As a new approach of blind source separation (BSS), independent component analysis (ICA) is a recently developed method in which the processed objects are mixed signals from linear combination of the original data, and the goal is to separate the source from the mixtures and the components separated are statistically independent, or as independent as possible. The ICA method are used in communication, sound and image processing, biology medicine, earthquake signals processing, even the finance date analysis etc. At present, the application domain of ICA is taken on end on end the enlarged trend. In the paper, the basic theory and algorithm of ICA are briefly introduced, and then the simulation instantaneous signals are denoised. The five signals such as two sines sweep signals and one-sine signals and one sawtooth wave signals and one stochastic noise are mixed together, and then the mixed signals are separated with ICA. The results show that the five signals can be exactly separated with ICA, and the ICA method has high ability to separate the instantaneous signals from their mixtures, consequently providing an effective technology for the pretreatment of signals to fault diagnosis of mechanical equipment.
展开▼
