Hybridized Multi-Layer Microgrid Formation & Optimization Technique Using Supervised Learning (ANFIS) for Resiliency Enhancement in T&D Power System

This research focuses on optimizing microgrid operations and enhancing power system resiliency in networks with a high penetration of renewable energy sources. It emphasizes the role of grid-forming inverters and robust control mechanisms for stable system performance under varying operating conditions. A robust transmission and distribution (T&D) co-simulation platform was developed to model and optimize the interaction between distribution and transmission systems, achieving effective economic dispatch, cost minimization, and loss reduction. The study also introduced an Adaptive Neuro-Fuzzy Inference System (ANFIS) for forecasting and optimization, which demonstrated fast convergence and improved computational efficiency. Through simulations, the framework successfully handled grid-tied and islanded operations, allowing for dynamic microgrid formation and reconnection, load control, and three-phase optimal power flow (3-φ OPF) analysis across the network. Two case studies were implemented — a 14-Bus IEEE transmission system with 15-Bus distribution, and a 14-Bus T/S system with a modified unbalanced 123-Bus IEEE distribution network — modeled in MATLAB/Simulink and optimized via GAMS. The ANFIS-based optimization improved system resiliency and operational efficiency, yielding a 68% reduction in distribution network operation cost, along with significant reductions in active and reactive power losses (up to 3.23% and 16.7%, respectively). The AutorunABA1.0 co-simulation platform, integrating Machine Learning, MATPOWER OPF, and GAMS, enabled efficient data exchange, faster computation, and optimal resource allocation under multiple constraints such as voltage, frequency, load priority, and line capacity. The study demonstrates that the proposed integrated simulation and optimization framework can effectively manage microgrid formation, DER allocation, and T&D coordination, thereby achieving both economically optimal and resilient power system operation.