An MIP-Based Heuristic Decomposition Approach for Distributed Energy Resource Scheduling

Zusammenfassung The decentralization of the power infrastructure poses new challenges for power plant operators and aggregators. A large number of power plants must be coordinated to successfully participate in energy markets. This requires to take uncertain exogenous influences, such as weather and market price, into account when determining operational schedules of energy resources and quantity-price tuples for bids to energy and reserve markets. Stochastic programming techniques are commonly applied to these decision problems. But, the resulting models are usually large-scale and hard to solve. We propose a temporal decomposition heuristic to speed-up the solution process of such models. The heuristic is based on mixed-integer programming formulations. Therefore, it is flexible with respect to the kinds of power plants and market environments that can be considered. We validate our findings on a realistic benchmark set of large stochastic programs and confirm its suitability to significantly reduce computational effort while retaining near-optimal solution quality.