LSTM-ANN based DTC of Induction Motor with 5 Level Inverter

Sat endra Singh, Assistant Professor, Mechanical Engineering, Vivekananda Global University, Email Id-singh.satendra@vgu.ac.in, Jaipur, India
A Kannagi, Associate Professor, Department of Computer Science and Information Technology, Jain (Deemed to be University), Email Ida. kannagi@jainuniversity.ac.in, Bangalore, India
Ms.LaxmiGoswami, Assistant Professor, Department of Electrical Engineering, Sanskriti University, Email Id- laxmigoswami.ee@sanskriti.edu.in, Mathura, Uttar Pradesh, India
Dhananjay Kumar Yadav, Assistant Professor, Maharishi School of Engineering & Technology, Maharishi University of Information Technology, Email Id- dhananjay.rishabh@gmail.com, Uttar Pradesh, India

^* Corresponding Author: singh.satendra@vgu.ac.in

Abstract

Medium voltage drives especially induction motor drives are mostly using in industries for many applications. The precious intelligent control of induction motors are high required in nuclear power generation units. Direct torque control (DTC) of induction motors is commonly selected for achieving smooth speed torque characteristics as compared with other existing control techniques. However, conventional DTC suffers from many issues because of maintaining constant reference flux in all the speed regions. Hence two ANN controllers based novel DTC is implemented in this paper. Three phase five level neutral point clamped (5L-NPC) inverter is used to drive the induction motor. The dc-link is established through 36 pulse rectifier. Moreover, photovoltaic (PV) system along with battery bank is also integrated to provide continuous power supply during off grid. An unchanging dc-link voltage is achieved by utilizing a bidirectional dc-dc circuit that is connected between the dc-link and the battery. The proposed control of dc-dc bidirectional circuit can further reduce the ripples and oscillations in dc-link produced by 36 pulse rectifier during unbalanced grid voltages. Realistic responses are presented in this paper by establishing hardware – in the – loop (HIL) with the help of OPAL-RT modules. The HIL based results are discussed under various case studies in this paper.