Globally Optimal Distributed Operation of Integrated Electric and Heating Systems

Abstract

Unit commitment (UC) is a key player in the coordinated operation of integrated energy systems. However, the participation of multiple market entities with widely different characteristics in large-scale energy systems has urged the critical need for the application of a distributed scheme to the UC problem. The NP-hard UC problem is a challenging mixed-integer programming problem. The presence of a large number of binary variables in the UC subproblems, which are solved by each participating entity after implementing the UC decomposition, fails to guarantee the convergence and the optimality of existing solution methods. To bridge this gap, this paper proposes a distributed method, using logic-based Benders decomposition (LBBD), for the UC problem in a typical multi-entity system, i.e., integrated electric and heating system (IEHS). By searching the branch and bound tree of the district heating system (DHS) subproblem, the lower bound of its objective function is rigorously derived as a valid Benders cut to ensure the convergence to global optimal results. This distributed method is suitable for both deterministic and robust UC solutions. Numerical simulations are conducted on two test systems to demonstrate the performance of the proposed model and its distributed solution method.