| Abstract [eng] |
This paper examines the method of using electricity market futures for cost optimization and practical application in specific areas. The aim of the paper is to create a prototype that, using machine learning algorithms, would help predict electricity prices and demand and conclude electricity futures contracts, thus reducing the costs of electricity consumption. The paper first analyzes the liberalized electricity market, its structure and new technologies which were developed in regards by liberalized market. The liberalized electricity market opens the way for a wider use of electricity futures contracts. It has been established that financial instruments allow reducing the risk of electricity price fluctuations and facilitate cost planning. It has also been established that supervised machine learning algorithms are most suitable for price forecasting and concluding futures contracts. After properly training these algorithms with historical data, it is possible to predict electricity prices and costs in the short term; the longer the period, the greater the error. The developed prototype predicts electricity prices, estimates future electricity demand and creates future electricity contracts if they are predicted to bring profit. The functionality of the prototype is tested in two areas: electric vehicle charging station networks and irrigation systems. The study shows that the prototype is able to reduce electricity costs and improve resource planning. The results of the work show that the prototype could be applied in other areas where it is important to properly manage electricity costs. The liberalized electricity market and the availability of futures contracts for a wider part of the market can help create more sustainable solutions in various areas, achieving energy efficiency and economic stability. |
