Home – Home – Jornals – Avtometriya 2026 number 5

This paper addresses the problem of multifactor efficiency analysis of complex technological processes under conditions of physical heterogeneity and high dimensionality of experimental data. The aim of the study is to develop an algorithm for the comparative analysis of operating conditions based on the Data Envelopment Analysis (DEA) method, providing interpretable efficiency assessments and identifying potential directions for process optimization. The catalytic methane pyrolysis process is considered as the object of investigation. An algorithm integrating preliminary statistical analysis with a set of DEA models (CCR, BCC, Super-Efficiency, and SBM) has been developed. The proposed approach enables efficiency evaluation, analysis of scale properties, and quantitative determination of input and output parameter reserves. A decision-making matrix integrating comparative efficiency scores, scale efficiency indicators, and parameter reserves into a unified analytical tool is proposed. The matrix allows ranking of operating conditions, identification of inefficient process regimes, and determination of priority directions for optimization and scaling. The proposed approach was validated using experimental data obtained from an operating methane pyrolysis unit, including 546 measurements collected under 25 technological operating conditions.