The problem of increasing the intelligence and efficiency of the TPP automated control system on the basis of modern software and hardware complexes by controlling the current energy indicators (CEI) of each unit, unit and station as a whole is considered. The main causes of the discrepancy between the current and regulatory CEI are identified and recommendations are given for their prompt elimination. It is shown that the existing algorithms for calculating the CEI are limited to fixing the CEI values with a significant error due to the error of the initial information and the impossibility of combining the forward and reverse balances of individual units, the unit and the station with each other. To increase the reliability of the initial information, it is proposed to use an intelligent algorithm to obtain estimates of the values of technological parameters based on measurement results using a system of material and thermal balance equations. An indicator of the reliability of estimates in the proposed algorithm using neural technologies is the fulfillment of material and energy balances. The description of the algorithm for selecting a variable operational interval for calculating the CEI and the methodology for developing control actions to eliminate the difference between the current and normative CEI are given.