Application of Gaussian Process modeling to analysis of functional unreliability

Gaussian Process Models (GPMs) have been used widely in many ways [1]. The present application uses a GPM for emulation of a system simulation code. Such an emulator can be applied in several distinct ways, discussed below. Most applications illustrated in this paper have precedents in the literature; the present paper is an application of GPM technology to analysis of the functional unreliability of a passive containment cooling system, which was previously analyzed [2] using an artificial neural network (ANN), and later [3, 4] by a method called "Alternating Conditional Expectations" (ACE). The present exercise enables a comparison of both the processes and the results. In this paper, (1) the original quantification of functional unreliability using ANN [2], and the later work using ACE [3], is reprised using GPM; (2) additional information provided by the GPM about uncertainty in the limit surface, generally unavailable in other representations, is discussed briefly; (3) a simple forensic exercise is performed, analogous to the inverse problem of code calibration, but with an accident management spin: given an observation about containment pressure, what can we say about the system variables?