Study On Filtering Methods And Error Chain In The Boost Glide Vehicle Warning And Detection

Boost glide vehicle(BGV),which is with long range,low flying height,high speed,and strong maneuvering capability,and whose trajectory and impact point can not be estimated by the traditional ballistic missile defense system(BMDS),brings great difficulties to the early warning and threat assessment of BMDS.Based on existing detection equipments of BMDS,the filtering and fitting algorithm of boost phase trajectory,the estimation and prediction algorithm of glide phase trajectory,and the construction method of estimation error chain are studied from the perspective of BGV flight dynamics.The specific content includes:Calculation methods of the mean square error lower bound and the optimal process noise covariance matrix are presented,which are used for the analysis of the BGV maneuver trajectory estimation.Through the study of maneuvering target tracking problem,it is concluded that the estimation of the maneuvering target state is biased and the traditional error lower bound method is not suitable for the evaluation of maneuvering target tracking algorithm.Based on the minimum mean square error criterion,the recursive algorithm of the mean square error lower bound of the state estimation of the maneuvering target is constructed.By analyzing the quantitative relationships among the unknown input,the estimation bias,and the process noise,the general form of the optimal gain matrix when the estimation is biased is derived,Under the frameworks of UKF and EKF,the mean square error lower bounds for the state estimation of maneuvering targets and the variation of the optimal process noise covariance matrix are achieved.This method provides theoretical basis and analysis method for the parameters design and performance evaluation of the BGV trajectory estimation algorithm.Based on binary observation,the trajectory estimation of BGV boost phase and the early warning information generation method are studied.When doing the real-time estimation of BGV boost trajectory with no prior information,the process noise variance adaptation becomes a difficulty in the algorithm design.Based on the quantitative relationships among the unknown input,the estimation bias,and the process noise,the expression of the process noise covariance matrix is constructed with the real time unknown input estimation value,which turns out to be a construction scheme of the process noise variance without any priori information.In order to improve the accuracy of trajectory estimation for BGV boost phase furthermore,a trajectory improvement strategy based on spline fitting and bidirectional filtering is proposed.The traditional two-body model is not suitable for BGV impact point prediction due to BGV's special trajectory characteristics.By comparing the ballistic parameters of BGV with those of the minimum energy trajectory,the special trajectory of BGV is identified.Simultaneously,based on the energy of BGV,the early warning information such as range,impact point,and time-to-go are generated with certain accuracies for the first time threat assessment.Based on the radar observation,the trajectory estimation and prediction method of BGV glide phase are studied.Based on the dynamic characteristics of BGV glide phase,the state equation of the traditional maneuvering reentry vehicle is improved,and the state prediction model of BGV glide phase is formed.Based on the model,an adaptive filtering algorithm called “variable structure interacting multiple model based on expected-mode augmentation” is proposed to solve the problems of estimating precision and computing speed in the trajectory estimation algorithm of the glide phase.Based on the trajectory characteristics of BGV glide phase and the estimation results of BGV glide motion parameters and aerodynamic parameters,the early warning information such as time-to-go,flight height,range,and reachable domain of BGV glide phase are generated under the assumption of equilibrium glide,for the second threat assessment.Based on the state prediction model of BGV glide phase,the trajectory prediction and error propagation characteristics of BGV glide phase are analyzed by using the unscented transform method,and the evolution principle of the trajectory prediction error of BGV glide phase is obtained.The construction and analysis of the BGV trajectory estimation error chain.The concept and construction process of the BGV trajectory estimation error chain are proposed.The simulation and evaluation system of the BMDS against the BGV attack is constructed.Two calculation methods of the error chain based on simulation system are presented.Through numerical simulation and analysis of the BGV trajectory estimation error chain under different sensor-deployment conditions,the validity of the general construction method of the error chain is verified,and so are the correctness of the error chain and the optimal results of the lower bound of the mean square error.The construction of the error chain provides an important analysis method for problems of the BMDS against BGV such as the situation awareness level assessment,sensor collaborative optimization deployment and so on.