The article discusses ways to solve engineering problems in the study of technological processes using methods of system analysis. The essence of this method is to study the technology as a cybernetic system with an assessment of the" reactions” of this system to external influences formed during an active experiment. At the same time, optimization problems are solved analytically. Analytical optimization is based on two main principles. The regression equations obtained as a result of processing experimental data and testing statistical hypotheses are models that adequately describe real processes. Each of these equations is an algebraic function of several variables, to which methods of mathematical analysis are applicable, including the study of extremums of functions in partial derivatives. The next step is to develop a process algorithm and develop computer programs that allow you to select the composition and predict the properties of the product. As an engineering interpretation, it is possible to construct optimized nomograms that allow solving both direct and inverse problems; that is, predicting the result or selecting technological factors. The research methods described in the article are implemented in the study of technologies of cellular concrete, foam concrete, cement-polymer concrete and products made of mineral wool and foam glass. As an example, the article considers the optimization of the selection of the composition of fine-grained concrete reinforced with chopped glass fiber. The implementation of the developed method allowed us to determine the optimal value of the determining parameters, including the consumption of fiber and plasticizer, as well as to form a method for studying the properties of products.