Font Size: a A A

Research On Moving Target Acoustic DOA Navigation Tracking Algorithm Based On Particle Filter

Posted on:2020-10-26Degree:MasterType:Thesis
Country:ChinaCandidate:Y M GongFull Text:PDF
GTID:2428330620962436Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
As a basic technology,navigation tracking technology is applied to many areas such as driverless,target tracking and agricultural irrigation.There are many types of basic navigation tracking technologies,but each technology has its own scope of application,and different navigation tracking technologies should be selected for the job background.With the in-depth study of the sound direction estimation technology,the navigation tracking of the target direction through sound has important research value.The Direction of Arrival(DOA)estimation technique for sound through sound array sensors is relatively mature,but most of the methods are based on the assumption that the target position is stationary or the target position is slowly changing.Therefore,the DOA estimation technique directly has a large error in the direction navigation of the moving target,especially when the trend direction of the target shows a nonlinearity,the algorithm will fail.Aiming at these problems,this paper proposes a particle-based moving target acoustic DOA navigation tracking method.The system model in the particle filter algorithm combines the observation information of the sound source with the behavior information of the moving object to overcome the limitation of the traditional DOA estimation algorithm only for the stationary target,and realize the dynamic navigation and tracking of the moving target trajectory.The main contents and results of this paper are as follows.Firstly,the paper analyzes the principle of particle filter algorithm and traditional acoustic DOA estimation method.In this part,the sound signal acquisition mode is determined by studying the L-type acoustic sensor array signal model.Through the theoretical and experimental research on the law of sound propagation,the gap between the actual sound attenuation law and the theory is found,which paves the way for the sound source signal selection in the acoustic DOA navigation tracking algorithm.The feasibility of multi-signal classification(MUSIC)for non-coherent multi-signal direction estimation is verified by theoretical analysis and experimental simulation.Secondly,the moving object acoustic DOA navigation tracking algorithm based on particle filter is implemented.The effectiveness of the algorithm is verified by the simulation form.Thirdly,in multi-target navigation,based on the influence of the interference signal source on the algorithm,the spatial spectral function of the MUSIC algorithm is substituted for the particle filter algorithm likelihood function [1],and an improved acoustic algorithm combining particle filter algorithm and MUSIC algorithm is proposed.DOA navigation tracking algorithm.The feasibility of the improved acoustic DOA navigation algorithm is verified by simulation,which provides a basis for multi-target direction estimation navigation.Finally,the article builds a typical motion scene of a moving target through MATLAB,and analyzes the relationship between the spatial position coordinates and the spatial direction angle of the target by using the moving trajectory of the moving target in space.In the typical scenario,the effectiveness of the proposed acoustic DOA navigation tracking algorithm for moving target navigation tracking is verified.This paper mainly studies the moving target navigation tracking technology through sound DOA.Through theoretical and experimental simulations,the paper proposes that the navigation tracking algorithm can improve the dynamic navigation tracking performance of moving targets,and improve the navigation tracking adjustment ability when the target direction changes nonlinearly.It has strong robustness and potential application value.
Keywords/Search Tags:Navigation tracking, Direction of Arrival, Moving target, Particle filtering, System model
PDF Full Text Request
Related items