Analysis and Research for Airplane Cockpit Sound’ICA Denoise Based on Blind Source Separation Principle
The characteristic of cockpit sound recorded by CVR is the key evidence in investigating accident causes for wrecked airplane. But many factors from inside and outside cockpit affects the analysis results, especially noise. And some analysis methods (such as FFT, WT, CZT, and so on) were impacted by internal and external environmental sound in aircraft cockpit. To solve the problems, further analysis and research for typical cockpit Sound’ ICA de-noising based on blind source separation principle. Firstly, the principle of de-noising analysis based on ICA are made in details ,including blind source separation and analysis of OGWE, maximum ratio of signal to noise of blind source separation algorithm; Then, process of cabin sound de-noising analysis based on ICA is done. Research indicates the de-noising of typical cockpit sounds is identical with the de-noising based on wavelet packet analysis in Matlab; more easy and reliable for the research and analysis of typical cockpit sound.
Helen Zhang, Gang Shen and David Jin
D. L. Cheng et al., "Analysis and Research for Airplane Cockpit Sound’ICA Denoise Based on Blind Source Separation Principle", Advanced Materials Research, Vols. 204-210, pp. 209-215, 2011