Coordinated Planning of Transmission Network Expansion and Distribution Network Modernization With Microgrids Under Non-Uniform Discrete Choices

Abstract

The paper proposes a coordinated planning method to reduce redundant costs for distribution network modernization with microgrids considering the practical configuration of candidate capacities. We first use the values of binary decision variables to represent whether to choose corresponding practical candidate capacities, whereby the investment choices are definitely feasible in engineering. Then, a bi-level stochastic model that incorporates combinatorial uncertain scenarios in spatial-temporal-event dimensions is formulated. To address the intractable large-scale mixed integer lower-level models, we also provide a clustering method in iterations to project scenarios onto all possible binary decisions. The number of binary decision variables and scenario-related state variables in lower-level models is thus significantly reduced without losing accuracy at the same time. Further, we extend enhanced Benders decomposition to a hot start nested form to be compatible with both inner-loop mixed integer linear subproblems and outer-loop linear security assessment subproblems. Mathematically, the proposed method can rapidly converge to practical and optimal choices for candidate capacities within finite iterations. Finally, we validate the total cost savings from coordinated planning and the optimality, rapidity, and scalability of the proposed method on integrated IEEE systems.