摘要
为了提高变压器有源降噪算法收敛的速度,保证算法稳态降噪量,本文针对变压器噪声特性,提出一种改进的基于窄带有源降噪的变压器有源降噪算法。该算法以经典的并行参考输入的窄带有源降噪技术为基础,结合波形合成的方法,由系统内部合成参考信号,无需采用参考送话器。并且引入了对残余误差信号中各频率信号的估计,使用了基于sigmoid函数的变步长策略,使算法在保持低稳态误差的同时提高变压器有源降噪算法收敛的速度。本文搭建了基于DSP的实验平台,用实地采集的变压器噪声信号进行了模拟变压器噪声环境下的降噪实验,在误差送话器处取得6.4dB的总体降噪效果。
In order to improve the convergence speed of transformer active noise reduction algorithm and ensure the steady-state noise reduction of the algorithm,an improved transformer active noise reduction algorithm based on narrow-band active noise control is proposed in this paper.The algorithm is based on multi-frequency notch filter technology,classical narrowband active noise reduction technology with parallel reference input,and combined with waveform synthesis method.The reference signal is synthesized from the system interior without reference microphone.Furthermore,the estimation of each frequency signal in residual error signal is introduced,and the variable step size strategy based on sigmoid function is combined to improve the convergence speed of active noise reduction algorithm for transformer while maintaining low steady-state error.An experimental platform based on DSP is built,and the noise reduction experiment under the simulated transformer noise environment is carried out with the transformer noise signal collected in the field.The overall noise reduction effect of 6.4dB is achieved at the error microphone.
作者
陈兴
王青云
Chen Xing;Wang Qingyun(School of Electric Power Engineering,Nanjing Institute of Technology,Nanjing 211167)
出处
《电气技术》
2019年第11期39-45,共7页
Electrical Engineering
关键词
变压器噪声
窄带有源降噪
变步长
transformer noise
narrow-band active noise control
variable step size
