作者: Eric Robert Bechhoefer
DOI:
关键词: False alarm 、 Data mining 、 Point (geometry) 、 Value (computer science) 、 Kalman filter 、 Connection (mathematics) 、 Component (UML) 、 Engineering
摘要: Disclosed are techniques used in connection with determining a health indicator (HI) of component, such as that an aircraft component. The HI is determined using condition indicators (CIs) which parameterize characteristics about component minimizing possibility false alarm. Different algorithms disclosed may be one or more CIs. normalized CI value. Techniques also described selecting particular CIs provide for maximizing separation between classifications. Given at point time predicting future state the Kalman filter. estimating data values alternative to performing acquisitions, when there no pre-existing data.