Most of the models proposed in the literature try to model failure rate functions and failure intensities by means of a single model with a reduced number of parameters. If this approach can be effective for controlled situations or simple models it may happen that new models should be introduced in order to describe more accurately the behavior of the cases where non-monotonic failure rate or intensity functions can be experienced. Examples of non-monotonic failure intensity functions are observed in the early life of industrial plants. Reasons for a non-monotonic failure intensity can be various and different in nature, for instance the learning curve of peculiar technical aspects, up to possible managerial decisions on the organization of resources during the commissioning and early production phases. A statistical model of a non-monotonic failure intensity for a complex repairable system can be of some guidance to take informed decisions on the time and resources dedicated to the commissioning and operational phases. Under these conditions new models should be introduced in order to describe more accurately the behavior of the plant where non-monotonic failure rate or intensity functions can be experienced. In this paper the authors compare two Non Homogenous Poisson Process (NHPP) based on piecewise functions with a newly proposed one relying on a modified Weibull probability density function which can be analytically solved.
|Titolo:||Large plants failures modeling under variable commissioning scheduling|
|Citazione:||Addabbo, T., Fort, A., Marino, R., Mugnaini, M., Vignoli, V., Michelassi, C., et al. (2017). Large plants failures modeling under variable commissioning scheduling. In 2017 IEEE International Symposium on Systems Engineering, ISSE 2017 - Proceedings (pp.332-336). New York : Institute of Electrical and Electronics Engineers Inc..|
|Appare nelle tipologie:||4.1 Contributo in Atti di convegno|