Distributed Assimilation of Grid Conditions and Load Integration via Social Learning

The continuously evolving landscape of the modern electric grid has equipped traditionally passive power elements with communication and intelligent actuation capabilities. This cyber-physical nature of the grid can be leveraged to empower individual power components with the ability to react to changing grid conditions in a collaborative and adaptive manner. Adaptive inferencing of grid conditions used in local decision-making by individual nodes offloads concentrated computational overheads otherwise incurred for centralized large-scale coordination and data processing in the grid. As such, in this paper, we employ social learning constructs for allowing individual end nodes to iteratively exchange signals containing local information to reach a consensus/belief regarding the general grid condition in real-time. This belief at equilibrium is then used by participating power entities to make adaptive local integration decisions that heed physical grid constraints. We present both theoretical and empirical studies highlighting how collaborative opinion formation in the grid can be utilized to improve grid conditions in the adaptive integration of diverse power entities.