An Improved Independent Component Analysis by Reference Signals and Its Application on Source Contribution Estimation

To estimate the contribution of main vibration and noise sources of vehicles, a source contribution estimation method based on an improved independent component analysis (ICA) algorithm is proposed. A measure of the similarity between the independent components and the reference signals with given characteristics is introduced. A widely used contrast function, namely kurtosis, is enhanced by the measure to obtain an improved contrast function. By means of fixed-point iteration and deflation approach, the improved contrast function is optimized and the improved ICA algorithm is attained. The contribution is computed by the reduced energy in each extraction, the reduction of the energy corresponds to the contribution of the extracted independent components. The correspondence relationship between the independent components (ICs) and source signals can be obtained by the signal characteristics. The effectiveness of the proposed algorithm is verified by numerical simulation and experiment.