Blind Sound Source Localizing Of Vehicles With Time Difference Of Arrivals

In recent years,noise emission has been considered an important part of new emissions in the technical world.Therefore,it is vital to detect,locate and separate specific signals to stop or reducethe noise emissions.Engineers,researchers,car companies and designers are facedwith various challenges in the production of cars with minimum noise emissions,therefore it is necessary totry to locate and extract sound noise emissions.In this dissertation,different methods for Blind Sound Source Localizing(BSSL)has been investigated from previous studied methods.Different car horn sound types and their characteristics were studied.Time Difference of Arrivals(TDOA)was established to be the best sound source localizing method upon the requirements considered.TDOA methodisdescribedand its pros and cons are discussed.General Cross-Correlation(GCC)method is studiedfor reaching out to best estimation of TDOAs.A band pass filter is used for pre-processing of the signal to gain better signal resolution on the frequency domain that was investigated.This filtering is chosen based on the demands of TDOA method;moreoverthe advantages and disadvantages of this filtering method is discussed.A computer program code written based on studied algorithm for source localizing method and a prototype instrument based on the requirements of the methodology has been designed and used for experimental evaluation.The results of both simulation and experimentfor both one and two sound sources wereinvestigated separately.The obtained results show that the intended sound sources could be located with an acceptable amount of precision.Moreover,the accuracy of located positions has direct relation with frequency bandwidth and signal to noise ratio.In addition,the error in both mentioned parts were analyzed.The presented figures show the type and amount of errors with respect to different rounds of simulation and experimental implementations.