Other topics

Cascading failures cause severe threats to the security of complex power system operations. My research has been focusing on two aspects: 1) to establish rigorous theoretic foundations for cascading failure analysis; and 2) to develop fast algorithms for online analysis.

The first issue is about how to guarantee the “confidence” of cascading failure analysis. My research mathematically characterizes when a cascading failure can be modeled as a Markov process. We also developed a sequential importance sampling (SIS) method to enhance the analysis efficiency 2-3 orders of magnitudes[1]. We further revealed the semi-analytical relationship between component failure probability and the blackout risk, simplifying time-consuming Monte Carlo simulations to simple matrix multiplication[2]. In addition, we proposed an adaptive sampling method to guarantee confidential results [3].

The second issue is about how to speed up the analysis with high accuracy. First, we converted the original physical grid into a virtual influential graph. Then we built a PageRank-based algorithm on this graph to identify the most vulnerable lines [4] and estimate the blackout risk[5]. It is shown that this method can accurately screen out the most influential lines (compared to elaborated simulations) in a provincial power grid with more than 1000 lines within 10 minutes, enabling the ONLINE application of cascading outage analysis.

The core algorithms has been applied to develop the cascading failure simulation and analysis system for Henan, Sichuan, and Guangdong provincial power grids. Due to my remarkable contribution, I was awarded the First Prize of Natural Sciences Award by the Ministry of Education of China in 2015, and the Second Prize of Scientific and Technical Progress Award of Sichuan Province.

[1] J. Guo, F. Liu*, J. Wang, J. Lin, and S. Mei, Toward Efficient Cascading Outage Simulation and Probability Analysis in Power Systems, IEEE Transactions on Power Systems, 2018, 33(3), 2370–2382

[2] J. Guo, F. Liu*, J. Wang, M. Cao, and S. Mei, Quantifying the Influence of Component Failure Probability on Cascading Blackout Risk, IEEE Transactions on Power Systems, 2018, 33(5), 5671–5681

[3] F. Liu*, J. Guo, X. Zhang, Y. Hou, and S. Mei, Mitigating the Risk of Cascading Blackouts: A Data Inference Based Maintenance Method, IEEE Access, 2018, 6, 39197–39207

[4] Z Ma, C Shen, F Liu*, S Mei, Fast Screening of Vulnerable Transmission Lines in Power Grids: a PageRank-based Approach, IEEE Transactions on Smart Grid, 2018, 10 (2), 1982-1991

[5] Z. Ma, F. Liu*, C. Shen, Z. Wang, and S. Mei, Fast Searching Strategy for Critical Cascading Paths Toward Blackouts, IEEE Access, 2018, 6, 36874–36886