Sound Source Depth Identifiable Three-Dimensional Focused Beamforming

Previously beamforming with planar microphone array can only identify the position of planar source in two dimension, such as the x, y coordinate or azimuth of the sound source, it can’t identify the depth of three-dimensional (3D) sound source, namely the beamforming with planar array based acoustical holography can not accurately identified the 3D coordinates of sound sources, definitely, it can’t recognize the sound source located in front of the planar array or behind. To break the limitations of the existing methods and technology, sound source depth identifiable three-dimensional focused beamforming is presented for the 3D source identification and localization in this thesis. B&K sixty channel slice wheel array and self-made reconfigurable umbrella shaped 3D microphone arrayarray has been used to vertify this method with simulation and experiment.The main contents of this thesis include:1) The conventional beamforming with microphone array on the sound source identification and localization is derived theoretically to show the limitations of this method, and that is proved by experiment on slice wheel array. Simulation results show the conventional beamforming methods can only identify the sound source in the case of the depth of the sound source is provided, that means this kind of methods can’t automatically recognize the depth of the sound source.2) Sound source depth identifiable three-dimensional focused beamforming is proposed to identify and localize the 3D sound source, and the performance of this method is simulated systemically. At the same time the factors which affect the performance of this method is evaluated by setting the localization error and the half-power beam width (HPBW) as key indexes.3) The method is confirmed experimentally in full anechoic chamber. The small speakers are selected as target sound sources in the experiment.2D measurement array is B&K sixty channel slice wheel array. The pressure is gained by wheel array in the chamber to verify the availability and resolution of this method which is analyzed by the theoretical research and simulation studies. The experiment results indicate that this method is capable to locate and identify sound source in 3D space. While the frequency is above 1500 Hz and the depth of the sound source is under 1.5m, the error keep within 20% in all direction.4) The reconfigurable umbrella shaped 3D three-dimensional microphone arrayis designed and developed by my own. The factors which affect the performance of this array is evaluated by HPBW among numerical simulation and experiment. The factors include total microphone number M, microphone spacing dj change mode, sound source frequency f, sound source depth zo, the opening angle of the array. The comparation is performed between B&K sixty channel slice wheel array and the self-made reconfigurable umbrella shaped 3D microphone array. The results show that the reconfigurable umbrella shaped 3D microphone array combined with the method has a good capability of three-dimensional sound source identification and localization.