Transmission Line Protection based on Fine-Tuning CatBoost Fault Classification Technique

Transmission line challenges are complex because they involve many different factors and are subject to significant fluctuations. These complexities make it crucial to implement precise protective measures. Effective protection ensures that electrical power flows without interruption. To address these challenges, a new classification algorithm is proposed, which is based on CatBoost. This algorithm works in conjunction with a Wavelet Transform-based decomposition system. These tools help in accurately assessing the condition of the transmission lines and responding to any issues that may arise. For practical application, the Indian standard system serves as the simulation model. This system operates at a voltage rating of 220 kV and spans a length of 120 km, with a frequency of 50 Hz. The decomposition frequency is set at 4 kHz, and to perform the Wavelet decomposition method, 80 samples are taken per cycle. A comprehensive data collection occurs within the transmission line using the MULTIRUN block of PSCADTM software. This software is instrumental in gathering the necessary data for analysis. The developed CatBoost algorithm is implemented in MATLAB, ensuring that hyperparameters are optimized for better performance. Once data is collected from PSCADTM, it is then imported into MATLAB for further processing. The CatBoost algorithm is employed to validate the data and assess the performance of the transmission line protection system. The results from this classification reveal that the proposed algorithm significantly outperforms traditional methods in protecting transmission lines.