Multiphase flow monitoring in oil pipelines
Neural networks, and related statistical pattern recognition techniques, appear to be well suited to the solution of a wide range of monitoring and diagnostic problems. In many applications, it is difficult or impossible to perform first-principles modelling of the system under consideration. If, however, sufficiently large quantities of labelled training data can be made available, then a statistical approach becomes feasible.
KeywordsMultiphase Flow Phase Fraction Beam Line Multilayer Perceptron Hide Unit
Unable to display preview. Download preview PDF.
- K Rafa, T Tomoda and R Ridley Proc. Energy Sources Technology Conference and Exhibition, ASME (1989) 89-Pet-7.Google Scholar
- M S Beck and A Plaskowski, Cross Correlation Flowmeters, (Adam Hilger, Bristol, 1987).Google Scholar
- JS Watt, HW Zastawny, MD Rebgetz, PE Hartley and WK Ellis, Nuclear Techniques in the Exploration and Exploitation of Energy and Mineral Resources, (IAEA, Vienna, 1991) p. 481.Google Scholar
- CM Bishop, invited review article: Neural Networks and their Applications to be published in Reviews of Scientific Instruments (1993).Google Scholar
- CM Bishop, Neural Networks for Pattern Recognition, (Oxford University Press, 1994).Google Scholar
- C M Bishop ‘Curvature Driven Smoothing: A Learning Algorithm for Feedforward Networks’, to be published in IEEE Transactions on Neural Networks (1993).Google Scholar
- A R Webb (1991) ‘Functional Approximation by Feedforward Networks: A Least-squares Approach to Generalisation’ RSRE Memorandum 4453, R.S.R.E., St. Andrews Road, Malvern, Worcs, WR14 3PS, U.K.Google Scholar