Apply Wavelet-ICA Filter for Feature Extraction

Independent component analysis (ICA) is a new effective technique for separation of statistically independent sources existing simultaneously in observations. Generally, ICA requires that the number of sensors should be no less than the number of independent sources to ensure enough information for separation of all sources. In some practical applications, this requirement of ICA is not met and we are interested in separation of only one source. A new method is proposed in this paper that requires only one transducer to acquire signal for separation of a specific feature. It employs wavelet transform as preprocessing and applies ICA to process the decompositions subsequently to find the latent feature, which is named as wavelet-ICA filter. The effectiveness of wavelet-ICA filter is demonstrated by applying it to both simulated signals and vibration signals collected from a gearbox for periodic impulse detection.