Research On Underwater Target Co-Location And Tracking Algorithm

With the increasing emphasis on marine resources in the world,marine military and hydroacoustic engineering technologies have developed rapidly.In modern naval battles,precise positioning and tracking of targets is a necessary prerequisite for target strikes such as unmanned submarines and underwater robots.However,traditional underwater single-node detection systems have been unable to achieve precise positioning and tracking purposes.A multi-node collaborative detection system based on water came into being.This thesis focuses on a cooperative positioning and tracking algorithm in underwater multi-node layout mode,aiming to effectively improve target positioning and tracking accuracy.Firstly,based on the target echo signals collected by different hydrophones in the array,a mathematical model of the received target signal is established.The multiple signal classification(MUSIC)algorithm is used to achieve the target angle estimation under a single node.On this basis,combined with Gauss-Newton iteration method,a target co-localization algorithm combining multi-node target angle information is designed.The simulation results show that the designed cooperative positioning algorithm improves the target positioning accuracy under low SNR.In addition,the influence of different formation forms and number of nodes on the accuracy of coordinated target positioning is discussed.Aiming at the unknown speed of sound in underwater environment,the theoretical analysis of the influence of sound velocity on the cooperative positioning performance based on MUSIC is carried out.Then,based on the positioning accuracy and time efficiency of the fusion co-location algorithm based on Gauss-Newton iteration method,a cooperative algorithm based on particle swarm optimization is designed.The influence of different node numbers and node layout patterns on target co-location is discussed.The numerical simulation analysis shows that the cooperative positioning algorithm based on particle swarm optimization has improved target positioning accuracy,algorithm optimization success rate and average iteration number.Finally,considering the different motion models that may occur in the actual application,the extended Kalman filter based on multi-node fusion and the multi-model extended Kalman filter algorithm are designed to coordinate the target and predict the possible motion trajectory of the target.The simulation analysis proves that the improved fusion tracking algorithm can achieve effective tracking of the target.