Analyzing The Impact Of Renewable Energy Penetration On Critical Clearing Time Using Recurrent Neural Network (RNN) With Critical Trajectory Method

Hardiyanto, Firki (2026) Analyzing The Impact Of Renewable Energy Penetration On Critical Clearing Time Using Recurrent Neural Network (RNN) With Critical Trajectory Method. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022221043-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Peningkatan penetrasi energi terbarukan, khususnya sumber berbasis inverter, menyebabkan penurunan inersia pada sistem tenaga modern. Rendahnya inersia melemahkan kestabilan transien dan secara signifikan memengaruhi Critical Clearing Time (CCT), yaitu waktu maksimum yang diperbolehkan untuk membersihkan gangguan sebelum sistem kehilangan sinkronisme. Seiring bertambahnya penetrasi energi terbarukan, nilai CCT menjadi lebih pendek dan semakin sensitif terhadap kondisi gangguan.Metode konvensional untuk menghitung CCT melalui simulasi domain waktu mampu memberikan hasil yang akurat,
namun membutuhkan komputasi yang sangat besar sehingga kurang praktis untuk analisis cepat ataupun aplikasi real-time. Untuk mengatasi keterbatasan tersebut, penelitian ini mengusulkan pendekatan hibrida yang mengintegrasikan Critical Trajectory Method (CTM) dengan Recurrent Neural Network (RNN) yang diimplementasikan secara manual. CTM digunakan untuk menghasilkan nilai acuan CCT, sedangkan RNN dilatih untuk memprediksi CCT berdasarkan variabel dinamis pasca-gangguan, jenis gangguan, dan parameter VirtualSynchronous Generator (VSG). Sebanyak 540 sampel data dihasilkan dari berbagai skenario penetrasi energi terbarukan dan kondisi gangguan. Setelah proses penyempurnaan data, model RNN yang diusulkan menunjukkan akurasi prediksi yang tinggi dengan nilai kesalahan
pengujian yang sangat rendah. Hasil penelitian ini menunjukkan bahwa kerangka RNN–CTM mampu mengestimasi CCT secara efisien tanpa memerlukan simulasi nonlinier penuh,
sehingga menawarkan alat analisis yang cepat dan andal untuk penilaian kestabilan pada sistem tenaga berinersia rendah yang didominasi energi terbarukan.
================================================================================================================================
The increasing penetration of renewable energy, especially inverter-based resources, has reduced the overall inertia of modern power systems. Lower inertia weakens transient stability and significantly affects the Critical Clearing Time (CCT), which represents the maximum time available to clear a disturbance before loss of synchronism. As renewable penetration increases, CCT becomes shorter and more sensitive to system disturbances. Conventional CCT estimation through time-domain simulation is accurate but
computationally intensive, making it impractical for fast or real-time applications. To address this issue, this study proposes a hybrid approach that integrates the Critical Trajectory Method (CTM) with a manually implemented Recurrent Neural Network (RNN). The CTM is used to
generate CCT reference values, while the RNN is trained to predict CCT based on post-fault dynamic variables, disturbance type, and Virtual Synchronous Generator (VSG) parameters. A dataset samples was generated from various renewable penetration and disturbance scenarios. After data refinement, the proposed RNN model achieved high prediction accuracy with very low testing errors. The results show that the RNN–CTM framework is capable of estimating CCT efficiently without requiring full nonlinear simulations, offering a fast and reliable tool for stability assessment in low-inertia, renewable-dominated power systems.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Critical Clearing Time, Recurrent Neural Network, Critical Trajectory Method, Penetrasi Energi Terbarukan, Stabilitas Transien Critical Clearing Time, Recurrent Neural Network, Critical Trajectory Method, Renewable Energy Penetration, Transient Stability
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3226 Transients (Electricity). Electric power systems. Harmonics (Electric waves).
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Firki Hardiyanto
Date Deposited:	29 Jan 2026 08:25
Last Modified:	29 Jan 2026 08:25
URI:	http://repository.its.ac.id/id/eprint/131003

Actions (login required)

View Item