Multistream sensor fusion-based prognostics model for systems with single failure modes

Abstract

Advances in sensor technology have facilitated the capability of monitoring the degradation of complex engineering systems through the analysis of multistream degradation signals. However, the varying levels of correlation with physical degradation process for different sensors, high-dimensionality of the degradation signals and cross-correlation among different signal streams pose significant challenges in monitoring and prognostics of such systems. To address the foregoing challenges, we develop a three-step multi-sensor prognostic methodology that utilizes multistream signals to predict residual useful lifetimes of partially degraded systems. We first identify the informative sensors via the penalized (log)location-scale regression. Then, we fuse the degradation signals of the informative sensors using multivariate functional principal component analysis, which is capable of modeling the cross-correlation of signal streams. Finally, the third step focuses on utilizing the fused signal features for prognostics via adaptive penalized (log)-location-scale regression. We validate our multi-sensor prognostic methodology using simulation study as well as a case study of aircraft turbofan engines available from NASA repository.

Degradation Modeling and Remaining Useful Life Prediction of Aircraft Engines Using Ensemble Learning

“…This result is partially consistent with that of Ref. [26] where three measurement variables (T50, htB, and W31) were selected according to the lognormal distributions of the measurement data. In addition, 11 variables were selected in Ref.…”

Data‐driven dimension reduction in functional principal component analysis identifying the change‐point in functional data

“…Over the last few decades Functional Data Analysis (FDA) has become a topic of great interest in statistics, with applications to, for instance global warming and climate change [9], financial market data analysis [14], medicine [15], or engineering systems safety monitoring [7]. In many applications it is important to analyze a sequence of functional data that collected over time.…”

Aircraft Engine Prognostics Based on Informative Sensor Selection and Adaptive Degradation Modeling with Functional Principal Component Analysis

“…The mean function µ(t) describes the common degradating trend of all the engines from an engine population, and the first few FPCs reflect the main varying modes of the population degradation process. As in [27,34], the measurement errors e ij are practically assumed to be i.i.d with the normal distribution N(0, σ 2 ) and are also independent of FPC scores in this study. Also, note that no pre-specified parametric form is Sensors 2020, 20, 920 8 of 21 needed to be assumed for µ(t) and the φ k (t)s, but they are adaptively derived from the engines HI in the following.…”

“…To identify multiple stages before failure, Chehade et al studied informative sensor selection and fusion using statistical hypothesis testing [26]. Besides, the task of informative sensor selection was integrated with the statistical modeling in [23] and [27] for engine prognostics.…”

“…Through fusion of the selected informative sensors into a composite HI, many statistical methods have been proposed for prognostics of complex systems, including aircraft engines. However, most of the statistical approaches are based on parametric models, such as a linear model [16,23,28], exponential model [17,27], quadratic model [24], etc. These methods are effective when the engine degradation pattern can be described with a specific model, but there are numerous cases where the degradation trajectory may not be appropriately fitted by a parametric model [1,3,16].…”