统计对称噪声中叶片振声信号提取的探讨

针对统计对称噪声系统对称点噪声统计特性相对稳定的特点 ,提出了利用对称镜像点的信号降低噪声 ,克服不稳定因素影响的改进方法。数值仿真结果表明 ,该方法克服了非平稳变化的影响 ,提高了实时分析能力。

Study on Extracting Blade Acoustic Signal Masked in Statistic Symmetrical Noise

The shortcoming of synchronous average method is taking too much time to avoid the effects of flow tumbling, so the shortcoming limits the application of vibration acoustic method. In order to overcome the influence of unstable factors, in this paper,a revised method has been studied,especially,for the case that the noise source is of statistical symmetry.In this case it presumes that the signal of noise acquired at symmetrical positions is of statistical symmetry, with the nearly same expected value, the background noise could be eliminated with two signals acquired at two symmetrical positions. The experimental and numerical simulation have been made, the results show the effects of the method is quite well with much less time than that used in STARS system, and the flow tumbling has been overcome. The ratio of signal to noise has been greatly improved.