A multi-disaster-scenario distributionally robust planning model for enhancing the resilience of distribution systems

作     者：Zhang, Gang Zhang, Feng Zhang, Xin Wu, Qiuwei Meng, Ke 

作者机构：Shandong Univ Key Lab Power Syst Intelligent Dispatch & Control Minist Educ 17923 Jingshi Rd Jinan 250061 Peoples R China Cranfield Univ Sch Water Energy & Environm Energy & Power Theme Cranfield MK43 0AL Beds England Tech Univ Denmark DTU Ctr Elect Power & Energy CEE Dept Elect Engn DK-2800 Lyngby Denmark Univ New South Wales Sch Elect Engn & Telecommun Sydney NSW 2052 Australia 

出 版 物：《INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS》 (国际电力与能源系统杂志)

年 卷 期：2020年第122卷

页      面：106161-106161页

核心收录：

学科分类：0808[工学-电气工程] 08[工学] 

基　　金：The present study was supported by the Deutsche Forschungsgemeinschaft (DFG) via Grants SFB 301 and SFB 494 and by special funding from the Ministry of Science of the Land Nordrhein–Westfalen 

主　　题：Distributionally robust method Resilience Network planning Power distribution system Stochastic programming Uncertainty 

摘      要：Resilience oriented network planning provides an effective solution to protect the distribution system from natural disasters by the pre-planned line hardening and backup generator allocation. In this paper, a multidisaster-scenario based distributionally robust planning model (MDS-DRM) is proposed to hedge against two types of natural disaster-related uncertainties: random offensive resources (ORs) of various natural disasters, and random probability distribution of line outages (PDLO) that are incurred by a certain natural disaster. The OR uncertainty is represented by the defined probability-weighted scenarios with stochastic programming, and the PDLO uncertainty is modeled as the moment based ambiguity sets. Moreover, the disaster recovery strategies of network reconfiguration and microgrid formation are integrated into the pre-disaster network planning for resilience enhancement in both planning and operation stages. Then, a novel primal cut based decomposition solution method is proposed to improve the computational efficiency of the proposed model. In particular, the equivalent reformulation of the original MDS-DRM is first derived to eliminate the PDLO-related variables. Then, the reformulation problem is solved by the proposed primal cut based decomposition method and linearization techniques. Finally, Simulation results are demonstrated for IEEE 13-node, 33-node and 135-node distribution systems to validate the effectiveness of the proposed method in enhancing the disaster-induced network resilience.