Water Quality Prediction Model Based on GRU hybrid network

Abstract

In view of the characteristics of randomness, non-linearity, randomness and interdependence of water quality data in water environment, in order to improve the prediction accuracy and prediction efficiency of water quality prediction model, a hybrid water quality prediction model based on convolutional neural network (CNN) combined with gated recurrent neural network (GRU) is proposed. First, the potential characteristics between water quality continuous data are extracted efficiently through CNN network. Then, based on the potential characteristics, a GRU network with temporal memory capability is used to accurately predict water quality data. Finally, the real monitoring data of Shanghai Jinze Reservoir is combined, and a water quality prediction model based on CNN-GRU is established. The experimental results show that the hybrid prediction model proposed in this paper has higher prediction accuracy than the traditional SVR water quality prediction model and the standard GRU network water quality prediction model.