An Algorithm For Tracking Multiple Targets In A Cluttered Environment

The data processing techniques for track-while-scan radar is applied in development of some type of radar system. Target motion model, Kalman filtering and predication techniques, data association algorithmic, measured data processing and implementation techniques are studied by using of multi-target tracking and filtering theory. It puts the stress on the study of adaptive filtering and predication algorithms and data association algorithms according to the requirements of actual radar tracking.Combining practical engineering, an adaptive Kalman filtering algorithm in spherical coordinates is presented. This algorithm eliminates the transformation of non-linear coordinates, reduces the computation load and resolves the tracking of different maneuvering models. By selecting properly various of gates and association criterion being adaptable to engineering application, a practical data association algorithm is presented, which increases the probability of correct association and reduces the computation load of association. Additionally, a method to assess tracking quality is put forward on the basis of predecessors' research results, and a tracking quality management method with multi-parameter score accumulation is analyzed, it features perfect engineering practicability, good commonality/universality and easy to be analyzed.Finally, effectiveness of the method presented in this paper is verified by computer simulation and practical test.Kalman filtering in spherical coordinates, adaptive α-β filtering, track quality management, maneuvering target tracking, track-while-scan radar...