Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot

Xu Wang; Yanfang Liu; Naiming Qi; Zhihao Tang

doi:10.1109/CCDC52312.2021.9602425

Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot

Xu Wang
, Yanfang Liu^*
, Naiming Qi
, Zhihao Tang

^*Corresponding author for this work

School of Astronautics

Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

For complicated dynamic systems, it is still a great challenge to build a sufficiently precise model to predict their long-term behavior, as there are always nonlinearities too complicated to model. In this paper, a probabilistic ensembles neural networks (PENN) is proposed to predict the long-term dynamic behavior. PENN combines the probabilistic deep network dynamics model with sampling-based uncertainty propagation to deal with the aleatoric and epsitemic uncertainties in dynamic system estimation. Simulation demonstrates the effectiveness of the PENN. The prediction error is about 0.05 deg for one step prediction and 10 deg for 100 steps propagation.

Original language	English
Title of host publication	Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	2586-2590
Number of pages	5
ISBN (Electronic)	9781665440899
DOIs	https://doi.org/10.1109/CCDC52312.2021.9602425
State	Published - 2021
Event	33rd Chinese Control and Decision Conference, CCDC 2021 - Kunming, China Duration: 22 May 2021 → 24 May 2021

Publication series

Name	Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021

Conference

Conference	33rd Chinese Control and Decision Conference, CCDC 2021
Country/Territory	China
City	Kunming
Period	22/05/21 → 24/05/21

Keywords

Dynamics Model
Long-Term Prediction
Probabilistic Deep Network
Probabilistic Ensembles Neural Netrworks

Access to Document

10.1109/CCDC52312.2021.9602425

Cite this

Wang, X., Liu, Y., Qi, N., & Tang, Z. (2021). Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot. In Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021 (pp. 2586-2590). (Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CCDC52312.2021.9602425

Wang, Xu ; Liu, Yanfang ; Qi, Naiming et al. / Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot. Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 2586-2590 (Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021).

@inproceedings{18e209fccc58485baa7c86e91d133694,

title = "Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot",

abstract = "For complicated dynamic systems, it is still a great challenge to build a sufficiently precise model to predict their long-term behavior, as there are always nonlinearities too complicated to model. In this paper, a probabilistic ensembles neural networks (PENN) is proposed to predict the long-term dynamic behavior. PENN combines the probabilistic deep network dynamics model with sampling-based uncertainty propagation to deal with the aleatoric and epsitemic uncertainties in dynamic system estimation. Simulation demonstrates the effectiveness of the PENN. The prediction error is about 0.05 deg for one step prediction and 10 deg for 100 steps propagation.",

keywords = "Dynamics Model, Long-Term Prediction, Probabilistic Deep Network, Probabilistic Ensembles Neural Netrworks",

author = "Xu Wang and Yanfang Liu and Naiming Qi and Zhihao Tang",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 33rd Chinese Control and Decision Conference, CCDC 2021 ; Conference date: 22-05-2021 Through 24-05-2021",

year = "2021",

doi = "10.1109/CCDC52312.2021.9602425",

language = "英语",

series = "Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "2586--2590",

booktitle = "Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021",

address = "美国",

}

Wang, X, Liu, Y , Qi, N & Tang, Z 2021, Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot. in Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021. Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021, Institute of Electrical and Electronics Engineers Inc., pp. 2586-2590, 33rd Chinese Control and Decision Conference, CCDC 2021, Kunming, China, 22/05/21. https://doi.org/10.1109/CCDC52312.2021.9602425

Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot. / Wang, Xu; Liu, Yanfang ; Qi, Naiming et al.
Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021. Institute of Electrical and Electronics Engineers Inc., 2021. p. 2586-2590 (Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot

AU - Wang, Xu

AU - Liu, Yanfang

AU - Qi, Naiming

AU - Tang, Zhihao

PY - 2021

Y1 - 2021

N2 - For complicated dynamic systems, it is still a great challenge to build a sufficiently precise model to predict their long-term behavior, as there are always nonlinearities too complicated to model. In this paper, a probabilistic ensembles neural networks (PENN) is proposed to predict the long-term dynamic behavior. PENN combines the probabilistic deep network dynamics model with sampling-based uncertainty propagation to deal with the aleatoric and epsitemic uncertainties in dynamic system estimation. Simulation demonstrates the effectiveness of the PENN. The prediction error is about 0.05 deg for one step prediction and 10 deg for 100 steps propagation.

AB - For complicated dynamic systems, it is still a great challenge to build a sufficiently precise model to predict their long-term behavior, as there are always nonlinearities too complicated to model. In this paper, a probabilistic ensembles neural networks (PENN) is proposed to predict the long-term dynamic behavior. PENN combines the probabilistic deep network dynamics model with sampling-based uncertainty propagation to deal with the aleatoric and epsitemic uncertainties in dynamic system estimation. Simulation demonstrates the effectiveness of the PENN. The prediction error is about 0.05 deg for one step prediction and 10 deg for 100 steps propagation.

KW - Dynamics Model

KW - Long-Term Prediction

KW - Probabilistic Deep Network

KW - Probabilistic Ensembles Neural Netrworks

UR - https://www.scopus.com/pages/publications/85125196345

U2 - 10.1109/CCDC52312.2021.9602425

DO - 10.1109/CCDC52312.2021.9602425

M3 - 会议稿件

AN - SCOPUS:85125196345

T3 - Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021

SP - 2586

EP - 2590

BT - Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 33rd Chinese Control and Decision Conference, CCDC 2021

Y2 - 22 May 2021 through 24 May 2021

ER -

Wang X, Liu Y , Qi N, Tang Z. Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot. In Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021. Institute of Electrical and Electronics Engineers Inc. 2021. p. 2586-2590. (Proceedings of the 33rd Chinese Control and Decision Conference, CCDC 2021). doi: 10.1109/CCDC52312.2021.9602425

Probabilistic Ensembles Neural Networks Model for Long-Term Dynamic Behavior Prediction of a Robot

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this