Learning of inverse kinematics using a neural network with efficient weights tuning ability

Fusaomi Nagata, Shota Inoue, Satoru Fujii, Akimasa Otsuka, Keigo Watanabe

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Generally, in making a neural network learn nonlinear relations properly, desired training set are used. The training set consists of multiple pairs of an input vector and an output one. Each input vector is given to the input layer for forward calculation, and the corresponding output vector is compared with the vector yielded from the output layer. Also, weights are updated using a back propagation algorithm in backward calculation. The time required for the learning process of the neural network depends on the number of total weights in the neural network and the one of the input-output pairs in the training set. In the proposed learning process, after the learning is progressed e.g., 200 iterations, input-output pairs having had worse errors are extracted from the original training set and form a new temporary set. From the next iteration, the temporary set is applied instead of the original set. In this case, only pairs with worse errors are used for updating the weights until the mean value of errors reduces to a level. After the learning conducted using the temporary set, the original set is applied again instead of the temporary set. It is expected by alternately applying the above two types of sets for iterative learning that the convergence time can be efficiently reduced. The effectiveness is proved through simulation experiments using a kinematic model of a leg with four-DOFs.

Original languageEnglish
Title of host publication2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1042-1046
Number of pages5
ISBN (Print)9784907764487
DOIs
Publication statusPublished - Sep 30 2015
Event54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015 - Hangzhou, China
Duration: Jul 28 2015Jul 30 2015

Other

Other54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015
CountryChina
CityHangzhou
Period7/28/157/30/15

Fingerprint

Inverse kinematics
Tuning
Neural networks
Backpropagation algorithms
Kinematics
Experiments

Keywords

  • Efficient weights tuning
  • Inverse kinematics
  • Leg with multi-DOFs
  • Neural network
  • Temporary training set

ASJC Scopus subject areas

  • Control and Systems Engineering

Cite this

Nagata, F., Inoue, S., Fujii, S., Otsuka, A., & Watanabe, K. (2015). Learning of inverse kinematics using a neural network with efficient weights tuning ability. In 2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015 (pp. 1042-1046). [7285331] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SICE.2015.7285331

Learning of inverse kinematics using a neural network with efficient weights tuning ability. / Nagata, Fusaomi; Inoue, Shota; Fujii, Satoru; Otsuka, Akimasa; Watanabe, Keigo.

2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015. Institute of Electrical and Electronics Engineers Inc., 2015. p. 1042-1046 7285331.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nagata, F, Inoue, S, Fujii, S, Otsuka, A & Watanabe, K 2015, Learning of inverse kinematics using a neural network with efficient weights tuning ability. in 2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015., 7285331, Institute of Electrical and Electronics Engineers Inc., pp. 1042-1046, 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015, Hangzhou, China, 7/28/15. https://doi.org/10.1109/SICE.2015.7285331
Nagata F, Inoue S, Fujii S, Otsuka A, Watanabe K. Learning of inverse kinematics using a neural network with efficient weights tuning ability. In 2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015. Institute of Electrical and Electronics Engineers Inc. 2015. p. 1042-1046. 7285331 https://doi.org/10.1109/SICE.2015.7285331
Nagata, Fusaomi ; Inoue, Shota ; Fujii, Satoru ; Otsuka, Akimasa ; Watanabe, Keigo. / Learning of inverse kinematics using a neural network with efficient weights tuning ability. 2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015. Institute of Electrical and Electronics Engineers Inc., 2015. pp. 1042-1046
@inproceedings{c4aa625f323d431883a04d001499e4d2,
title = "Learning of inverse kinematics using a neural network with efficient weights tuning ability",
abstract = "Generally, in making a neural network learn nonlinear relations properly, desired training set are used. The training set consists of multiple pairs of an input vector and an output one. Each input vector is given to the input layer for forward calculation, and the corresponding output vector is compared with the vector yielded from the output layer. Also, weights are updated using a back propagation algorithm in backward calculation. The time required for the learning process of the neural network depends on the number of total weights in the neural network and the one of the input-output pairs in the training set. In the proposed learning process, after the learning is progressed e.g., 200 iterations, input-output pairs having had worse errors are extracted from the original training set and form a new temporary set. From the next iteration, the temporary set is applied instead of the original set. In this case, only pairs with worse errors are used for updating the weights until the mean value of errors reduces to a level. After the learning conducted using the temporary set, the original set is applied again instead of the temporary set. It is expected by alternately applying the above two types of sets for iterative learning that the convergence time can be efficiently reduced. The effectiveness is proved through simulation experiments using a kinematic model of a leg with four-DOFs.",
keywords = "Efficient weights tuning, Inverse kinematics, Leg with multi-DOFs, Neural network, Temporary training set",
author = "Fusaomi Nagata and Shota Inoue and Satoru Fujii and Akimasa Otsuka and Keigo Watanabe",
year = "2015",
month = "9",
day = "30",
doi = "10.1109/SICE.2015.7285331",
language = "English",
isbn = "9784907764487",
pages = "1042--1046",
booktitle = "2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Learning of inverse kinematics using a neural network with efficient weights tuning ability

AU - Nagata, Fusaomi

AU - Inoue, Shota

AU - Fujii, Satoru

AU - Otsuka, Akimasa

AU - Watanabe, Keigo

PY - 2015/9/30

Y1 - 2015/9/30

N2 - Generally, in making a neural network learn nonlinear relations properly, desired training set are used. The training set consists of multiple pairs of an input vector and an output one. Each input vector is given to the input layer for forward calculation, and the corresponding output vector is compared with the vector yielded from the output layer. Also, weights are updated using a back propagation algorithm in backward calculation. The time required for the learning process of the neural network depends on the number of total weights in the neural network and the one of the input-output pairs in the training set. In the proposed learning process, after the learning is progressed e.g., 200 iterations, input-output pairs having had worse errors are extracted from the original training set and form a new temporary set. From the next iteration, the temporary set is applied instead of the original set. In this case, only pairs with worse errors are used for updating the weights until the mean value of errors reduces to a level. After the learning conducted using the temporary set, the original set is applied again instead of the temporary set. It is expected by alternately applying the above two types of sets for iterative learning that the convergence time can be efficiently reduced. The effectiveness is proved through simulation experiments using a kinematic model of a leg with four-DOFs.

AB - Generally, in making a neural network learn nonlinear relations properly, desired training set are used. The training set consists of multiple pairs of an input vector and an output one. Each input vector is given to the input layer for forward calculation, and the corresponding output vector is compared with the vector yielded from the output layer. Also, weights are updated using a back propagation algorithm in backward calculation. The time required for the learning process of the neural network depends on the number of total weights in the neural network and the one of the input-output pairs in the training set. In the proposed learning process, after the learning is progressed e.g., 200 iterations, input-output pairs having had worse errors are extracted from the original training set and form a new temporary set. From the next iteration, the temporary set is applied instead of the original set. In this case, only pairs with worse errors are used for updating the weights until the mean value of errors reduces to a level. After the learning conducted using the temporary set, the original set is applied again instead of the temporary set. It is expected by alternately applying the above two types of sets for iterative learning that the convergence time can be efficiently reduced. The effectiveness is proved through simulation experiments using a kinematic model of a leg with four-DOFs.

KW - Efficient weights tuning

KW - Inverse kinematics

KW - Leg with multi-DOFs

KW - Neural network

KW - Temporary training set

UR - http://www.scopus.com/inward/record.url?scp=84960153753&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84960153753&partnerID=8YFLogxK

U2 - 10.1109/SICE.2015.7285331

DO - 10.1109/SICE.2015.7285331

M3 - Conference contribution

AN - SCOPUS:84960153753

SN - 9784907764487

SP - 1042

EP - 1046

BT - 2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -