The Research Of The Target Location Method And The Realization Of System Based On Acoustic Sensor Network

The technique of passive acoustic detecting has virtue of secret, unlimited, precise,low-cost and strong anti-interference ability. So, it has been the very important means ofspying and detecting in modern war and has been widely used in civil and martial field. Inthe martial application field, its characteristic especially is applicable to detect and localizetarget, for example, helicopter or tank. In the civil application field, indoor localization is anapplication. So many scholars have researched the technique of passive acoustic localization,our country also begin and enhance scientific research. Combining sensor network’sflexibility and acoustic localization technique can improve localization effort.The essence of passive location technology based on acoustic sensor network is thecombination of locating method and algorithm. Locating method and algorithm are the keypoints of passive location technology. They affect the positioning accuracy together. Firstly,the traditional methods of localization has been studied and analyzed in the paper, which arebased on different measuring information for target arrival signal, such as the timedifference of arrival(TDOA), the direction of arrival(DOA) and the difference of dopplerfrequency(DD). Then, focusing on the methods based on DOA, three basic types ofpositioning algorithm have been studied, simulated and analyzed, such as the linear leastsquares(LLS) algorithm, the nonlinear least squares(NLS) algorithm and the maximumlikelihood estimation(ML) algorithm. The study can provide technical references for thedesign of target localization system based on acoustic sensor network.On the basis of the study on the positioning algorithm and the researching on theprinciple of passive direction-finding localization, the particle swarm optimization (PSO)-hemispherical iterated algorithm which is based on least squares error estimation ispresented. Hemispherical iterative method need to have an initial point, and the result oftarget localization is highly sensitive to initial point. So, the paper uses PSO in the initialvalue selection to solve the problem. At last, passive localization posed in this thesis issimulated and analyzed in Matlab.In the environment of Visual C++6.0, a target localization system based on acousticsensor network is developed based on study on localization module and theoretical analysisof algorithm. The system consists of two parts basically, fusion positioning and targetsimulator. The display software adopt the modular design with modules independent, the software reliability enhanced, and the software interface more friendly, attractive and easyto operate. After the test in the lab, the display software is validated as competent, stable andsuitable for application. The functions satisfy system’s demand.