Search

Although the use of a series capacitor increases the transmission line (TL) capacity, it inverts the voltage/current and creates a sub-harmonic frequency that results in unintended operation of the protection devices, which further decreases the transient stability under fault conditions. Restoring the transient stability requires an accurate diagnosis of faults in TLs, which further requires exploration of in-depth features. This imposes challenges for the existing approaches as they demand a large amount of data. In this paper, a self-attentive weight-sharing capsule network (WSCN) is proposed to achieve robust and high classification performance while utilizing a limited amount of data. The capsule network, with the weight-sharing mechanism delivers robust performance in detecting and classifying the faults in the TL domain. Simultaneously, the self-attention layer highlights the more dominant features that make the network work upon the limited data effectively. A Western-System-Coordinating-Council WSCC 9-bus and 3-machine test model, which was modified with the series capacitor was studied to quantify the robustness of the self-attention WSCN. The results were compared with those of others networks to confirm the explicit classification ability of the proposed network. A study on real-world data obtained from power network was also conducted to validate the proposed self-attention WSCN. [ABSTRACT FROM AUTHOR]

• Development of a new approach for transmission line faults detection and classification based on the convolutional neural network combined with wavelet transform which makes the network to efficiently train and extract the relevant features from the three phase faulty data. • Integration of the self attention mechanism with proposed CNN to identify the particular type of fault more accurately that enables the model to perform the precise classification. • Variation of input signals like voltage signal, current signal, and combined with voltage and current signal are employed at various sampling frequency to verify the performance of proposed model. • Noises are considered to prepare the input data to confirm the robustness of proposed SAT-CNN model. • Performance comparison between proposed and some other state-of-the-art model is done to confirm the high performance of proposed SAT-CNN model. This paper introduces a novel self-attention convolutional neural network (SAT-CNN) model for detection and classification (FDC) of transmission line faults. The transmission lines continuously experience the number of shunt faults and its effect in the practical system rises the instability, line restoration cost and damages the load. Therefore, a robust and precise model is needed to detect and classify the faults for the rapid restoration of faulty phases. In this paper, we propose a SAT-CNN framework with time series imaging based feature extraction model for FDC of a transmission line. To ensure the noise immunity performance, the discrete wavelet transform (DWT) has been used to denoise the faulty voltage and current signals. The effectiveness of the proposed SAT-CNN framework is tested by varying the input signals namely voltage, current, and combined voltage and current signal, under the various sampling frequencies. The robustness of the proposed model is verified by adding the noises to the input data. Results show that the proposed model is capable to perform precise classification and detection of transmission line faults with high accuracy. A comparison between the proposed and other state-of-the-art FDC model is also studied to show the superiority of the proposed SAT-CNN model. [ABSTRACT FROM AUTHOR]