Morphological Component Analysis based Detection of Transient Behaviour in Environmental Vibration for the Watt Balance

The detection of stationary and non-stationary noise in environmental vibration data is an important issue when considering the precision of the Watt balance, an electromechanical apparatus for the new definition of the kilogram in the international system of Units (SI). The method used in this paper is Morphological Component Analysis (MCA). The MCA is a novel method which allows us to separate features in an observation when these features present different morphological aspects. Based on the nature that the transient events and the stationary noise have different morphological properties, different wavelet dictionaries and discrete sin/cosine dictionaries are selected to match and separate them. The experiment results illustrate that the MCA method can successfully separate the transient events from vibration signal.