Rail transit vehicle base and infrastructure engineering equipment data asset creation method

Abstract

The invention discloses a method for creating data assets of rail transit vehicle bases and infrastructure engineering equipment. The method comprises the following steps of: digitally acquiring full-life-cycle data of operation and maintenance equipment in a rail-mounted area of an existing line and a newly-built line; preprocessing including data cleaning and space-time alignment is carried out on the collected rail-mounted area operation and maintenance equipment data; performing equipment data asset construction including digital identity card generation and three-dimensional digital base map generation on the preprocessed rail-mounted area operation and maintenance equipment data; and performing intelligent application including fault diagnosis and maintenance decision generation, operation and maintenance equipment health degree prediction and data asset value evaluation based on the obtained data equipment assets. According to the method, a full-life-cycle data asset architecture is constructed, the requirements of digital twinning on'full-stage, high-correlation and traceable 'data are met, and intelligent operation and maintenance decision making is supported.

Description

Technical Field

[0001] This invention relates to the field of digital technology for rail transit, specifically to a method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment. Background Technology

[0002] The current digital development of rail transit exhibits two major trends: First, newly built lines generally adopt BIM (Building Information Modeling) technology to achieve collaborative design, construction, and operation and maintenance, but this largely remains at the "model visualization" level and is not deeply integrated with equipment management systems. Second, existing lines face the need for "digital transformation," but their historical data (paper drawings, discrete electronic documents) suffers from extremely low utilization rates due to heterogeneous formats and aging media, making it difficult to support digital twin applications. The industry's data assetization rate is less than 30%, and the seamless integration of data throughout the entire lifecycle has become a core bottleneck.

[0003] The specific shortcomings or deficiencies of the existing technology are as follows:

[0004] 1. Deficiencies in existing data governance: Existing solutions only address the "data collection" problem and do not provide effective solutions for OCR recognition errors of paper drawings and the completion of discrete data, thus failing to form complete data assets;

[0005] 2. Data disconnect in new lines: The BIM model and equipment management system use a custom interface, which has poor compatibility and data cannot be automatically synchronized to the operation and maintenance system, resulting in the loss of data value;

[0006] 3. Insufficient data correlation dimensions: Equipment maintenance data (such as fault records and replaced parts) are stored separately from infrastructure status data, lacking a three-dimensional correlation of "space-time-fault", which cannot support root cause analysis;

[0007] 4. Low spatiotemporal fusion accuracy: The existing multi-source data coordinate transformation error exceeds 5mm, which cannot meet the "millimeter-level mapping" requirement of digital twins, resulting in insufficient data utilization.

[0008] 5. Incomplete full lifecycle coverage: Traditional solutions only cover a single stage, such as the "construction period" or the "operation and maintenance period," and do not form a full lifecycle data asset from "equipment delivery - construction and installation - operation and maintenance - scrapping," which cannot support the long-term application of digital twins. Summary of the Invention

[0009] This application provides a method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment to solve the above-mentioned technical problems existing in the prior art.

[0010] According to a first aspect, one embodiment provides a method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment, the method comprising:

[0011] Digitally collect data on the entire lifecycle of track maintenance equipment for both existing and newly built lines;

[0012] The collected data on track area operation and maintenance equipment is preprocessed, including data cleaning and spatiotemporal alignment.

[0013] The preprocessed track area operation and maintenance equipment data is used to construct equipment data assets, including digital ID generation and 3D digital base map generation.

[0014] Based on the obtained data, intelligent applications are developed for equipment assets, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment.

[0015] Furthermore, digital collection of the entire lifecycle data of track maintenance equipment on existing and newly built lines will be conducted, specifically including:

[0016] Data collection for maintenance equipment in existing track areas, including:

[0017] Historical data supplementation includes: extracting model and serial number information from the nameplate of the photographic equipment; extracting information from paper reports using OCR recognition technology; and scanning the entire length of existing rail sections using a laser profilometer.

[0018] Real-time data acquisition, including:

[0019] For testing equipment: install a 4G / 5G module on the equipment to upload testing data in real time; connect the equipment to a portable data logger to synchronously store testing data and collection timestamps;

[0020] For maintenance equipment: Install multiple sensors on the equipment to collect equipment status data and link it to the corresponding rail section;

[0021] The auxiliary data collection includes: embedding RFID tags in the rail sleepers at preset intervals to store the rail code, material and laying time. When the inspection and maintenance equipment is in operation, the RFID reader automatically identifies the rail and avoids data mismatch with the section.

[0022] Furthermore, digital collection of the entire lifecycle data of track maintenance equipment on existing and newly built lines will be conducted, specifically including:

[0023] Data collection for operation and maintenance equipment in the track area of ​​newly built railway lines, including:

[0024] BIM collaborative data collection includes: establishing associated nodes for rail sections and maintenance equipment in the BIM model of newly built lines, and directly writing equipment parameters and construction period inspection data into the BIM model attribute column;

[0025] Construction monitoring interface development includes: developing standardized interfaces to connect to the detection modules of rail laying equipment, collecting initial rail gauge and smoothness data in real time, and automatically synchronizing them to the equipment data management system;

[0026] The equipment embedding design includes: pre-installing data acquisition modules on maintenance equipment at the factory to automatically upload equipment startup time, operation duration, and consumable consumption, eliminating the need for on-site manual recording.

[0027] Furthermore, the collected track area operation and maintenance equipment data undergoes preprocessing, including data cleaning and spatiotemporal alignment, specifically including:

[0028] Data processing includes:

[0029] Wavelet transform algorithm is used to remove waveform noise, and feature points are automatically identified by thresholding method;

[0030] Outliers are removed using the 3σ principle, and then the data is smoothed using a moving average to improve accuracy.

[0031] Maintenance data processing includes:

[0032] The OCR+NLP algorithm is used to parse unstructured text data, extract key information, and convert it into structured data.

[0033] The Kalman filter algorithm is used to correct the acquisition delay and avoid data mismatch with the operation process;

[0034] Data completion processing includes:

[0035] For missing detection data, an LSTM prediction model is used to complete the data.

[0036] Furthermore, the collected track area operation and maintenance equipment data undergoes preprocessing, including data cleaning and spatiotemporal alignment, specifically including:

[0037] Spatiotemporal alignment processing includes:

[0038] Spatial alignment: Taking into account the linear extension characteristics of the rail, a linear segmented registration algorithm is adopted to divide the rail into multiple segments according to mileage. Multiple feature points are selected in each segment, and the coordinates within the segment are fitted by the least squares method to ensure the accurate correspondence between the equipment operation position and the rail segment.

[0039] Time alignment: Synchronize and associate the equipment operation time with the rail status data timestamp to establish a mapping relationship between the equipment operation time and the rail status data timestamp.

[0040] Furthermore, the preprocessed track area operation and maintenance equipment data is used to construct equipment data assets, including generating digital identity cards and 3D digital base maps. Specifically, this includes:

[0041] Establish digital ID cards for track area maintenance equipment according to preset coding rules, and link them with equipment calibration records, maintenance history, and work records; the digital ID card format for track area maintenance equipment is TT-SS-EE, where TT is the equipment type, SS is the equipment number, and EE is the home base;

[0042] A digital ID card for the rail section is established according to the preset coding rules and associated with inspection records, maintenance records and equipment. The digital ID card format for the rail section is LL-MM-CC-FF, where LL is the line number, MM is the mileage range, CC is the rail material and FF is the year of laying.

[0043] Furthermore, the preprocessed track area operation and maintenance equipment data is used to construct equipment data assets, including generating digital identity cards and 3D digital base maps. Specifically, this includes:

[0044] 3D digital base map generation, including:

[0045] The system is based on the BIM rail model of the new line and the laser scanning 3D model of the existing line, and is superimposed with the spatial data of the track maintenance equipment and the equipment-rail section correlation data.

[0046] Storage and Access: A hybrid storage method of BIM model + time series database is adopted, in which the BIM model is used to store spatial data, and the time series database is used to store equipment inspection / maintenance time series data, supporting real-time viewing and data export on the web.

[0047] Furthermore, based on the obtained data and equipment assets, intelligent applications are implemented, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment. Specifically, these include:

[0048] The fault diagnosis and maintenance decision generation includes:

[0049] Establish a four-dimensional knowledge graph linking "defects - detection equipment - repair equipment - repair solutions";

[0050] When the detection equipment detects a defect, it automatically matches similar cases in the knowledge graph, obtains recommended maintenance equipment and operation parameters, generates a maintenance work order, and pushes it to the equipment operation terminal simultaneously.

[0051] Furthermore, based on the obtained data and equipment assets, intelligent applications are implemented, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment. Specifically, these include:

[0052] The health prediction of the maintenance equipment includes: predicting the equipment lifespan and health status using a deep learning model based on the acquired equipment maintenance data.

[0053] Furthermore, based on the obtained data and equipment assets, intelligent applications are implemented, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment. Specifically, these include:

[0054] The valuation of the data assets includes:

[0055] A comprehensive asset valuation is conducted based on the established asset valuation index system.

[0056] For rail maintenance equipment, the asset value assessment index system includes primary and secondary indicators. The primary indicators include data integrity, data availability, decision contribution, and cost savings.

[0057] The secondary indicator of data integrity includes the rail inspection data integrity rate, which is calculated as follows: Rail inspection data integrity rate = (Number of rail sections with actual inspection records / Total number of sections) × 100%;

[0058] The secondary indicator of data availability includes the device data access frequency, which is calculated as: monthly access frequency / total device data volume.

[0059] The secondary indicator of decision-making contribution includes the improvement value of rail damage detection rate, which is calculated as: (current detection rate - manual detection rate set value) × 100%.

[0060] The secondary indicator of cost savings includes the rail maintenance cost reduction rate, which is calculated as: Rail maintenance cost reduction rate = (Set maintenance cost - Current maintenance cost) / Set maintenance cost × 100%.

[0061] According to a second aspect, one embodiment provides a data asset creation system for rail transit vehicle depots and infrastructure engineering equipment, the system comprising:

[0062] The data acquisition layer is used to digitally collect data on the entire lifecycle of track maintenance equipment for existing and newly built lines.

[0063] The data processing layer is used to perform preprocessing on the collected track area operation and maintenance equipment data, including data cleaning and spatiotemporal alignment.

[0064] The asset construction layer is used to construct equipment data assets from the pre-processed track area operation and maintenance equipment data, including digital ID generation and 3D digital base map generation.

[0065] The intelligent application layer is used for intelligent applications based on the obtained data and equipment assets, including fault diagnosis and maintenance decision generation, health prediction of operation and maintenance equipment, and data asset value assessment.

[0066] According to a third aspect, one embodiment provides an electronic device, the device comprising: a processor and a memory;

[0067] The memory is used to store one or more program instructions;

[0068] The processor is configured to run one or more program instructions to perform the steps of a method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in any of the preceding claims.

[0069] According to a fourth aspect, one embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of a method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in any of the preceding claims.

[0070] This application provides a method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment, which has the following beneficial effects:

[0071] 1. Significantly improved data utilization: Through spatiotemporal fusion algorithms and correlation models, data utilization has been effectively improved, and multi-source data correlation analysis has been formed;

[0072] 2. Optimized data processing efficiency: The parsing of existing paper drawings has been reduced from 2 hours / sheet by traditional manual entry to 0.5 seconds / sheet;

[0073] 3. Improved accuracy of maintenance decisions: The accuracy of maintenance decision-making models based on knowledge graphs is significantly improved compared to existing single-sensor prediction solutions, while reducing equipment downtime due to failures.

[0074] 4. Reduced lifecycle costs: Through full lifecycle data traceability, the waste rate of prefabricated components and the workload of operation and maintenance personnel can be effectively reduced, saving operation and maintenance costs;

[0075] 5. Digital twin support capability: Achieve millimeter-level spatiotemporal alignment (registration error < ±1mm), meet the requirements of digital twins for "real-time mapping and accurate simulation", and provide core data asset support for intelligent operation and maintenance of rail transit. Attached Figure Description

[0076] Figure 1 A flowchart illustrating a method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment, as provided in one embodiment of the present invention;

[0077] Figure 2 This is an overall architecture diagram of a data asset creation system for rail transit vehicle depots and infrastructure engineering equipment, provided as an embodiment of the present invention. Detailed Implementation

[0078] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of this application. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, certain operations related to this application are not shown or described in the specification. This is to avoid obscuring the core parts of this application with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0079] Furthermore, the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. At the same time, the steps or actions in the method description can be rearranged or adjusted in a manner obvious to those skilled in the art. Therefore, the various orders in the specification and drawings are only for the clear description of a particular embodiment and do not imply a necessary order, unless otherwise stated that a particular order must be followed.

[0080] The first embodiment of this invention provides a method for creating data assets of rail transit vehicle depot and infrastructure engineering equipment, employing a "four-layer architecture (data acquisition layer, data processing layer, asset construction layer, intelligent application) + dual-mode construction (transfer learning OCR model, equipment health prediction model)" technical approach. The following is a combination of... Figure 1 Please provide a detailed explanation.

[0081] Definition of the scope of rail transit vehicle depot and infrastructure engineering equipment:

[0082] Taking rail inspection and maintenance equipment as an example, rail transit vehicle depot and infrastructure engineering equipment can be divided into three categories, which are the core objects of digitalization and data asset management:

[0083]

[0084] like Figure 1 As shown, in step S100, the entire lifecycle data of the track maintenance equipment of existing and newly built lines is digitally collected.

[0085] Data Acquisition Layer: Taking rail inspection and maintenance equipment as an example, to address the differences in rail operation and maintenance scenarios between existing and newly built lines, a dual-mode acquisition mechanism of "reverse reconstruction + forward delivery" is adopted to solve the problems of "missing historical data and real-time data gaps."

[0086] S110, Digital Data Acquisition (Reverse Reconstruction) of Existing Rail Maintenance Equipment

[0087] Historical data supplementation:

[0088] For old rail inspection equipment without electronic records (such as flaw detectors manufactured before 2010), the equipment nameplate (extracting model number and serial number) and paper inspection report are photographed using an industrial camera (OCR recognition of damage records, using the original CRNN-OCR model, with an accuracy rate of ≥95%).

[0089] For existing rail sections, a laser profilometer (accuracy ±0.1mm) is used to scan the entire length of the rail to collect wear trend data for the past 5 years (historical change curves are fitted by the wear difference between adjacent sections, with a completion rate ≥85%).

[0090] Real-time data acquisition:

[0091] Inspection equipment: Equip ultrasonic flaw detectors with 4G / 5G modules to upload damage data in real time (including GPS positioning, error ≤5m, corresponding rail mileage K value); connect laser profilometers to portable data loggers to synchronously store wear data and acquisition timestamps;

[0092] Maintenance equipment: Install torque sensors (range 0-500 N·m, sampling rate 10 Hz) and vibration sensors (range ±20 g) on ​​the main shaft of the rail grinding machine to collect equipment load data during the grinding process and associate it with the specific rail section being ground;

[0093] Auxiliary data collection: RFID tags (storing rail code, material, and laying time) are buried every 50m along the rail sleepers. When inspection and maintenance equipment is in operation, the RFID reader automatically identifies the rail and avoids data mismatch with the section.

[0094] S120, Digital Data Acquisition of Rail Maintenance Equipment for New Line (Forward Delivery)

[0095] BIM Collaborative Data Acquisition: In the newly built BIM model of the line (LOD400 accuracy), establish associated nodes for rail sections (divided by mileage, such as K0+000-K0+500) and maintenance equipment (such as grinding machines and flaw detectors). Equipment parameters (model, calibration cycle) and construction period inspection data (such as geometric parameters during rail laying) are directly written into the BIM model attribute column.

[0096] Construction monitoring interface: Develop a standardized interface (based on the MQTT protocol, with a transmission delay of < 100ms) to connect to the detection module of rail laying equipment (such as a rail laying machine) to collect the initial gauge and smoothness data of the rails in real time and automatically synchronize them to the equipment data management system;

[0097] Equipment embedded point design: When maintenance equipment (such as rail grinding machines) leaves the factory, a data acquisition module is pre-installed with a built-in SIM card, which can automatically upload the equipment's start-up time, operation time, and consumable consumption (such as the wear of grinding wheels), without the need for on-site manual recording.

[0098] Furthermore, the data acquisition layer can adopt the following alternatives: LiDAR scanner → UAV photogrammetry: suitable for large areas of existing lines (such as outdoor tracks of vehicle depots), improving acquisition efficiency; BIM platform → CIM platform (City Information Model): suitable for cross-line collaborative scenarios, can integrate data assets of multiple lines, and realize city-level rail transit data management.

[0099] like Figure 1 As shown, in step S200, the collected rail maintenance equipment data undergoes preprocessing including data cleaning and spatiotemporal alignment.

[0100] The data processing layer, addressing the characteristics of rail equipment data—"multi-source heterogeneous and spatiotemporally dispersed"—focuses on resolving data cleaning, correlation, and alignment issues.

[0101] S210, a heterogeneous data cleaning engine (focusing on the characteristics of rail equipment data).

[0102] Data processing:

[0103] Flaw detection waveform data: Wavelet transform algorithm (such as db4 wavelet basis) is used to remove waveform noise from ultrasonic flaw detector (signal-to-noise ratio improved by 30%), and the damage feature points are automatically identified by threshold method (threshold is set to 1.5 times the normal waveform amplitude);

[0104] Geometric parameter data: The track gauge and level data collected by the track geometry measuring instrument are used to remove outliers (such as out-of-tolerance data caused by equipment vibration) using the 3σ principle, and then the data is smoothed by moving average (window size 5m) to improve accuracy;

[0105] Maintenance data processing:

[0106] Unstructured text: Use OCR+NLP algorithms to parse rail maintenance work orders (such as "2024-05-10 Grinding rails from K1+300 to K1+500, vertical wear reduced to 0.3mm"), extract key information of "time-equipment-section-maintenance effect", and convert it into structured data;

[0107] Equipment status data: For the torque and vibration data of the grinding machine, the Kalman filter algorithm is used to correct the acquisition delay (delay error < 0.5s) to avoid data mismatch with the operation process;

[0108] Data completion: For missing rail inspection data (such as wear data of a certain section in Q2 2023), the LSTM prediction model is used to complete the data based on adjacent quarterly data (Q1, Q3) and equipment operation records (such as whether grinding was performed in Q2). The completion accuracy is ≥80%.

[0109] S220, Spatial-Temporal Alignment Algorithm for Rail Scenes

[0110] Spatial Alignment: Addressing the "linear extension" characteristic of rails, the original ICP algorithm was improved to a "linear segmented registration algorithm"—dividing the rail into 10m segments based on mileage, selecting three feature points for each segment (such as the rail head apex and rail web midpoint), and fitting the segment coordinates using the least squares method (source coordinates: equipment-acquired coordinates; target coordinates: BIM model rail coordinates). The registration error is ≤ ±1mm, ensuring accurate correspondence between "equipment operating position - rail segment".

[0111] Time alignment: Establish a mapping relationship between "equipment operation time - rail status time" - such as the rail grinding machine operation time (2024-05-10 09:00-11:00), synchronously link the laser profilometer detection data before grinding (08:30) and the re-inspection data after grinding (11:30), forming a time closed loop of "inspection-maintenance-re-inspection".

[0112] Furthermore, the data processing layer can adopt the following alternatives: Improved ICP algorithm → NDT (Normal Distribution Transform) algorithm: suitable for scenarios with low point cloud density (such as severely occluded areas), improving computation speed; CRNN-OCR → Transformer-OCR (such as ViT-OCR): suitable for scenarios with blurred nameplates (such as severely aged existing line equipment), improving recognition accuracy.

[0113] like Figure 1 As shown, in step S300, the preprocessed track area operation and maintenance equipment data is used to construct equipment data assets, including generating digital ID cards and generating three-dimensional digital base maps.

[0114] The asset construction layer specifically includes:

[0115] S310, Dual-Dimensional Digital Identity System

[0116] Dimension 1: Digital ID Card for Rail Maintenance Equipment (Encoding Rule: 8 digits, format "TT-SS-EE")

[0117] TT (2 digits): Equipment type (01 = flaw detector, 02 = grinder, 03 = laser profilometer).

[0118] SS (3 digits): Equipment number (e.g., 001 = the first flaw detector);

[0119] EE (3 digits): Home base (e.g., 005 = a certain vehicle depot base);

[0120] Related data: Equipment calibration records (e.g., calibration on 2024-03-15, error ≤0.1mm), maintenance history (replacement of flaw detector probe on 2024-01-20), and work records (work on segment K1+300-K1+500 on 2024-05-10).

[0121] Dimension 2: Digital ID Card for Rail Sections (Encoding Rule: 12 digits, format "LL-MM-CC-FF")

[0122] LL (2 digits): Line number (01 = Line 1);

[0123] MM (6 digits): Mileage range (e.g., 01300 = K1 + 300);

[0124] CC (2 digits): Rail material (01=U71Mn);

[0125] FF (2 digits): Year of installation (24 = 2024);

[0126] Related data: Inspection record (2024-05-08 No flaws detected, wear 0.5mm), maintenance record (2024-05-10 Grinding, wear reduced to 0.3mm), equipment association (01-001-005 flaw detector, 02-003-005 grinding machine repair).

[0127] S320, 3D Digital Base Map for Rail Maintenance

[0128] Construction logic: Based on the BIM rail model of the newly built line and the laser-scanned 3D model of the existing line, two types of data are overlaid:

[0129] 1) Spatial data of rail maintenance equipment: such as the movement trajectory of the flaw detector during operation (based on GPS+IMU positioning) and the working range of the grinding machine (calculated through equipment parameters, such as a grinding width of 200mm).

[0130] 2) Equipment-rail correlation data: The status of rail sections is marked with different colors (green = normal, yellow = slight wear, red = damage). Clicking on a section can view the model of the testing equipment and the operation record of the maintenance equipment.

[0131] Storage and Access: A hybrid storage system of "BIM model + time series database" is adopted—the BIM model (glTF format, file size reduced by 60% after lightweighting) stores the data, while the InfluxDB time series database stores the device detection / maintenance time series data (such as rail wear data updated once per hour), supporting real-time viewing and data export on the web.

[0132] like Figure 1 As shown, in step S400, intelligent applications are performed based on the obtained data equipment assets, including fault diagnosis and maintenance decision generation, operation and maintenance equipment health prediction, and data asset value assessment.

[0133] The intelligent application layer specifically includes:

[0134] S410, Rail Fault Diagnosis and Repair Decision

[0135] Knowledge Graph Construction: Establish a four-dimensional relationship graph of "rail defects - detection equipment - maintenance equipment - maintenance solutions". Example relationship:

[0136] Defect: Vertical wear of the rail ≥1mm;

[0137] Inspection equipment: Laser profilometer (to confirm the location of wear);

[0138] Repair equipment: Rail grinding machine (model HG450, grinding parameters: speed 3000r / min, grinding depth 0.2mm);

[0139] Repair plan: Grind in 3 passes, and re-inspect with a profilometer after each pass;

[0140] Intelligent decision-making process: When the detection equipment finds a defect (such as the flaw detector finding a rail flaw) → the system automatically matches similar cases in the knowledge graph (matching degree ≥ 90%) → recommends maintenance equipment (such as turnout welding repair equipment) and operation parameters → generates a maintenance work order and pushes it to the equipment operation terminal simultaneously.

[0141] S420, Health Prediction of Maintenance Equipment

[0142] Prediction model: Based on the LSTM algorithm, taking a rail grinding machine as an example, the input data includes: equipment operation time (cumulative grinding hours), grinding amount (cumulative grinding length of rail), spindle torque fluctuation value, vibration amplitude → output "remaining tool life" and "next maintenance time";

[0143] Application results: Prediction accuracy ≥85%, avoiding rail grinding quality defects caused by excessive tool wear (defect rate reduced from 15% to 5%), and reducing downtime due to sudden equipment failures (reduced from 4 hours / time to 1.5 hours / time).

[0144] S430, Data Asset Valuation (for rail maintenance scenarios)

[0145]

[0146] Furthermore, the intelligent application layer adopts the following alternatives: LSTM health prediction → GRU (Gated Recurrent Unit): fewer model parameters, shorter training time, and basically the same prediction accuracy, suitable for edge devices with limited computing power; Value assessment index → ​​EVA (Economic Value Added) model: adds "data asset revenue" index (such as data external service revenue), suitable for market-oriented operation scenarios.

[0147] Application example:

[0148] I. Taking the creation of data assets for an existing project as an example, the specific steps are as follows:

[0149] Step 1: Data Collection → Scan with Faro Focus S70, photograph the equipment nameplate with an industrial camera, and read the equipment number with RFID;

[0150] Step 2: Data processing → CRNN-OCR recognition of nameplates, improved ICP algorithm to align point cloud (registration error ±0.8mm), and completion of missing data;

[0151] Step 3: Build Assets → Generate a 20-digit digital ID and build a 3D digital base map;

[0152] Step 4: Intelligent Application → Deploy the knowledge graph decision model and connect to the fault case library.

[0153] II. Taking the creation of data assets for a newly established project as an example, the specific steps are as follows:

[0154] Step 1: BIM Modeling → Create a LOD400 model using Revit and embed asset data points (including component IDs and manufacturing information).

[0155] Step 2: Interface Development → Develop an MQTT interface using the Python Flask framework to connect to the construction monitoring system (settlement monitor, stress sensor).

[0156] Step 3: Data Traceability → Build a prefabricated component traceability system (based on a MySQL database) to record data throughout the entire process of "factory delivery - transportation - installation - operation and maintenance";

[0157] Step 4: Acceptance and delivery → Synchronize the BIM model and traceability data to the operation and maintenance system (CMMS) and complete interface testing.

[0158] The core algorithm used is as follows:

[0159] 1. Spatiotemporal alignment algorithm

[0160] The coordinate transformation is achieved using an improved ICP algorithm, and the loss function is defined as follows:

[0161]

[0162] The parameters in the formula are defined as follows:

[0163]

[0164] 2. Transfer learning OCR model

[0165] CRNN network feature extraction formula:

[0166]

[0167] The parameters in the formula are defined as follows:

[0168]

[0169] Character recognition probability:

[0170]

[0171] The parameters in the formula are defined as follows:

[0172]

[0173] 3. Equipment health prediction model

[0174] LSTM hidden state update equation:

[0175]

[0176]

[0177]

[0178]

[0179]

[0180] The parameters in the formula are defined as follows:

[0181]

[0182] Examples of key parameter values ​​for the above algorithm and model are as follows:

[0183]

[0184] The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment proposed in this invention has the following advantages:

[0185] 1. Construction of a closed-loop data system for "inspection-maintenance-status": Breaking through the separation problem in the traditional operation and maintenance of rail transit vehicle depots and infrastructure, where "inspection data is stored in equipment and maintenance data is stored in work orders", by using digital identity cards and spatiotemporal alignment, the system achieves full-link data connectivity of "which equipment is inspected → which section of rail has a defect → which equipment is repaired → what is the status after repair";

[0186] 2. Reverse reconstruction of historical data of existing rail transit vehicle depots and infrastructure engineering equipment: For the detection data of existing lines that have no electronic records, an innovative "laser profilometer supplementary collection + LSTM trend fitting" scheme is adopted to supplement the defect data of the past 5 years and solve the problem of "no basis" for operation and maintenance decisions caused by the lack of historical data.

[0187] 3. Forward delivery of equipment data and BIM for new rail transit vehicle depots and infrastructure projects: Equipment data points are embedded during the construction phase, and the inspection data during the construction period is automatically synchronized to the BIM model, avoiding the "disconnection between equipment data and model" after the new line is put into operation, and increasing the data assetization rate from 30% to 85%.

[0188] Corresponding to the aforementioned method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment, this invention also discloses a system for creating data assets of rail transit vehicle depots and infrastructure engineering equipment, such as... Figure 2 As shown, it specifically includes:

[0189] The data acquisition layer is used to digitally collect data on the entire lifecycle of track maintenance equipment for existing and newly built lines.

[0190] The data processing layer is used to perform preprocessing on the collected track area operation and maintenance equipment data, including data cleaning and spatiotemporal alignment.

[0191] The asset construction layer is used to construct equipment data assets from the pre-processed track area operation and maintenance equipment data, including digital ID generation and 3D digital base map generation.

[0192] The intelligent application layer is used for intelligent applications based on the obtained data and equipment assets, including fault diagnosis and maintenance decision generation, health prediction of operation and maintenance equipment, and data asset value assessment.

[0193] It should be noted that for a detailed description of the data asset creation system for rail transit vehicle depots and infrastructure engineering equipment provided in the embodiments of the present invention, please refer to the relevant description of the data asset creation method for rail transit vehicle depots and infrastructure engineering equipment provided in the embodiments of this application, which will not be repeated here.

[0194] In addition, embodiments of the present invention also provide an electronic device, the device comprising: a processor and a memory; the memory being used to store one or more program instructions; the processor being used to execute one or more program instructions to perform the steps of a method for creating data assets of rail transit vehicle depot and infrastructure engineering equipment as described in any of the preceding embodiments.

[0195] It should be noted that for a detailed description of an electronic device provided in the embodiments of the present invention, please refer to the relevant description of a method for creating data assets of rail transit vehicle depot and infrastructure engineering equipment provided in the embodiments of this application, which will not be repeated here.

[0196] In addition, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in any of the preceding claims.

[0197] It should be noted that for a detailed description of the computer-readable storage medium provided in the embodiments of the present invention, please refer to the relevant description of the method for creating data assets of rail transit vehicle depot and infrastructure engineering equipment provided in the embodiments of this application, which will not be repeated here.

[0198] Those skilled in the art will understand that all or part of the functions of the various methods in the above embodiments can be implemented by hardware or by computer programs. When all or part of the functions in the above embodiments are implemented by computer programs, the program can be stored in a computer-readable storage medium, which may include: read-only memory, random access memory, disk, optical disk, hard disk, etc., and the program is executed by a computer to achieve the above functions. For example, the program can be stored in the memory of a device, and when the program in the memory is executed by the processor, all or part of the above functions can be achieved. In addition, when all or part of the functions in the above embodiments are implemented by computer programs, the program can also be stored in a server, another computer, disk, optical disk, flash drive, or external hard drive, etc., and can be downloaded or copied to the memory of a local device, or the system of the local device can be updated. When the program in the memory is executed by the processor, all or part of the functions in the above embodiments can be achieved.

[0199] The above examples illustrate the present invention only to aid in understanding it and are not intended to limit the scope of the invention. Those skilled in the art can make various simple deductions, modifications, or substitutions based on the principles of this invention.

Claims

1. A method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment, characterized in that, The method includes: Digitally collect data on the entire lifecycle of track maintenance equipment for both existing and newly built lines; The collected data on track area operation and maintenance equipment is preprocessed, including data cleaning and spatiotemporal alignment. The preprocessed track area operation and maintenance equipment data is used to construct equipment data assets, including digital ID generation and 3D digital base map generation. Based on the obtained data, intelligent applications are developed for equipment assets, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment.

2. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, Digital collection of full lifecycle data for track maintenance equipment on existing and newly built lines, specifically including: Data collection for maintenance equipment in existing track areas, including: Historical data collection includes: extracting model and serial number information from the equipment nameplate; extracting information from paper reports using OCR recognition technology; and scanning the entire length of existing rail sections using a laser profilometer. Real-time data acquisition, including: For testing equipment: install a 4G / 5G module on the equipment to upload testing data in real time; connect the equipment to a portable data logger to synchronously store testing data and collection timestamps; For maintenance equipment: Install multiple sensors on the equipment to collect equipment status data and link it to the corresponding rail section; The auxiliary data collection includes: embedding RFID tags in the rail sleepers at preset intervals to store the rail code, material and laying time. When the inspection and maintenance equipment is in operation, the RFID reader automatically identifies the rail and avoids data mismatch with the section.

3. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, Digital collection of full lifecycle data for track maintenance equipment on existing and newly built lines, specifically including: Data collection for operation and maintenance equipment in the track area of ​​newly built railway lines, including: BIM collaborative data collection includes: establishing associated nodes for rail sections and maintenance equipment in the BIM model of newly built lines, and directly writing equipment parameters and construction period inspection data into the BIM model attribute column; Construction monitoring interface development includes: developing standardized interfaces to connect to the detection modules of rail laying equipment, collecting initial rail gauge and smoothness data in real time, and automatically synchronizing them to the equipment data management system; The equipment embedding design includes: pre-installing data acquisition modules on maintenance equipment at the factory to automatically upload equipment startup time, operation duration, and consumable consumption, eliminating the need for on-site manual recording.

4. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, The collected track area maintenance equipment data undergoes preprocessing, including data cleaning and spatiotemporal alignment, specifically including: Data processing includes: Wavelet transform algorithm is used to remove waveform noise, and feature points are automatically identified by thresholding method; Outliers are removed using the 3σ principle, and then the data is smoothed using a moving average to improve accuracy. Maintenance data processing includes: The OCR+NLP algorithm is used to parse unstructured text data, extract key information, and convert it into structured data. The Kalman filter algorithm is used to correct the acquisition delay and avoid data mismatch with the operation process; Data completion processing includes: For missing detection data, an LSTM prediction model is used to complete the data.

5. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, The collected track area maintenance equipment data undergoes preprocessing, including data cleaning and spatiotemporal alignment, specifically including: Spatiotemporal alignment processing includes: Spatial alignment: Taking into account the linear extension characteristics of the rail, a linear segmented registration algorithm is adopted to divide the rail into multiple segments according to mileage. Multiple feature points are selected in each segment, and the coordinates within the segment are fitted by the least squares method to ensure the accurate correspondence between the equipment operation position and the rail segment. Time alignment: Synchronize and associate the equipment operation time with the rail status data timestamp to establish a mapping relationship between the equipment operation time and the rail status data timestamp.

6. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, The preprocessed track area maintenance equipment data is used to construct equipment data assets, including generating digital identity cards and 3D digital base maps. Specifically, this includes: Establish digital ID cards for track area maintenance equipment according to preset coding rules, and link them with equipment calibration records, maintenance history, and work records; the digital ID card format for track area maintenance equipment is TT-SS-EE, where TT is the equipment type, SS is the equipment number, and EE is the home base; A digital ID card for the rail section is established according to the preset coding rules and associated with inspection records, maintenance records and equipment. The digital ID card format for the rail section is LL-MM-CC-FF, where LL is the line number, MM is the mileage range, CC is the rail material and FF is the year of laying.

7. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, The preprocessed track area maintenance equipment data is used to construct equipment data assets, including generating digital identity cards and 3D digital base maps. Specifically, this includes: 3D digital base map generation, including: The system is based on the BIM rail model of the new line and the laser scanning 3D model of the existing line, and is superimposed with the spatial data of the track maintenance equipment and the equipment-rail section correlation data. Storage and Access: A hybrid storage method of BIM model + time series database is adopted, in which the BIM model is used to store spatial data, and the time series database is used to store equipment inspection / maintenance time series data, supporting real-time viewing and data export on the web.

8. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, Based on the obtained data on equipment assets, intelligent applications are implemented, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment. Specifically, these include: The fault diagnosis and maintenance decision generation includes: Establish a four-dimensional knowledge graph linking "defects - detection equipment - repair equipment - repair solutions"; When the detection equipment detects a defect, it automatically matches similar cases in the knowledge graph, obtains recommended maintenance equipment and operation parameters, generates a maintenance work order, and pushes it to the equipment operation terminal simultaneously.

9. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, Based on the obtained data on equipment assets, intelligent applications are implemented, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment. Specifically, these include: The health prediction of the maintenance equipment includes: predicting the equipment lifespan and health status using a deep learning model based on the acquired equipment maintenance data.

10. The method for creating data assets of rail transit vehicle depots and infrastructure engineering equipment as described in claim 1, characterized in that, Based on the obtained data on equipment assets, intelligent applications are implemented, including fault diagnosis and maintenance decision generation, equipment health prediction, and data asset value assessment. Specifically, these include: The valuation of the data assets includes: A comprehensive asset valuation is conducted based on the established asset valuation index system. For rail maintenance equipment, the asset value assessment index system includes primary and secondary indicators. The primary indicators include data integrity, data availability, decision contribution, and cost savings. The secondary indicator of data integrity includes the rail inspection data integrity rate, which is calculated as follows: Rail inspection data integrity rate = (Number of rail sections with actual inspection records / Total number of sections) × 100%; The secondary indicator of data availability includes the device data access frequency, which is calculated as: monthly access frequency / total device data volume. The secondary indicator of decision-making contribution includes the improvement value of rail damage detection rate, which is calculated as: (current detection rate - manual detection rate set value) × 100%. The secondary indicator of cost savings includes the rail maintenance cost reduction rate, which is calculated as: Rail maintenance cost reduction rate = (Set maintenance cost - Current maintenance cost) / Set maintenance cost × 100%.

Images