| Search / Korean Journal of Chemical Engineering |
HWAHAK KONGHAK,
Vol.39, No.3, 361-367, 2001
Vol.39, No.3, 361-367, 2001
오류역전파 신경망 이론을 이용한 UV/TiO2/H2O2 시스템의 제거 효율 예측
Prediction of Removal Efficiency in UV/TiO2/H2O2 System using Error Back-propagation Neural Network
오류역전파 신경망을 광촉매 현탁방식 UV/TiO2/H2O2 시스템의 운전 변수인 포름산의 초기농도, 산화제의 투입량, 촉매량 및 반응용액의 순환유속에 대하여 학습시켜 시스템의 포름산 제거 효율 예측 모델을 구성하였다. 또한 구성된 모델로 다양한 조건에서 UV/TiO2/H2O2 시스템의 포름산 제거 효율을 예측하여 실험결과와 비교하였다. 4개의 층으로 구성된 오류역전파 신경망은 학습상수(η)와 모멘텀(α)이 각각 0.9 및 0.7인 경우 시스템의 포름산 제거 효율을 가장 정확히 학습하였으며, 포름산의 초기 농도 및 산화제 투입량, 촉매량의 변화에 따른 제거 효율의 변화를 정확히 예측하였다. 이로부터 오류역전파 신경망이 UV/TiO2/H2O2 시스템의 제거 효율을 예측하기 위한 모델로 이용가능함을 알 수 있었다.
A prediction model of removal efficiency in UV/TiO2/H2O2 system was constructed using the error back propagation neural network with generalized delta rule. The network was trained to produce removal efficiency according to initial concentrations of formic acid, dosages of oxidant and concentrations of catalyst. The reliability of removal efficiency prediction using the constructed neural network was verified by comparing the experimental data of the system for various conditions. The error back propagation network with four layer learned exactly the removal efficiency curve of the system when step size coefficient(η) was 0.9 and momentum(α) was 0.7. It offered reasonable prediction of removal efficiency curve for the various initial concentrations of formic acid, dosages of oxidant and concentrations of catalyst. It was found that the error back propagation neural network is the proper model to predict the removal efficiency of UV/TiO2/H2O2 system.
[References]
- Pruden AL, Ollis DF, J. Catal., 82, 404, 1983
- Kraeutler B, Bard AJ, J. Am. Chem. Soc., 100(19), 5985, 1978
- Matthews RW, J. Chem. Soc.-Faraday Trans., 80, 457, 1984
- Piccinini P, Minero C, Vincenti M, Pelizzetti E, Catal. Today, 39(3), 187, 1997
- Augugliaro V, Loddo V, Marci G, Palmisano L, Lopezmunoz MJ, J. Catal., 166(2), 272, 1997
- Maurino V, Minero C, Pelizzetti E, Piccinini P, Serpone N, Hidaka H, J. Photochem. Photobiol. A-Chem., 109, 177, 1997
- Sabate J, Anderson MA, Kikkawa H, Edwards M, Hill CG, J. Catal., 127, 167, 1991
- Nogueira RFP, Jardim WF, Solar Energy, 56(5), 471, 1996
- Tennakone K, Tilakaratne CTK, Kottegoda IRM, J. Photochem. Photobiol. A-Chem., 87, 177, 1995
- Byrne JA, Eggins BR, Brown NMD, McKinney B, Rouse M, Appl. Catal. B: Environ., 17(1-2), 25, 1998
- Sirisuk A, Hill CG, Anderson MA, Catal. Today, 54(1), 159, 1999
- Butters, et al., U.S. Patent, 5,462,674, 1995
- Butters, et al., U.S. Patent, 5,554,300, 1996
- Butters, et al., U.S. Patent, 5,589,078, 1996
- Wasserman PD, "Neural Computing(Theory and Practice)," Van Nostrand Reinhold, NY, 1989
- Khanna T, "Foundations of Neural Networks," Addison-Wesley, NY, 1990
- Caudill M, Butler C, "Understanding Neural Network(computer explorations)," MIT Press, 1992
- Ungar LH, Powell BA, Kamens SN, Comput. Chem. Eng., 14, 561, 1990
- Bhat N, McAvoy TJ, Comput. Chem. Eng., 14, 573, 1990
- Ydstie BE, Comput. Chem. Eng., 14, 583, 1990
- Psichogios DC, Ungar LH, Ind. Eng. Chem. Res., 30, 2564, 1990
- Lee WJ, Kim SJ, Shul YG, Lee TH, HWAHAK KONGHAK, 33(5), 610, 1995
- Himmelblau DM, Korean J. Chem. Eng., 17(4), 373, 2000
- Sohn SH, Oh SC, Yeo YK, Korean J. Chem. Eng., 16(3), 382, 1999
- Matthews RW, Water Res., 20, 569, 1986
- Linsebigler AL, Lu GQ, Yates JT, Chem. Rev., 95(3), 735, 1995
- Sakata T, "Photocatalysis: Fundamentals and Applications," ed. Serpone, N. and Pelizzetti, E., John Wiley & Sons, Inc., New York, 311, 1989
- Turchi CS, Ollis DF, J. Catal., 122, 178, 1990
- Widrow B, Stearns S, "Adaptive Signal Processing," Prentice-hall, Englewood Liffs, NJ, 1985
- Leonard JA, Kramer MA, Comput. Chem. Eng., 14, 337, 1990
- Rumelhart DE, Hilton GE, Williams RJ, "Parallel Distributed Processing," Cambridge, MA, 1986
참고문헌정보(참고문헌, 인용문헌)는 화학공학소재연구정보센터에 의해 제공되고 있습니다.