Methods of Handling the Tolerance and Test-Point Selection Problem for Analog-Circuit Fault Diagnosis
Chenglin Yang,Shulin Tian,Bing Long,Fang Chen
DOI: https://doi.org/10.1109/tim.2010.2050356
IF: 5.6
2011-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:Soft-fault diagnosis and tolerance are two challenging problems in analog-circuit fault diagnosis. Although many analog faults can be diagnosed theoretically, they cannot be accurately diagnosed due to the influence of component tolerance. This paper proposes approaches to the tolerance-handling method in soft-fault diagnosis. First, the slope fault model and its theoretical proof are presented. In linear analog circuits, the voltage equation between two nodes is linear and can be expressed by a point-slope-form equation in which the point is determined by the nominal voltage values on the two selected nodes, and the slope, which is invariant, can be used as the fault model, which is applicable to both hard (open or short) and soft (parametric) faults. Then, the parameter tolerance is taken into consideration. The point-slope-form equation is governed by a point and a slope value. The tolerance influences the fault diagnosis by shifting both the point and the nominal slope value. These influences can be mitigated in two ways: 1) The point is measured when an actual circuit is under steady state and free of fault, and 2) the maximum slope deviation is readily obtained by using optimization theory if the equation governing the voltage between two nodes is available; otherwise, a threshold proportion coefficient is used to calculate the maximum deviation. Based on these methods, a novel test-point selection approach for the fault dictionary technique is proposed. The effectiveness of the proposed approaches is verified by both simulated and experimental results.