航空科学技术2024，Vol.35Issue(4)：97-103,7.DOI:10.19452/j.issn1007-5453.2024.04.013
基于数据驱动的机动目标跟踪
Maneuvering Target Tracking Based on Data-driven
摘要
Abstract
Target tracking is the key to accurate target strike,especially for maneuvering targets,which can improve the estimation accuracy of target state and achieve precise target strike tasks.This paper proposes a data-driven maneuvering target tracking algorithm based on neural network theory for the tracking problem of maneuvering targets.By generating a large amount of flight data with different maneuvering modes,a data sample library for target maneuvering is established.A deep neural network based on bidirectional input is used to train a large amount of different data,generate models of maneuvering trajectories,and match trajectory sequence data with different maneuvering models to achieve trajectory prediction for any maneuvering target,thereby achieving tracking of maneuvering targets and effectively improving the tracking accuracy of maneuvering targets.The simulation results validate the effectiveness of data-driven target tracking algorithms,provides new ideas and methods for predicting maneuvering target trajectories,and improves the timeliness and accuracy of trajectory prediction.关键词
神经网络/数据驱动/机动目标/目标跟踪/轨迹预测Key words
neural networks/date-driven/maneuvering target/target tracking/trajectory prediction分类
信息技术与安全科学引用本文复制引用
马云红,李欣怡,王泽霖,曹天航,王永坤..基于数据驱动的机动目标跟踪[J].航空科学技术,2024,35(4):97-103,7.基金项目
航空科学基金(2020002005300) Aeronautical Science Foundation of China(2020002005300) （2020002005300）
