Enhancing resilience analysis of power systems using robust estimation. (June 2019)

• Record Type: Journal Article
• Authors: Shen, Lijuan; Tang, Loon Ching
• Is Part Of: Reliability engineering & system safety. Volume 186(2019)
• Page Start: 134
• Page End: 142
• Publication Date: 2019-06
• Subjects: Heavy-tailed -- Maximum likelihood -- Minimum distance estimation -- Log-log plot -- Disruption; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2019.02.022
• Languages: English
• ISSNs: 0951-8320

Abstract

Highlights: We propose a robust estimation procedure for the power law distribution. The proposed estimators are more robust in the presence of contaminated data. The method is illustrated with the blackout data from the U.S. power grid. Abstract: It has been well-recognized that the distribution of the blackout size of a power grid system has a heavy tail. The power-law distribution is a popular model for the heavy-tail phenomenon, and it is widely used in power system disruptions. However, there are significant reporting errors in the disruption data reported in public available databases, such as the database of the Electric Disturbance Events (OE-417) maintained by the US Department of Energy. Traditional inference techniques such as the maximum likelihood estimation can be sensitive to such contaminated data due to the reporting errors. In this paper, we propose a robust estimation procedure for the power-law distribution based on the minimum distance estimation method. A comprehensive simulation is used to evaluate the performance of the proposed method, and compare the performance with the existing maximum likelihood method. It is found that the proposed method outperforms the existing maximum likelihood method in the presence of contaminated data. We apply the proposed method to the blackout data from Jan 2002 to Aug 2016 based on the OE-417 database.