Forecast-Driven and Scenario-Based Building Energy Management Using a Stochastic Optimization Approach

Buildings are essential components of power grids, and their energy performance directly affects overall power system operation. This paper presents a novel stochastic optimization framework for building energy management systems, aiming to enhance buildings' energy performance and facilitate their effective integration into emerging intelligent power grids. In this method, solar power generation and building electricity demand forecasts are combined with historical data, leveraging statistical characteristics to generate probability matrices and corresponding scenarios with associated probabilities. These scenarios are then used to solve the stochastic optimization problem, optimizing building energy flow while accounting for existing uncertainties. The results demonstrate that the proposed methodology effectively manages inherent uncertainties while maintaining performance and outperforming rule-based and custom-built reinforcement learning-based solutions.