Feature relevance estimation for vibration-based condition monitoring of an internal combustion engine
Journal Title: TecnoLógicas - Year 2017, Vol 20, Issue 39
Abstract
Condition monitoring of Internal Combustion Engines (ICE) benefits cost-effective operations inthe modern industrial sector. Because of this, vibration signalsare commonly monitored as part of a non-invasive approach to ICE analysis. However, vibration-based ICE monitoring poses a challenge due tothe propertiesof this kind of signals. They arehighly dynamic and non-stationary, let alone the diverse sources involved in the combustion process. In this paper, we propose a feature relevance estimation strategy forvibration-based ICE analysis. Our approach is divided intothree main stages: signal decomposition using an Ensemble Empirical Mode Decomposition algorithm, multi-domain parameter estimation from time and frequency representations, and a supervised feature selection based on the Relief-Ftechnique. Accordingly, we decomposedthe vibration signals by using self-adaptive analysis to represent nonlinear and non-stationary timeseries. Afterwards, time and frequency-based parameters were calculatedto code complex and/or non-stationary dynamics. Subsequently, we computeda relevance vector index to measure the contribution of each multi-domain feature to the discrimination of different fuel blend estimation/diagnosis categories forICE. In particular, we worked with an ICE dataset collected from fuel blends under normal and fault scenarios at different engine speeds to test our approach. Our classification resultspresented nearly 98% of accuracy after using a k-Nearest Neighbors machine. They reveal the way our approach identifies a relevant subset of features for ICE condition monitoring.One of thebenefitsis thereduced number of parameters.
Authors and Affiliations
José Alberto Hernández-Muriel, Andrés Marino Álvarez-Meza, Julián David Echeverry-Correa, Álvaro A. Orozco-Gutiérrez, Mauricio Alexánder Álvarez-López
Keywords
Related Articles
- EP ID EP348405
- DOI 10.22430/22565337.698
- Views 87
- Downloads 0
Technical-economic feasibility analysis of micro grids integrating fuel cells in non-interconnected zones in Colombia
The sustainable power solutions provided tonon-interconnected zonesshould bebased on the renewable energy resources available in each area. Thispresents a challenge for the techno-economic feasibility of alternativ...
Transmission network expansion planning considering weighted transmission loading relief nodal indexes
This paper presents a model and a solution approach for the transmission network expansion planning (TNEP) problem that integrates security constraints given by weighted transmission loading relief (WTLR) indexes. Such i...
Photovoltaic Potential Estimation by Means of Data Mining in Four Colombian Cities
This work analyzes the photovoltaic potential of four cities in Colombia—Bogotá, Cúcuta, Manizales, and Pasto—using information collected in situ and data mining strategies. The methodology of this study is based on clus...
The roles of nanotechnology and internet of nano things in healthcare transformation
Healthcare, as a basic human right, did not remain immune to innovative technologies. Technological progress has significantly contributed to high-quality, on-time, acceptable and affordable healthcare. Since their ap...
Analysis of the efficiency of a virtualoptical multiplexing method, by using theta modulation
We present an efficient virtual-optical multiplexing method for compressing information contained in gray level images. The efficiency is measured using three parameters: quality factors of...