ウェーブレットによる自動車加速音の瞬時相関解析と聴感評価に関する検討(機械力学,計測,自動制御)

概要

論文の詳細を見る
Recently, many researchers who work with time-frequency analysis using wavelet transform have focused on analyzing wavelets that are derived using a mathematical approach. In the present analysis, a measured signal is adopted as the wavelet, and we analyze the correlation between acoustic signals in the car cabin and suction noise signals by applying the proposed system. Because traditional calculation of correlation repeats the averaging procedure, the original signal must be stationary. Consequentially, a technique for separating and identifying noise from each part of the engine is used for noise source contribution analysis. To apply the method to time-varying signals, the concept of an instantaneous correlation factor (ICF) is introduced, and we prove that a dominant feature of the correlation can be estimated by the ICF. The time-varying correlation for the noise source contribution analysis of an accelerating car is analyzed. And the fundamental experiment about its subjective evaluation in that case is also conducted.
社団法人日本機械学会の論文
2006-07-25