飞机噪声特性数据的方向谱集成及支持向量机预测

针对飞机声源理论模型和NPD数据的不足,给出了一种基于频谱轨迹同步测试数据回归飞机噪声辐射规律的方法。该方法将飞机的噪声声场看成是以飞机质点为中心,距离R为半径的飞机噪声频谱球面,获取飞机噪声源数据的过程就是回归球面上各点方向谱的过程。剖析了飞机噪声特性数据的方向谱集成原理,给出了测试数据的几何关系计算及频谱重建方法,建立了飞机噪声方向谱的支持向量机回归模型,实现了飞机噪声声源数据的连续预测,为全面了解飞机噪声辐射规律,开展飞机噪声环境影响评价及噪声污染防治提供了科学直观的数据基础。

Directivity Spectrum Integration and Support Vector Machine PredictionBased on Aircraft Noise Characteristic Data

A method to regress aircraft noise radiation orderliness is presented aimed at making up the deficiencies of acoustic model and NPD data. This methed regards the sound field of aircraft noise as a spectrum sphere which with the center of aircraft centroid and the radius of distance R, the process of obtaining basal noise data is just the work of regressing the directivity spectrum on the sphere surface. This paper analyzes the integration theory, gives the way of calculating geometrical parameters and rebuilding the measured spectrums, establishs a support vector machine regression model by using the measured date as samples, and realizes the prediction with an example. The research result can give us scientific and intuitionistic data for understanding the radiation orderliness entirely, actualizing aircraft noise evaluation and noise pollution prevention.