Fault mode optimization method based on fault propagation probability model

Abstract

The invention discloses a fault mode optimization method based on a fault propagation probability model. Firstly, the fault sample size of each module of the system is allocated, and then the fault propagation probability model of the module is established for each module to obtain the fault propagation among each fault mode. probability, and then according to the fault propagation probability and the out-degree of the component node where the fault mode is located, the fault diffusion intensity among the fault modes is calculated, and according to the fault diffusion intensity, each fault mode is searched for the path with the maximum fault diffusion probability when the original fault occurs, and the The number of subsequent failure modes included in the path is used as the number of fault diffusion, and the failure modes with the number of fault diffusion greater than the predetermined threshold are screened out, and the priority is selected; finally, the selected failure mode set of each module is collected to obtain the selected failure mode set of the system. The invention optimizes the fault mode based on the number of fault diffusion, performs key sampling on the propagation fault, and improves the reliability of the testability verification scheme.

Description

technical field [0001] The invention belongs to the technical field of testability verification, and more specifically relates to a fault mode optimization method based on a fault propagation probability model. Background technique [0002] Testability verification is based on the results of fault detection, isolation, and indication. If only the faults that occur naturally in the product are detected, it will not meet the sample size requirements specified in the design. Therefore, it is necessary to artificially introduce faults, that is, fault simulation. Before conducting test validation, the overall workflow of the validation needs to be determined to determine the test sample size, indicate fault injection, and data recording and evaluation. [0003] To carry out testing verification, the sample size of the test must first be selected, and the selection of the sample size is random selection. If n samples are selected for the test, the test will fail F times. A posit...

AI Technical Summary

Problems solved by technology

Random sampling often results in propagating failure modes not being sampled

Propagating fault means that the failure rate of a failure mode is low, but once it occurs, it will spread to other components, resulting in a large scale of impact

If the testable design for propagating faults is not complete, once it occurs without proper detection and isolation, the risk to the user will be the same as that of common faults. r sum on par

Images