Perancangan Kontroler Neural Network Untuk Sistem Kontrol Kecepatan Motor BLDC

Percival, David (2024) Perancangan Kontroler Neural Network Untuk Sistem Kontrol Kecepatan Motor BLDC. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022201242-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (2MB) | Request a copy

Abstract

Penelitian ini bertujuan merancang kontroler neural network untuk mengoptimalkan kecepatan motor BLDC dan mengetahui performa motor BLDC yang di kontrol dengan kontroler neural network dengan kecepatan motor BLDC ideal. Neural network muncul sebagai solusi yang potensial. Penelitian ini memfokuskan pada analisis perbandingan antara kontroler Neural Network dan metode tradisional dalam konteks pengaturan kecepatan motor DC.
Metode yang digunakan pada penelitian ini menggunakan pelatihan neural network pada matlab dan membandingkan hasil kecepatan motor BLDC yang dikontrol dengan neural network. Parameter yang di kontrol pada motor ini adalah, eponch, learning rate, validasi yang dilakukan, dan performance metric yang digunakan mean square error.
Pada penelitian ini didapatkan hasil simulasi berupa kecepatan motor BLDC yang dikontrol dengan neural network memiliki kecepatan 0.8 ms tanpa overshoot dan kecepatan motor BLDC ideal tercapai dalam 0.2 ms tanpa overshoot. Terdapat perbedaan sebesar 0.6 ms antara kecepatan motor BLDC dengan kontroler neural network dan kecepatan motor BLDC ideal. jika dibandingkan dengan pengontrolan menggunakan PID didapatkan hasil sebesar 1.5 sekon dan overshoot sebesar 0.213%. dan dapat disimpulkan pengontrolan menggunakan neural network lebih baik dari pada PID.

================================================================

This research aims to design a neural network controller to optimize BLDC motor speed and determine the performance of a BLDC motor controlled with a neural network controller with ideal BLDC motor speed. Neural networks emerged as a potential solution. This research focuses on comparative analysis between Neural Network controllers and traditional methods in the context of DC motor speed regulation.
The method used in this research uses neural network training in Matlab and compares the speed results of BLDC motors controlled with a neural network. The parameters controlled on this motorbike are eponch, learning rate, validation carried out, and the performance metric used is mean square error.
In this research, the results showed that the speed of the BLDC motor controlled using a neural network had a speed of 0.8 ms without overshoot and the ideal BLDC motor speed was achieved in 0.2 ms without overshoot. There is a difference of 0.6 ms between the speed of a BLDC motor with a neural network controller and the ideal BLDC motor speed. When compared with control using PID, the results were 1.5 seconds and an overshoot of 0.213%. and it can be concluded that control using a neural network is better than PID.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Neural Network, Brushless DC Motor, Controller, Neural Network, Brushless DC Motor, Kontroler
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2681.O85 Electric motors, Brushless.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Percival David
Date Deposited:	31 Jul 2024 08:14
Last Modified:	31 Jul 2024 08:14
URI:	http://repository.its.ac.id/id/eprint/110664

Actions (login required)

View Item