フィードフォワード型能動制御による重機騒音の低減効果

書誌事項

タイトル別名
  • THE REDUCTION EFFECT OF CONSTRUCTION MACHINERY NOISE BY FEEDFORWARD TYPE ACTIVE NOISE CONTROL
  • フィードフォワードガタ ノウドウ セイギョ ニ ヨル ジュウキ ソウオン ノ テイゲン コウカ

この論文をさがす

抄録

 The construction machineriy noises have a sound peak in low frequency region(approximately less than 100 Hz). Active noise control (in the following, ANC) is one of the methods to reduce the low frequency noise. ANC is a system that reduce noises by interference between noise and sound wave from a speaker that has reverse phase against the noise. In recent years, ANC is occasionally applied to reduce noise of construction machineries like backhoes and rough terrain cranes.<br> As usual, the control method of the ANC is feedback that has only error microphone, or feedforward that has the reference microphone and the error microphone. In the construction sites, there are many kinds of noise sources like backhoes, cranes, power generator, roadheader, etc., therefore the ANC using feedback control may lose the target noise. Also, most of construction machineries move around in the construction site during operation, for safety it is hard to set up two microphones on the machineries for feedforward control.<br> So we developed an active noise control system for noise of construction machineries that can be set up on moving noise source like backhoes and can control stably. We considered about the applicability of the feedforward control which uses only reference signal, uses no Ĉ. Then, we proposed control methods of the ANC which estimate a predominant frequency and calibrate an amplitude ratio and a phase delay in time domain.<br> To estimmate a dominant frequancy in time domain, the proposed control method uses the optimized filter of LMS algorism. The LMS algorism minimizes the difference of input signal and the band-pass- filtered signal in the ANC system. The coefficient of the optimized filter is updated by each sampling time, the estimated frequency at n+1(n means a time number. n=1, 2...) is mainly derived from the estimated freqency at n, and the ratio of the coefficient of the optimized filter at n+1 to the coefficient of the optimized filter at n.<br> As for the calibration of an amplitude ratio and a phase delay in time domain, the proposed control method uses correction values of amplitude and phase delay in each frequency, which are obtained in advance from the comparison o of the input signal and the output signal of the ANC system. When the control starts, at first, the system takes out the correction value of the estimated frequency. Next, the system subtracts the number of data of the correction value from the number of data of one wavelength for the estimated frequency. At last, the system takes out the past data at the point of calculated difference from the start point, and the data multiply by the correction value of amplitude is the output signal.<br> To confirm the usefulness of the proposed method, we carried out numerical simulations. In the simulation, we used some fixed sin waves of 50Hz to 100Hz, and fluctuated sin waves such as 40Hz to 50Hz or 40Hz to 100Hz, as input signal. As a result, the proposed methods can control stably and the control speed of proposed method is faster than that of conventional feedforward control and conventional feedback control.<br> Then, we carried out experiments in a laboratory to verify the effectiveness of ANC system using proposed method. Some fixed sin waves of 50Hz to 100Hz and construction machinery noises are used as input signal. As a result, the ANC reduced fixed sine waves by maximum about 20dB, and the ANC reduced construction machinery noise by about 10dB.

収録刊行物

詳細情報 詳細情報について

問題の指摘

ページトップへ