Predicting the Characteristics of Biofouling Mass Based on RBF Network

Abstract:

A new prediction model of material chemical character effects on biofouling mass was built based on RBF network, in which there were four input vectors, which were carbon content, hydrogen content and oxygen content of the solid materials and flow rate, and one output vectors, which was the average amount of biofouling formed on the solid surface. Firstly, creating the sample database and normalizing all samples. Secondly, training the model based on the training samples to obtain the optimal prediction model, then, predicting the training samples. Comparing with experimental results, the accuracy of the RBF model is 95.5%. Besides, the model was tested by poly (ethylene terephthalate), and the predicted and actual results are consistent. Thus, the construction of the predictive model is reasonable and feasible.

• Periodical: Advanced Materials Research (Volume 818)
• Pages: 246-250
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.818.246
• Online since: September 2013
• Authors: Xiao Qiang Wen
• Keywords: Biofouling, Material Chemical Character, Prediction, RBF Network

References

[1] Heng S Q, Yang G H, Wu L F, et al. Application of Neural Network in Prediction Thermal Resistance of Corrosion Fouling. Total Corrosion control, Vol. 18 (2004), p.11.

[2] Fan S S, Wang H N. Prediction of Fouling in Condenser Based on Multi-Model Combination. Chinese Journal of Sensors and Actuators, Vol. 18 (2005), p.225.

[3] Fan S S. Gray theory based prediction for condenser fouling. Journal of Electric Power Science and Technology, Vol. 22 (2007), p.12.

[4] Zhao B, Yang S R, Liu F. Experimental Study on Dynamic Simulation of Cooling Water Fouling Resistance Prediction by Support Vector Machine. Proceedings of the CSEE, Vol. 30 (2010), p.92.

[5] Zhang Y, Wang H N. Prediction of condenser fouling based on locally weighted partial leastsquares regression algorithm. Chinese Journal of Scientific Instrument, Vol. 31 (2010), p.299.

[6] Peng M F, Shen M, He Y G. Analog Circuit Diagnosis Using RBF Network and D-S Evidential Reasoning. Transactions of China Electro Technical Society, Vol. 24 (2009), p.7.

[7] Jin X. Using RBF to Enable Circuit Emulation Service over internet. Computer Systems and Applications, 2009, p.17.

[8] Yu R H. Study on influence factors of biofouling in cooling water system. Dalian, Dalian University of Technology, (2000).