NORMALIZATION OF SPECIFIC CONSUMPTION OF ENERGY IN THE INDUSTRY WITH USED NEURO-FUZZY MODELING

Science-based standards provide specific energy consumption benchmark for determining the level of energy pro-duction, and normative method of controlling the power is one of the most effective and versatile. At the same time, the pace of development of new technologies, market fluctuations and other destabilizing factors stipulate for the dynamic changes in the standards, respectively, methods of analysis should provide results in "real time". The comparative study of "classical" methods of valuation rate costs of electricity on the basis of the correlation-regression analysis with a neuro-fuzzy modelling was conducted in the paper. The promising use of this approach for modeling the power consumption, which is caused by increasing the accuracy of the calculation and adaptation speed at the expense of self-learning was proved. Thus, the implementation of multi neuro-fuzzy model based on the most influential factors is provided by deviation of the calculated data from the experimental at the level of ± 5%. The conditions for minimizing the specific energy consumption in a wide range of destabilizing factors of influence, as well as optimal modes of operation of the enterprise from the point of view of energy saving were determined.