M.Sc. Tezi Görüntüleme | |||||||||||||||||||||
|
|
||||||||||||||||||||
| Summary: Researches about aqustic source positioning have been increased past few years such as service robot applications, video – conference applications and military based applications like sniper positioning and aqustic watching. In this thesis TDOA which is commonly used in aqustic source positioning and has high sensitivity method based simulations and experiments have been performed. For an aqustic source positioning process firstly the system tested and passed with the analytic (forced) method which is already present in the literature. Than the solution performed with PSO algorithm has lot less process complexity as compared with present analytic (forced) methods. By this thesis we had a new iterative approach to TDOA based aqustic source positioning.In this study, four identical microphone for sensing the sound waves, ADC (AD9201) which has 20 MHz sampling rate for converting the analog sound information to digital, are used. For data acquisition and signal processing, a VHDL coded Xilinx brand Spartan 2E family FPGA with (CY7C68001) USB chip for transferring the digital sound data to computer is used. Aqustic sound signals taken by the microphones have been transferred to the computer and TDOA values have been calculated between the two microphones by using Generalized Cross-Correlation Approach (GCC). Hyperbola curve has been drawn amid to these two microphones by using the TDOA information between the microphones. The conjunction point calculation of hyperbola curves for all microphone peers, which gives the aqustic source position, has been made by using both analytic (forced) method and PSO algorithm. Key Words: Aqustic source positioning, Microphone arrays, TDOA, PSO Learning, Time processing, cross correlation | |||||||||||||||||||||