Rezumat
CZU 621.785.6.048.7:004.89
DOI https://doi.org/10.52577/eom.2026.62.4.82
This study is devoted to the digital design of the high-frequency induction surface hardening processes for the internal surfaces of У8-6МА2 pump cylinders using artificial intelligence (AI) methods. The focus is placed on the development of an intelligent model that combines machine learning algorithms and fuzzy logic for optimizing heating and cooling regimes. The proposed approach enables the consideration of nonlinear relationships between process parameters (frequency, temperature, inductor travel speed, and cooling intensity) and the target material properties. The modeling results demonstrate a significant improvement in the mechanical and tribological properties of the cylinder liner: surface hardness increased to 48–52 HRC, wear coefficient reduced by 30–35%, and operational lifetime of the equipment extended. The integration of digital design with AI ensures the selection of optimal HFС parameters with minimal energy consumption and an environmental impact. The scientific novelty of the work lies in the integration of AI methods into conventional induction hardening processes, opening prospects for the creation of intelligent control systems to enhance the reliability and durability of oil and gas pumping equipment.
Keywords: digital design, high-frequency induction, artificial intelligence, fuzzy logic, machine learning, У8-6МА2 pump, cylinder liner, mechanical properties, tribological properties, surface hardening.