Dynamic Incentive Design for Demand Response in Energy System MicroGrids

The integration of renewable energy sources into microgrids introduces variability and unpredictability, requiring innovative solutions to balance supply and demand.Demand Response (DR) programs incentivize consumers to adjust energy usage during critical periods, reducing grid stress and improving efficiency.This paper presents a dynamic DR incentive framework that leverages day-ahead forecasts to optimize price discounts while maintaining occupant comfort.Using contextual multi-armed bandits (MAB) and mixed-integer linear programming (MILP), our approach dynamically adapts to fluctuating renewable generation, energy prices, and consumer preferences.Unlike traditional rule-based methods, it optimizes incentives to maximize consumer participation and grid stability while minimizing reliance on costly diesel generators.Simulation results demonstrate significant improvements in financial sustainability and operational efficiency, highlighting the potential of data-driven incentive designs for resilient and adaptive energy systems. CCS Concepts Applied computing Multi-criterion optimization and decision-making; Theory of computation Integer programming; Hardware Smart grid.