Assessing the MobileNet50 Convolutional Neural Network's Performance in Identifying and Classifying Electrical Defects

The growing complexity of modern electrical systems necessitates the adoption of robust and efficient techniques for fault detection and classification. This study explores the application of the MobileNet50 Convolutional Neural Network (CNN) to address a critical challenge in the realm of electrical engineering. The primary aim of this investigation is to assess the classifier's performance in accurately identifying the presence of electrical issues and categorizing them into distinct classes. A diverse dataset encompassing various failure scenarios is employed to evaluate the model's effectiveness. This research seeks to scrutinize the efficacy of the MobileNet50 CNN in identifying and classifying electrical problems. Leveraging the MobileNet50 CNN's reputation for resilience and adaptability in handling intricate datasets, we analyze a dataset containing diverse electrical fault situations. The key objectives of this study involve evaluating the classifier's effectiveness and efficiency in accurately detecting flaws and subsequently classifying them into predefined categories. Performance metrics such as accuracy, precision, recall, and F1 score are employed to assess the model's performance. Additionally, the study investigates the impact of hyperparameter adjustments on enhancing the classifier's performance. The findings presented in this research contribute valuable insights into the potential use of MobileNet50 CNN in the identification and classification of electrical faults. These insights hold significant implications for the advancement of fault diagnostic systems within the field of electrical engineering.