Font Size: a A A

Research On Robot Sound Source Localization Method Based On Beamforming

Posted on:2022-06-10Degree:MasterType:Thesis
Country:ChinaCandidate:Z X HanFull Text:PDF
GTID:2518306533994499Subject:Electronic information
Abstract/Summary:PDF Full Text Request
Sound source localization technology uses time delay,beamforming and other algorithms to process the sound signals collected by microphone array to get the direction of sound source target.It has a wide and potential application prospect in sonar,noise detection,fault diagnosis,intelligent products and other fields.With the rapid development of robotics technology,the combination of sound source localization technology and robotic platform can better realize the localization of sound source targets.On the one hand,robot sound source localization can assist the robot to perceive the direction of the sound source target in the surrounding environment,on the other hand,it can also use the characteristics of robot movement to measure the relative distance of the sound source target,which can solve the problem that it is difficult to determine the distance of sound source target in passive sound source location.Therefore,the research of robot sound source localization method has important theoretical significance and application value.For the reason that the accuracy of this method needs to be improved,the minimum variance distortionless response(MVDR)algorithm is used to locate the target,which has higher localization accuracy than the conventional beamforming algorithm(CBF),at the same time,an improved calculation method of cross-shaped microphone array is proposed to solve the problem of excessive calculation when the beamforming algorithm is used on a two-dimensional microphone array.The main contents of this paper are as follows:(1)The common microphone array structures are analyzed,an improved cross microphone array calculation method is proposed to reduce the amount of calculation and improve the beam resolution of microphone array.(2)The daily used sound source localization algorithms are researched,selects the beamforming algorithms to compute the direction of the target,and studies the angle of two dimensions estimation problem caused by the improved cross array calculation method,the time delay estimation(TDE)algorithm is also used to adjust the direction of the robot so that the robot can measure the distance by moving.(3)The whole embedded hardware platform is built,including the construction of cross microphone array and the design of data acquisition module.(4)The software design of the system is implemented on the embedded platform,including A/D conversion of sound signal,digital filtering and fast fourier transform(FFT).The sound source localization algorithm program is designed to estimate the angle of the sound source target,based on the angle estimation,the robot measure the distance program by moving is designed,and the GUI is designed for easy operation.At the end of this paper,the accuracy of the data acquisition part of the robot sound source localization system was tested,and alter the angle and distance of the target many times to experiment,the test results verify the feasibility of the system designed in this paper.
Keywords/Search Tags:sound source localization, robot, MVDR, distance measurement
PDF Full Text Request
Related items