A Novel Approach for Machinery Health Prognostics Using Statistical Tools

摘要

Condition based maintenance of machinery is being much talked about in thernengineering sector of defense and commercial industry. A lot of expenditure is generallyrnincurred on condition monitoring of machinery to avoid unexpected downtimes andrnfailures vis-à-vis optimizing machinery operation. The concept is ever evolving due torntechnological advancements as well as with the emergence of unique nature of defects.rnVibration Analysis is a potent tool of condition monitoring for prediction and diagnosticsrnof machinery failures. Presently, time and frequency spectra are being widely used forrndefect diagnostics of machinery. However, they require signal conditioning to eliminaternnoise and to enhance resolution of spectrum. Extensive research in the area of signalrnprocessing has been undertaken to refine time and frequency spectra. Notwithstandingrnapplication of statistical tools for analysis of various defects in machinery using conditionrnmonitoring data can be a viable option. Research in this area, where statistical modelsrnhave been applied, revealed encouraging results. In this paper, we have modeled bearingrnvibration data by applying time varying Markov Switching Auto Regressive methodrnwhich was found very helpful in estimating RUL of machinery.