Bilayer Distributed Optimization for Robust Microgrid Dispatch With Coupled Individual-Collective Profits

Abstract

This paper addresses the economic dispatch problem in microgrids with multi-entities and uncertainties. A novel robust bilayer distributed optimization (BDO) method is put forward to coordinate the individual and collective profits. In the upper-layer of the BDO model, each entity in the microgrids develops a two-stage recourse-based robust dispatch model that minimizes the individual operation cost. This robust model determines the objectives and schemes in both the basic and worst-case scenarios, and provides more practical optimal schedules. According to the generation margin and load curtailment in each entity, a multi-objective bargain dispatch is conducted in the lower-layer coordinating center to co-optimize the tie-line plans, thus getting the lowest collective cost in a cooperative manner through regional power complementarity. An iterative descent algorithm is put forward for the efficient solution of the nonconvex BDO model. Therein a nested column and constraint generation algorithm with equilibrium constraints is developed to tackle the upper-layer robust optimization problems with binary recourses and non-independent uncertainties. The lower-layer bargain model is refined as a mixed-integer linear programming via a complementarity approach for the cooperation of multi-entities. Numerical results of case studies demonstrate the availability and superiority of the robust BDO method and the solution algorithm.