Purpose

In this article, a new ANC algorithm is proposed to attenuate the noise of industrial fans. It was named MC-FXLMS/F algorithm. It has showed to have a higher level of attenuation in comparison to other classical algorithms found in the literature. It is seen that the performance of the MC-FXLMS/F is superior when applying specific levels for the reference signal and also considering a substantial number of filter coefficients.

Methods

In the literature, it is possible to find different methods to perform the attenuation of the noise generated by the blades pass frequency, being one of the methods the active noise control. As it is possible to predict the periodicity of noise generated by fans, it is possible to use synthesized signals to be used as a reference signal for an active noise control system. In this work, two classical ANC algorithms, namely the FXLMS and the C-FXLMS/F were also evaluated and implemented. They were compared to the new algorithm. in terms of efficiency (for different types of synthesized and narrowband reference signals). The power spectral density results were presented for variations in the signal levels with and without active noise control.

Results

The results showed the applicability of using pure tones to attenuate specific frequencies of the noise spectrum. The algorithms C-FXLMS/F and MC-FXLMS/F were shown to be more sensitive to high levels of the reference signal in contrast to the FXLMS algorithm that benefits from higher levels.

中文翻译：

---

改进的ANC消除工业风扇噪声的算法

目的

在本文中，提出了一种新的ANC算法来衰减工业风扇的噪声。它被称为MC-FXLMS / F算法。与文献中发现的其他经典算法相比，它具有更高的衰减水平。可以看出，当为参考信号应用特定电平并考虑大量滤波器系数时，MC-FXLMS / F的性能会更高。

方法

在文献中，有可能找到不同的方法来执行由叶片通过频率产生的噪声的衰减，这是主动噪声控制的方法之一。由于可以预测风扇产生的噪声的周期性，因此可以将合成信号用作有源噪声控制系统的参考信号。在这项工作中，还评估并实现了两种经典的ANC算法，即FXLMS和C-FXLMS / F。将它们与新算法进行了比较。在效率方面（针对不同类型的合成和窄带参考信号）。功率谱密度结果针对有和没有有源噪声控制的信号电平变化给出。

结果

结果表明使用纯音调来衰减噪声频谱的特定频率的适用性。与受益于较高级别的FXLMS算法相比，C-FXLMS / F和MC-FXLMS / F算法对高电平参考信号更敏感。