Sh.Sh. Ziyaev, N.I. Yunusov, U.Kh. Dzhalolov

Annotation:

This article examines the issue of improving the efficiency of thermal management in a diesel internal combustion engine (ICE) under dynamic loads. The authors identify the shortcomings of conventional PID controllers associated with high inertia and nonlinearity of thermal processes. An adaptive neuro-fuzzy control system (ANFIS) is proposed as a solution, combining the flexibility of fuzzy logic with the learning ability of neural networks. During the study, a mathematical model of the cooling system was developed and an algorithm for tuning membership functions based on a hybrid learning method was synthesized. A comparative analysis of the modeling results showed that the use of a neurofuzzy controller reduces the time to reach the operating temperature and minimizes the amplitude of temperature fluctuations during sudden load changes. The obtained results can be used in the development of advanced microprocessor-based control systems for electronic control units (ECUs) to improve fuel efficiency and extend the service life of power units.

Keywords: diesel ICE, cooling system, neuro-fuzzy logic, ANFIS, adaptive control, temperature stabilization, mathematical modeling.