Multi-Stage Adaptive Stochastic-Robust Scheduling Method With Affine Decision Policies for Hydrogen-Based Multi-Energy Microgrid

Abstract

Zero-carbon clean energy such as hydrogen has been developed rapidly to reduce carbon emissions, gradually promoting them as the main energy supply for multi-energy microgrids (MEMGs), which motivates the deployment of hydrogen-based MEMG (H-MEMG). The main difficulty of the H-MEMG scheduling problem is how to handle source/load uncertainties for ensuring the solution feasibility and economics in the actual operation. To this end, this paper proposes a novel multi-stage adaptive stochastic-robust optimization (MASRO) approach, which combines the ideas of stochastic programming and multi-stage robust optimization. The established model has the objective of the expected operation cost and ensures the solution feasibility by designed constraints rather than the “minmax” structure. Specifically, first, affine policies are adapted to describe the complex relationship between decision variables and uncertainty realizations; Second, an affine policy-based solution approach is proposed for the MASRO H-MEMG scheduling model. Then, the complex conversion relationship and coupling constraints are reformulated, and a tractable mixed-integer linear programming (MILP) model is established; Third, based on the solved affine functions, the real-time rolling and nonrolling economic dispatch models are proposed to respectively pursue the economic and computational requirements, and both can guarantee solution robustness and nonanticipativity. Numerical tests are implemented on a real H-MEMG, verifying that the proposed method could guarantee the feasibility and economic efficiency of actual H-MEMG operations.