BAG: A Linear-Nonlinear Hybrid Time Series Prediction Model for Soil Moisture

Guoying Wang, Lili Zhuang, Lufeng Mo (), Xiaomei Yi, Peng Wu and Xiaoping Wu
Additional contact information
Guoying Wang: College of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, China
Lili Zhuang: College of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, China
Lufeng Mo: College of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, China
Xiaomei Yi: College of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, China
Peng Wu: College of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, China
Xiaoping Wu: School of Information Engineering, Huzhou University, Huzhou 313000, China

Agriculture, 2023, vol. 13, issue 2, 1-20

Abstract: Soil moisture time series data are usually nonlinear in nature and are influenced by multiple environmental factors. The traditional autoregressive integrated moving average (ARIMA) method has high prediction accuracy but is only suitable for linear problems and only predicts data with a single column of time series. The gated recurrent unit neural network (GRU) can achieve the prediction of time series and nonlinear multivariate data, but a single nonlinear model does not yield optimal results. Therefore, a hybrid time series prediction model, BAG, combining linear and nonlinear characteristics of soil moisture, is proposed in this paper to achieve the identification process of linear and nonlinear relationships in soil moisture data so as to improve the accuracy of prediction results. In BAG, block Hankel tensor ARIMA (BHT-ARIMA) and GRU are selected to extract the linear and nonlinear features of soil moisture data, respectively. BHT-ARIMA is applied to predict the linear part of the soil moisture, and GRU is used to predict the residual series, which is the nonlinear part, and the superposition of the two predicted results is the final prediction result. The performance of the proposed model on five real datasets was evaluated. The results of the experiments show that BAG has a higher prediction accuracy compared with other prediction models for different amounts of data and different numbers of environmental factors.

Keywords: nonlinear characteristics; block Hankel tensor; autoregressive integrated moving average; gated recurrent unit; soil moisture prediction (search for similar items in EconPapers)
JEL-codes: Q1 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 (search for similar items in EconPapers)
Date: 2023
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2077-0472/13/2/379/pdf (application/pdf)
https://www.mdpi.com/2077-0472/13/2/379/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jagris:v:13:y:2023:i:2:p:379-:d:1058225

Access Statistics for this article

Agriculture is currently edited by Ms. Leda Xuan

More articles in Agriculture from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().