On the basis of the gas compressor units of compressor plants of a main gas pipeline mathematical models of the macro-level were generated for analysis and parametric optimization of combined energy-and-technology units. In continuation of the study these models was applied to obtain the regression dependencies. For this purpose, a numerical experiment was used which had been designed with the use of regression analysis mathematical tool, which assumes that the test results should represent independent, normally distributed, random variables with approximately equal variance. Herewith we study the dependence of the optimization criterion on the value of control parameters (factors). Planning, conducting and processing of results of the experiment was conducted in the following sequence: choice of the optimization criteria, selection of control parameters (factors), encoding factors, the matrix of experiment compiling, assessing significance of regression coefficients, testing the adequacy of the model and reproducibility of the experiments. As the optimization criteria the electricity capacity and efficiency of combined energy-technology units were adopted. As control parameters for the installation with a gas-expansion-and-generator machine the temperature of the fuel gas before the expander, the pressure of the fuel gas after the expander and the temperature of the air supplied to the compressor of the engine were adopted, while for the steam turbine the adopted optimization criteria were compression in the compressor of the engine, the steam consumption for the technology and the temperature of the air supplied to the compressor of the engine. The application of the outlined methodological approach makes it possible to obtain a simple polynomial dependence, which significantly simplify the procedures of analysis, parametric optimization and evaluation of efficiency in the feasibility studies of the options of construction of the energy facilities.