Multi-Period Fast Robust Optimization of Active and Reactive Power in Active Distribution Networks

Zhang, Jian; Cui, Mingjian; He, Yigang; Li, Fangxing (Fran)

doi:10.36227/techrxiv.12739280.v1

Multi-Period Fast Robust Optimization-200519.pdf (1.08 MB)

Download file

Multi-Period Fast Robust Optimization of Active and Reactive Power in Active Distribution Networks

preprint

posted on 2020-08-04, 07:09 authored by Jian Zhang, Mingjian CuiMingjian Cui, Yigang He, Fangxing (Fran) Li

Connecting a large number of intermittent distributed generators (DGs) and electric vehicles (EVs) to a distribution network increases uncertainty and seriously threatens the safety and economic operation of that distribution network. Robust optimization is an important method of hedging against the adverse effects of uncertainty. However, the existing literature on robust optimization of distribution networks has not taken into account constraints on the travel distances of regulating equipment. In response to this deficiency, this paper considers constraints on the cycle period of energy storage systems (ESSs), travel distances of switchable capacitors and reactors, and load tap changers (OLTCs) and step voltage regulators (SVRs). A two-stage multi-period mixed-integer second-order cone robust model for coordinated active and reactive power optimization of a distribution network is formulated based on branch flow equations. There are not any dummy variables in the second stage model. In contrast to the deficiency of the low computational rate of the column-and-constraint generation (CCG) method, an iterative solution to the first and second stage models is proposed based on the cutting plane. Increases to the variables and constraints are not needed to solve the first stage model. For the second stage multi-period model, only the model of each single period needs to be solved first. Then the results of the two models are accumulated. Overall, the computational speed is significantly enhanced. The capabilities of the proposed method are validated by three simulation cases.