Extended Targets Tracking Based On Gaussian Process In Clutter

With the rapid development of sensor technology,the sensor resolution is more and more high,the target signal often crosses multiple sensor resolution units,and the target contour information can not be ignored any more.The traditional tracking technology which regards the target as a point is facing great challenge.Extended target tracking(ETT)not only considers the motion characteristics of the target,but also considers the contour characteristics of the target,It can estimate jointly the target motion state and contour state through signal processing and data processing.ETT has important research significance and application value in the military and civil fields such as high resolution radar/sonar detection,remote sensing and unmanned driving.Aiming at the problem of ETT in clutter environment,several extended target tracking algorithms based on Gaussian process are proposed in this paper.The main work is as follows:1.For the problem of extended target tracking in clutter-free environment,the Gaussian process extended Kalman filter method is simulated and analyzed.Firstly,the state space model of extended target is established based on the Gaussian process,and the whole measurement of each frame is augmented to form the augmentation measurement matrix.Secondly,the extended Kalman filter is used to update the state and covariance of extended target.Finally,compared with the Random Matrix(RM)method in simulation,the algorithm performance is analyzed systematically.2.In order to solve the problem of extended target tracking in clutter,a Gaussian process probabilistic data association(GP-PDA)method is proposed.Firstly,the joint validation gate of extended target is established based on the Gaussian process to select the validated measurements.Secondly,the association events of all extended targets are enumerated with validated measurements,and the probability of association events is calculated.Finally,the probabilistic data association algorithm is used to update the state and covariance of extended target.Simulation verifies the effectiveness of the algorithm.3.Aiming at the problem of multi-maneuvering extended target tracking in clutter,a variable structure multiple model joint probabilistic data association method based on Gaussian process is proposed.Firstly,the adaptive model set is constructed by expecting model augmentation method,and each extended target state is initialized.Secondly,based on the Gaussian process,the joint validation gate of extended target is established to select the valid measurement and form the confirmation matrix.Then,splitting confirmation matrix to get the association matrix and hence to solve the associated event probability.Finally,the data association problems of multiple extended targets are further solved by updating each extended target state and covariance.