Reliability simulation of a multi-state Wind Turbine Generator using SHyFTOO

Abstract

Accurate reliability analysis modelling requires appropriate stochastic formalisms, which can capture the relevant operation and degradation characteristics of the system under analysis. Physical, working and environmental conditions can be very relevant, but their inclusions in a stochastic model is challenging. Among reliability analysis methodologies, a recent formalism named Hybrid Dynamic Fault Tree, which emerges from dynamic reliability theory, may be a suitable candidate to accomplish in this task. In this paper, a wind turbine generator case study has been chosen to demonstrate the potential capabilities of this modeling approach and motivate the use of this formalism to other practitioners. The main novelty of the proposed model is the integration of the wind speed, an independent exogenous physical variable, as a trigger to modify the parameters of the probability density of failures and the aging of components. The Hybrid Dynamic Fault Tree of the proposed case study has been coded using the SHyFTOO, an easy-to-use library developed under the MATLAB® framework. Achieved results show that the Hybrid Dynamic Fault Tree is a valid formalism that should be used to improve the modelling of a multi-state system when working and operative conditions cannot be disregarded.