Novel testability modelling and diagnosis method considering the supporting relation between faults and tests

Abstract

Establishing an effective testability model is key to testability design and fault diagnosis. Under the general assumptions that test circuits are completely reliable and all test signals can be accurately transmitted to display devices, traditional testability modelling methods are effective. However, these assumptions cannot hold for many practical applications, thereby eventually leading to incorrect fault diagnosis results. To overcome these problems, a novel extended testability modelling method that considers the supporting relation between faults and tests (SRFT) is presented. The blocking and enabling relations between faults and tests are analysed from the perspective of test information transmission and test failure. Accordingly, the element design, mathematical expression, and model type design were elaborated in detail. A matrix fusion algorithm and a diagnosis knowledge construction algorithm are proposed herein to express the supporting relation in a diagnostic matrix, wherein the third state of the test is introduced to describe the test failure. According to the supporting relation information in the diagnostic matrix, three new testability indices are proposed to evaluate the system testing capability. Our case study results clearly demonstrate the superiority of the proposed method, which can accurately describe the supporting relation and help obtain accurate diagnostic results.