the image.

[0196] In one example, out-of-focus blur restoration can be performed on on-site images of target industrial equipment. For instance, when the on-site image of the target industrial equipment has concentric circle blur caused by inaccurate focus, a deconvolution method based on the point spread function can be used for image restoration.

[0197] In one example, noise suppression processing can be applied to images of industrial equipment. For instance, when images of the target industrial equipment contain significant noise due to low lighting, a trained neural network model can directly predict a clean image, effectively removing noise while preserving details.

[0198] In one example, a search request can combine multiple input formats simultaneously.

[0199] For example, a user can first input "high-temperature section of boiler room" via voice, then manually add the text label "pressure reducing valve," or simultaneously upload a photo of the area's characteristic environment. By logically combining the voice-converted text, the manually entered text, and the text description generated by image recognition, a comprehensive query condition is constructed. Based on this comprehensive query condition, it is matched against the structured records of industrial field equipment in the database, thereby accurately locating a list of valve matching results that meet all the characteristics, for the user to compare and confirm.

[0200] It is understood that the search request is not limited to the above-described form. In possible implementations, the form of the search request can be diverse, including but not limited to text input, voice input, photo recognition input, or any combination thereof listed in the above embodiments. Those skilled in the art should understand that any input format capable of converting a user's intention to locate industrial field equipment into a processable and retrieval-performing format, as long as its achieved technical effect is "rapid location of industrial field equipment based on a search request," falls within the protection scope of this invention.

[0201] In one example, a user's search history can be automatically recorded, including search criteria, search time, and search results. Users can quickly repeat previous searches from the history or view previous search results to confirm the target industrial field equipment.

[0202] S302: Match the search request with the structured records of industrial field equipment in the database.

[0203] In one example, when the search request is a text description, a matching query is performed between the preprocessed text description and the logically derived fields of the structured records of industrial field equipment in the database.

[0204] In one example, the database can be matched with structured records of industrial field equipment using methods such as exact matching, fuzzy matching, or semantic matching.

[0205] It is understandable that exact matching is used when the industrial field equipment number or exact name is known, and the returned matching results completely contain the query conditions; fuzzy matching is used when only the department's industrial field equipment information is known, and it can return all structured records of industrial field equipment containing the query keywords, and supports synonym expansion and spelling correction; semantic matching is used for queries described using natural language, and it can understand the semantic intent of the query and return semantically related structured records of industrial field equipment.

[0206] In one example, when the search request is for image recognition, the system analyzes user-uploaded images of industrial equipment. A convolutional neural network model extracts high-dimensional vector representations of the equipment's appearance features (such as valve type, color, markings, and surrounding objects). The extracted high-dimensional vector representations are then compared to the feature vectors of industrial equipment images stored in the system (using methods such as cosine similarity or Euclidean distance). A similarity score is calculated between the image being recognized and each candidate industrial equipment image in the database, and the matching results are returned in descending order of score. If the database stores multiple images of the same industrial equipment from different angles, the score corresponding to the image with the highest similarity to the image being recognized is selected as the final score for the structured record of that industrial equipment.

[0207] In one example, when the search request is based on speech recognition, the received user voice input is converted into a text string. This converted text string undergoes the same preprocessing steps as the "text description" (such as removing spaces and debouncing). A fuzzy matching query is then performed between the preprocessed text string and logically derived fields of the structured records of industrial field equipment in the database.

[0208] S303: Returns a list of matching results.

[0209] In one example, after executing a matching query based on a search request, the returned matching results are presented to the user in a list format combining text and images. Each matching result also includes a thumbnail, a text description of the industrial field equipment, and tags.

[0210] In one example, the matching results list can be sorted by relevance score, which takes into account multiple factors such as keyword matching degree, field weight, and creation time. Records with higher relevance scores are listed first, making it easier for users to quickly locate them.

[0211] In one example, the list of matching results can be filtered according to user-defined criteria, such as displaying only industrial field equipment records for specific systems or regions.

[0212] In one example, the matching results list supports multiple view modes. For instance, the list view displays structured records of candidate industrial field equipment in a table format, with each row showing columns of key information for the industrial field equipment, and supports sorting and pagination; the card view displays structured records of candidate industrial field equipment in a card format, with each card containing a thumbnail of the industrial field equipment and key information for easy browsing; and the map view marks candidate industrial field equipment on a power plant floor plan or 3D model, visually showing the location distribution of the industrial field equipment, and users can click on map markers to view details of the industrial field equipment.

[0213] S304: The user browses the list of matching results to confirm the target industrial field equipment.

[0214] In one example, by clicking on any item in the matching results list, a high-resolution image of the industrial field equipment corresponding to the structured record of that equipment, along with complete record information (such as detailed functional descriptions, location features, creation time, etc.), is displayed. This enables a direct comparison between the matching results and the field image, ultimately helping the user to accurately identify and locate the target industrial field equipment.

[0215] In one example, when identifying the target industrial field equipment, location navigation services are supported, such as displaying the navigation path from the current location to the target industrial field equipment, estimating the walking distance and time, etc., through the inertial sensor and Bluetooth beacon positioning technology of the mobile terminal.

[0216] In one example, when multiple candidate industrial field devices exist, users can select multiple candidate records for side-by-side comparison. The comparison view simultaneously displays complete information such as the image, text description, and tags of the selected industrial field device, and highlights the differences between different industrial field devices to facilitate user differentiation and confirmation. For image search scenarios, the results display module also provides thumbnail comparisons between the image to be searched and the images of candidate industrial field devices to help users intuitively judge the matching results.

[0217] In one example, the matching results list can be further filtered and refined. When the initial search yields too many or insufficiently precise results, the user can add filtering conditions to the current results, such as further specifying the region, industrial field equipment type, or operating status. The system will then continue filtering the current results to provide a more accurate candidate list.

[0218] In one example, users can also edit and modify the structured records of industrial field equipment in the database (such as updating descriptions and adding tags). Users can modify and update the stored structured records of industrial field equipment. For example, users can update the image information of industrial field equipment, such as adding new angle photos or replacing old photos with poor quality; users can also update the text descriptions of industrial field equipment, such as correcting errors in location descriptions, adding missing operating instructions, and adding new tag categories.

[0219] For example, the original structured record of the valve was "Equipment number V-3105, system #3 boiler main steam system, located at a high point on the east side of the boiler room". Now, the valve location has been changed from "high point on the east side of the boiler room" to "high point on the west side of the boiler room", and the data management module updates the valve number to "Equipment number V-3105, system #2 boiler main steam system, located at a high point on the west side of the boiler room".

[0220] In one example, users can also delete structured records of industrial field equipment in the database, allowing them to remove unwanted records. Deletion can be done single-record or in batches. To delete a single record, the user selects the target record and confirms the deletion. To delete a batch of records, users can set deletion conditions, such as deleting records created within a specific time period or a specific region. Data deletion supports both soft and hard deletion modes. Soft deletion marks the record as deleted, preserving the data for recovery; hard deletion physically erases the data record and associated image files, making it unrecoverable.

[0221] For example, the original structured record of the valve was "Equipment number V-3105, system #3 boiler main steam system, located on the high ground east of the boiler room". Now, if the valve is to be removed, the valve record "Equipment number V-3105, system #2 boiler main steam system, located on the high ground west of the boiler room" will be hard deleted.

[0222] In this way, by storing on-site images and environmental features of industrial equipment in a database to create a visualized digital archive, and enabling fast and accurate retrieval through search requests, the shortcomings of maintenance personnel finding remote industrial equipment, such as time-consuming and error-prone locations, are solved. The search results (such as precise descriptions of industrial equipment and on-site photos) are displayed to the operators for confirmation, eliminating misidentification, improving operational safety and accuracy, and significantly improving valve locating efficiency and operational accuracy.

[0223] In one example, different industrial field devices connected to the same local area network can synchronize structured records across multiple devices, such as... Figure 7 As shown, the specific methods include:

[0224] S401: Broadcast detected online devices on the local area network.

[0225] When a user clicks the "LAN Synchronize" button on the main interface, the device initiating the data synchronization sends a broadcast message to a designated port on the local area network (e.g., 255.255.255.255). This message contains the device's name, IP address, and system version information. Other devices running the same system on the same local area network automatically reply with a message containing their own information upon receiving the broadcast. The industrial field device initiating the data synchronization collects these replies and displays all synchronizable devices in an online list, allowing the user to select which industrial field device to receive the data synchronization from.

[0226] S402: Establish a long-lived connection with the discovered peer device based on the Transmission Control Protocol.

[0227] After the user selects an industrial field device from the list to receive data synchronization, the two devices establish a reliable long-term connection via Transmission Control Protocol (TCP). Once the connection is successfully established, a data exchange identifier is created between the industrial field device initiating the data synchronization and the industrial field device receiving the data synchronization. This data exchange identifier includes a unique ID from all digital records of industrial field devices in the databases of both devices, along with their corresponding "last update timestamp."

[0228] S403: Exchange the unique identifier and last update timestamp of structured records in industrial field equipment to determine the incremental data list.

[0229] After obtaining the unique IDs and timestamp lists of all digital records on the other device, the timestamps of the local records are compared one by one with the timestamps of the remote records. Through this comparison, the system quickly identifies which records exist on the other device but not on its own (i.e., newly added records) and which records have a local version that is older than the one on the other device (i.e., records that need to be updated). A concise "incremental data list" is generated, where each item is data that actually needs to be synchronized, avoiding the resource waste and time loss caused by full synchronization.

[0230] S404: Transfer file data based on the incremental data list

[0231] Based on the generated incremental data list, the system begins transmitting structured text data, including record IDs, descriptions, tags, and timestamps. Once all text data has been transmitted and confirmed to be correct, the system proceeds to the image file transmission stage.

[0232] In one example, to improve transmission stability and interference resistance, each image data is divided into fixed-size (e.g., 1MB) blocks for "chunk transmission" during transmission. If a data block fails to transmit due to network fluctuations, the system will automatically retry, up to a maximum of three times.

[0233] In one example, during data synchronization, if the system finds that the ID of a certain industrial field device digital record on the other device is exactly the same as the ID of an industrial field device digital record in the local database, but the content is different, the system will automatically read the "last update timestamp" of both industrial field device structured records. According to preset rules, the system retains the version with the updated timestamp as the final industrial field device digital record. After the entire synchronization process is complete, the system will display a prompt informing the user which conflicts were automatically resolved and which version was ultimately retained. If the user believes the automatic selection was incorrect, they can still manually revert to the old version.

[0234] In one example, the entire digital record synchronization process does not necessarily have to be continuous and uninterrupted. Users can pause or cancel synchronization at any time during the process using buttons on the interface. Data that has been transferred and written to the local database will be automatically retained by the system. When the user initiates synchronization with the same device again, the system will automatically identify any unfinished synchronization tasks based on the timestamp of the previous synchronization completion, achieving "resume interrupted transmission" and synchronizing only the remaining unfinished data blocks.

[0235] S405: Data integrity verification, confirming synchronization is complete.

[0236] After all industrial field device digital records have been transmitted, the system automatically performs data integrity verification. The receiving device calculates the MD5 hash value of all received digital records, especially image files, and compares this value with the MD5 hash value calculated by the sending end before transmission. Only when the hash values ​​of all data blocks match perfectly will the system confirm that the synchronized industrial field device digital records are complete and tamper-proof. After successful verification, the system merges the newly received industrial field device digital records into the local database, updates the user interface display, and finally sends a "synchronization complete" notification to the user.

[0237] In this way, point-to-point synchronization via local area network enables the sharing of digital archives among multiple devices without the need for an external network. It allows for one-click import of accumulated digital records of industrial field equipment data. Inspection and maintenance personnel can divide the work to record digital records of industrial field equipment in different areas, and after synchronization, a unified digital archive of industrial field equipment is formed throughout the plant, enabling rapid knowledge transfer and collective accumulation. Simultaneously, multi-device synchronization also provides multiple copies of the digital records of industrial field equipment, effectively avoiding the risk of data loss due to the failure of a single device.

[0238] Example 3

[0239] Figure 8This is a schematic diagram of the structure of an electronic device according to an example embodiment of the present disclosure. The electronic device includes a memory, a processor, and a computer program stored in the memory and used to run on the processor. When the processor executes the computer program, it implements the positioning assistance method of any of the above embodiments. Figure 8 The electronic device 600 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0240] like Figure 8 As shown, the electronic device 600 can be manifested as a general-purpose computing device, such as a server device. The components of the electronic device 600 may include, but are not limited to: at least one processor 601, at least one memory 602, and a bus 603 connecting different system components (including memory 602 and processor 601).

[0241] Bus 603 includes a data bus, an address bus, and a control bus.

[0242] The memory 602 may include volatile memory, such as random access memory (RAM) 6021 and / or cache memory 6022, and may further include read-only memory (ROM) 6023.

[0243] The memory 602 may also include a program tool 6025 (or utility) having a set (at least one) program module 6024, such program module 6024 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0244] The processor 601 executes various functional applications and data processing by running computer programs stored in the memory 602, such as the positioning assistance method provided in any of the above embodiments.

[0245] Electronic device 600 can also communicate with one or more external devices 604 (e.g., keyboard, pointing device, etc.). This communication can be performed via input / output (I / O) interface 605. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 606. Network adapter 606 communicates with other modules of electronic device 600 via bus 603. It should be understood that, although... Figure 8 As not shown, other hardware and / or software modules may be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.

[0246] It should be noted that although several units / modules or sub-units / modules of the electronic device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.

[0247] Example 4

[0248] This disclosure also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the positioning assistance method provided in any of the above embodiments.

[0249] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0250] Example 5

[0251] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements any of the above-described positioning assistance methods.

[0252] The program code for executing the computer program product disclosed herein can be written in any combination of one or more programming languages. The program code can be executed entirely on a user device, partially on a user device, as a stand-alone software package, partially on a user device and partially on a remote device, or entirely on a remote device.

[0253] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of the present invention is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but all such changes and modifications fall within the scope of protection of the present invention.

Claims

1. An industrial field equipment positioning auxiliary system, characterized in that, include: The information acquisition module is used to collect image data and corresponding text description data of industrial field equipment; The data storage module is used to associate and store the image storage path corresponding to the image data with the text description data as a structured record. The structured record includes at least a record identifier, an image storage path, and text description content. The intelligent retrieval module is used to receive search terms input by the user and perform matching retrieval on the structured records based on the search terms; The results display module is used to display the retrieved structured records in a graphic and textual format to help users identify the target device.

2. The system according to claim 1, characterized in that, The text description data includes at least one of the following: device function description, system information, location feature description, and user-defined tags.

3. The system according to claim 2, characterized in that, The data storage module also includes logically derived fields, which are generated based on the text description content and the user-defined tags with a preset separator. The intelligent search module performs the matching search based on the logically derived fields.

4. The system according to claim 3, characterized in that, The structured record also includes a creation timestamp, and the data storage module can also be used to build a database index based on the logically derived field and the creation timestamp; The intelligent search module is used to perform searches based on the database index and sort the search results in reverse order of creation time.

5. The system according to claim 1, characterized in that, The intelligent search module can also be used to remove spaces from the search terms entered by the user and perform a search on the text description content through wildcard matching.

6. The system according to claim 1, characterized in that, The intelligent retrieval module may also include a debounce unit, specifically used to trigger a retrieval operation when a user input pause exceeds a preset delay threshold.

7. The system according to claim 1, characterized in that, The results display module displays the search results in the form of a list of images and text. Each list item includes at least a device thumbnail, key text descriptions, and tag information. In response to the user's selection of a list item, display the corresponding full image and all text description data.

8. The system according to claim 1, characterized in that, The system also includes a local area network (LAN) point-to-point data synchronization module, used to achieve structured record synchronization among multiple devices in an environment without external network coverage. The data synchronization module includes: The device discovery unit discovers peer devices within the local area network based on User Datagram Protocol (UDP) broadcast. The connection establishment unit establishes a long-lived connection with the discovered peer devices based on the transmission control protocol; An incremental synchronization unit is used to exchange the unique identifier and last update timestamp of the structured record in each device to determine an incremental data list, which represents the structured record that needs to be updated by the device. The conflict handling unit is used to retain the newer version based on the timestamp and issue a prompt to the user when version conflicts exist for the same structured record on different devices.

9. The system according to claim 8, characterized in that, The incremental synchronization unit can also be used to divide the structured record into data blocks of a preset size for transmission, and to perform a limited number of retries for data blocks that fail to be transmitted. And / or, The incremental synchronization unit can also be used to retain the transmitted data blocks when the network is interrupted and the data synchronization process is terminated, and to continue transmitting the unfinished data from the point of interruption during the next synchronization.

10. A method for assisting in the positioning of equipment in an industrial field, characterized in that, Includes the following steps: Collect image data and corresponding text description data of industrial field equipment; The image storage path corresponding to the image data is associated with the text description data and stored as a structured record. The structured record includes at least a record identifier, an image storage path, and text description content. Receive user-inputted search terms and perform matching and retrieval on the structured records based on the search terms; The retrieved structured records are displayed in a graphic and textual format to help users identify the target device.

Images