Paper Abstract and Keywords |
Presentation |
2014-05-29 14:30
[Poster Presentation]
A proposal for an agricultural environmental control system based on wireless sensor networks and machine learning Yukimasa Kaneda, Hirofumi Ibayashi, Yuya Suzuki (Shizuoka Univ.), Naoki Oishi (Research Institute of Agric.), Hiroshi Mineno (Shizuoka Univ.) ASN2014-22 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In recent years , agricultural support using ICT has been actively conducted.An example of such system is monitoring environmental information by building a sensor network in the facilities gardening. Then predicting the environment after a certain time from the sensing data, and to control the control equipment.However , the quality of the wireless communication is reduced in the harsh environments of horticulture , control failure of control equipment or loss of measurement data can be considered.In addition , in order to perform appropriate control according to the prediction , it is necessary to make a prediction with high accuracy at high speed .
In this study , we suggest the environmental control system smart and reliable .By validating failure that may occur , and take appropriate measures with respect to that portion and enabled a reliable sensor networks 99.68% data delivery ratio .Further, by performing learning using the SVR which is one of machine learning for prediction , focusing on the characteristics of time series data , it was confirmed that compared with the SVR which performed the usual study , reduction of prediction error about 17% is performed . |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Agriculture Support / Robust Sensor Network / Support Vector Regression (SVR) / Time-series Data Prediction / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 114, no. 65, ASN2014-22, pp. 75-76, May 2014. |
Paper # |
ASN2014-22 |
Date of Issue |
2014-05-22 (ASN) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
ASN2014-22 |
|