Research On Passive Acoustic Localization Technology Based On Microphone-array

Passive acoustic identification and localization is a new technology subject coveringmany fields, such as acoustics, electronics, signal detection, signal processing, softwaredesign and so on, which enjoys wide application in military field, assisted listeningdevices, video conference system，security monitoring field, robot localization and otherfields. However, because of the specificity of acoustic signal and complexity of acousticenvironment in reality, the study of theoretical algorithm simulation is in an active statewhile there are various problems in the engineering application field such as high cost,large computation quantity, low accuracy and instantaneity which result in limitedapplication examples. From the practical application angle of improving localizationaccuracy under low sampling frequency and lowering hardware cost, this paper designsand achieves a set of passive acoustic localization system with five-microphone array,and provides support for achieving high-accuracy localization using this method to someextent, as well as gives proof for the possibility of applying this method to embeddedfacility.Firstly, this paper carries out mathematical deduction and simulation for acousticlocalization model with five-microphone array and algorithm, and probe into the keytechnique in passive acoustic localization, i.e. time delay estimation. Aiming at the tissuethat the localization accuracy using generalized cross correlation to estimate time delay islimited by the sampling frequency of hardware system, this paper proposes to utilize up-sampling method to conduct sampling frequency transition in order to improvelocalization accuracy, and proceeds to simulation confirmation with MATLAB.Secondly, the paper designs the hardware part of the passive acoustic localizationsystem, including the selection of microphone, the drawing of signal preprocessingcircuits and welding process of PCB, selection of MCU and ADC and their initialization,related power module and voltage transition module, test and calibration of the wholehardware circuits.Finally, the paper gives experiments in indoor environment under the condition ofweak noise, strong noise and different array distance respectively, and discusses thelocalization range through experiments. Affected by hardware, reverberation and otherfactors, the experiment results have some extent of errors comparing with theoreticalresults. Then the paper modifies the localization algorithm, and the modified results showthat the acoustic localization system designed in this paper is simple, of low cost andavailable, which gives proof to its engineering realization of this algorithm.