A spaceflight power test data whole-process traceability management method and system

Abstract

The present application relates to a kind of aerospace power test data whole-process traceability management method and system, it is related to data traceability management field, the present application is generated by deploying distributed high-precision data acquisition system and using IRIG-B hardware clock synchronization, tamper-evident original physical reference data packet, introduce lightweight sandbox intercept agent, automatically solidify dynamic grid partition parameter, boundary condition, material physical property curve file and algorithm version Hash Such as implicit intermediate state parameter, adopt directed acyclic graph to dynamically build whole life cycle data blood relationship atlas and through block chain hash chain, finally through visual traceability audit interactive interface, realize one-key to call out intermediate state parameter, expand physical property dynamic curve and model source code, and support sandbox one-key re-calculation verification.

Description

Technical Field

[0001] This invention relates to the field of data traceability and control technology, and in particular to a method and system for full-process traceability and control of aerospace power test data. Background Technology

[0002] In aerospace propulsion tests, such as liquid rocket engine firing, solid rocket motor ground ignition, and turbopump and attitude control propulsion system hot-fire tests, data traceability and control are crucial for ensuring zero-risk test results, accurate fault location, and controlled product technical status. Aerospace propulsion tests are characterized by high risk, multi-physics coupling, and severe transient responses, placing extremely high demands on the rigor of data acquisition and post-test auditing.

[0003] However, in existing aerospace propulsion test data processing workflows, raw physical quantities (such as millisecond-level dynamic pressure, transient temperature, and high-frequency strain) often require multiple levels of complex calculations, filtering, downsampling, and algorithmic encapsulation (such as thrust reconstruction algorithms, flow conversion models, digital twin real-time simulation data, and inverse heat transfer solution models). During post-fault troubleshooting or quality audits, auditors or experts often only see processed secondary and tertiary characteristic data (such as average heat flux density, peak thrust, and steady-state flow), losing the evolution path and intermediate state parameters between the raw physical quantities and the final calculation results (i.e., the black box and untraceability of intermediate state parameters).

[0004] For example, in the analysis of test data for a certain type of high-thrust liquid rocket engine, researchers discovered that the calculated heat flux density of a key cross-section exceeded the standard. Tracing the source of this calculation result revealed that the index was not a directly measured value, but rather a secondary characteristic data calculated using instantaneous temperature data collected by wall thermocouples combined with a reverse heat transfer mathematical model. Because traditional data acquisition and archiving systems do not record all elements of the calculation model, such as the boundary condition settings, mesh generation parameters, and material property parameters dynamically changing with temperature (i.e., implicit intermediate parameters), different analysts used different versions of the solution script to calculate different heat flux results. Furthermore, it was impossible to prove which process fully reflected the actual physical state at the time, significantly prolonging the cycle of fault location and quality restoration. To address the aforementioned technical deficiencies, a solution is proposed. Summary of the Invention

[0005] The purpose of this invention is to generate an immutable original physical reference data package by deploying a distributed high-precision data acquisition system and using IRIG-B hardware clock synchronization. A lightweight sandbox interception agent is introduced to automatically solidify implicit intermediate state parameters such as dynamic mesh partitioning parameters, boundary conditions, material property curve files, and algorithm version hashes. A directed acyclic graph is used to dynamically construct a full life cycle data lineage map and upload it to the blockchain via hashing. Finally, through a visual traceability and auditing interactive interface, intermediate state parameters can be retrieved with one click, dynamic property curves and model source code can be expanded, and one-click recalculation verification in the sandbox is supported.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for full-process traceability and control of aerospace propulsion test data, comprising the following steps: S1. Deploy a distributed high-precision data acquisition system on the aerospace power test bench. Synchronously acquire high-frequency raw physical quantities and test environment metadata through a unified IRIG-B hardware clock source, and integrate them into an unalterable raw physical reference data package. S2. When performing filtering, downsampling and advanced algorithm encapsulation on the original physical reference data package, a lightweight sandbox interception agent is introduced. The lightweight sandbox automatically captures and solidifies all intermediate state parameters in the calculation process. The intermediate state parameters include dynamic mesh generation parameters, boundary condition setting basis, function curve files of material property parameters, Git version hash value of algorithm script and version number of the numerical calculation library on which it depends. S3. Using the original physical data as the root node, the feature data, intermediate state parameters, and algorithm model generated by multi-level calculations are used as child nodes. A directed acyclic graph is used to dynamically construct a full life cycle data lineage graph, and the data is uploaded to the chain through cryptographic hashing. S4. Provides a visual traceability and auditing interactive interface. When experts question a certain secondary or tertiary feature data, the intermediate state parameters on which the feature data was generated can be retrieved, and the dynamic curve of the material properties and the source code of the inverse heat transfer model at that time can be expanded with one click. It also supports one-click rerun in the sandbox environment.

[0007] Furthermore, the specific process for generating the original physical baseline data packet is as follows: S101. In the distributed high-precision data acquisition system, a high-precision hardware clock synchronization source conforming to the IRIG-B standard is deployed. Each front-end sensor in the distributed high-precision data acquisition system directly receives the clock synchronization signal through the hardware bus and stamps the high-frequency raw physical quantities with an absolute timestamp. S102. While acquiring high-frequency raw physical quantities, automatically capture and associate the calibration metadata obtained by the front-end sensor from the national metrology institution before the test. The calibration metadata includes the sensor calibration coefficient matrix, zero-point drift correction table, hardware channel physical interface number, and the unique physical device ID of the data acquisition device. Bind the calibration metadata to the high-frequency raw physical quantities at the underlying level to obtain the raw data stream. S103. Within the time window before and after the ignition command is triggered by the distributed high-precision data acquisition system, synchronously capture the test environment metadata recorded by the test bench control system. The test environment metadata includes meteorological conditions, gas supply pressure, and initial preset state of valves. Inject the test environment metadata as a structured tag into the raw data stream. S104. The high-frequency raw physical quantities, calibration metadata, hardware channel configuration and test environment metadata that have been time aligned are packaged together to obtain the initial physical reference data package. A unique digital fingerprint is calculated on the initial physical reference data package using a cryptographic hash algorithm, and the digital fingerprint is written into the blockchain node to finally generate an immutable raw physical reference data package.

[0008] Furthermore, the lightweight sandbox specifically includes: The environment context isolation module is used to create an independent isolated runtime space for each advanced algorithm encapsulation. This isolated runtime space is independent of the host system's global environment variables, ensuring that the numerical computing library version, dynamic link library, and interpreter environment on which the algorithm depends are fully controlled and can be accurately recorded. The dynamic hijacking and interception proxy module is a lightweight proxy program deployed at the edge of a lightweight sandbox. It is used to monitor the algorithm script's reading operations on external files and writing operations on output result files in real time, achieving seamless interception of data streams without intruding on the original business source code. Runtime status probes, embedded within numerical computation loops and solvers, are used to capture implicit intermediate parameters in real time during computation execution. The version and source fingerprint generator is used to obtain the unique version identifier of the currently executing algorithm script in real time, and package it with all captured intermediate state parameters to generate an intermediate state dataset with timestamps and cryptographic hash verification.

[0009] Furthermore, the specific process for capturing intermediate state parameters is as follows: S201. When an analyst starts the reverse heat transfer calculation script in the lightweight sandbox environment, the environment context isolation module is activated. The version and source fingerprint generator automatically queries the version control system to obtain the Git version hash value of the current calculation script, and at the same time scans and locks the specific version number of the numerical calculation library loaded inside the lightweight sandbox. S202. During the calculation process, it is necessary to continuously call the function curve file of material property parameters. The function curve file reading request is intercepted by the dynamic hijacking and interception proxy module, and a complete binary copy of the function curve file is copied and saved to the persistent temporary storage area of ​​the lightweight sandbox. S203. When the calculation enters the core iterative solution stage, the runtime state probe embedded in the solver is triggered, automatically capturing the dynamic mesh generation parameters generated at that moment, and capturing the boundary condition setting basis under the current step in real time. S204. When the calculation process is successfully completed, the lightweight sandbox will package all the intermediate state parameters automatically captured above into a unified package.

[0010] Furthermore, the specific process of constructing a full lifecycle data lineage map is as follows: S301. Define the original physical reference data packet generated in S1 as the root node of the entire directed acyclic graph. The root node fully encapsulates the high-frequency timing physical quantities, sensor calibration coefficient matrix and test environment metadata that are aligned with the IRIG-B hardware clock. S302. The dynamic mesh generation parameters, boundary condition setting basis, function curve files of material property parameters, Git version hash value of algorithm script and version number of the numerical computing library it depends on are each used as an independent intermediate state child node. S303. Utilizing the topological structure of a directed acyclic graph, the root node, intermediate state child nodes, algorithm model child nodes, and the finally generated second-level and third-level feature data nodes are connected by directed edges. The direction of the directed edges follows the causal time sequence of data flow and computation. S304. Perform cryptographic hash calculations on the topology of the entire directed acyclic graph, the data content of each node, and the relationships between nodes to generate a unique global lineage fingerprint, and write the global lineage fingerprint into the underlying storage of the blockchain.

[0011] Furthermore, the specific process of retrieving the intermediate state parameters upon which the generation of the feature data depends is as follows: S401. When an expert clicks on a questionable secondary feature data node on the visual audit interactive interface, based on the topological connection relationship of the directed acyclic graph, the reverse addressing algorithm is automatically executed to lock the target node of the secondary feature data node in the full life cycle data lineage graph, and traverses all its direct parent nodes and indirect dependent nodes in reverse to generate the range of underlying intermediate state parameters involved in the secondary feature data node. S402. While performing reverse addressing, the cryptographic hash fingerprints of the target node and its associated child nodes are automatically extracted and compared in real time with the global lineage fingerprints fixed in the underlying storage of the blockchain. After the verification is passed, it is confirmed that the data link has compliant legal-level anti-tampering effect, and then the corresponding intermediate state parameters are authorized to be loaded and decrypted. S403. Real-time display of the dynamic mesh generation parameters used in the current calculation, the basis for setting boundary conditions, the function curve file of material property parameters, as well as the Git version hash value of the algorithm script used to perform the calculation and the version number of the numerical calculation library it depends on. S404, the visual traceability audit interface provides a one-click expansion function, which presents the dynamic curves of material properties in a visual chart format.

[0012] This invention also provides a full-process traceability and control system for aerospace propulsion test data, including a data acquisition unit, a data interception unit, a map construction unit, and a visualization query unit, wherein: The data acquisition unit is used to deploy a distributed high-precision data acquisition system on the aerospace power test bench. It synchronously acquires high-frequency raw physical quantities and test environment metadata through a unified IRIG-B hardware clock source and integrates them into an unalterable raw physical reference data package. The data interception unit is used to introduce a lightweight sandbox interception agent when performing filtering, downsampling and advanced algorithm encapsulation on the original physical reference data packet. The lightweight sandbox automatically captures and solidifies all intermediate state parameters in the calculation process. The graph construction unit is used to dynamically construct a full life cycle data lineage graph using the original physical data as the root node, and the feature data, intermediate state parameters, and algorithm models generated by multi-level calculations as child nodes, and then upload it to the blockchain through cryptographic hashing. The visualization query unit provides a visual traceability and auditing interactive interface. When an expert questions a certain secondary or tertiary feature data, the intermediate state parameters on which the feature data was generated can be retrieved, and the dynamic curve of the material properties and the source code of the inverse heat transfer model at that time can be expanded with one click. It also supports one-click rerun in the sandbox environment.

[0013] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: This method and system for full-process traceability and control of aerospace propulsion test data generates unalterable original physical reference data packages by deploying a distributed high-precision data acquisition system and using IRIG-B hardware clock synchronization. It introduces a lightweight sandbox interception agent to automatically solidify implicit intermediate state parameters such as dynamic mesh partitioning parameters, boundary conditions, material property curve files, and algorithm version hashes. It uses a directed acyclic graph to dynamically construct a full lifecycle data lineage map and uploads it to the blockchain via hashing. Finally, through a visual traceability and auditing interactive interface, it enables one-click retrieval of intermediate state parameters, expansion of material property dynamic curves and model source code, and supports one-click sandbox recalculation verification. Attached Figure Description

[0014] Figure 1 A schematic diagram of the overall method flow of the present invention is shown; Figure 2 A schematic diagram of the overall structure of the present invention is shown. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: like Figure 1 As shown, a method for full-process traceability and control of aerospace propulsion test data includes the following steps: S1. Deploy a distributed high-precision data acquisition system on the aerospace power test bench. Synchronously acquire high-frequency raw physical quantities and test environment metadata through a unified IRIG-B hardware clock source, and integrate them into an unalterable raw physical reference data package. The specific process for generating the original physical baseline data packet is as follows: S101. In the distributed high-precision data acquisition system, a high-precision hardware clock synchronization source conforming to the IRIG-B standard is deployed. Each front-end sensor in the distributed high-precision data acquisition system directly receives the clock synchronization signal through the hardware bus and marks the high-frequency raw physical quantities (such as millisecond-level dynamic pressure fluctuations, thermocouple transient millivolt voltage signals, etc.) with an absolute timestamp. S102. While acquiring high-frequency raw physical quantities, automatically capture and associate the calibration metadata obtained by the front-end sensor from the national metrology institution before the test. The calibration metadata includes the sensor calibration coefficient matrix, zero-point drift correction table, hardware channel physical interface number, and the unique physical device ID of the data acquisition device. Bind the calibration metadata to the high-frequency raw physical quantities at the underlying level to obtain the raw data stream. S103. Within the time window before and after the ignition command is triggered by the distributed high-precision data acquisition system, synchronously capture the test environment metadata recorded by the test bench control system. The test environment metadata includes meteorological conditions, gas supply pressure, and initial preset state of valves. Inject the test environment metadata as a structured tag into the raw data stream. S104. The high-frequency raw physical quantities, calibration metadata, hardware channel configuration and test environment metadata that have been time aligned are packaged together to obtain the initial physical reference data package. A unique digital fingerprint is calculated on the initial physical reference data package using a cryptographic hash algorithm (such as the national cryptographic standard SM3 or SHA-256), and the digital fingerprint is written into the blockchain node to finally generate an immutable raw physical reference data package.

[0017] S2. When performing filtering, downsampling and advanced algorithm encapsulation on the original physical reference data package, a lightweight sandbox interception agent is introduced. The lightweight sandbox automatically captures and solidifies all intermediate state parameters in the calculation process. The intermediate state parameters include dynamic mesh generation parameters, boundary condition setting basis, function curve files of material property parameters, Git version hash value of algorithm script and version number of the numerical calculation library on which it depends. Lightweight sandboxes specifically include: The environment context isolation module is used to create an independent isolated runtime space for each advanced algorithm encapsulation. This isolated runtime space is independent of the host system's global environment variables, ensuring that the numerical computing library version, dynamic link library, and interpreter environment on which the algorithm depends are fully controlled and can be accurately recorded. The dynamic hijacking and interception proxy module is a lightweight proxy program deployed at the edge of a lightweight sandbox. It is used to monitor in real time the algorithm script's reading operations on external files (such as physical property parameter curve files and mesh configuration files) and writing operations on output result files, so as to achieve seamless interception of data flow without intruding on the original business source code. Runtime status probes are embedded within numerical computation loops and solvers to capture implicit intermediate parameters in real time during the computation execution phase (such as the number of elements and node coordinates of dynamic mesh generation, the basis for real-time setting of boundary conditions, and dynamic function tables of material property parameters as a function of temperature). The version and source fingerprint generator is used to obtain the unique version identifier of the currently executing algorithm script in real time, and package it with all captured intermediate state parameters to generate an intermediate state dataset with timestamps and cryptographic hash verification.

[0018] The specific process of capturing intermediate state parameters is as follows: S201. When an analyst starts the reverse heat transfer calculation script in the lightweight sandbox environment, the environment context isolation module is activated. The version and source fingerprint generator automatically queries the version control system to obtain the Git version hash value of the current calculation script, and at the same time scans and locks the specific version number of the numerical calculation library loaded inside the lightweight sandbox. S202. During the calculation process, it is necessary to continuously call the function curve file of material property parameters. The function curve file reading request is intercepted by the dynamic hijacking and interception proxy module, and a complete binary copy of the function curve file is copied and saved to the persistent temporary storage area of ​​the lightweight sandbox. S203. When the calculation enters the core iterative solution stage, the runtime state probe embedded in the solver is triggered, automatically capturing the dynamic mesh generation parameters generated at that moment (such as the number of mesh cells, density distribution pattern and boundary node coordinates), and capturing the boundary condition setting basis under the current step in real time (such as the real-time constraint value of the convective heat transfer coefficient). S204. When the calculation process is successfully completed, the lightweight sandbox will package all the intermediate state parameters automatically captured above into a unified package.

[0019] S3. Using the original physical data as the root node, the feature data, intermediate state parameters, and algorithm model generated by multi-level calculations are used as child nodes. A directed acyclic graph is used to dynamically construct a full life cycle data lineage graph, and the data is uploaded to the chain through cryptographic hashing. The specific process of constructing a full lifecycle data lineage map is as follows: S301. Define the original physical reference data packet generated in S1 as the root node of the entire directed acyclic graph. The root node fully encapsulates the high-frequency timing physical quantities, sensor calibration coefficient matrix and test environment metadata that are aligned with the IRIG-B hardware clock. S302. The dynamic mesh generation parameters, boundary condition setting basis, function curve files of material property parameters, Git version hash value of algorithm script and version number of the numerical computing library it depends on are each used as an independent intermediate state child node. S303. Utilizing the topological structure of a directed acyclic graph, the root node, intermediate state child nodes, algorithm model child nodes, and the finally generated second-level and third-level feature data nodes are connected by directed edges. The direction of the directed edges follows the causal time sequence of data flow and computation. S304. Perform cryptographic hash calculations on the topology of the entire directed acyclic graph, the data content of each node, and the relationships between nodes to generate a unique global lineage fingerprint, and write the global lineage fingerprint into the underlying storage of the blockchain.

[0020] S4. Provides a visual traceability and auditing interactive interface. When experts question a certain secondary or tertiary feature data, the intermediate state parameters on which the feature data was generated can be retrieved, and the dynamic curve of the material properties and the source code of the inverse heat transfer model at that time can be expanded with one click. It also supports one-click rerun in the sandbox environment.

[0021] The specific process for retrieving the intermediate state parameters upon which the generation of the feature data depends is as follows: S401. When an expert clicks on a questionable secondary feature data node (such as the calculated value of heat flux density exceeding the standard) on the visual audit interactive interface, based on the topological connection relationship of the directed acyclic graph, the reverse addressing algorithm is automatically executed to lock the target node of the secondary feature data node in the full life cycle data lineage graph, and traverses all its direct parent nodes and indirect dependent nodes in reverse to generate the range of underlying intermediate state parameters involved in the secondary feature data node. S402. While performing reverse addressing, the cryptographic hash fingerprints of the target node and its associated child nodes are automatically extracted and compared in real time with the global lineage fingerprints fixed in the underlying storage of the blockchain. After the verification is passed, it is confirmed that the data link has compliant legal-level anti-tampering effect, and then the corresponding intermediate state parameters are authorized to be loaded and decrypted. S403 displays in real time the dynamic meshing parameters used in the current calculation (such as the number of mesh cells and node distribution), the basis for setting boundary conditions (such as convection heat transfer coefficient constraints), the function curve file of material property parameters, and the Git version hash value of the algorithm script that was executing the calculation and the version number of the numerical computing library it depends on. S404, the visual traceability audit interface provides a one-click expansion function, which presents the dynamic curves of material properties in a visual chart format.

[0022] Example 2: like Figure 2 As shown, a full-process traceability and control system for aerospace propulsion test data includes a data acquisition unit, a data interception unit, a map construction unit, and a visualization query unit, wherein: The data acquisition unit is used to deploy a distributed high-precision data acquisition system on the aerospace power test bench. It synchronously acquires high-frequency raw physical quantities and test environment metadata through a unified IRIG-B hardware clock source and integrates them into an unalterable raw physical reference data package. The data interception unit is used to introduce a lightweight sandbox interception agent when performing filtering, downsampling and advanced algorithm encapsulation on the original physical reference data packet. The lightweight sandbox automatically captures and solidifies all intermediate state parameters in the calculation process. The graph construction unit is used to dynamically construct a full life cycle data lineage graph using the original physical data as the root node, and the feature data, intermediate state parameters, and algorithm models generated by multi-level calculations as child nodes, and then upload it to the blockchain through cryptographic hashing. The visualization query unit provides a visual traceability and auditing interactive interface. When an expert questions a certain secondary or tertiary feature data, the intermediate state parameters on which the feature data was generated can be retrieved, and the dynamic curve of the material properties and the source code of the inverse heat transfer model at that time can be expanded with one click. It also supports one-click rerun in the sandbox environment.

[0023] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value.

[0024] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. In the two embodiments provided in this application, it should be understood that the disclosed apparatus and system can be implemented in other ways; for example, the apparatus embodiments described above are merely illustrative, and the division of modules is merely a logical functional division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed; another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the apparatus or module can be electrical, mechanical or other forms. The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for full-process traceability and control of aerospace propulsion test data, characterized in that, Includes the following steps: S1. Deploy a distributed high-precision data acquisition system on the aerospace power test bench. Synchronously acquire high-frequency raw physical quantities and test environment metadata through a unified IRIG-B hardware clock source, and integrate them into an unalterable raw physical reference data package. S2. When performing filtering, downsampling and advanced algorithm encapsulation on the original physical reference data package, a lightweight sandbox interception agent is introduced. The lightweight sandbox automatically captures and solidifies all intermediate state parameters in the calculation process. The intermediate state parameters include dynamic mesh generation parameters, boundary condition setting basis, function curve files of material property parameters, Git version hash value of algorithm script and version number of the numerical calculation library on which it depends. S3. Using the original physical data as the root node, the feature data, intermediate state parameters, and algorithm model generated by multi-level calculations are used as child nodes. A directed acyclic graph is used to dynamically construct a full life cycle data lineage graph, and the data is uploaded to the chain through cryptographic hashing. S4. Provides a visual traceability and auditing interactive interface. When experts question a certain secondary or tertiary feature data, the intermediate state parameters on which the feature data was generated can be retrieved, and the dynamic curve of the material properties and the source code of the inverse heat transfer model at that time can be expanded with one click. It also supports one-click rerun in the sandbox environment.

2. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, The specific process of data acquisition in S1 is as follows: S101. In the distributed high-precision data acquisition system, a high-precision hardware clock synchronization source conforming to the IRIG-B standard is deployed. Each front-end sensor in the distributed high-precision data acquisition system directly receives the clock synchronization signal through the hardware bus and stamps the high-frequency raw physical quantities with an absolute timestamp. S102. While acquiring high-frequency raw physical quantities, automatically capture and associate the calibration metadata obtained by the front-end sensor from the national metrology institution before the test. The calibration metadata includes the sensor calibration coefficient matrix, zero-point drift correction table, hardware channel physical interface number, and the unique physical device ID of the data acquisition device. Bind the calibration metadata to the high-frequency raw physical quantities at the underlying level to obtain the raw data stream.

3. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, The specific process of generating the original physical reference data packet is as follows: Within the time window before and after the ignition command is triggered by the distributed high-precision data acquisition system, the test environment metadata recorded by the test bench measurement and control system is captured synchronously. The test environment metadata includes meteorological conditions, gas supply pressure, and initial preset state of valves. The test environment metadata is injected into the raw data stream as a structured tag. The initial physical benchmark data package is obtained by integrating the time-aligned high-frequency raw physical quantities, calibration metadata, hardware channel configuration, and test environment metadata. A unique digital fingerprint is calculated on the initial physical benchmark data package using a cryptographic hash algorithm, and the digital fingerprint is written into the blockchain node to finally generate an immutable raw physical benchmark data package.

4. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, Lightweight sandboxes specifically include: The environment context isolation module is used to create an independent isolated runtime space for each advanced algorithm encapsulation. This isolated runtime space is independent of the host system's global environment variables, ensuring that the numerical computing library version, dynamic link library, and interpreter environment on which the algorithm depends are fully controlled and can be accurately recorded. The dynamic hijacking and interception proxy module is a lightweight proxy program deployed at the edge of a lightweight sandbox. It is used to monitor the algorithm script's reading operations on external files and writing operations on output result files in real time, achieving seamless interception of data streams without intruding on the original business source code. Runtime status probes, embedded within numerical computation loops and solvers, are used to capture implicit intermediate parameters in real time during computation execution.

5. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, The lightweight sandbox also includes a version and source fingerprint generator, which is used to obtain the unique version identifier of the currently executing algorithm script in real time and package it with all captured intermediate state parameters to generate an intermediate state dataset with timestamps and cryptographic hash verification.

6. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, The specific process of capturing intermediate state parameters is as follows: S201. When an analyst starts the reverse heat transfer calculation script in the lightweight sandbox environment, the environment context isolation module is activated. The version and source fingerprint generator automatically queries the version control system to obtain the Git version hash value of the current calculation script, and at the same time scans and locks the specific version number of the numerical calculation library loaded inside the lightweight sandbox. S202. During the calculation process, it is necessary to continuously call the function curve file of material property parameters. The function curve file reading request is intercepted by the dynamic hijacking and interception proxy module, and a complete binary copy of the function curve file is copied and saved to the persistent temporary storage area of ​​the lightweight sandbox. S203. When the calculation enters the core iterative solution stage, the runtime state probe embedded in the solver is triggered, automatically capturing the dynamic mesh generation parameters generated at that moment, and capturing the boundary condition setting basis under the current step in real time. S204. When the calculation process is successfully completed, the lightweight sandbox will package all the intermediate state parameters automatically captured above into a unified package.

7. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, The specific process of constructing a full lifecycle data lineage map is as follows: S301. Define the original physical reference data packet generated in S1 as the root node of the entire directed acyclic graph. The root node fully encapsulates the high-frequency timing physical quantities, sensor calibration coefficient matrix and test environment metadata that are aligned with the IRIG-B hardware clock. S302. The dynamic mesh generation parameters, boundary condition setting basis, function curve files of material property parameters, Git version hash value of algorithm script and version number of the numerical computing library it depends on are each used as an independent intermediate state child node. S303. Utilizing the topological structure of a directed acyclic graph, the root node, intermediate state child nodes, algorithm model child nodes, and the finally generated second-level and third-level feature data nodes are connected by directed edges. The direction of the directed edges follows the causal time sequence of data flow and computation. S304. Perform cryptographic hash calculations on the topology of the entire directed acyclic graph, the data content of each node, and the relationships between nodes to generate a unique global lineage fingerprint, and write the global lineage fingerprint into the underlying storage of the blockchain.

8. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, The specific process for retrieving the intermediate state parameters upon which the generation of the feature data depends is as follows: S401. When an expert clicks on a questionable secondary feature data node on the visual audit interactive interface, based on the topological connection relationship of the directed acyclic graph, the reverse addressing algorithm is automatically executed to lock the target node of the secondary feature data node in the full life cycle data lineage graph, and traverses all its direct parent nodes and indirect dependent nodes in reverse to generate the range of underlying intermediate state parameters involved in the secondary feature data node. S402. During reverse addressing, the cryptographic hash fingerprints of the target node and its associated child nodes are automatically extracted and compared in real time with the global lineage fingerprints fixed in the underlying storage of the blockchain. After the verification is passed, it is confirmed that the data link has compliant legal-level anti-tampering effect, and then the corresponding intermediate state parameters are authorized to be loaded and decrypted.

9. The method for full-process traceability and control of aerospace propulsion test data according to claim 1, characterized in that, The visual traceability and auditing interactive interface is used to display in real time the dynamic mesh generation parameters used in the current calculation, the basis for setting boundary conditions, the function curve file of material property parameters, the Git version hash value of the algorithm script that performed the calculation at that time, and the version number of the numerical calculation library it depends on. The visual traceability audit interface also provides a one-click expand function, which presents the dynamic curves of material properties in a visual chart format.

10. A full-process traceability and control system for aerospace propulsion test data, characterized in that, The method for full-process traceability and control of aerospace propulsion test data, as described in claims 1-9, includes a data acquisition unit, a data interception unit, a map construction unit, and a visualization query unit, wherein: The data acquisition unit is used to deploy a distributed high-precision data acquisition system on the aerospace power test bench. It synchronously acquires high-frequency raw physical quantities and test environment metadata through a unified IRIG-B hardware clock source and integrates them into an unalterable raw physical reference data package. The data interception unit is used to introduce a lightweight sandbox interception agent when performing filtering, downsampling and advanced algorithm encapsulation on the original physical reference data packet. The lightweight sandbox automatically captures and solidifies all intermediate state parameters in the calculation process. The graph construction unit is used to dynamically construct a full life cycle data lineage graph using the original physical data as the root node, and the feature data, intermediate state parameters, and algorithm models generated by multi-level calculations as child nodes, and then upload it to the blockchain through cryptographic hashing. The visualization query unit provides a visual traceability and auditing interactive interface. When an expert questions a certain secondary or tertiary feature data, the intermediate state parameters on which the feature data was generated can be retrieved, and the dynamic curve of the material properties and the source code of the inverse heat transfer model at that time can be expanded with one click. It also supports one-click rerun in the sandbox environment.

Images