Research on Water Quality Change and Prediction Based on Wavelet Analysis —— A Case Study of Guohe River

Abstract

Utilizing the monthly monitoring data of water quality indexes in the Guohe River Basin from 2005 to 2018 (a total of 168 months),this paper explores the application of wavelet analysis and neural network in the water quality research of the river basin.Wavelet analysis is employed to clarify the multi-scale change rules of water quality indexes in the Guohe River Basin.The main influencing factors of water quality are selected by principal component analysis,and a wavelet neural network model is established to predict the main influencing factors.The results show that each water quality index presents multi-scale oscillation,and there are three main change periods of about 8 months,20 months,and 30 months. At present,the main factor affecting the water quality of the Guohe River Basin is the pollution factor represented by chemical oxygen demand.The curve fitting is good between the chemical oxygen demand predicted by the wavelet neural network and the measured values with the mean relative error (MRE) and root mean square error (RMSE) of 8.4% and 1.5,respectively.This indicates the good stability and high prediction precision of the model.The application of wavelet neural networks provides a new idea for the study of water pollution in river basins.

Keywords

• water quality change
• water quality prediction
• wavelet analysis
• BP neural network
• Guohe River Basin