The Comparison of Four Different Groundwater Level Prediction Models in Baoding City

DOI

10.2991/aeece-15.2015.101How to use a DOI?

Keywords

BP neural network prediction model; regression analysis prediction model; time series prediction model; Markov prediction model; Baoding City; groundwater level.

Abstract

This paper has quantitatively predicted the groundwater level variation in Baoding City based on BP neural network model, regression analysis model, time series model, and Markov model. It can be concluded that the prediction accuracy of BP neural network model and time series model is the highest, and their average relative prediction deviations are 3.5% and 2.3%. The prediction accuracy of regression analysis model is lowest, and the average relative prediction deviation is 14.69%. The range of average relative prediction deviation of Markov model is from 1.8% to 4.3%. The prediction accuracy of Markov model is high. Markov model can only predict the specific state of groundwater level, which is an interval value, rather than a specific value. Therefore, the reliability of Markov model can be improved by expanding the scope of the forecast on the premise of definitely meeting the requirements of actual work. This study provides scientific support towards groundwater level prediction of Baoding City, and has practical application value for urban water planning of Baoding City.

Copyright

© 2015, 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  - Qi Wang
AU  - Tingshan Tian
AU  - Changqing Li
PY  - 2015/09
DA  - 2015/09
TI  - The Comparison of Four Different Groundwater Level Prediction Models in Baoding City
BT  - Proceedings of the International Conference on Advances in Energy, Environment and Chemical Engineering
PB  - Atlantis Press
SP  - 501
EP  - 507
SN  - 2352-5401
UR  - https://doi.org/10.2991/aeece-15.2015.101
DO  - 10.2991/aeece-15.2015.101
ID  - Wang2015/09
ER  -