Target trajectory prediction model of long short-term memory network based on attention mechanism

Aiming at the problems of low prediction accuracy and poor generalization of traditional recurrent neural networks in target trajectory prediction, an attention-mechanism-based target trajectory prediction model (ATT-LSTM) was proposed. The model introduces the attention mechanism into the LSTM network and focuses on the learning of key track information to improve the model's prediction performance. The simulation results show that the model has higher prediction accuracy and better generalization than the existing model.

I. Introduction

With the development of science and technology and modern warfare, various types of air and space threat targets such as quick strike weapons, fourth-generation fighter jets, hypersonic cruise missiles are emerging constantly. They have various types, different movement characteristics, and some have high maneuverability, which brings challenges to battlefield target tracking. Target tracking is mainly divided into two parts: trajectory prediction and filtering update. The accuracy of trajectory prediction directly affects the performance of filtering update. Therefore, it is particularly urgent to find a rapid and accurate trajectory prediction method.