Prediction of outlet air temperature of direct air condenser based on BP neural network. (August 2019)

• Record Type: Journal Article
• Authors: Bai, Jianyun; Meng, Xinyu; Qu, Yan; Feng, Geng
• Is Part Of: Journal of physics. Volume 1303(2019)
• Publication Date: 2019-08
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1303/1/012014
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: At present, the dynamic parameter monitoring technology of the direct air cooling system is relatively weak. According to the study of outlet air flow temperature characteristics, this paper proposes a direct air condensing unit outlet air flow temperature prediction model based on BP neural network for studying dynamic parameters of direct air cooling units. The historical data of each factor is collected from the on-site DCS system. After the data is preprocessed, the BP neural network is used to establish the air temperature prediction model of the air condenser. Finally, the model is validated by the data that does not participate in the training. The model can predict the outlet air temperature of the air condenser under different working conditions, providing a basis for the next step of developing dynamic parameter monitoring technology, and also provide a reference for the cleaning cycle of the air condenser.