Digital twinning-oriented rail train braking system full life cycle data management and application system and method

Abstract

The invention relates to a digital twinning-oriented rail train braking system full life cycle data management and application system and method, and relates to the technical field of rail traffic intelligent manufacturing and intelligent operation and maintenance. The system comprises a unified data access and processing layer, a digital twin model layer, an intelligent application layer and a unified data service and interface layer which are connected in sequence. According to the digital twinning-oriented rail train braking system full life cycle data management and application system and method disclosed by the invention, the fundamental change of data driving is realized; a'digital brain 'of the braking system is constructed; the business is enabled to be full-chain, and remarkable value is created; and a new mode of closed-loop optimization is created.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing and intelligent operation and maintenance technology for rail transit, and in particular to a system and method for full life-cycle data management and application of rail train braking system oriented towards digital twins. Background Technology

[0002] The braking system of rail trains is a core system ensuring operational safety, characterized by high complexity and extremely stringent reliability requirements. With the deepening of digital transformation, massive amounts of heterogeneous data are generated throughout its entire lifecycle. However, the current management and application of this data suffer from the following systemic deficiencies:

[0003] Data silos abound, lacking integration and connectivity: The braking system involves multiple independent subsystems (such as the BCU braking control unit, TCMS train network control system, fault diagnosis system, and MRO maintenance and operation system). Data from each system uses different formats, protocols, and is stored in a scattered manner. CAE simulation data from the design phase, BOM and quality inspection data from the manufacturing phase, real-time sensor data from the operation phase, and work order records from the maintenance phase are isolated from each other, making correlation analysis impossible. This creates a severe "data silo" effect, preventing the release of the comprehensive value of the data.

[0004] The data lacks spatiotemporal dimensions and has weak correlation: existing operational data mostly only contains simple timestamps, severely lacking accurate and automatic correlation with the actual operating location of trains (mileage markers). This makes it impossible to deeply couple and analyze fault events and performance degradation with specific track conditions (such as gradient and curve radius), geographical environment (such as tunnels and bridges), and weather conditions, greatly limiting the ability to perform data mining and decision support based on environmental context.

[0005] The application model is superficial and fails to provide forward-looking empowerment: Current data applications are basically limited to post-event traceability and simple statistical reports, such as querying historical alarms and calculating component replacement frequency. Data has not been effectively used for pre-event virtual simulation verification, in-event intelligent decision support, and closed-loop iterative optimization of product design, and cannot support advanced intelligent application scenarios such as predictive maintenance and solution optimization.

[0006] Lack of high-fidelity system-level digital mirror: Current technology lacks a digital twin that can span the entire lifecycle of the braking system, deeply integrate mechanisms and data, and interact and evolve synchronously with the physical entity in real time. This results in a lack of a unified, accurate, and dynamic digital model foundation for in-depth analysis, simulation testing, and optimization decision-making.

[0007] Therefore, the existing technology system is unable to meet the urgent needs of modern rail transit for high safety, high reliability, high efficiency and low cost operation and maintenance, which restricts the pace of intelligent upgrading of the industry. Summary of the Invention

[0008] The present invention aims to solve the technical problems in the prior art by providing a system and method for full life-cycle data management and application of a rail train braking system oriented towards digital twins.

[0009] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0010] A data management and application system for the entire lifecycle of a rail train braking system based on digital twins, comprising, in sequence: a unified data access and processing layer, a digital twin model layer, an intelligent application layer, and a unified data service and interface layer; wherein:

[0011] The unified data access and processing layer is used to connect to and collect heterogeneous data sources from all stages of the braking system's lifecycle;

[0012] The digital twin model layer includes a model management and online update engine. The digital twin model layer is used to receive real-time and historical data from the full lifecycle data lake through the model management and online update engine, drive model operation and continuously calibrate model parameters, and realize high-fidelity virtual mapping and state prediction of physical entities.

[0013] The intelligent application layer is used to provide a variety of advanced application function modules based on the full lifecycle data lake and digital twin model;

[0014] The unified data service and interface layer is used to provide standardized data access APIs, model calling services, simulation computing services, and visualization component services for upper-layer applications and external systems in a microservice architecture, ensuring the system's openness, scalability, and ease of integration.

[0015] In the above technical solution, the heterogeneous data sources in the unified data access and processing layer include: simulation models and data in the design phase, BOM (Bill of Materials) and quality inspection reports in the manufacturing phase, real-time data and fault logs of vehicle sensors in the operation phase, and maintenance records and component replacement history in the maintenance phase.

[0016] In the above technical solution, the digital twin model layer also includes: a static model library and a dynamic model library;

[0017] The static models in the static model library include: the three-dimensional geometric model of the braking system, the assembly relationship model, and the BOM (Bill of Materials).

[0018] The dynamic models in the dynamic model library include: mechanistic models based on physicochemical rules and data-driven models trained on historical data.

[0019] In the above technical solution, the intelligent application layer includes: a control strategy simulation test module, a predictive maintenance decision module, and an asset performance panoramic dashboard and optimization module;

[0020] The control strategy simulation test module is used to load the new algorithm onto the digital twin and run the simulation by inputting a piece of real data of historical low adhesion conditions to quickly verify the effectiveness and safety of the new algorithm;

[0021] The predictive maintenance decision module is used to automatically generate early warning work orders when the degradation model in the digital twin predicts that the performance will exceed the tolerance range after a certain period of time, and recommends to check in the next planned maintenance to avoid online failures.

[0022] The asset performance dashboard and optimization module are used to query data and then make decisions.

[0023] A method for managing and applying the full lifecycle data of a rail train braking system based on digital twins, applicable to the aforementioned system for managing and applying the full lifecycle data of a rail train braking system based on digital twins, includes the following steps:

[0024] Step S1, Data Access and Processing: The unified data access and processing layer accesses data from various sources; after standardization, the data is stored in the full lifecycle data lake;

[0025] Step S2, Digital Twin Construction and Operation: The model management and online update engine in the digital twin model layer continuously acquires data from the full lifecycle data lake to drive and calibrate the model;

[0026] Step S3, Smart Application Execution, specifically includes:

[0027] Step S31, Control Strategy Simulation Test: In the control strategy simulation test module, the new algorithm is loaded onto the digital twin, and a piece of real data from a historical low-adhesion working condition is input for simulation to quickly verify the effectiveness and safety of the new algorithm.

[0028] Step S32, Predictive Maintenance Decision: When the degradation model in the digital twin predicts that the performance will exceed the tolerance range after a certain period of time, the predictive maintenance decision module will automatically generate an early warning work order and recommend checking in the next planned maintenance to avoid online failures.

[0029] Step S33, Asset Performance Panorama Dashboard: Use the Asset Performance Panorama Dashboard and the optimization module to query data and make decisions.

[0030] The present invention has the following beneficial effects:

[0031] The present invention relates to a system and method for full lifecycle data management and application of a digital twin-based rail train braking system:

[0032] It has achieved a fundamental data-driven transformation: by unifying spatiotemporal benchmarks and integrating data across the entire domain, it has completely solved the problem of data silos, transforming data from scattered, low-value records into high-quality, deeply mineable strategic assets, and providing a solid foundation for intelligentization.

[0033] A "digital brain" for the braking system has been constructed: the created dynamic digital twin is not only a static model, but also a "digital brain" capable of real-time perception, simulation prediction, and decision support, realizing deep interaction and intelligent collaboration between the digital and physical worlds, which is a major upgrade of the technical architecture.

[0034] Empowering the entire business chain and creating significant value: It provides powerful tools for R&D innovation, operation and maintenance transformation, and management decision-making, enabling trial and error in a virtual environment, prediction of future risks, and optimization of global resources, resulting in multiple benefits such as shortened R&D cycles, reduced operation and maintenance costs, and improved security.

[0035] It has pioneered a new closed-loop optimization model: it has established an intelligent feedback closed loop of "operation-maintenance-design-manufacturing", so that product optimization and improvement are no longer based on experience, but driven by real operating data, thereby maximizing the value of the product throughout its entire life cycle. Attached Figure Description

[0036] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0037] Figure 1 This is a schematic diagram of the overall architecture of the digital twin-oriented rail train braking system full lifecycle data management and application system of the present invention. Detailed Implementation

[0038] The inventive concept of this invention is as follows:

[0039] The data lake construction method based on the "vehicle-time-mileage" three-dimensional unification: The whole life cycle data management and application system of the rail train braking system for digital twins of this invention uses the kilometerpost as one of the core spatiotemporal references in the data fusion system, which elevates the train position information to a dimension of equal importance with time information, realizes the precise correlation between data and physical line environment, and lays the foundation for in-depth analysis.

[0040] The "mechanism-data" dual-driven dynamic digital twin modeling and updating mechanism: The whole life cycle data management and application system for the rail train braking system of this invention combines the physical rule-based mechanism model with the machine learning-based data-driven model to construct the digital twin, and designs an online model update engine to continuously calibrate the model using real-time data to ensure the high fidelity of the digital twin throughout its entire life cycle.

[0041] A closed-loop empowerment ecosystem for the entire lifecycle: This invention's data management and application system for the entire lifecycle of a rail train braking system, geared towards digital twins, transcends simple data management, constructing a closed-loop empowerment system covering the entire chain of "design-manufacturing-operation-maintenance-redesign." It creatively applies "past tense" data from the operation and maintenance phase to "future tense" optimization in the design simulation phase, maximizing the value of the data.

[0042] The following is in conjunction with the appendix Figure 1 The present invention will be described in detail below.

[0043] The present invention relates to a digital twin-based rail train braking system full lifecycle data management and application system, comprising, in sequence: a unified data access and processing layer, a digital twin model layer, an intelligent application layer, and a unified data service and interface layer.

[0044] The unified data access and processing layer is used to connect to and collect heterogeneous data sources from all stages of the braking system's lifecycle. These heterogeneous data sources include simulation models and data from the design phase, BOM (Bill of Materials) and quality inspection reports from the manufacturing phase, real-time data and fault logs from onboard sensors from the operation phase, and maintenance records and component replacement history from the maintenance phase. The unified data access and processing layer has a built-in data cleaning and transformation engine, and innovatively assigns all data a unified vehicle unique identifier (VID), a timestamp accurate to milliseconds, and a mileage post based on route coordinates (KilometerPost). After processing, the data is written into the full lifecycle data lake for storage.

[0045] The digital twin model layer receives real-time and historical data from the full lifecycle data lake through the model management and online update engine, drives model operation, and continuously calibrates model parameters to achieve high-fidelity virtual mapping and state prediction of physical entities. The digital twin model layer includes: a model management and online update engine, a static model library, and a dynamic model library. The static and dynamic model libraries each contain pre-built models of various braking systems. The digital twin model layer is the core of the system. Static models in the static model library include: a 3D geometric model of the braking system, an assembly relationship model, and a Bill of Materials (BOM). Dynamic models in the dynamic model library include: mechanistic models based on physicochemical rules (such as brake cylinder pressure-force conversion models and brake pad friction thermodynamic models) and data-driven models trained based on historical data (such as LSTM-based brake pad wear and degradation prediction models).

[0046] The intelligent application layer is used to provide a variety of advanced application function modules based on the full lifecycle data lake and digital twin model; the intelligent application layer includes: control strategy simulation and testing module, predictive maintenance decision module, and asset performance panoramic dashboard and optimization module;

[0047] The unified data service and interface layer is used to provide standardized data access APIs, model calling services, simulation computing services, and visualization component services for upper-layer applications and external systems in a microservice architecture, ensuring the system's openness, scalability, and ease of integration.

[0048] The present invention provides a method for full lifecycle data management and application of a digital twin-based rail train braking system. The applicable system is the same as the present invention's system for full lifecycle data management and application of a digital twin-based rail train braking system. The method includes the following steps:

[0049] Step S1, Data Access and Processing:

[0050] The unified data access and processing layer receives data from various sources. This data includes design data, manufacturing data, operational data, and maintenance data. For example, specific data received includes: real-time pressure and speed data from TCMS, with precise mileage stamps added using positioning system information; and records from the maintenance system indicating that "XX brake pads were replaced on [date]".

[0051] After standardization, the data is stored in a full lifecycle data lake. A typical data record format is: {VID:T001,Time:2023-10-2708:00:00.000,KM:1234.5, Data:BrakeCylinderPressure =345kPa}.

[0052] Step S2, Digital Twin Construction and Operation:

[0053] The model management and online update engine continuously draws data from the full lifecycle data lake to drive and calibrate the model. For example, real brake pad wear data is used to continuously optimize the data-driven degradation model, making its predictions increasingly accurate. At this point, the digital twin is no longer static, but a dynamic model that can "age" and "change" in sync with the physical entity.

[0054] Step S3, Smart Application Execution, specifically includes:

[0055] Step S31, Control Strategy Simulation Test: In the control strategy simulation test module, load the new algorithm onto the digital twin and input a piece of real data of historical low adhesion conditions (such as ground sliding data in rainy weather) for simulation to quickly verify the effectiveness and safety of the new algorithm. After confirming that there are no errors, deploy it to the real vehicle.

[0056] Step S32, Predictive Maintenance Decision: The system detects that the response time parameter of a vehicle's brake cylinder begins to drift slowly. The degradation model in the digital twin predicts that its performance will exceed the tolerance range after 30 days. The predictive maintenance decision module automatically generates an early warning work order and recommends inspection during a future planned maintenance, thus avoiding online failures.

[0057] Step S33, Asset Performance Panorama Dashboard: Use the Asset Performance Panorama Dashboard and the optimization module to query data and make decisions.

[0058] For example, it can be found that supplier B's brake pads have an average lifespan 20% longer than supplier A's, but the price is only 5% higher. Based on this, more data-supported purchasing decisions can be made.

[0059] In addition, the design department can discover that a certain type of brake cylinder wears out faster in cold regions, which can provide direction for material improvements in the next generation of products.

[0060] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A data management and application system for the entire lifecycle of a rail train braking system based on digital twins, characterized in that, This includes, in sequence: a unified data access and processing layer, a digital twin model layer, an intelligent application layer, and a unified data service and interface layer; among which: The unified data access and processing layer is used to connect to and collect heterogeneous data sources from all stages of the braking system's lifecycle; The digital twin model layer includes a model management and online update engine. The digital twin model layer is used to receive real-time and historical data from the full lifecycle data lake through the model management and online update engine, drive model operation and continuously calibrate model parameters, and realize high-fidelity virtual mapping and state prediction of physical entities. The intelligent application layer is used to provide a variety of advanced application function modules based on the full lifecycle data lake and digital twin model; The unified data service and interface layer is used to provide standardized data access APIs, model calling services, simulation computing services, and visualization component services for upper-layer applications and external systems in a microservice architecture, ensuring the system's openness, scalability, and ease of integration.

2. The system for full lifecycle data management and application of a rail train braking system based on digital twins as described in claim 1, characterized in that, The heterogeneous data sources in the unified data access and processing layer include: simulation models and data in the design phase, bill of materials (BOM) and quality inspection reports in the manufacturing phase, real-time data and fault logs from vehicle sensors in the operation phase, and maintenance records and component replacement history in the maintenance phase.

3. The system for full lifecycle data management and application of a rail train braking system based on digital twins as described in claim 1, characterized in that, The digital twin model layer also includes: a static model library and a dynamic model library; The static models in the static model library include: the three-dimensional geometric model of the braking system, the assembly relationship model, and the BOM (Bill of Materials). The dynamic models in the dynamic model library include: mechanistic models based on physicochemical rules and data-driven models trained on historical data.

4. The system for full lifecycle data management and application of a rail train braking system based on digital twins as described in claim 1, characterized in that, The intelligent application layer includes: a control strategy simulation and testing module, a predictive maintenance decision-making module, and an asset performance overview and optimization module; The control strategy simulation test module is used to load the new algorithm onto the digital twin and run the simulation by inputting a piece of real data of historical low adhesion conditions to quickly verify the effectiveness and safety of the new algorithm; The predictive maintenance decision module is used to automatically generate early warning work orders when the degradation model in the digital twin predicts that the performance will exceed the tolerance range after a certain period of time, and recommends to check in the next planned maintenance to avoid online failures. The asset performance dashboard and optimization module are used to query data and then make decisions.

5. A method for managing and applying the full lifecycle data of a rail train braking system based on digital twins, wherein the applicable system is the rail train braking system for managing and applying the full lifecycle data of a rail train braking system based on digital twins as described in any one of claims 1-4, characterized in that, Includes the following steps: Step S1, Data Access and Processing: The unified data access and processing layer accesses data from various sources; after standardization, the data is stored in the full lifecycle data lake; Step S2, Digital Twin Construction and Operation: The model management and online update engine in the digital twin model layer continuously acquires data from the full lifecycle data lake to drive and calibrate the model; Step S3, Smart Application Execution, specifically includes: Step S31, Control Strategy Simulation Test: In the control strategy simulation test module, the new algorithm is loaded onto the digital twin, and a piece of real data from a historical low-adhesion working condition is input for simulation to quickly verify the effectiveness and safety of the new algorithm. Step S32, Predictive Maintenance Decision: When the degradation model in the digital twin predicts that the performance will exceed the tolerance range after a certain period of time, the predictive maintenance decision module will automatically generate an early warning work order and recommend checking in the next planned maintenance to avoid online failures. Step S33, Asset Performance Panorama Dashboard: Use the Asset Performance Panorama Dashboard and the optimization module to query data and make decisions.

Images