基于盲源分离方法的声振降噪算法研究

Research of Sound vibrations algorithm based on blind source separation

本文研究一种可以分离固体火箭发动机试验中的环境噪声的方法,该方法实现的技术途径是基于盲源分离思路.完成了模型建立与FastICA-AIC新算法设计,并在固体发动机上进行搭载试验,进行振源信号与噪声信号的分离.本文解决了源信号与源信号个数未知,且源信号与分离后的信号个数不一定一致的情况下,声振降噪的难题,具有智能化、广泛性、适应性等优点,可以更有效的进行发动机状态诊断与监测,对提升测试技术有较大的帮助.

This paper mainly study a kind of method for separated the environmental noise of solid rocket motor,based on blind source separation.In this paper,we have completed the model and the new algorithm FastICA-AIC.Carry out load test on solid motor,and separate vibration source signal from noise signal.In this paper,the problem of acoustic vibration noise reduction is solved when the number of source signals and the number of source signals are unknown,and the number of source signals and the number of separated signals are not necessarily the same.It has the advantages of intelligence,extensiveness and adaptability,and it can more effectively conduct engine state diagnosis and monitoring,which is of great help to improve the testing technology.