面向时变机动目标的自适应航迹整体估计方法OA

In the complex maneuvering target tracking scenario with unknown time-varying target motion model,the use of a fixed state-space model will significantly reduce the tracking performance due to model mis-match.In view of the above problem,this paper proposes an adaptive tracking method based on trajectory-orient-ed estimation estimation.In this method,the trajectory-oriented estimation of the target is modeled as a polynomi-al time series;the time polynomial model is established to characterize the evolution process of its motion state;the polynomial coefficients are recursively estimated in matrix form according to the minimum deviation sum of squares criterion,so as to achieve the global estimation of the full-time track at each sampling time,and further-more,the asymptotic decoupling estimation method for polynomial coefficients is proposed to reduce the computa-tional complexity.On this basis,the polynomial models of different degrees are used to construct a multi-channel parallel trajectory-oriented estimation estimator group,and adaptive channel selection is carried out according to the real-time error analysis results to deal with the single model mismatch problem caused by the complex motion pattern of time-varying maneuvering targets.Simulation results show that the proposed method can realize the se-lection of adaptive tracking channels and model parameter adjustment when the real motion state of the target can-not be ascertained,and that compared with the traditional recursive filtering method,this trajectory estimation method has better tracking performance.