Abstract
This paper presents a model-in-the-loop (MIL) validation framework for integrating and testing a future-aware tertiary control layer for the Center for Microgrid Research (CMR) Microgrid. The framework enables closed-loop interaction among forecasting, optimization, and real-time digital simulation, providing a realistic and low-risk environment for evaluating control performance before field deployment. A long short-term memory (LSTM) model is developed for short-term load and photovoltaic (PV) forecasting using real CMR Microgrid data and incorporated into the tertiary control algorithm for future-informed decision-making. The tertiary control layer is integrated with a real-time digital model of the microgrid to allow bidirectional data exchange and execution of control commands based on real-time system feedback. The proposed framework is validated under various operating conditions, including grid-connected, islanded, and transitional modes. Results demonstrate high forecasting accuracy, effective interpretation of system states, optimized DER operation, and seamless communication within the MIL environment.