CN-121979492-A - Scene-driven model-based system architecture weighing method and device

CN121979492ACN 121979492 ACN121979492 ACN 121979492ACN-121979492-A

Abstract

The invention relates to the field of system architecture design, analysis and trade-off, in particular to a scene-driven model-based system architecture trade-off method and device. The method comprises the steps of 1, constructing a scene evaluation table, 2, building a system efficiency tree, 3, carrying out system architecture modeling based on a model, and 4, carrying out multi-quality characteristic analysis of the system architecture based on the model, and identifying sensitive points and risk points. And 5, modifying system architecture elements based on the sensitive points, generating a new architecture, and repeating the steps 3-4, wherein the step 5 is alternative system architecture quantitative evaluation and visual analysis. Based on user requirements and use case scenes, the method considers the interaction influence between the field characteristics of the system and the multi-quality characteristics of the system, combines quantitative evaluation and visual analysis, improves the confidence level of the trade-off of alternative architecture of the system, reduces the influence of the traditional engineering experience, non-technical factors and the like on the trade-off of the architecture, and reduces the subjectivity of an evaluation result.

Inventors

LI ZELIN
YUE YAZHOU
LI WEITONG
WANG SHUYANG
CHANG HUA
WANG WENHAO
WANG YONG
WU FANGFANG

Assignees

中国航空工业集团公司西安飞行自动控制研究所

Dates

Publication Date: 20260505
Application Date: 20251224

Claims (10)

1. A scene-driven model-based system architecture balancing method is characterized in that the method is used for selecting one from a plurality of alternative architectures as an optimal system architecture, wherein the alternative system architecture is a functional architecture, a logic architecture or a physical architecture, and the method comprises the following steps: Step 1, collecting requirements of stakeholders and use scenes of a system, expanding the use scenes, and forming a scene evaluation table based on MOE/KPP of a use scene capturing system; Step 2, analyzing system requirements, constraints and system environments based on the scene evaluation table generated in the step 1, decomposing each MOE/KPP into MOP and TPM, and constructing a system efficiency tree T= { { MOEs/KPPs }, { MOPs }, { TPMs }, wherein MOEs/KPPs are MOE/KPP sets, the number is i, MOPs are MOP sets, the number is j, TPMs are TPM sets, and the number is k; Modeling the alternative architecture of the system, including the deployment of software resources, hardware resources and software in hardware of the system, to form an alternative system architecture model Archi = { a 1 ,a 2 ,…,a n , where n is the number of alternative architectures }; Step 4, modeling the system related attribute of the alternative architecture model based on the system efficiency tree established in the step 2, and carrying out analysis based on the multi-quality characteristics of the alternative architecture to generate an analysis result x= { x 1 ,x 2 ,…,x k of the multi-quality characteristics of the system, wherein x k is an analysis result corresponding to a system TPM k ; And 5, carrying out quantitative evaluation on the alternative system architecture according to the analysis result of the alternative architecture in the step 4, and completing optimization of the system architecture.
2. The method according to claim 1, wherein for the host system, step 1 specifically comprises: step 11, collecting the demands from stakeholders and the use scenes of the system, and expanding the initial scenes; Step 12, capturing key indexes of task demands of a system in the whole life cycle in a standing manner from a user angle based on scenes, namely MOEs, wherein the MOEs are criteria for successful system operation, and the number of the MOEs can be changed according to different scales and complexity of the system, but is more suitable in 10-50; for the subsystem, step 1 specifically includes: And identifying key performance parameters, namely KPPs, according to a demand set distributed to the subsystem by the system, and recording the key performance parameters in a scene evaluation table.
3. The method of claim 2, wherein step 12 further comprises, for the host system: And (5) prioritizing the identified MOEs according to the high, medium and low importance degrees, and recording the priority in a scene evaluation table.
4. The method of claim 2, wherein in step1, the MOEs of the host system and KPPs of the recognition subsystem are captured, specifically, the MOEs and KPPs are recognized in a model-based manner by establishing system use cases and use case analysis by using SysML language.
5. The method of claim 1, wherein step 2 comprises: step 21, MOPs identification, namely decomposing MOEs in the scene evaluation table generated in the step 1 based on the functional architecture of the system to form MOPs; step 22, TPM identification, namely decomposing the MOPs in the step 21 based on the physical architecture of the system to form TPMs; Step 23, based on MOEs/KPPs, MOPs, TPMs, using MOEs/KPPs as a father node, MOPs as child nodes, TPMs as leaf nodes, and forming a system efficiency tree according to user-defined grouping categories.
6. The method of claim 1, wherein step 3 comprises: and establishing an alternative system architecture model, namely describing software resources, hardware resources and deployment of software in hardware of the system through a system modeling language to form the system architecture model.
7. The method of claim 1, wherein step 4 comprises: step 41, expanding an alternative system architecture model based on the constructed system efficiency tree, and adding the attribute related to multi-quality characteristic analysis; And 42, carrying out quality characteristic analysis based on the expanded alternative system architecture model, wherein the analysis items cover all branches in the system efficiency tree, completing the branches which cannot be covered by the model analysis in a manual mode, and combining the analysis results based on the model analysis and the manual mode to form a complete analysis result x= { x 1 ,x 2 ,…,x k }.
8. The method of claim 1, wherein step 42 further comprises: and (3) identifying sensitive points and risk points of the alternative system architecture in the analysis process, returning to the step (3) for modifying architecture elements to generate a new alternative system architecture based on the sensitive points, repeating the step (4), and recording the identified risk points in a risk point list.
9. The method of claim 1, wherein step 5 comprises: Step 51, performing weight distribution on TPMs in the efficacy tree, wherein w= { W 1 ,w 2 ,…,w k }, sum { W i }, wherein i=1 to k } = 1; Step 52, defining a normalization function for qualitative, quantitative, dimensional and dimensionless results, and normalizing the analysis results of each TPMs to obtain a normalized result u i : u i =f(x i ), wherein 0< = u i < = 1; Step 53, the quantitative evaluation of the alternative system architecture is that according to the weight distributed by the efficacy tree and the normalized result u i , the quantitative evaluation result M of the architecture is calculated: step 54, alternative system architecture comparison analysis, determining a preferred architecture Archi Excellent (excellent) : archi Excellent (excellent) =max{M i , where i=1, 2, n }.
10. A scenario driven model-based system architecture balancing apparatus for selecting one of a plurality of alternative architectures as an optimal system architecture, the alternative system architecture being a functional architecture, a logical architecture, or a physical architecture, the apparatus comprising: the scene evaluation table construction module is used for collecting the requirements of stakeholders and the use scenes of the system, expanding the use scenes and forming a scene evaluation table based on MOE/KPP of the use scene capturing system; the system efficiency tree construction module is used for analyzing system requirements, constraints and system environments based on a scene evaluation table, decomposing each MOE/KPP into MOP and TPM, and constructing a system efficiency tree T= { { MOEs/KPPs }, { MOPs }, { TPMs }, wherein MOEs/KPPs are MOE/KPP sets, the quantity is i, MOPs are MOP sets, the quantity is j, TPMs are TPM sets, and the quantity is k according to user-defined grouping categories; The system architecture model construction module is used for modeling an architecture alternative of the system, and comprises software resources, hardware resources and deployment of software in hardware of the system to form an architecture model Archi = { a 1 ,a 2 ,…,a n of the system, wherein n is the number of the architecture alternatives; The system multi-quality characteristic analysis module is used for modeling the system related attribute of the alternative architecture model based on the system efficiency tree, carrying out analysis of multi-quality characteristics based on the alternative architecture, and generating an analysis result x= { x 1 ,x 2 ,…,x k of the system multi-quality characteristics, wherein x k is an analysis result corresponding to a system TPM k ; And the quantitative evaluation module is used for carrying out quantitative evaluation on the alternative system architecture according to the analysis result of the alternative architecture to finish the optimization of the system architecture.

Description

Scene-driven model-based system architecture weighing method and device Technical Field The invention relates to the field of system architecture design, analysis and trade-off, in particular to a scene-driven model-based system architecture trade-off method and device. Background With the rapid development of information technology, the software scale and the complexity of the system are continuously improved, and the requirements on the quality, the cost and the progress of the system are increasingly strict, especially in the field of large-scale complex systems. In the complex system development process, the quality characteristics of the system are mainly determined by the architecture design, and the rationality of the architecture design directly influences the function realization, performance index and long-term evolution capability of the system. In the forward research and development flow of the system, the diversity of the system design can be effectively increased through abstractions of different levels of the system at different stages. This diversity exploration helps to find the best solution to meet stakeholder needs and scenarios. Specifically, in the system demand definition and analysis stage, multiple alternative functional architectures are usually generated according to the demands and application scenarios of interested parties, while in the architecture definition and analysis stage, software resources, hardware resources and deployment modes of the software and hardware resources need to be comprehensively considered, so that multiple alternative physical architectures are derived. However, as architecture diversity increases, how to make scientific and reasonable trade-offs between these alternative functional and physical architectures becomes an important challenge. In the actual development process, two key questions are usually required to be answered, namely, which system architecture meets the system requirements, and which system architecture is more suitable for specific development environments and application scenes. The resolution of these problems directly affects the quality of the final system and the efficiency of subsequent development. Currently, in the framework weighing process of a complex system, engineering experience and domain knowledge are generally relied on, and quantitative analysis and explicit description methods are lacked. The traditional trade-off mode has the following problems that the trade-off process is high in subjectivity and difficult to form consistency and traceability, the selection of sub-optimal or even wrong architecture design schemes may be caused due to the lack of comprehensive evaluation of different architecture schemes, and the conflict and coordination between multi-dimensional requirements (such as performance, reliability and safety) in a large-scale complex system cannot be effectively met. Therefore, how to build a scientific, quantitative and explicit architecture trade-off method becomes a key problem to be solved in the current complex system research and development field. Disclosure of Invention The invention aims to: The scene-driven model-based system architecture balancing method and device are provided, based on user requirements and use case scenes, the interaction influence between the field characteristics of the system and the multi-quality characteristics of the system is considered, the quantitative evaluation analysis is combined, the confidence level of the system alternative architecture balancing is improved, the influence of traditional engineering experience based, non-technical factors and the like on the architecture balancing is reduced, and the subjectivity of an evaluation result is reduced. The technical scheme is as follows: A scene-driven model-based system architecture balancing method is used for selecting one from a plurality of alternative architectures as an optimal system architecture, wherein the alternative system architecture is a functional architecture, a logic architecture or a physical architecture, and the method comprises the following steps: Step 1, collecting requirements of stakeholders and use scenes of a system, expanding the use scenes, and forming a scene evaluation table based on MOE/KPP of a use scene capturing system; Step 2, analyzing system requirements, constraints and system environments based on the scene evaluation table generated in the step 1, decomposing each MOE/KPP into MOP and TPM, and constructing a system efficiency tree T= { { MOEs/KPPs }, { MOPs }, { TPMs }, wherein MOEs/KPPs are MOE/KPP sets, the number is i, MOPs are MOP sets, the number is j, TPMs are TPM sets, and the number is k; Modeling the alternative architecture of the system, including the deployment of software resources, hardware resources and software in hardware of the system, to form an alternative system architecture model Archi = { a 1,a2,…,an, where n is the number of alternative arc