This research investigates the accuracy of fault classification using deep learning models, specifically Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNNs). These models are applied to the 161kV transmission line from Aboadze Thermal Power Station through Takoradi, Tarkwa, and Prestea to New Obuasi in Ghana. The study aims to evaluate the performance of these models in accurately detecting and classifying faults to ensure the reliability and efficiency of the power supply. The methodology includes data collection from the transmission line, training LSTM and CNN models, and assessing their accuracy in fault classification. Results demonstrate high accuracy in fault prediction and classification, supporting effective maintenance and reducing power outages. The study concludes with recommendations for improving fault classification systems using deep learning techniques.