Research And Implementation Of Sound Source Localization Algorithm Based On TDOA

In recent years,sound source localization technology has been a hot topic,which has been valued by many experts and scholars at home and abroad.The principle of sound source localization is to extract relevant characteristic parameters(such as time difference of arrival,angle of arrival of arrival,intensity of signal of arrival,etc.)from the received sound source information to solve the position coordinates of the sound source point.With the development of microphone array processing technology,microelectronic technology,computer technology and other related technologies,people have more and more application requirements for sound source localization technology.The application of sound source localization technology has not only been limited to the military field,but it has been widely used in various fields such as traffic monitoring,industrial fault detection,and security monitoring.The research focus of this paper is the research and implementation of sound source localization algorithm based on Time Difference of Arrival(TDOA),including algorithm research and hardware system implementation.The algorithm research part mainly contains three contents.The first part is a detailed introduction to the basic principle of TDOA and two steps of TDOA which are delay estimation and positioning solution.Time delay estimation is an algorithm that solves the time delay TDOA from the sound source to different microphones in the received signal.The mainstream idea is to calculate the time delay through Generalized Cross-Correlation(GCC).The positioning solution is an algorithm that uses time delay and the receiver position to solve the sound source position.The commonly used algorithm is the Newton iteration method;The second part proposes a two-step time delay estimation algorithm with low computationalcomplexity.Sample the envelope of the received signal with a low sampling frequency,perform GCC calculation and a preliminary calibration on the time axis according to the obtained first step delay.Obtain two speech signals with rough time delay error,intercept a small segment of the signal,and perform GCC operation to interpolate the obtained cross-correlation function to obtain the second step delay.Finally,the results of the two delays are combined to obtain the true delay of the signal.In the Matlab environment,the effects of various parameters of the two-step GCC are simulated,the time complexity comparison and anti-noise performance comparison between the algorithm and the traditional GCC are analyzed;The third part is to follow the two steps of delay estimation and positioning solution,study the various factors that affect the final positioning accuracy in the sound source localization system,and put forward effective improvement measures for different influencing factors.Influencing factors include noise,reverberation,low sampling frequency in the delay estimation step,and delay estimation error,microphone position error,and microphone array topology in the position resolution step.Relevant simulation experiments are carried out for various influencing factors and improvement measures.Simulation results show that the various improvement measures introduced can effectively suppress the corresponding influencing factors and improve the positioning accuracy.The hardware system implementation mainly includes three parts:microphone array,STM32 single-chip microcomputer,and PC.They are responsible for sound signal collection,sound signal storage,and algorithm calculation functions of the sound source localization system respectively.Thus,the hardware implementation of the sound source localization function is completed.Then the functional test of the hardware system is carried out.The test proves that the hardware system can achieve sound source localization under indoor environment,and the optimal positioning accuracy is within 1 m.