Title :
The neural network method of parameter recognition for the horizontal well testing
Author :
Jiang, Meng ; Ze, Hu
Author_Institution :
Sch. of Oil & Gas Eng., Chongqing Univ. of Sci. & Technol., Chongqing, China
Abstract :
The horizontal well is a very complex issue, only two situations considered, which are homogeneous and dual media infinite formation in the paper. According to self-organization, self-learning and adaptive characteristics of neural network, mathematical models built, then BP neural network classifier model constructed in order to recognize horizontal well testing parameters, at last, one example on-site given, has proved that the neural network method of parameter recognition for the horizontal well testing is very significant.
Keywords :
backpropagation; geotechnical engineering; mechanical engineering computing; mechanical testing; neural nets; unsupervised learning; BP neural network classifier model; adaptive characteristics; horizontal well testing parameter; mathematical model; parameter recognition; self-learning characteristics; self-organization characteristics; Artificial neural networks; Mathematical model; Media; Reservoirs; System identification; Testing; Training; horizontal well testing; neural network; parameter recognition;
Conference_Titel :
Business Management and Electronic Information (BMEI), 2011 International Conference on
Print_ISBN :
978-1-61284-108-3
DOI :
10.1109/ICBMEI.2011.5918006
