Pengaturan Kecepatan Motor Induksi Tiga Phasa Menggunakan Metode Direct Torque Control Berbasis Artificial Neural Network Pada Mobil Listrik

Nasir, Andi Fadel Muhammad (2025) Pengaturan Kecepatan Motor Induksi Tiga Phasa Menggunakan Metode Direct Torque Control Berbasis Artificial Neural Network Pada Mobil Listrik. Diploma thesis, Institut Teknologi Sepuluh Nopember.

Text 5022201247-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Mobil listrik merupakan solusi strategis dalam mengurangi ketergantungan terhadap bahan bakar fosil dan dampak lingkungan. Salah satu tantangan utama dalam pengembangannya adalah pengaturan kecepatan motor induksi tiga fasa secara presisi dan adaptif. Penelitian ini mengusulkan sistem kendali kecepatan berbasis Direct Torque Control (DTC) yang dikombinasikan dengan Artificial Neural Network (ANN), serta diintegrasikan dengan metode kontrol Proportional-Integral-Derivative (PID). Pada pengujian tanpa beban, kontrol PID-ANN menunjukkan waktu pencapaian steady state yang sangat cepat (0,0246–0,1 detik) tanpa overshoot, sementara PID-SMC membutuhkan waktu lebih lama (0,6–0,7 detik), namun memberikan akurasi lebih tinggi dengan nilai RMSE lebih rendah (contoh pada 1460 RPM: PID-ANN = 2,9425; PID-SMC = 0,0323). Pada pengujian berbeban, PID-ANN tetap lebih cepat (0,04–0,3 detik), sementara PID-SMC menunjukkan kestabilan lebih tinggi (0,45–0,7 detik) dengan error dan RMSE lebih kecil (misalnya, 1168 RPM: PID-ANN = 2,2013; PID-SMC = 0,0373). Pada simulasi kendaraan mulai melaju, PID-ANN mencapai steady state dalam 0,036 detik dengan RMSE = 0,705 RPM, sedangkan PID-SMC memerlukan 0,520 detik, namun dengan SSE = 0,0034% dan RMSE = 0,7279 RPM. Dalam simulasi variasi kontur jalan (menanjak, datar, menurun), PID-ANN menunjukkan respon adaptif lebih cepat (misalnya, steady state pada 0,501 detik, ESS = 0,6 RPM, RMSE = 0,3609 RPM), sedangkan PID-SMC membutuhkan waktu lebih lama (0,71 detik), namun lebih unggul dalam akurasi jangka panjang (ESS = 0,001 RPM, RMSE = 0,2733 RPM). Hasil ini membuktikan bahwa DTC berbasis ANN mampu meningkatkan performa sistem penggerak motor induksi, dengan keunggulan ANN pada respon cepat, performa yang adaptif dan responsif dalam pengaturan kecepatan motor induksi, terutama pada kondisi dinamis, dan dapat menjadi solusi efektif untuk sistem penggerak mobil listrik.
==========================================================================================================================================
Electric vehicles represent a strategic solution to reduce fossil fuel dependency and mitigate environmental impact. One of the major challenges in their development is achieving precise and adaptive speed regulation of three-phase induction motors. This study proposes a speed control system based on Direct Torque Control (DTC) integrated with an Artificial Neural Network (ANN) and enhanced by a Proportional-Integral-Derivative (PID) control scheme. Under no-load conditions, the PID-ANN controller achieved a significantly faster steady-state response (0.0246–0.1 seconds) without overshoot, whereas the PID-SMC controller required a longer duration (0.6–0.7 seconds) but yielded higher accuracy with lower RMSE values (e.g., at 1460 RPM: PID-ANN = 2.9425; PID-SMC = 0.0323). Under loaded conditions, PID-ANN remained faster (0.04–0.3 seconds), while PID-SMC provided improved stability (0.45–0.7 seconds) with lower error and RMSE (e.g., at 1168 RPM: PID-ANN = 2.2013; PID-SMC = 0.0373). During vehicle start-up simulations, PID-ANN reached steady state in 0.036 seconds with RMSE = 0.705 RPM, whereas PID-SMC required 0.520 seconds, resulting in SSE = 0.0034% and RMSE = 0.7279 RPM. In road contour simulations (incline, flat, decline), PID-ANN demonstrated faster adaptive response (e.g., steady state at 0.501 seconds, ESS = 0.6 RPM, RMSE = 0.3609 RPM), while PID-SMC took longer (0.71 seconds) but achieved superior long-term accuracy (ESS = 0.001 RPM, RMSE = 0.2733 RPM). These results confirm that ANN-based DTC effectively enhances the performance of induction motor drive systems, offering rapid response, adaptability, and robustness in dynamic conditions, making it a viable solution for electric vehicle propulsion systems.

Item Type:	Thesis (Diploma)
Uncontrolled Keywords:	Artificial Neural Network, Direct Torque Control, Motor Induksi, Mobil Listrik, Artificial Neural Network, Direct Torque Control, Electric Vehicle, Induction Motor.
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2692 Inverters T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2785 Electric motors, Induction. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK4055 Electric motor
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Andi Fadel Muhammad Nasir
Date Deposited:	29 Jul 2025 03:02
Last Modified:	29 Jul 2025 03:02
URI:	http://repository.its.ac.id/id/eprint/122657

Actions (login required)

View Item