Research On The Method Of Large Dynamic Range Acoustic Signal Measurement And Sound Source Localization

Acoustic signals are ubiquitous in human life.How to obtain effective sound source information and use acoustic signals to achieve signal evaluation has important application value in the field of acoustic analysis.Among them,efficient sensing of weak sound signals in high noise environments,signal characterization and evaluation,and sound source localization are of great importance in various fields such as noise control,military operations,industrial production,and everyday life.Aiming at the above difficult problems,this thesis focuses on the research on acoustic signal measurement evaluation and positioning methods that meet distributed measurement applications and can realize signal detection in strong noise environments,including 1)large dynamic range acoustic signal acquisition methods to achieve high-precision signal acquisition;2)Acoustic signal analysis,processing and evaluation,through the study of acoustic signal analysis and processing algorithm to achieve effective signal information extraction and sound level measurement;3)Sound source feature analysis and localization algorithm research,based on the background of distributed sensor network measurement,by extracting signal Feature component A robust localization algorithm based on information fusion is proposed.In order to adapt to the acoustic signal pickup in the strong noise environment,this thesis proposes a large dynamic range acoustic signal segment amplification acquisition method based on the unified comparison voltage,and designs the circuit for simulation verification.According to the maximum allowable input voltage of the ADC,the signal is segmented according to different degrees of attention.Each segment corresponds to different magnifications,output in reverse order.Finally,the data is synthesized and restored according to the magnifications of different segments.This method can not only ensure the linear amplification of large and small signals,but also improve the resolution of the ADC;in the analysis,processing and evaluation of acoustic signals,the Variational Mode Decomposition(VMD)method is applied to reduce the influence of environmental noise,highlight the effective acoustic signals,and focus on the double line in the frequency weighting process.To overcome the shortcomings of the characteristic transformation method,a digital filter design method based on Chaotic Weighted Particle Swarm Optimization(CW-PSO)is proposed.The simulation results show that the algorithm is more suitable for the ideal curve,and finally the sound level measurement and calculation is completed through octave analysis.In the feature analysis and extraction of the sound source signal,the signal decomposition method based on Adaptive Iterative Filtering(ALIF)is used,and the simulation verifies that the algorithm has stronger robustness in both stationary and nonstationary signals;The component is used as the basic beam,and based on the distributed measurement application scenario,a sound source localization algorithm based on the combination of dominant frequency characteristic energy and phase-transform-weighted subjectable response power(SRP-PHAT)beamforming is proposed.By building a simulation localization model Experiments are carried out to verify that the method can reconstruct the energy distribution of the sound field while positioning,and has a high accuracy,which provides a method idea for practical engineering applications.