Development of neural network for control production process in oil and gas fields

For a long time there has been a tendency to increase field productivity, therefore, increasing oil recovery is the main task for fuel and energy complex. Currently, neural networks are increasingly used in various industries. The advantage of neural networks is to work with a large amount of data, however, it must have sufficient data sets collected and prepared for its operation, thereby achieving high decision accuracy. When developing oil and gas fields, the main task is to ensure maximum production from an economic and physical point of view. Oil production at oil and gas fields varies in volume, complexity, operating conditions, etc., therefore, it is necessary to find the optimal production conditions for each field. At the moment, the main problems in oil production at oil and gas fields are: the long processing time of data collected from wells, the increased risks of operating these wells, as well as the low amount of oil produced. The main objective of this study is to develop a control method us in artificial intelligence to control the production process in oil and gas fields, taking in to account all factors, in order to maximize oil production. In the course of this study, direct transmission to the neural network was obtained, which allows oil to be extracted at oil and gas fields, taking into account all factors. The resulting neural network, without reconfiguring weighted connections, generates output signals when applied to the input to the network. The resulting neural network expresses the patterns that are present in the input data. This network turns out to be the functional equivalent of some model of dependencies between variables.