Industrial Water Requirement Forecast Model and Application Based on PCA-BP Neutral Network

Abstract:

ndustrial water requirement is usually influenced by multiple factors, among which have close relations, which can cause distortion of some forecasting results, and weaken the suitability of some formulas. To this end, we adopt BP neural network model, use principal component analysis (PCA) to analyze the relationships between variables of the model to solve the related issues between the various factors, in order to set up forecast model of industrial water requirement and take Zhengzhou City as an example to analyze. The results showed that this method could take the influenced factors of industrial water requirement into consideration comprehensively, offer higher and more accurate forecasts, and provide the reference framework for integrated planning of regional water resources and long-term water supply planning.

• Periodical: Applied Mechanics and Materials (Volumes 325-326)
• Pages: 1479-1482
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.325-326.1479
• Online since: June 2013
• Authors: Han Xiao, Jin Ling Hu, Guo Li Wang, Ben De Wang
• Keywords: BP Neural Network (BPNN), Principal Component Analysis (PCA), Water for Industrial Uses

References

[1] He Gang, Wu Zening and Hu Caihong. Industrial Water Requirement Forecast Model Based on Quota and Quantitative analysis: Journal of Water Resources and Water Engineering, 2008. 19(2): 60-62.

[2] Long Xunjian, Qian Ju and Liang Chuan. BP Neutral Network based on PCA and Its Applicationof Water Requirement Forecast: Journal of Chengdu University of Technology, 2010. 37(2): 206-210.

[3] Yu Jianying, He Xuhong, in: Data Statistical Analysis and the Application of SPSS, chapter, 9, 291-310, the People's Postal & Telecommunications Press (2003).

[4] Yang Jiangang, in: Artificial Neural Networks Practical Course, chapter, 3, 44-45, Zhejiang University Press (2003).