Research On Nonlinear Target Tracking Method Of USV In Complex Measurement Environment

Along with the continuous development of the national economy,as well as the exploration of the ocean and the development of marine resources continue to deepen,unmanned surface vehicle(USV)has gradually become a research hotspot all over the world.USV have the characteristics of high intelligence and good maneuverability,so that they are widely used in many fields,such as marine cargo transportation,marine disaster warning and even military defense.In order to achieve accurate identification and tracking of surrounding objects,unmanned ships must have strong environmental perception,so target tracking has become its key technical support.However,the characteristics of USV and its complex using environment may cause the measurements of the sensor contain outlier noise,which will make it difficult for traditional target tracking estimation algorithms to obtain high-precision estimation results.Moreover,the target tracking problem of USV requires that the estimation algorithm can be used for the nonlinear system,which limits the application of the classic Kalman filter algorithm.Therefore,this paper studies the target tracking problem of USV in a complex measurement environment,and the main related work includes the following three parts:1)USV target tracking model and measurement noise analysis.The characteristics of the USV and its complex using environment are analyzed.Combined with the mathematical model of USV target tracking,a complex measurement error model of the USV sensor system is established.Through the analysis of the measurement error,the mechanism of the measurement error affecting the performance of the target tracking algorithm is revealed and verified by simulation experiments.Simulation results show that in complex measurement environments,the performance of the target tracking algorithm will be seriously degraded,and a robust target tracking algorithm must be adopted to improve the performance.2)Research on robust nonlinear target tracking algorithm.Several classic nonlinear estimation algorithms and commonly used robust methods are introduced firstly.Then a new robust estimation algorithm is proposed based on the solution framework of the CKF algorithm,and the difference between the proposed estimation algorithm and the existing estimation algorithm is analyzed.Simulation experiments verify that the proposed algorithm has good robustness,and Wilcoxon signed rank test verifies that the proposed method has significant differences from the estimation results of the remaining algorithms.3)Mathematical tools are used to theoretically analyze the performance difference between target tracking algorithms.Simulation experiments show that the performance of the proposed algorithm has a significant difference from the remaining algorithms,but does not reveal the essence of this difference.Therefore,the influence function is introduced firstly,and then the influence functions of several algorithms are derived and calculated.The calculation results show that the proposed algorithm has good robustness.Subsequently,the ill-conditioned matrix and conditional number are introduced.The simulation experiments verified that the proposed algorithm has good numerical stability.