A Cooperative Target Tracking Method With Single Hydrophone Base On Kalman Filter

With the development of Chinese autonomous underwater vehicles,the demand for locating and tracking underwater vehicle is also increasing.At present,the development of submersibles is miniaturized and convenient,and the research focus of submersible tracking will also shift to low cost and simplicity.Submersibles such as AUV usually move at a long distance and at a constant speed when driving to the work area,without the need for high-precision tracking.However,when it enters the working state,they often need to perform maneuvering movements,and the required tracking accuracy is higher.Therefore,from the perspective of low cost,this paper only relies on a single hydrophone to study the target tracking technology under different precision requirements.In this paper,two tracking modes are designed.Firstly,two tracking methods under different observation conditions are established based on the extended kalman filter.By analyzing the observable conditions,it is determined that the "s" shaped observation track can achieve target tracking.When the target is not in the working state,but is traveling at a slow speed toward the work area,a non-speed measurement mode is adopted to reduce the communication overhead,and tracking of the slow-speed-change moving target is only achieved by ranging.Aiming at the impact of initial point selection on tracking,a two-dimensional search virtual long baseline method was used to determine the initial position to ensure tracking convergence.Then the relationship between tracking accuracy and track and the relationship between tracking accuracy and data error are discussed through simulation.Finally,the sea trial data processing shows that the algorithm can achieve low-precision tracking of targets.When the target enters the maneuvering working state,speed measurement mode is adopted by increasing the observed target speed.First from the perspective of algorithm optimization,an adaptive algorithm that can simultaneously estimate the state and measure noise online is designed.Then comparisons are made through several other kalman filtering algorithms by simulations to illustrate the advantages of the adaptive algorithm in terms of accuracy and stability.Finally,the lake test data is processed.The results show that as the adaptive algorithm estimates the noise gradually and gradually,the error does not exceed0.9m,which indicates that the adaptive algorithm can track the maneuvering target.