Self-organizing map based adaptive protection scheme for IEEE benchmark networked microgrids: a comprehensive framework for multi-modal operation

Purpose — This paper presents a novel adaptive protection scheme for networked microgrids (NMGs) based on the Self-Organizing Map (SOM) clustering technique, addressing critical challenges in multi-microgrid protection coordination. Design/Methodology/Approach — An SOM-based framework automatically classifies eight distinct NMG operating conditions and adapts relay settings in real time. The method is validated comprehensively on the IEEE benchmark test system for networked microgrids using PSCAD/EMTDC simulations across normal, fault, disturbance, and emergency scenarios. Findings — The proposed scheme achieves 94.7% classification accuracy on test data, reduces fault clearing times by an average of 34.3%, decreases miscoordination events by 45%, and improves the overall protection reliability index by 10.7% compared to conventional overcurrent protection. Practical Implications — The framework includes hardware specifications, commissioning procedures, and economic analysis demonstrating a 1.75-year payback period, providing a viable roadmap for real-world deployment. Originality/Value — This is the first comprehensive implementation and validation of SOM-based adaptive protection on the standardized IEEE benchmark NMG system, covering all major operating modes and transition states within a unified framework.