skip to main content

Title: Prediction of Transition-State Energies of Hydrodeoxygenation Reactions on Transition-Metal Surfaces Based on Machine Learning
Authors:
; ; ; ; ; ; ; ; ;
Award ID(s):
1632824
Publication Date:
NSF-PAR ID:
10182074
Journal Name:
The Journal of Physical Chemistry C
Volume:
123
Issue:
49
Page Range or eLocation-ID:
29804 to 29810
ISSN:
1932-7447
Sponsoring Org:
National Science Foundation
More Like this
  1. When a transition fault test set leaves undetected transition faults because of logic redundancies, test constraints, or the existence of hard-to-detect faults, it leaves transition fault sites uncovered. For the case where multicycle tests are used, this paper explores the possibility of covering the sites of undetected transition faults by using tests for what are referred to as optimistic unspecified transition faults. For this discussion, a standard transition fault is associated with an extra delay of a single clock cycle. An unspecified transition fault captures in a single fault the behaviors of transition faults of different durations. Because faults with different durations may be detectable or undetectable independently by a multicycle test, an unspecified transition fault may be detected even if the standard transition fault at the same site is undetectable. This effect is enhanced with optimistic unspecified transition faults. The paper describes an iterative test compaction procedure for multicycle tests that supplements the set of standard transition faults with optimistic unspecified transition faults to cover the sites of undetected standard transition faults.