Development of direct torque control with neural network for speed change in 3-phase induction motor

• Main Author: Mohammad Gunawan, Muhammad Faqih Muhsin
• Format: Thesis
• Language: English; English; English
• Published: 2019
• Subjects: TK5101-6720 Telecommunication. Including telegraphy, telephone, radio, radar, television
• Online Access: http://eprints.uthm.edu.my/577/1/24p%20MUHAMMAD%20FAQIH%20MUHSIN%20MOHAMMAD%20GUNAWAN.pdf; http://eprints.uthm.edu.my/577/2/MUHAMMAD%20FAQIH%20MUHSIN%20MOHAMMAD%20GUNAWAN%20COPYRIGHT%20DECLARATION.pdf; http://eprints.uthm.edu.my/577/3/MUHAMMAD%20FAQIH%20MUHSIN%20MOHAMMAD%20GUNAWAN%20WATERMARK.pdf; http://eprints.uthm.edu.my/577/

This project proposes a technique to develop an improved direct torque control (DTC) with neural network in order to control the speed of the three-phase induction motor. This project will developed using MATLAB software to simulate the outcome for the module designed. The NN controller have to be designed and tuned properly. This project focus on the simulation of modelling the three-phase induction motor, implementing the neural network control and comparing the proposed model to the conventional DTC. This researched proposed an ability to control a three-phase induction motor and estimates the speed of motor. NN controller is being used and implemented throughout this project to control the speed of an induction motor. This DTC control works if there is a changes in loads, the inverter switch for the motor will changes as well. Thus, a NN structure has been designed to improve the conventional DTC system. This system is being tested with load disturbance of 0.2Nm and 0.8Nm to verify its effectiveness of NN controller compared to conventional DTC controller. Previously, conventional control system used PID controller for DTC. This project proposed an improvement for this type of DTC control by using NN. The NN controller improves the torque response as well as speed response for NN for 13ms faster compared to PID-DTC.