Search for Author, Title, Keyword
A study of the Inverse Gaussian Process with hazard rate functions-based drifts applied to degradation modelling
 
More details
Hide details
 
Online publication date: 2022-07-04
 
 
Publication date: 2022-07-04
 
 
Eksploatacja i Niezawodność – Maintenance and Reliability 2022;24(3):590-602
 
HIGHLIGHTS
  • Flexible hazard rate functions are considered as drifts in the inverse Gaussian process.
  • Degradation trajectories are characterized by the hazard rate functions-based drifts.
  • Random effects are considered to individually characterize the degradation trajectories.
  • Illustrative case studies are analysed to demonstrate the capability of proposed models.
KEYWORDS
ABSTRACT
The stochastic modelling of degradation processes requires different characteristics to be considered, such that it is possible to capture all the possible information about a phenomenon under study. An important characteristic is what is known as the drift in some stochastic processes; specifically, the drift allows to obtain information about the growth degradation rate of the characteristic of interest. In some phenomenon’s the growth rate cannot be considered as a constant parameter, which means that the rate may vary from trajectory to trajectory. Given this, it is important to study alternative strategies that allow to model this variation in the drift. In this paper, several hazard rate functions are integrated in the inverse Gaussian process to describe its drift in the aims of individually characterize degradation trajectories. The proposed modelling scheme is illustrated in two case studies, from which the best fitting model is selected via information criteria, a discussion of the flexibility of the proposed models is provided according to the obtained results.
 
CITATIONS (1):
1.
The Wiener Process with a Random Non-Monotone Hazard Rate-Based Drift
Luis Alberto Rodríguez-Picón, Luis Carlos Méndez-González, Luis Asunción Pérez-Domínguez, Héctor Eduardo Tovanche-Picón
Mathematics
 
eISSN:2956-3860
ISSN:1507-2711
Journals System - logo
Scroll to top