ANALYSIS OF INPUT PARAMETERS FOR EVALUATION OF THE LEVEL OF THE EXTERNAL MACRO-INFORMATION ENVIRONMENT OF OIL AND GAS EXTRACTION FACTORIES BY USING NEURONETWORKS

Abstract

The article highlights the opportunities of applying mathematical models, namely, neural network technologies to analyze the level of the external macro-informational environment of oil and gas production factories. In the conditions of a dynamic natural and socio-economic environment, the activity of oil and gas production factories requires the creation of balanced systems of organizational management of both production and information processes at the enterprise. A significant amount of information, the presence of complex or uncertain dependencies between it and data streams circulating in the factory, is a precondition for the use of neural network methods for the analysis of information for the activities of oil and gas production enterprises and the use of the obtained results for the improvement of economic activity. The main stages of the process of building a neural network are considered. The first stage of creating a neural network involves the formation of a system of input data. Based on the model of the information environment of the oil and gas production enterprise, a system of information characteristics of the external macro information environment of the oil and gas production factory was formed and systematized, which includes: information about the state of resource provision with explored oil and gas reserves and information on natural and climatic conditions; information about the price situation for oil and gas on world markets; information about the state of the world economy and markets; information about innovation and globalization processes; information about global political, environmental, demographic and socio-cultural processes in countries where oil and gas production enterprises may operate. This system of characteristics is justified by their functional purpose, reflects the specifics of oil and gas production enterprises and was formed on the basis of the analysis of literary sources and scientific works in this direction. A coded value is set for each information characteristic, depending on the range of normative values and the direction of change in dynamics. The formed system of characteristics is ready for neural network training, which is the next stage of research.