Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM

Aiming at the trajectory and motion characteristics of weakly constrained non-cooperative targets, a LSTM-based encoder-decoder multi-step trajectory prediction technology (EDMTP) is proposed. The introduction of first-order difference processing reduces the time dependence of the trajectory data, a...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autor principal:	Li Qingyong, He Bing, Zhang Xianyang, Zhu Xiaoyu, Liu Gang
Formato:	article
Lenguaje:	ZH
Publicado:	Editorial Office of Aero Weaponry 2021
Materias:	\|trajectory prediction\|lstm\|encoder-decoder\|supervised learning\|multi-step prediction Motor vehicles. Aeronautics. Astronautics TL1-4050
Acceso en línea:	https://doaj.org/article/1e16884d9df14a83bd1d0ef52927d6fc
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:1e16884d9df14a83bd1d0ef52927d6fc
record_format	dspace
spelling	oai:doaj.org-article:1e16884d9df14a83bd1d0ef52927d6fc2021-11-30T00:13:49ZEncoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM1673-504810.12132/ISSN.1673-5048.2020.0175https://doaj.org/article/1e16884d9df14a83bd1d0ef52927d6fc2021-04-01T00:00:00Zhttps://www.aeroweaponry.avic.com/fileup/1673-5048/PDF/2020-00175.pdfhttps://doaj.org/toc/1673-5048Aiming at the trajectory and motion characteristics of weakly constrained non-cooperative targets, a LSTM-based encoder-decoder multi-step trajectory prediction technology (EDMTP) is proposed. The introduction of first-order difference processing reduces the time dependence of the trajectory data, and obtains a trendless trajectory. Constructing an input and output trajectory data pair, transforming the prediction problem into a supervised learning problem, the change of model performance in the multi-step prediction process is studied to realize end-to-end trajectory prediction. Simulation results show that this method can extract more trajectory features from historical trajectory data, and has obvious advantages in multi-step trajectory prediction. Compared with the trajectory prediction algorithms of KFTP and HMMTP, the error growth rate of EDMTP decrease by 2.18% and 3.52% year-on-year, respectively, and achieves better trajectory prediction results.Li Qingyong, He Bing, Zhang Xianyang, Zhu Xiaoyu, Liu GangEditorial Office of Aero Weaponryarticle\|trajectory prediction\|lstm\|encoder-decoder\|supervised learning\|multi-step predictionMotor vehicles. Aeronautics. AstronauticsTL1-4050ZHHangkong bingqi, Vol 28, Iss 2, Pp 49-54 (2021)
institution	DOAJ
collection	DOAJ
language	ZH
topic	\|trajectory prediction\|lstm\|encoder-decoder\|supervised learning\|multi-step prediction Motor vehicles. Aeronautics. Astronautics TL1-4050
spellingShingle	\|trajectory prediction\|lstm\|encoder-decoder\|supervised learning\|multi-step prediction Motor vehicles. Aeronautics. Astronautics TL1-4050 Li Qingyong, He Bing, Zhang Xianyang, Zhu Xiaoyu, Liu Gang Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM
description	Aiming at the trajectory and motion characteristics of weakly constrained non-cooperative targets, a LSTM-based encoder-decoder multi-step trajectory prediction technology (EDMTP) is proposed. The introduction of first-order difference processing reduces the time dependence of the trajectory data, and obtains a trendless trajectory. Constructing an input and output trajectory data pair, transforming the prediction problem into a supervised learning problem, the change of model performance in the multi-step prediction process is studied to realize end-to-end trajectory prediction. Simulation results show that this method can extract more trajectory features from historical trajectory data, and has obvious advantages in multi-step trajectory prediction. Compared with the trajectory prediction algorithms of KFTP and HMMTP, the error growth rate of EDMTP decrease by 2.18% and 3.52% year-on-year, respectively, and achieves better trajectory prediction results.
format	article
author	Li Qingyong, He Bing, Zhang Xianyang, Zhu Xiaoyu, Liu Gang
author_facet	Li Qingyong, He Bing, Zhang Xianyang, Zhu Xiaoyu, Liu Gang
author_sort	Li Qingyong, He Bing, Zhang Xianyang, Zhu Xiaoyu, Liu Gang
title	Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM
title_short	Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM
title_full	Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM
title_fullStr	Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM
title_full_unstemmed	Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM
title_sort	encoder-decoder multi-step trajectory prediction technology based on lstm
publisher	Editorial Office of Aero Weaponry
publishDate	2021
url	https://doaj.org/article/1e16884d9df14a83bd1d0ef52927d6fc
work_keys_str_mv	AT liqingyonghebingzhangxianyangzhuxiaoyuliugang encoderdecodermultisteptrajectorypredictiontechnologybasedonlstm
_version_	1718406850025619456

Encoder-Decoder Multi-Step Trajectory Prediction Technology Based on LSTM

Ejemplares similares