APPLICATION OF NEURAL NETWORK MODELS TO DETERMINE THE FLOW RATE OF COMPONENTS OF A MULTIPHASE FLOW FROM A WELL
K.I. Budnikov
Institute of Automation and Electrometry, Siberian Branch, Russian Academy of Sciences, Novosibirsk, Russia
Keywords: oil wells, multiphase flow meters, machine learning, neural networks
Abstract
Due to its high productivity and demand for products, the oil industry occupies a special place in the modern industry. Increased attention is paid to monitoring its condition and the performance of the exploited areas. Exploration of new fields and drilling of producing wells require a lot of time and financial costs. To reduce them, advanced technologies are currently widely used, including various artificial intelligence methods, for example, to assess the percentage of fractions in the oil flow from the well. The paper presents the results of studying a number of models based on neural networks of different architectures for predicting the flow rate of components of a multiphase flow from a well. To conduct the study, data received from sensors of multiphase flow meters based on a Venturi pipe and an X-ray flow meter are used.
|
