The Research Of Maneuvering Target Tracking System With Adaptive Kalman Prediction Technology

With the rapid development of science and technology and the modern battlefield environment is becoming increasingly complex, more and more modern target maneuver performance is complex and changeable, and people on the tracking performance requirements are increasingly high, therefore, makes the modern maneuvering target tracking technology has become both in military and civil use is a research hotspot. In this paper, from the tracking target image acquisition, algorithm analysis and theoretical modeling aspects of maneuvering target tracking technology is studied, the main work of this paper are:analysis of the tracking target image acquisition method, based on obtaining the trajectory on the nonlinear characteristics of modern target tracking system, analyzes several nonlinear filtering methods commonly used extended Kalman filtering algorithm, unscented Kalman filtering algorithm and the volume of Kalman filter algorithm is deduced in theory, analysis and simulation provides the corresponding theoretical basis for research on adaptive algorithm for these three types of nonlinear filtering algorithm in high dimensional environment for nonlinear state estimation accuracy performance and stable performance for the subsequent nonlinear maneuvering target tracking system.Describes the current statistical model and adaptive target tracking algorithm theory and performance, and points out that the maneuvering frequency initial target maneuver performance reflects the current statistical model of the value is difficult to determine, and which does not change with target maneuvering performance and adaptive adjustment resulting in current statistical model for maneuvering target performance change description ability, reduce the overall accuracy of tracking target tracking system, put forward a kind of maneuvering frequency adaptive algorithm based on detection of motor performance, the establishment of maneuvering frequency adaptive regulation model, analysis and Simulation of the target maneuver acceleration in repeated change of circumstances, the algorithm of maneuvering frequency adaptive adjustment ability, and with the traditional fixed maneuvering frequency parameters of the current statistical model of the algorithm are compared analysis.Pointed out that the current statistical model using Rayleigh distribution describing the target acceleration correction has certain limitations, which makes the current statistical model of maneuvering and non maneuvering weak tracking performance is poor, put forward a kind of adaptive filtering algorithm with maximum acceleration, a bell shaped membership of fuzzy maximum acceleration correction function, and in order to enhance the improvement algorithm of acceleration the tracking capability with time-varying acceleration extremism fading factor regulating the. Analysis and simulation verified when the target maneuver with the existence of the weak maneuvering and non maneuvering motion, simultaneously also has the effect of improved tracking algorithm for target maneuver acceleration mutation of the situation, and with the traditional current statistical model and the improved algorithm has carried on the comparative analysis of literature.Finally, combining the filter method and modeling theory mentioned above were predicted to estimate the occluded target movement situation, and has carried on the simulation experiment, verified the feasibility of the model and algorithm in this paper.