Production Prediction Model Of Complex Industrial Processes Based On GRU Neural Network
Yongming M Han, Zilan Du, Zhiqiang Geng, Yajie Wang, Feng Xie, Kai Chen · 2020
Raw materials consumption and many physical and chemical reactions in the industrial process are not conducive to analyze the production process. Therefore, the ethylene production prediction model based on gated recurrent unit (GRU) is brought forth in this paper, which can be used to accurately predict ethylene production capacity in complex industrial processes. Using the total consumption of raw materials, fuel, water, steam and electricity as the input of the network architecture and the ethylene production as the output, GRU network is set as the kernel to build the production prediction model. Form experiment results comparing with the back propagation (BP) neural network and the long short-term memory (LSTM), GRU can better predict ethylene production, thus reducing production energy consumption and improving production efficiency.