The Improved LMS Adaptive Filtering Algorithm And Its Application On Sound Power Measurement

The background noise from microphone and signal conditioning circuit which are necessary for the sound power measurement limit the measuring dynamic range and affect the measurement results.But only accurate measurements and analysis of noise can.provide effective basis for noise control.When the acoustic source being measured is weak,the acoustic source being measured will be damged if the background noise is not well controlled.Then the sound power measurement system can't accurate analysis,even can not work normally.For the problem that some bands attenuation is insufficient and the attenuation error beyonds the tolerance range of level 1 of GB3240-1982 when A/C frequency weighting,the thesis proposes an improved LMS adaptive filtering algorithm to filter the measured signals,weaken the influence of background noise on the measured signals and improve the SNR.Then the thesis builds the sound power measurement system based on the improved LMS algorithm to improve the performance of the measurement system.Firstly,adaptive filtering technology and its pplications in acoustic measurement are introduced.Then an improved LMS algorithm is proposed based on the disadvantages of LMS adaptive filtering algorithm when it is applied in the sound power measurement.By constructing two adaptive coefficient ?(n)and ?(n),the thesis set up a new nonlinear function between step and error.The improved LMS algorithm has a faster convergence speed in the case of guaranteeing the accuracy and improves the real-time performance.The simulation results on MATLAB show that when the measured signal is polluted seriously,the algorithm can reduce the noise effectively.Moreover,the useful signals can be accurately recovered within a wide band range.Secondly,combining with the improved LMS algorithm,a sound power measurement system is set up on the LabVIEW based on pre-polarization capacitance measurement microphone,signal conditioning circuit and zhongtai industrial EM-9118B-18 synchronous data acquisition card.The theory of sound power determination by sound pressure method is expounded.Then the detailed system structure and the working process of the sound power measurement system based on improved LMS adaptive filtering algorithm are introduced.And the realization principle of sound power measurement based on improved LMS algorithm is introduced in detail,and the corresponding LabVIEW implementation program of the key part is given.Finally,in view of the sound power measurement system based on improved LMS adaptive filtering algorithm,the thesis tests the system using the input whose sound pressure level is close to the background noise.The results prove that when the sound pressure level of input is close to 30dB,after adaptive filtering by the improved LMS algorithm,the attenuation value error of A/C frequency weighting is smaller,and meet the the requirements of GB3240-1982 grade 1 tolerance range.Especially,the problem of insufficient attenuation is solved when the frequency is lower than 31.5Hz.