Perancangan Dan Simulasi Sistem Kontrol Kecepatan Motor Induksi Tiga Fasa Menggunakan ANN-PID

Okta Huda Winarso, Nelvin (2026) Perancangan Dan Simulasi Sistem Kontrol Kecepatan Motor Induksi Tiga Fasa Menggunakan ANN-PID. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5009211042-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Motor induksi merupakan jenis motor listrik yang banyak digunakan pada aplikasi industri karena konstruksinya yang sederhana dan keandalannya yang tinggi. Namun, karakteristik nonlinier dan variasi dinamika sistem akibat perubahan setpoint serta gangguan beban menyebabkan pengendali PID konvensional sering mengalami keterbatasan dalam menghasilkan respon yang cepat dan akurat. Oleh karena itu, penelitian ini bertujuan untuk merancang dan menganalisis sistem kontrol kecepatan motor induksi menggunakan pengendali PID berbasis Artificial Neural Network (ANN-PID) sebagai metode penalaan parameter PID yang adaptif, serta membandingkan kinerjanya dengan PID konvensional. Perancangan sistem dilakukan melalui simulasi dengan memanfaatkan ANN untuk menyesuaikan parameter PID secara dinamis. Evaluasi performa dilakukan pada tiga skenario perubahan setpoint kecepatan, yaitu 500–900–1300 RPM, 1300–900–500 RPM, dan 900–1300–500 RPM, serta pada
pengujian gangguan beban (disturbance) dengan variasi torsi mekanik. Parameter performa yang dianalisis meliputi rise time, settling time, maximum overshoot, error steady state, dan recovery time. Hasil simulasi menunjukkan bahwa pengendali ANN-PID secara umum mampu meningkatkan akurasi pengendalian kecepatan motor induksi. ANN-PID menghasilkan error steady state yang lebih kecil dibandingkan PID konvensional pada hampir seluruh skenario pengujian, seperti pada skenario 500–900–1300 RPM di mana error steady state pada 900 RPM berhasil diturunkan dari 0,13% menjadi 0,01%. Selain itu, pada kecepatan menengah dan tinggi, ANN-PID mampu mempercepat respon transien dengan rise time dan settling time yang lebih singkat, serta menunjukkan kemampuan disturbance rejection yang lebih baik melalui recovery time yang lebih cepat dan overshoot yang lebih rendah pada sebagian besar kondisi gangguan. Namun demikian, pada perubahan setpoint yang bersifat ekstrem, baik PID konvensional maupun ANN-PID masih menghasilkan overshoot yang sangat besar, bahkan melebihi 100%, khususnya pada transisi penurunan kecepatan yang tajam. Hal ini menunjukkan bahwa meskipun ANN-PID efektif dalam meningkatkan performa steady state dan adaptivitas sistem, metode ini masih memiliki keterbatasan dalam meredam respon transien agresif dan memerlukan pengembangan lebih lanjut.
===================================================================================================================================
Induction motors are widely used in industrial applications due to their simple construction and high reliability. However, the nonlinear characteristics of the system and dynamic variations caused by setpoint changes and load disturbances often limit the performance of conventional PID controllers in achieving fast and accurate speed control. Therefore, this study aims to design and analyze an Artificial Neural Network–based PID (ANN-PID) speed control system for an induction motor, where the ANN is utilized as an adaptive tuning mechanism for PID parameters, and to compare its performance with a conventional PID controller. The control system was designed and evaluated through simulation by implementing ANN to dynamically adjust the PID parameters. Performance evaluation was conducted under three speed setpoint change scenarios, namely 500–900–1300 RPM, 1300–900–500 RPM, and 900–1300–500 RPM, as well as under load disturbance conditions with varying mechanical torque. The evaluated performance parameters include rise time, settling time, maximum overshoot, steady-state error, and recovery time. Simulation results indicate that the ANN-PID controller generally improves the speed control performance of the induction motor. The ANN-PID consistently produces a lower steady-state error compared to the conventional PID controller across most test scenarios. For example, in the 500–900–1300 RPM scenario, the steady-state error at 900 RPM was reduced from 0.13% to 0.01%. In addition, at medium and high speeds, the ANN-PID demonstrates faster transient responses with shorter rise time and settling time, as well as superior disturbance rejection capability, characterized by faster recovery time and lower overshoot under most disturbance conditions. However, under extreme setpoint transitions, both the conventional PID and ANN- PID controllers still exhibit very large overshoot values, exceeding 100%, particularly during sharp speed reduction. This indicates that although the ANN-PID controller effectively enhances steady-state accuracy and system adaptability, it still has limitations in suppressing aggressive transient responses and requires further improvement.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Motor Induksi, Artificial Neural Network, Self Tuning PID, Kontrol Kecepatan, Kontrol Adaptif Induction Motor, Artificial Neural Network, Self-Tuning PID, Speed Control, Adaptive Control
Subjects:	T Technology > T Technology (General) > T57.62 Simulation T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2785 Electric motors, Induction. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Nelvin Okta Huda Winarso
Date Deposited:	04 Feb 2026 08:49
Last Modified:	04 Feb 2026 08:49
URI:	http://repository.its.ac.id/id/eprint/132102

Actions (login required)

View Item