Application of neural networks in Geotechnical Engineering
การประยุกต์ใช้ Neural Networks ในงานวิศวกรรมธรณีเทคนิค
Saravut Jaritngam, D. of Civil Eng., F. of Eng., PSU.
Somchai Chuchom, D. of Industrial Eng., F. of Eng., PSU.
Anthony T. C. Goh, Nanyang Technological U., Singapore
Teh C. I., Nanyang Technological U., Singapore
Kai S. Wong, Nanyang Technological U., Singapore
Corresponding e-mail : jsaravut@ratree.psu.ac.th
Published : Civil Engineering Magazine, The Engineering Institute of Thailand, Vol. 4,
October-December, 1999 : 23-30
Key words : neural networks, braced excavation, piles, Geotechnical Engineers
Geotechnical Engineers often have to solve complex problems involving a number of the interacting factors. This paper demonstrates the use of back-propagation neural networks to solve complex nonlinear problems. An overview of the neural networks methodology is presented. Two practical examples are presented to demonstate the potential of this approach for capturing nonlinear interactions between variables in complex geotechnical engineering systems. The first example de-monstrates the use of the neural network to predict the maximum wall deflection for braced excava-tion systems in clay. The second example relates to the prediction of the ultimate load capacity of piles from stress wave data.
BACK