There is a growing interest from industry, government, and academia for prognostics and health management of engineering systems and critical components. Prognostics aim to predict the time when an engineering system or a critical component will no longer perform its intended functionality. Health management is to take a measure to respond to the anticipation of failures and minimize economic loss, and then prevent any unexpected accidents. Thanks to advances of sensor systems, a large amount of direct and indirect health monitoring data may be available. Direct health monitoring data mean that data can be directly used as health indicators to assess the current health condition of engineering systems and critical components. Indirect health monitoring data indicate that some transformations of data should be conducted to construct health indicators for condition assessment of engineering systems and critical components. Thus, signal processing and data mining algorithms are required to preprocess indirect health monitoring data prior to the use of any prognostic algorithms. After preprocessing is conducted, a health indicator is constructed to describe how far the current health condition of engineering systems and critical components deviate from their expected normal health conditions. Once a health indicator is available, prognostic algorithms are developed to extrapolate the current health condition to future health conditions and then predict the remaining useful life.
|Translated title of the contribution||Avances en pronósticos y gestión del estado del sistema|
|Original language||English (US)|
|Number of pages||300|
|State||Published - 31 Mar 2019|
CACES Knowledge Areas
- 116A Computer Science