Multi-Source Information Fusion Based on Neural Networks in Air Quality Forecasting

DOI

10.2991/ecae-17.2018.35How to use a DOI?

Keywords

multi-source information fusion; air quality forecasting; time series; BP neural network; NARX neural network

Abstract

To forecast the air quality accurately, the model of air quality using multi-source information fusion technology based on neural network is proposed. The back propagation (BP) neural network models with time-series and no time-series training samples, the nonlinear auto-regressive (NARX) neural network with time-series training sample are respectively established on the MATLAB platform. The daily data of NO2, O3, PM10 and AQI are predicted using the models respectively. The conclusions are as follows: the three models with reliability, high prediction accuracy for air quality forecasting are successfully established. The accuracy of NARX with dynamic feedback capability is higher than BP neural network, while the BP neural network of larger non time-series training sample is of higher prediction accuracy.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Xiaoqiang Zhao
AU  - Yubing Chen
AU  - Qiang Gao
AU  - Dan Deng
PY  - 2017/12
DA  - 2017/12
TI  - Multi-Source Information Fusion Based on Neural Networks in Air Quality Forecasting
BT  - Proceedings of the 2017 2nd International Conference on Electrical, Control and Automation Engineering (ECAE 2017)
PB  - Atlantis Press
SP  - 164
EP  - 168
SN  - 2352-5401
UR  - https://doi.org/10.2991/ecae-17.2018.35
DO  - 10.2991/ecae-17.2018.35
ID  - Zhao2017/12
ER  -