engleski

Bank of Adaptive Filters for Fault Detection in Multiple Regime Process Using Unsupervised Training

Change detection is kind of adaptive filtering for non-stationary signals, and is the tool in fault detection and diagnosis. In this paper bank of adaptive filter is used for fault detection in multiple regime process. Bank of filters is trained by proposed unsupervised method suitable for periodic signals like vibrations from rotational machinery.

adaptive filter; fault detection; unsupervised training

nije evidentirano