Research On Localization Technologies Of Short-time Impulsive Acoustic Source

Acoustic source localization technology has been widely used in industry, medical, military and office due to its advantages of high accuracy and no limitation of visibility. High-accuracy acoustic localization has always been the key problem of wireless location technologies. Short-time impulsive acoustic signals, such as crash, applause and gunshot, are widespread in the world, which come with emergencies, and last for milliseconds. High-accuracy passive impulsive acoustic source localization technology is the method of apperceiving outside environment for robot, intelligent vehicle, and human-computer interaction systems. It has important research and application value to research effective localization algorithms and develop locating device.After analyzing the character of short-time impulsive acoustic signal and common localization algorithms, this paper focused researched impulsive acoustic localization algorithms based on TDOA and subspace, and the implementation of the algorithms was also researched. Developed the short-time impulsive device on TMS320VC5502 platform, and tested its performance of localization. The main content is as follows:(1) Analyzed the time-frequency character of short-time impulsive acoustic signal and the applicability of common-used signal detection methods. Considering the demand of efficiency and accuracy, determined the amplitude detection method, whose perform rate is fastest, as the scheme of signal detection. Studied the short-time localization algorithms based on TDOA and subspace, analyzed their locating performance by simulation. Proposed and realized a multi-frequency near-field source localization algorithm suitable for impulsive acoustic source;(2) Designed the short-time impulsive acoustic source localization signal processing flow, set up localization system hardware device, which can realize analog signal receiving and magnifying, multi-channel parallel AD sampling, digital signal processing and locating results displaying;(3) Studied the implementation of impulsive acoustic source localization algorithm based on TDOA and subspace, solved the problem of low operation efficiency of cross-correlation sequence calculation and 2-dimension spectral peak searching. Implemented the calculation of matrix eigenvalues and eigenvectors. Based on this, designed and realized the software of short-time impulsive acoustic localization.(4) Set up experiment environment to test the function of AD sampling, signal detecting, location calculating, operation efficiency and the locating accuracy of algorithms. Test results show that the localization system designed in this paper can achieve real-time positioning for impulsive acoustic source whose signal lasts only milliseconds. The accuracy in close range localization can reach centimeter.The main contribution and innovation of this paper include:(1) Proposed a hardware system realizable method for multi-frequency near-field source localization which is suitable for impulsive acoustic source. By applying orthogonal transformation to the receiving data of transducers, a complex observation matrix is acquired. The eigenvalues and eigenvectors of the complex covariance matrix are computed using similarity transformation method of eigenvalue decomposition, after which the noise subspace is established. Considering the frequency spectrum of source signals, a suitable spectral function is constructed for multi-frequency signals to search for the location of source. Expanded the 2D-MUSIC algorithm to localization system for short-time impulsive acoustic source.(2) Proposed the hardware implementation method of short-time impulsive acoustic source localization algorithm based on TDOA. Used amplitude detection to detect the target signal. Computed the time delay of impulsive signal by calculating cross-correlation sequence. Solved the hyperbolic equation using Chan algorithm. The time used for hardware to calculate the location is 0.06 s.The researches above provide solutions of wireless location for domestic robot, unmanned vehicles systems, provide technological reserve for human-computer interaction products such as interactive electronic whiteboard.