A method for failure prediction and health management based on agile systems engineering

Abstract

The application discloses a kind of based on agile system engineering's fault prediction and health management method, the method is based on SysML modeling, the concept of PHM system requirement, function, logic, physical architecture modeling analysis is fused into PHM architecture design method, the traceability between each model is established, when PHM system requirement changes, then can according to traceability, the influence range of requirement change is analyzed, with recursive and iterative mode to update system model from the system engineering point of view, support PHM system rapid iteration upgrade.The application supports the architecture of PHM system leads, based on model, agile development mode of data-driven, forms the standardization support environment suitable for PHM agile system engineering, supports the rapid iteration upgrade of PHM system, provides technical means for the use and maintenance support of equipment, to the reliability, availability, security, maintainability of equipment system, promote the innovation and development of equipment system have important significance.

Description

Technical Field

[0001] This invention belongs to the field of computer software technology and relates to a fault prediction and health management method, specifically a fault prediction and health management method based on agile systems engineering. Background Technology

[0002] Predictive Fault and Health Management (PHM) collects and processes various data from the system, using intelligent reasoning algorithms to diagnose and predict faults, assess the system's health status, and trigger optimal maintenance decisions based on available resources and constraints. PHM is of great significance for improving the safety, reliability, and maintainability of complex equipment, reducing life-cycle costs, and achieving autonomous and predictive maintenance.

[0003] Model-based systems engineering (MBSE) is the formal application of modeling methods to support activities such as system requirements, design, analysis, verification, and validation, which begin in the conceptual design phase and continue throughout the design and development phases and the entire lifecycle.

[0004] Agile Systems Engineering (aMBSE) uses system architecture models as the main work artifacts throughout the entire system development lifecycle. It adopts an architecture-led, model-based, and data-driven development model, supports highly scalable and simulable system architectures, and builds complex systems in an incremental and iterative manner.

[0005] SysML (System Modeling Language) is a general-purpose system modeling language that includes nine types of diagrams: module definition diagrams, internal module diagrams, use case diagrams, activity diagrams, sequence diagrams, state machine diagrams, parameter diagrams, package diagrams, and requirements diagrams. It supports the specification, analysis, design, and verification of complex systems that may include hardware, software, information, personnel, programs, and facilities. It allows for modeling of systems at the structural, behavioral, and requirements levels, and imposes mathematical or logical constraints on system models to facilitate understanding among engineers and designers from different knowledge backgrounds, eliminating comprehension barriers and ensuring the correctness, consistency, and completeness of the entire development process.

[0006] Existing PHM technologies largely focus on diagnostic and predictive models and algorithms. However, with the increasing complexity of high-tech equipment functions, structures, and technologies, PHM is no longer an isolated supplementary system supporting maintenance, but has become an indispensable system closely integrated with other systems such as structure, power, electromechanical, and environmental control. PHM must be implemented at the initial conceptualization, design, and development stages of engineering systems to integrate it into the overall system framework and meet overall operational requirements. Further research into PHM system development methodologies is necessary to capture its interaction with various systems, enabling accurate system-level diagnostics and predictions. In particular, the constantly evolving needs of stakeholders and system requirements necessitate a highly agile PHM system development process and architecture, enabling the tracking and validation of PHM requirements within a unified architecture and supporting rapid iterative upgrades. However, effective solutions are currently lacking. Summary of the Invention

[0007] To address the shortcomings of existing technologies in ensuring traceability, consistency, and reusability of PHM systems, the lack of systematic approaches to developing and deploying PHM systems, the absence of visual modeling support tools for systematically building PHM systems, and the difficulty in meeting the rapid development and upgrade requirements of PHM systems, this invention provides a fault prediction and health management method based on agile systems engineering. This method, based on SysML modeling, integrates the concepts of PHM requirements, functions, logic, and physical architecture modeling and analysis into the PHM architecture design methodology, supporting the construction of a PHM system development support platform environment.

[0008] The objective of this invention is achieved through the following technical solution:

[0009] A fault prediction and health management method based on agile systems engineering includes the following steps:

[0010] Step S1: Interact with stakeholders to determine the top-level requirements of PHM and generate PHM system design requirements;

[0011] Step S2: Design and generate the functional architecture of the PHM system;

[0012] Step S3: Design the execution logic architecture of the PHM system;

[0013] Step S4: Based on the fault impact analysis, design the physical architecture of the PHM system and assign functional elements to the system elements of the candidate physical architecture;

[0014] Step S5: Through the continuous transfer and mapping of the PHM system from requirements to functional architecture, logical architecture and then to physical architecture, the division, alignment and allocation from system requirements to architectural entities and then to system elements are realized;

[0015] Step S6: By analyzing the traceable links of the SysML model, analyze and examine the design elements and interfaces of the functional architecture, logical architecture, and physical architecture from functional, logical, and physical perspectives to verify the correctness of the model;

[0016] Step S7: Perform a change impact analysis based on traceability and iteratively update the PHM system.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] This invention integrates the principles of systems engineering into the entire lifecycle of PHM system development, supports architecture-led, model-based, and data-driven agile development models for PHM systems, forms a standardized support environment suitable for agile PHM systems engineering, supports rapid iterative upgrades of PHM systems, provides technical means for equipment use and maintenance support, and is of great significance for ensuring the reliability, availability, security, and maintainability of equipment systems and promoting the innovative development of equipment systems. Attached Figure Description

[0019] Figure 1 This is a flowchart of the fault prediction and health management method based on agile systems engineering according to the present invention;

[0020] Figure 2 Example of a use case diagram for PHM task objectives and functions;

[0021] Figure 3 Example of a PHM stakeholder requirements diagram;

[0022] Figure 4 Example of PHM functional requirements decomposition;

[0023] Figure 5 This serves as an example of PHM functional requirement traceability.

[0024] Figure 6 This is an example of a PHM domain model;

[0025] Figure 7 This is an example of a PHM functional hierarchy architecture;

[0026] Figure 8 Example of a PHM system activity diagram;

[0027] Figure 9 Example of a state machine diagram for a diagnostic module;

[0028] Figure 10 Example of PHM physical architecture;

[0029] Figure 11 Example of updated feature requirements;

[0030] Figure 12For the updated functional architecture diagram;

[0031] Figure 13 This is the updated functional architecture diagram. Detailed Implementation

[0032] The technical solution of the present invention will be further described below with reference to the accompanying drawings, but it is not limited thereto. Any modifications or equivalent substitutions to the technical solution of the present invention that do not depart from the spirit and scope of the technical solution of the present invention should be covered within the protection scope of the present invention.

[0033] This invention provides a fault prediction and health management method based on agile systems engineering. The method, based on SysML modeling, integrates the concepts of PHM system requirements, functions, logic, and physical architecture modeling and analysis into the PHM architecture design method, establishing traceability relationships between various models. When PHM system requirements change, the impact scope of the change can be analyzed based on the traceability relationships, and the system model can be updated recursively and iteratively from a systems engineering perspective, supporting rapid iterative upgrades of the PHM system. Figure 1 As shown, it includes the following steps:

[0034] Step S1: Interact with stakeholders to determine the top-level requirements of the PHM system and generate PHM system design requirements. Specific steps are as follows:

[0035] Step S101: Use use case diagrams to describe the goals of the PHM system and the external systems involved in achieving those goals, for example... Figure 2 As shown.

[0036] Step S102: Construct a requirements diagram of PHM stakeholders that supports the task objectives, as shown in the example below. Figure 3 As shown.

[0037] Step S103: Based on the requirements diagram constructed in step S102, further decompose the functional requirements of the PHM system. An example of PHM system functional requirement decomposition is shown below. Figure 4 As shown.

[0038] Step S104: Establish the traceability relationship between the functional requirements of the PHM system, for example... Figure 5 As shown.

[0039] Step S2: Design and generate the functional architecture of the PHM system. The specific steps are as follows:

[0040] Step S201: Construct a PHM domain model using a module definition diagram, modeling the functional elements related to the PHM system. The module diagram defines the system's black-box structure, focusing on the connections between components and external interfaces. An example is shown below. Figure 6 As shown.

[0041] Step S202: Decompose the functions and construct the functional architecture of the PHM system using a module definition diagram, as shown in the example below. Figure 7 As shown.

[0042] Step S3: Design the execution logic architecture of the PHM system. The specific steps are as follows:

[0043] Step S301: Create an activity diagram to represent task-level behaviors, describing how these activities execute PHM system functions according to a certain sequence and logical decisions, in order to visualize the PHM system operation process. Figure 8 This indicates the task-level behavior of the PHM system in performing top-level functions, with the aim of fulfilling task objectives, while describing the basic process in logical order.

[0044] Step S302: Visualize the dynamic changes of each module's state throughout the entire lifecycle of the PHM system using a state machine diagram. The state machine diagram of the diagnostic module is shown below. Figure 9 As shown.

[0045] Step S4: Based on the failure impact analysis, design the physical architecture of the PHM system and assign functional elements to the system elements of the candidate physical architecture. The specific steps are as follows:

[0046] Step S401: Apply fault tree analysis to assist in fault analysis and identify key physical components.

[0047] Step S402: Represent the physical architecture of the PHM system using a module definition diagram, such as... Figure 10 As shown.

[0048] Step S403: Convert the PHM system physical architecture constructed in step S402 into an XML file and import it into the downstream implementation tool to realize the transfer of the PHM system physical architecture to the downstream project, and further implement each functional module in the downstream project.

[0049] Step S5: Through the continuous transfer and mapping of the PHM system from requirements to functional architecture, logical architecture and then to physical architecture, the division, alignment and allocation of PHM system requirements to architectural entities and then to PHM system elements are realized.

[0050] Step S6: By analyzing the traceable links of the SysML model, analyze and examine the design elements and interfaces of the functional architecture, logical architecture, and physical architecture from functional, logical, and physical perspectives to verify the correctness of the PHM system model.

[0051] Step S7: Perform a change impact analysis based on traceability and iteratively update the PHM system. The specific steps are as follows:

[0052] Step S701: Perform a change impact analysis based on traceability to assess the stakeholders and scope of impact of the requirement change.

[0053] Step S702: If a requirement changes, identify all relevant “suspicious” elements through change impact analysis, review them, and determine whether to further modify the element.

[0054] Step S703: For all elements that need to be modified, update the corresponding requirements, functions, logic, and physical architecture models in a recursive and iterative manner.

[0055] Example:

[0056] When the system adds the requirement for bearing degradation and failure mode prediction and assessment, the fault prediction and health management process based on agile systems engineering is as follows:

[0057] First, determine that the target belongs to "predictive maintenance service", corresponding to Figure 2 The use case "Prognostic Assessment" in the text; tracing back to Figure 3 Functional Requirements; further break down the requirements, tracing back to Figure 4 Data acquisition (F1), data processing (F2), health state estimation (F4.1), and health state evaluation (F4.1) are all part of the data collection and processing framework. Detailed requirements are outlined below. Figure 4 The "F1-Data Acquisition" requirement gives rise to the "VibrationData Acquisition" requirement for acquiring bearing vibration data. The "F2-Data Processing" requirement gives rise to the "Vibration Data Transformation" requirement for converting time-domain vibration data into a spectrum. The "Health State Evaluate" requirement gives rise to the "Bearing Health State Evaluate" requirement for assessing bearing health status. Updated functional requirements include... Figure 11 As shown; each requirement is traced back to Figure 6 The PHM system domain model consists of "Data_Acquisition", "Data_Processing", and "Prognostic_Assessment"; and then traces back to... Figure 7 In the functional hierarchy architecture, update "F1-Data Acquisition", "F2-Data Processing", and "F4.1-Health State Estimation". Figure 7The functional hierarchy architecture is as follows: "F1-DataAcquisition" derives the function "Vibration Data Acquisition", "F2-Data Processing" derives the function "Vibration Data Transformation", "F4.1-Health State Estimation" derives "Health State Evaluate", and further derives the function "Bearing Health State Evaluate". The updated functional architecture diagram is as follows: Figure 12 As shown; next, tracing back to Figure 10 The physical architecture of the PHM system, and the physical architecture of the updated PHM system, such as Figure 13 As shown, the function "Vibration DataAcquisition" is derived under "F1-Data Acquisition", "Vibration DataTransformation" is derived under "F2.1-Data Manipulation Function", and "Health StateEvaluate" is derived under "F4.1-Health State Estimation", which in turn gives rise to the function "Bearing Health State Evaluate". In this embodiment, during the implementation phase, Fourier transform can be used to convert time-domain vibration data into a spectrum, thereby realizing Vibration DataTransformation; neural network models such as SOM can be used to classify and predict bearing failure modes, thereby realizing "Bearing Health State Evaluate".

[0058] It is evident that the traceability relationship between SysML models helps to quickly track the models of each node in the PHM system. When requirements change, only the model related to the changed requirements needs to be modified, while other parts remain unchanged.

Claims

1. A fault prediction and health management method based on agile systems engineering, characterized in that... The method includes the following steps: Step S1: Interact with stakeholders to determine the top-level requirements of PHM and generate PHM system design requirements; Step S2: Design and generate the functional architecture of the PHM system; Step S3: Design the execution logic architecture of the PHM system; Step S4: Based on the fault impact analysis, design the physical architecture of the PHM system and assign functional elements to the system elements of the candidate physical architecture. The specific steps are as follows: Step S401: Apply fault tree analysis to perform auxiliary fault analysis and identify key physical components; Step S402: Represent the physical architecture of the PHM system using a module definition diagram; Step S403: Convert the PHM system physical architecture constructed in step S402 into an XML file and import it into the downstream implementation tool to realize the transfer of the PHM system physical architecture to the downstream project, and further implement each functional module in the downstream project. Step S5: Through the continuous transfer and mapping of the PHM system from requirements to functional architecture, logical architecture and then to physical architecture, the division, alignment and allocation from system requirements to architectural entities and then to system elements are realized; Step S6: By analyzing the traceable links of the SysML model, analyze and examine the design elements and interfaces of the functional architecture, logical architecture, and physical architecture from functional, logical, and physical perspectives to verify the correctness of the model; Step S7: Perform a change impact analysis based on traceability and iteratively update the PHM system.

2. The fault prediction and health management method based on agile systems engineering according to claim 1, characterized in that... The specific steps of step S1 are as follows: Step S101: Use use case diagrams to describe the goals of the PHM system and the external systems involved in achieving those goals; Step S102: Construct a PHM stakeholder requirements map to support the task objectives; Step S103: Based on the requirements diagram constructed in step S102, further decompose the functional requirements of the PHM system; Step S104: Establish the traceability relationship between the functional requirements of the PHM system.

3. The fault prediction and health management method based on agile systems engineering according to claim 1, characterized in that... The specific steps of step S2 are as follows: Step S201: Construct a PHM domain model using a module definition diagram, and model the functional elements related to the PHM system; Step S202: Decompose the functions and construct the functional architecture of the PHM system using a module definition diagram.

4. The fault prediction and health management method based on agile systems engineering according to claim 1, characterized in that... The specific steps of step S3 are as follows: Step S301: Create an activity diagram to represent task-level behaviors, describing how these activities execute PHM system functions according to a certain sequence and logical decisions, in order to visualize the PHM system operation process; Step S302: Visualize the dynamic changes of the state of each module throughout the entire lifecycle of the PHM system based on the state machine diagram.

5. The fault prediction and health management method based on agile systems engineering according to claim 1, characterized in that... The specific steps of step S7 are as follows: Step S701: Perform a change impact analysis based on traceability to assess the stakeholders and scope of impact of the requirement change; Step S702: If a requirement changes, identify all relevant "suspect" elements through change impact analysis, review them, and determine whether to further modify the element. Step S703: For all elements that need to be modified, update the corresponding requirements, functions, logic, and physical architecture models in a recursive and iterative manner.

Images