The use of multi-agent systems for the short-term forecasting of power consumption of oil and gas industry enterprises

The electrotechnical complexes of the oil and gas industry are considered critical infrastructure, which requires compliance with energy saving and energy efficiency regulations. Due to the complexity of technical processes, oil and gas industry enterprises implement only technical measures requiring capital investment. The possibility of implementing an organizational measure which consists of forecasting power consumption of oil and gas industry enterprises is considered in this article. The article presents an example demonstrating the problems of forecasting such power consumption, namely the low frequency of the load curve and its high dispersion. In order to solve these problems, the possibility of using multi-agent systems to increase energy efficiency through short-term forecasting of power consumption by enterprises of the oil and gas industry is considered. The consumer agents (considering regulatory consumers), generation agents, and storage agents for building multi-agent systems are proposed. These agents address not only their own but also the overall target function of the system, to ensure a power balance and reduce the cost of power consumption. The possibility of reducing the expenses of the enterprise by forecasting the power consumption by consumer agents, finding the optimal schedule for the generation and accumulation of energy, and the inclusion of consumer-regulators is considered. The target function and input and output data were defined for each agent.