A Labview System with Novel and Advanced Prognostic Tools

We propose a portable data acquisition and prognostic system that contains both hardware and software with several innovative ideas. First, our hardware system consists of a high speed data acquisition card and a portable lunchbox PC. The portable lunchbox PC has advanced prognostics algorithms and user friendly Graphical User Interface (GUI) for displaying component fault status and trends. Second, the prognostics software has several innovative algorithms, which will be implemented in Labview. The first one is an adaptive physics based prognostic tool. The idea is motivated by damage mechanics, which associates the vibration amplitude and natural frequency of the vibration to the damage status. This idea has been experimentally proven to be very accurate in bearing failure prediction. The second prognostic tool is data driven and is a Hidden Markov Model (HMM) based approach that can predict the degraded state of the system. Tests using experimental data showed that the various degraded states can be correctly and unambiguously identified by the algorithm. The third tool is a hybrid one and was proven to achieve very high performance in the 2008 PHM Challenge. Finally, to further enhance the prognostic performance, we propose to apply Dempster Shafer algorithm to perform prognostic fusion.