A New Direct Torque Control of an Efficient and Cost-Effective Traction System Using Two Squirrel Cage Induction Motors Feed by a Single Inverter

Abstract

Reducing the financial cost of controlling electrical machinery is a major factor in industrial systems. The research is based on controlling dual squirrel cage induction motors (SCIMs) using direct torque command (DTC) based on neural algorithms (NAs). The proposed neural DTC (NDTC), developed from DTC and NAs, where NAs have a lot of outstanding advantages in many aspects, such as accuracy, robustness, and dynamic performance. However, the NDTC-based dual-SCIM system is a simple, robust technique, reliable and highly efficient system with a low cost of implementation. In this work, Matlab software was used to study the NDTC behavior of dual SCIMs by comparing the results obtained with DTC results in terms of response time, current quality, torque ripples, overshoot, … etc. Also, the performance of the NDTC is studied in case the parameters of the system change compared to DTC performance. The obtained results show the efficiency and performance of the NDTC of the dual SCIMs, and that this system is a practical alternative to controlling a group of motors in industrial systems. The ratio ripples between both strategies are 72.97%, 48.38% and 66.11% for current, flux, and torque. This means that the NDTC can be a good candidate for controlling machines.