Exponential-Based Testability Index Allocation and Feasibility Analysis Method

Abstract

Existing mainstream testability index allocation (TIA) methods have the problem that the allocation value cannot reflect the index changes with the failure rate. One of the main reasons is that most of the existing methods use linear allocation functions, which cannot reflect the promotion law of testability index. To address the issue, a nonlinear TIA method based on exponential function is proposed in this paper. Based on the basic mathematical model of the TIA problem, a testability allocation algorithm based on exponential function is designed, which can obtain more accurate fault detection rate (FDR) and fault isolation rate (FIR) values that meet the constraints. In addition, a feasibility analysis method based on an improved cost function model is proposed to achieve the purpose of comparing the proposed TIA method with the mainstream ones in terms of rationality. A control system is used to validate the proposed method in effectiveness and feasibility. Results show that the proposed TIA method can better obtain the allocated value of each index well, and can better reflect the law of the testability index changing with the failure rate.