Unsymmetrical Fault Prediction and Fault Level Evaluation for Inverter Systems in the South African Power Distribution Network

This study addresses the challenges of fault level contributions from wind turbine generators in South Africa's distribution grid compared with the Synchronous generator, mainly focusing on the limited fault contribution from inverter-based systems during grid faults. In this study, ANN and SVM models are developed to classify network faults with synchronous generators and Type 1 wind turbines. Improving fault level understanding with these models aids in reliable grid planning, making fault detection more precise and efficient than traditional calculation-based methods. Machine learning models like ANN and SVM are expected to outperform conventional symmetrical component analysis, especially in complex fault scenarios with a penetration of renewable plants.