{"id":247949,"date":"2026-09-11T12:22:45","date_gmt":"2026-09-11T09:22:45","guid":{"rendered":"https:\/\/azbuki.bg\/?p=247949"},"modified":"2026-09-11T12:22:45","modified_gmt":"2026-09-11T09:22:45","slug":"study-of-brushless-motor-control-methods-based-on-ai-algorithms","status":"publish","type":"post","link":"https:\/\/azbuki.bg\/en\/xxvii-international-scientific-conference-transport-2025\/study-of-brushless-motor-control-methods-based-on-ai-algorithms\/","title":{"rendered":"Study of Brushless Motor Control Methods Based on ai Algorithms"},"content":{"rendered":"

Roman Petrov, Nikita Yakovlev
\n<\/strong>Polytechnical institute, Yaroslav-the-Wise Novgorod State University, Veliky Novgorod, <\/em>Russia<\/em><\/p>\n

Szabolcs Kocsis
\n<\/strong>Sz\u00e9chenyi Istv\u00e1n University, Gy\u0151r, <\/em>Hungary<\/em><\/p>\n

Slavcho Bozhkov, Nencho Nenov, Ivan Milenov, Marian Mutafchiev,
\n<\/strong>Krasimir Krastanov, Daniela Todorova, Miryana Evtimova<\/strong><\/p>\n

Todor Kableshkov University of Transport, Sofia, Bulgaria<\/em><\/p>\n

https:\/\/doi.org\/10.53656\/isct-2025.15<\/a><\/p>\n

\"\"
\nPDF<\/a><\/p>\n

Pages 193-203<\/em><\/p>\n

Abstract.<\/strong> The paper is devoted to the study of controllers based on artificial intelligence: a fuzzy logic controller (FLC), an artificial neural network (ANN) based on a multilayer perceptron, and an adaptive neuro-fuzzy inference system (ANFIS). Traditional PID controllers demonstrate limited efficiency in controlling brushless motors under conditions of variable loads and disturbances due to the inability to adapt to the nonlinearities of the systems.\u00a0 A comprehensive DC motor model was developed in the MATLAB\/Simulink environment and tested under dynamic conditions (acceleration, deceleration, variable load). Speed\/torque overshoot metrics are used, as well as the time of setting. The results show that ANFIS provides excellent accuracy (overspeed is 45% lower compared to PID), but suffers from slower reactions. ANN provides an optimal balance by reducing overspeed by 7% and regulation time by 35% compared to PID, while maintaining real-time applicability. The FLC has excellent transient performance, but exhibits torque fluctuations. The study confirms that controllers with artificial intelligence significantly increase adaptability to disturbances and non-linearities, and ANNs are recommended for real-time systems (robotics, drones), and ANFIS for precision applications. The results provide engineers with some recommendations on the choice of control algorithms for modern electric drives that require resistance to disturbances and energy efficiency.<\/p>\n

Keywords: brushless motors; neural networks; fuzzy logic; hybrid systems; intelligent controllers; energy harvesting devices based on multiferroic materials<\/p>\n

 <\/p>\n

    \n
  1. Introduction<\/strong><\/li>\n<\/ol>\n

    The paper focuses on the application of artificial intelligence techniques, such as neural networks and fuzzy logic, to control brushless direct current (BLDC) motors. These techniques are expected to significantly improve the performance of these motors compared to traditional proportional-integral-derivative (PID) controllers.<\/p>\n

    In the era of rapidly developing information technologies, artificial intelligence (AI) is penetrating into a wide variety of industries, becoming a new stage of development in technology. One of these promising areas is engine control, something that no moving device can do without, whether it’s cars, airplanes, industrial installations, or energy systems.<\/p>\n

    The introduction of algorithms opens up new opportunities. The use of artificial intelligence algorithms (neural networks, fuzzy logic, hybrid systems) to control BLDC engines allows for the implementation of an intelligent speed controller (AI-ESC) capable of analyzing data in real time, adapting to real-world operating conditions, predicting failures, and significantly improving the key performance of the controlled device.<\/p>\n

    The authors of the paper [1] conclude that PID controller is a simple controller with a simple tuning method but with a moderate response and performance. Fuzzy controller is more complicated controller but with a good and more stable performance. Euro fuzzy controller is very complicated controller but with a very good performance.<\/p>\n

    In study [2] authors have compared the use of an intelligence controller on the BLDC motor using data extraction from one of the intelligence controllers. The test was carried out by two scenarios, and analyzing the results of the system response was carried out by graph and some statistical parameters such as MAE, IAE, ITAE, and ISE. We found in scenario 1, ANN has a better performance compared to other controllers with the MAE, IAE, ITAE, and ISE value of 31.3003; 105.6280; 208.0630; and 5,7289 e4, respectively.<\/p>\n

    In paper [3] authors presents an ANFIS controller for speed control of BLDC motor along with comparisons with other controllers, PID and Fuzzy controllers. ANFIS controller showed the best results.<\/p>\n

    The performance analysis of BLDC motor under transient and steady-state with the two different controllers such as PI and ANFIS is studied in work [4] by using the MATLAB\/Simulink 2021a version. From the results, it is clear that the ANFIS controller outperforms PI at all speeds and under various operating conditions.<\/p>\n

    Classical systems work based on precise mathematical models and linear algorithms. Such systems require fine tuning, are unable to account for external reimbursements, predict failures, or adapt to changes without human intervention.<\/p>\n

    The use of artificial intelligence algorithms (in particular, neural networks, fuzzy logic and their hybrids) to control brushless DC motors (BLDC) makes it possible to dynamically adjust the parameters of the brushless motor control system to changing environmental conditions more efficiently than the PID control method.<\/p>\n

    The article provides a comparative analysis of the effectiveness of artificial intelligence algorithms (neural networks, fuzzy logic, neuro-fuzzy systems) for controlling the speed of a brushless DC motor (BLDC) in order to confirm their advantages over classical approaches.<\/p>\n

     <\/p>\n

      \n
    1. Studied ai algorithms<\/strong><\/li>\n<\/ol>\n

      The following algorithms were chosen for study:<\/p>\n