CN-122019137-A - Fault injection parameter processing method and device based on test flow context

CN122019137ACN 122019137 ACN122019137 ACN 122019137ACN-122019137-A

Abstract

The application provides a fault injection parameter processing method and device based on test flow context, wherein the method obtains flow context information containing multi-step logic relationship by analyzing a test flow described by Automatic Test Markup Language (ATML); and performing collaborative analysis and dynamic decision-making according to the flow context information and the real-time resource state of the current test system, intelligently generating an adaptive fault injection parameter, and finally executing fault injection according to the generated parameter. The method overcomes the defects of fixed fault parameters and lack of flow adaptability in the comprehensive test in the prior art, realizes the transition from static parameter table lookup to dynamic context sensing decision, and remarkably improves the automation level of aviation test and test, test precision and adaptability to complex test scenes.

Inventors

REN CHAOXU
LU LINHAI
WEN QIANG
HAN HUIJIE
LI LIJIA
Lv zhongyu
ZHANG JUNXIA
PAN GUOQING

Assignees

北京航天测控技术有限公司

Dates

Publication Date: 20260512
Application Date: 20251229

Claims (10)

1. A fault injection parameter processing method based on a test flow context, the method comprising: Receiving test description of automatic test markup language ATML, obtaining signal semantic attribute of the test description by using the test description, and extracting flow context information of a test flow; Receiving a fault injection instruction, and determining fault injection parameters by using a constraint satisfaction problem CSP model according to the fault injection instruction and the flow context information; Calling a multi-objective optimizer to determine an optimal scheduling mode based on the real-time resource state and the flow context information, and distributing hardware resources for the fault injector; converting the fault injection parameters into equipment instructions executable by the hardware resources, and executing the equipment instructions to complete testing; And when the resource conflicts, generating feedback constraint, and re-executing the fault injection parameters, wherein the resource conflicts comprise that the change of the resource state affects the historical planning task or a newly inserted high-priority task exists.
2. The method of claim 1, wherein prior to said receiving a test description of the automatic test markup language, ATML, the method further comprises: And constructing a test signal body library, wherein the test signal body library comprises class hierarchy relations of aviation test signals and corresponding subclasses.
3. The method of claim 1, wherein receiving the test description of the automatic test markup language ATML, obtaining the signal semantic attribute of the test description using the test description, and extracting flow context information of a test flow, comprises: receiving an input Automatic Test Markup Language (ATML) file, and extracting test description in the ATML file, wherein the test description is used for describing a test signal; Calculating and rule reasoning based on semantic similarity are carried out on the test description and the test signal ontology library, and nonstandard signal expressions in the test description are mapped to signal classes corresponding to the test description to obtain a matching result; Creating a corresponding signal class instance according to the matching result; And filling all the attributes of the examples by using the ATML file and the test signal ontology library to obtain the signal semantic attributes of the test description, wherein standardized conversion is performed on non-standard signals.
4. The method of claim 1, wherein the receiving the fault injection instruction and determining the fault injection parameters using a constraint satisfaction problem CSP model based on the fault injection instruction and the flow context information comprises: Constructing a constraint satisfaction problem CSP model, wherein constraint conditions of the CSP model comprise intrinsic constraints of a fault mode, the signal semantic attribute, signal and time sequence constraints of the flow context information and feedback constraints; Receiving a fault injection instruction, and matching a fault mode corresponding to the fault injection instruction in a preset fault model library; Injecting current context information into the CSP model, and instantiating the fault mode to obtain a specific problem to be solved, wherein the current context information comprises the flow context information and the signal semantic attribute; and calling a CSP solver to solve the specific problem to obtain fault injection parameters.
5. The method of claim 1, wherein invoking the multi-objective optimization scheduler to determine an optimal scheduling manner based on the real-time resource status and the flow context information and allocating hardware resources for the failed injector comprises: constructing a multi-objective optimization scheduler for the resource allocation data; Calling the multi-objective optimization scheduler to process according to the fault injection parameters, the real-time resource state and the flow context information, so as to obtain an optimal scheduling mode; And according to the optimal scheduling mode, distributing hardware resources for the fault injector.
6. The method of claim 5, wherein after allocating hardware resources for the failed injector in the optimized scheduling manner, the method further comprises: During test execution, changes in resource status are continuously monitored.
7. The method of claim 6, wherein the step of generating a feedback constraint when resources collide, re-executing the fault injection parameters comprises: Generating new feedback constraint by adopting the multi-objective optimization scheduler, wherein the feedback constraint is the resource conflict or flow deviation information; Updating time constraint conditions in the CSP model according to the new feedback constraint; and re-deducing the fault injection parameters according to the new feedback constraint and the updated time constraint condition, and re-planning the task.
8. A fault injection parameter processing device based on a test flow context, the device comprising: The acquisition module is used for receiving the test description of the automatic test markup language ATML, obtaining the signal semantic attribute of the test description by using the test description, and extracting the flow context information of the test flow; the parameter determining module is used for receiving a fault injection instruction and determining fault injection parameters by utilizing a constraint satisfaction problem CSP model according to the fault injection instruction and the flow context information; The scheduling mode determining module is used for calling the multi-objective optimizer to determine an optimal scheduling mode based on the real-time resource state and the flow context information and distributing hardware resources for the fault injector; The conversion module is used for converting the fault injection parameters into equipment instructions executable by the hardware resources and executing the equipment instructions to complete testing; And the feedback module is used for generating feedback constraint when the resource conflicts, and re-executing the step of injecting the parameters to the fault, wherein the resource conflicts comprise that the change of the resource state affects the historical planning task or a newly inserted high-priority task exists.
9. The apparatus of claim 8, wherein the feedback module is configured to continuously monitor changes in the status of the resource during execution of the test.
10. The device according to claim 8, wherein the feedback module is specifically configured to generate a new feedback constraint using the multi-objective optimization scheduler, the feedback constraint being the resource conflict or flow deviation information, update a time constraint condition in the CSP model according to the new feedback constraint, and re-derive the fault injection parameter and re-plan a task according to the new feedback constraint and the updated time constraint condition.

Description

Fault injection parameter processing method and device based on test flow context Technical Field The application relates to the technical field of avionic test, in particular to a fault injection parameter processing method and device based on test flow context. Background Reliability testing of avionics, particularly Line Replaceable Units (LRUs), is critical to ensuring flight safety. The fault injection test is used as a core verification means to simulate a real fault and check the fault tolerance and recovery capability of the equipment. With the increase of test complexity, the adoption of standards such as Automatic Test Markup Language (ATML) to describe the test flow has become the basis for realizing test standardization and automation. Currently, the mainstream technical paradigm is typically "static parameter table query+fixed rule mapping" to automate the execution of ATML descriptions to fault injection. The specific implementation process comprises the steps of carrying out keyword matching or simple semantic similarity calculation depending on a predefined signal type library, and calling static and preset parameters from a database by adopting a fixed corresponding relation of a fault mode-parameter template, wherein the generation logic of the parameters is independent of a specific test flow. And the resource scheduling stage in the test process is limited to the allocation of hardware channels and simple time sequence ordering, and the decision is separated from the logic intention and semantic constraint of the test flow. Therefore, the technical scheme mainly has the problems of disjointing a static parameter system and a dynamic flow context, lacking self-adaptive capacity, incapability of obtaining feedback in real time according to flow execution, poor flow cooperativity, poor cooperativity between two links of parameter generation and resource scheduling, limited test depth and efficiency, and low test automation efficiency caused by conserved and fixed test parameters. Disclosure of Invention The application provides a fault parameter processing method and device, which are used for solving the comprehensive technical problems of low test automation degree, poor parameter adaptability and insufficient robustness in a complex multi-step test scene caused by the fact that the generation of fault injection parameters and the global context (comprising step logic, time sequence constraint and signal dependence relation) of a test flow are disjointed and cannot be dynamically cooperated with the real-time system resource state in the prior art. In a first aspect, the present application provides a fault injection parameter processing method based on a test flow context, including: Receiving test description of automatic test markup language ATML, obtaining signal semantic attribute of the test description by using the test description, and extracting flow context information of a test flow; Receiving a fault injection instruction, and determining fault injection parameters by using a constraint satisfaction problem CSP model according to the fault injection instruction and the flow context information; Calling a multi-objective optimizer to determine an optimal scheduling mode based on the real-time resource state and the flow context information, and distributing hardware resources for the fault injector; converting the fault injection parameters into equipment instructions executable by the hardware resources, and executing the equipment instructions to complete testing; And when the resource conflicts, generating feedback constraint, and re-executing the fault injection parameters, wherein the resource conflicts comprise that the change of the resource state affects the historical planning task or a newly inserted high-priority task exists. In one possible embodiment, before the receiving the test description of the automatic test markup language ATML, the method further comprises: And constructing a test signal body library, wherein the test signal body library comprises class hierarchy relations of aviation test signals and corresponding subclasses. In one possible implementation manner, the receiving the test description of the automatic test markup language ATML, obtaining the signal semantic attribute of the test description by using the test description, and extracting the flow context information of the test flow includes: receiving an input Automatic Test Markup Language (ATML) file, and extracting test description in the ATML file, wherein the test description is used for describing a test signal; Calculating and rule reasoning based on semantic similarity are carried out on the test description and the test signal ontology library, and nonstandard signal expressions in the test description are mapped to signal classes corresponding to the test description to obtain a matching result; Creating a corresponding signal class instance according to the matching result; And fill