Incipient Fault Detection and Diagnosis in Turbine Engines Using Hidden Markov Models

Abstract

Incipient fault detection and diagnosis in turbine engines is key to effective maintenance and improved availability of systems dependent on these engines. In this paper, we present a novel method for incipient fault detection and diagnosis using Hidden Markov Models (HMMs). In particular, we focus on engine faults that are manifest in transient operating conditions such as engine startup and acceleration. HMMs are stochastic signal models that are effective in modeling transient signals. They are developed with engine data collected under nominal operating conditions. Engine data representing different fault conditions are used to develop the fault HMMs; a separate model is developed for each of the faults. Once the nominal and fault HMMs are developed, new engine data collected from the engine are evaluated against the HMMs and a determination is made whether a fault is indicated. Here, we demonstrate our HMM-based fault detection and diagnosis approach on engine speed profiles taken from a real engine. Further, the effectiveness of the HMM-based approach is compared with a neural-network-based approach and a method based on using principal component analysis in conjunction with a neural network approach.

Ascertainment-adjusted parameter estimation approach to improve robustness against misspecification of health monitoring methods

“…Standard DHMMs have been used in the past in many applications [31,57,42,8,3,26] and in particular for noisy speech and character recognition [7,50,17,12]. In the context of PHM, DHMM has also been widely used, for instance in [25] for predictive modeling dedicated to intelligent maintenance in complex semiconductor manufacturing processes, in [29] for incipient fault detection and diagnosis in turbine engines, in [2] for failure isolation for cognitive robots and in [16] for anomaly detection in electronic systems.…”

Lean burn combustion monitoring strategy based on data modelling

“…Kalman filtering, polynomial curve fitting, and multivariate statistical analysis algorithms have been considered for detecting and isolating component fault [6]- [8]. Machine learning techniques such as Logistic Regression (LR), neural network (NN), Support Vector Machine (SVM), Hidden Markov Model (HMM), as well as combinations of these are more often adopted to not only discriminate fault from normal, but also diagnose possible defects [5], [9]- [12]. Despite widespread use in academic publications, application issues concerning how to select appropriate data set and model, as well as the impact of data and model selection on performance are not systematically addressed in the literature.…”

Symbolic Dynamic Analysis of Transient Time Series for Fault Detection in Gas Turbine Engines

“…However, model-based diagnostics may suffer from inaccuracy and loss of robustness due to low reliability of the transient models. This problem is partially circumvented through usage of data-driven diagnostics; for example, Menon et al [7] used hidden Markov models (HMMs) for transient analysis of gas turbine engines.…”