Reentry glide vehicle trajectory prediction method via multidimensional intention fusion

Abstract

To address the problems of difficulty in identifying the attack intention of reentry glide vehicle under the influence of no-fly zones and low accuracy of long-term trajectory prediction, a reentry glide vehicle trajectory prediction method via multidimensional intention fusion is proposed. Firstly, a time-varying parameter model set is constructed by estimating and predicting the aerodynamic parameters of non-cooperative vehicles. Secondly, considering the distance and angle relationship between the attack location, the no-fly zone and the vehicle, we construct the distance and angle dimensional intention cost function based on the artificial potential field method and the pseudo heading deviation, respectively, and propose the multidimensional intention fusion model based on the normalization method, which solves the problem of inference failure caused by the unreasonable coefficient setting in the traditional cost function. Finally, long-term trajectory prediction is achieved by inferring the vehicle's attack intent and parameter model via the Bayesian inference model and multidimensional intent fusion model. The simulation results show that the proposed method has the advantages of accurate intention inference, high trajectory prediction accuracy and short algorithm time compared with the existing methods.

Topics & Concepts