Performance Comparison Between Deep Learning Models for Fault Classification in Transmission Lines Using Time Series Data DOI Creative Commons
Mohsin Ali Tunio, Mohsin Ali Tunio, Muhammad Amir Raza

и другие.

Energy Science & Engineering, Год журнала: 2025, Номер unknown

Опубликована: Апрель 17, 2025

ABSTRACT Deep learning has become a vital tool for addressing complex challenges in power systems, particularly fault detection and classification transmission lines. This study presents comparative analysis of three advanced time‐series models like temporal convolutional networks (TCN), bidirectional long short‐term memory (BiLSTM), gated recurrent units (GRU) classification. Leveraging comprehensive data set encompassing diverse scenarios single‐phase to ground (AG), double line (ABG), three‐phase (ABC) from both simulated real data, the provides rigorous evaluation these models’ performance under realistic conditions. The results demonstrate that TCN achieves accuracy 99.9%, significantly outperforming BiLSTM (92.31%) GRU (95.27%), while also excelling precision, recall, F 1 score, training efficiency. Additionally, this incorporates feature extraction techniques discrete wavelet transform (CWT) establish new benchmarks findings underscore TCN's robustness handling dynamic nature signals its practical potential real‐time applications, contributing development more reliable efficient system systems.

Язык: Английский

Carbon neutrality and economic stability nexus: An integrated renewable energy transition to decarbonize the energy sector DOI
Muhammad Amir Raza, M.M. Aman, Laveet Kumar

и другие.

Energy Reports, Год журнала: 2025, Номер 13, С. 4586 - 4608

Опубликована: Апрель 15, 2025

Язык: Английский

Процитировано

0
Performance Comparison Between Deep Learning Models for Fault Classification in Transmission Lines Using Time Series Data DOI Creative Commons
Mohsin Ali Tunio, Mohsin Ali Tunio, Muhammad Amir Raza

и другие.

Energy Science & Engineering, Год журнала: 2025, Номер unknown

Опубликована: Апрель 17, 2025

ABSTRACT Deep learning has become a vital tool for addressing complex challenges in power systems, particularly fault detection and classification transmission lines. This study presents comparative analysis of three advanced time‐series models like temporal convolutional networks (TCN), bidirectional long short‐term memory (BiLSTM), gated recurrent units (GRU) classification. Leveraging comprehensive data set encompassing diverse scenarios single‐phase to ground (AG), double line (ABG), three‐phase (ABC) from both simulated real data, the provides rigorous evaluation these models’ performance under realistic conditions. The results demonstrate that TCN achieves accuracy 99.9%, significantly outperforming BiLSTM (92.31%) GRU (95.27%), while also excelling precision, recall, F 1 score, training efficiency. Additionally, this incorporates feature extraction techniques discrete wavelet transform (CWT) establish new benchmarks findings underscore TCN's robustness handling dynamic nature signals its practical potential real‐time applications, contributing development more reliable efficient system systems.

Язык: Английский

Процитировано

0