Research On Cooperative Localization And Tracking Technology Of Multi-Base Sonar

In recent years,with the development and advancement of vibration and noise reduction technology,the radiation noise level of large underwater structures continues to decrease,thus putting forward higher requirements on sonar detection technology.To improve the effectiveness of the sonar system,a new sonar working system needs to be researched,and the multi-base sonar system is born.Compared with the traditional single-base sonar,the multibase sonar system adopts the transceiver split mode of operation,which has the advantages of both active and passive sonar and has strong detection capability and concealment,and can make up for the shortcomings of small area and single function of single node.In this paper,the research of multi-base sonar positioning and tracking technology is carried out.In this thesis,we firstly studied the basic principle of localization of two-base sonar system,and on this basis,we modeled and analyzed the bearing only localization and time only localization(TOL)algorithms of two-base sonar system,further extended the model to multibase sonar system,studied the TOL algorithm,linear least squares algorithm,total least squares algorithm and constrained total least squares algorithm for localization of multi-base sonar system,theoretically deduced the localization error of the algorithms,and used geometric dilution of precision is used as an index to evaluate the positioning accuracy,and the effectiveness of the above algorithms is verified through simulation.Subsequently,the adaptive weight particle swarm optimization algorithm was designed to address the problem that the above positioning algorithm is vulnerable to the initial value selection,and the effects of the number of nodes and layout patterns on the positioning accuracy were analyzed.To solve the problem that the adaptive weight particle swarm optimization algorithm is easy to fall into the local optimal solution in the process of finding the optimal solution,the marine predator algorithm(MPA)algorithm is subsequently introduced,it is also verified by simulation that the MPA algorithm outperforms the adaptive weight particle swarm optimization algorithm in terms of the success rate,convergence times and the accuracy of the search.Then,linear Kalman filtering algorithm,extended Kalman filtering algorithm,unscented Kalman filtering algorithm and volumetric Kalman filtering algorithm are designed to accomplish the target tracking task.In addition,for the tracking problem of multimodal targets,the nonlinear filtering algorithm is combined with the interactive multimodal algorithm to make up for the model mismatch deficiency of the unimodal algorithm and effectively improve the target tracking accuracy.Finally,a relevant sea trial experimental study was conducted to track the uniform linear motion target using linear filtering algorithm and nonlinear filtering algorithm,and the experimental results verified the effectiveness of the nonlinear filtering algorithm and provided data support for the actual target tracking.