A fast fixed-point algorithm for independent component analysis
Neural Computation
Independent component analysis: algorithms and applications
Neural Networks
A maximum likelihood approach to single-channel source separation
The Journal of Machine Learning Research
Mono-component signal extraction and analysis of mechanical watch movements
Control and Intelligent Systems
Time-varying lapped transforms and wavelet packets
IEEE Transactions on Signal Processing
A new independent component analysis for speech recognition and separation
IEEE Transactions on Audio, Speech, and Language Processing
A novel optimization based method for separation of periodic signals
Digital Signal Processing
Hi-index | 0.00 |
This paper presents a method to combine signal decomposition with blind signal separation for separating one-dimensional mixed multi-component signal of mechanical watch movement. The method consists of three steps. First, a multi-component signal is decomposed into a set of redundant signals, called pseudo multi-channel mixtures, using the continuous wavelet transform (CWT). Then, independent component analysis (ICA) is used to acquire the independent components. Finally, a correlation criterion is applied to automatically select the source components. The new method can effectively separate a multi-component signal into a series of independent components corresponding to different sources. The effectiveness of this method is demonstrated by means of a computer simulation example. And the proposed method is applied for analyzing and diagnosing mechanical watch movements. It is found that the separated source components effectively reveal the insight of the mechanical watch movement, and can be used for fault diagnosis.