STUDIES ON PREDICTING PAPER MAKING WASTEWATER TREATMENT WITH BP NEURAL NETWORK
A papermaking wastewater treatment control system was set up in the lab, the experimental data was used to train BP neural network, and a neural network model was built for papermaking wastewater treatment. The result of using this model to simulate papermaking wastewater treatment indicated that the BP neural network had a good learning ability. The result of wastewater treatment was predicted with this model, and the relative error was 19%, which showed that the generalization ability of BP neural network was not so good. Howerver, when the neual network was trained with dynamic, the neural network model precdicted the result of wastewater treatment accurately, and the relative error was just 1.9%, which showed that the dynamic training BP neural network model was better than the BP neural network that was not being dynamic trained. The results of this paper lays the foundation to the computer control of papermaking wastewater treatment.
configuration software MCGS paper making wastewater BP neural network intelligent control
Wan Jinquan Zhang Yancong Ma Yongwen Wang Yan
College of Resource Science &Paper Engineering, South China University of Technology, Guang zhou, 510640, China
国际会议
广州
英文
766-769
2006-11-08（万方平台首次上网日期，不代表论文的发表时间）