一种新的基于陷波器的变电站噪声控制法

针对变电站噪声传统控制方法的缺陷,提出一种将人工神经网络与陷波器相结合用于控制变电站噪声的新方法.首先对自适应有源控制的对象变电站噪声进行理论分析;之后分析了2种传统的自适应有源噪声控制方法,即基于陷波器的有源控制法和基于人工神经网络的有源控制法.考虑到这2种方法的优缺点,利用人工神经网络算法代替陷波器中的最小均方（LMS）算法.最后,分别用这3种方法对变电站实测噪声信号进行处理,结果表明所提的方法不仅具有更好的稳定性,而且对变电站噪声有更好的抑制效果.

A Novel Notch Filter Based Method for Substation Noise Control

For deficiencies of conventional passive methods for substation noise control, a new method of substation noise control which combines artificial neural networks and notch filter is presented in this paper. Firstly, substation noise is theoretically analyzed. Then, two conventional adaptive and active noise control methods, i.e. notch filter based active control method and artificial neural network based active control method, are analyzed. Taking into account of the characteristics of these two methods, an artificial neural network algorithm is used instead of Least Mean Square （LMS） algorithm of notch filter. Finally, these three methods are comparatively evaluated by using actual noise signals from substation. The results show that the proposed method has better stability and suppression effects of substation noise compared with other two conventional methods mentioned.