Akademik Veri Yönetim Sistemi
1. ULUSLARARASI MÜHENDİSLİK BİLİMLERİ VE MULTİDİSİPLİNER YAKLAŞIMLAR KONGRESİ, İstanbul, Türkiye, 23 Şubat 2021, ss.85-97
Abstract Induction motors continue to be the most used electric machine in the industry with their
advantages such as simple internal structures and very low maintenance requirements. Induction motors
are used in many application areas of the industry together with driver systems. In addition to the various
motor control methods used in drivers, sinusoidal PWM inverters have an important place in motor
control. An induction motor fed by a sinusoidal PWM inverter is generally thought to have a constant
load, fixed winding step, and fixed total harmonic distortion. However, during the application, the
switching frequency of the PWM inverter must be adjusted according to the behavior of the Induction
motor at different loads, different harmonizations, and even different current and voltage values. Because in practically all Induction motor applications, optimal operating conditions and high efficiency
are required. In this study, the SPWM switching frequency classification and estimated decision trees
machine learning model were designed considering the operating conditions and structural parameters
of an Induction motor fed with SPWM, and the switching frequency of the SPWM inverter was estimated with minimal error. This work is a unique source of information and application that will contribute to the implementation and development of advanced control systems, where the SPWM switching
frequency can be automatically changed according to the load status and different structural characteristics of the engine.
Keywords: Induction Motors, Decision Trees (DT), Sinuzoidal PWM (SPWM)