Features of data collection and processing for construction of probable-statistical models of energy consumption

The article is devoted to the collection and processing of statistical data, which are the basis for constructing probability-statistical models of energy consumption. A list of the minimum amount of data is given, which includes statistics of annual energy costs by expense items; Volumes of production activity; List of implemented and planned energy-saving measures. The article presents the methods of data recovery in the presence of gaps and methods to reduce the size of primary information. To reduce the amount of data, a method for selecting the indicators of the most significant group of factors affecting the energy efficiency of production, and a method for reducing the size of data in the context of sectoral forecasting, by selecting the most energy-intensive industrial enterprises is viewed.