Distributed continuous-time unit commitment with energy storage in multi-area networks

Abstract

This paper proposes a distributed solution for multi-area unit commitment (UC) problem with continuous-time energy generation and storage, offering an enhanced operation tool that leverages the available operational flexibility resources via higher fidelity modeling to enable effective resource sharing among areas via coordinated continuous-time interconnection power exchange. The proposed methodology involves formulating a variational multi-area UC problem with energy storage where decision variables (including power, energy, and commitment statuses) are modeled as continuous-time trajectories and ramping is defined as the time-derivative of the respective power trajectory. The variational multi-area UC problem is then projected into Bernstein function space, leading to a mixed-integer linear programming (MILP) problem with Bernstein coefficients of dispatch and commitment status trajectories as decision variables. The function space-based multi-area UC problem is then decomposed into per-area UC sub-problems solved using a distributed algorithm. Implemented on two different test networks and compared against the benchmark centralized and traditional discrete-time solutions, the numerical results highlight the solution accuracy and efficacy of the proposed distributed method to achieve optimal decisions on interconnection power exchanges such that the energy and ramping needs of all participating areas are met.