Font Size: a A A

Research On Target Tracking And Trajectory Prediction Based On Angle Observation

Posted on:2019-11-29Degree:MasterType:Thesis
Country:ChinaCandidate:P W LuFull Text:PDF
GTID:2432330551461666Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
In modern warfare,attack helicopter is characterized by high maneuverability,strong firepower and wide range of uses.Its low altitude flight seriously threatens the ground units.Aiming at the needs of lightweight air defense weapons system,a method of target tracking and trajectory prediction based on angle observation is studied under the condition of only angular measurement,which provides an angle reference for attacking helicopters.The main work is as follows:(1)The performance and characteristics of attack helicopter are analyzed,and the corresponding flight path is designed according to the way of its low altitude flight,which provides simulation data for the angle tracking algorithm and experiment.The state models and observation equation of the target are established,and according to the state model of the target the single model angle tracking algorithm is designed.Then the maneuver frequency adaption is applied to the single model angle tracking algorithm.The performance of the algorithm is simulated,and the maneuver frequency adaption is suitable for the little error of observation in the angle tracking algorithm.(2)Considering the tracking performance degradation of single model algorithm when the error of observation is larger,an interacting multiple model angle tracking algorithm is designed,and the tracking results of different model sets are compared.According to the real-time property of the algorithm,support vector machine classifies the states in all directions and reduces the number of parallel computing models to achieve real-time angular tracking of the target.The performance of the algorithm is compared by simulation.The tracking error of the interacting multiple angular tracking algorithm is small but the computational complexity is large.However,the angular tracking algorithm based on SVM classifier can meet the real-time tracking with less tracking error.(3)Combined with angular tracking algorithm,the corresponding double loop angle prediction algorithm is designed to complete the target's angle prediction.Under different time of prediction and error of observation,the results of angle prediction by single model and multi model angle are compared.The results show that the prediction is more accurate in the stage of stable flight,and the angle error of prediction by multi model is smaller.(4)Design system test,build angle tracking semi-physical simulation test platform.The experiment shows that in the process of angle tracking,the target is located in the center of the observation screen.The experimental error is analyzed,and the experimental results of different angular tracking algorithms are compared.The results show that the proposed algorithm in this paper is feasible.
Keywords/Search Tags:angle tracking, double loop angle prediction, interacting multiple model, support vector machines, adaptive filter
PDF Full Text Request
Related items