Compressive System Identification for Multiple Line Outage Detection in Smart Grids

Mohammad Babakmehr, Farnaz Harirchi, Ahmed Al-Durra, S. M. Muyeen, Marcelo Godoy Simoes

Research output: Contribution to journalArticlepeer-review

22 Scopus citations


Real-time power line outage detection (POD) and localization is an important monitoring task for the modern smart grid. Reliable monitoring of power lines status plays a critical role in the system-wide blackout prevention. In this paper, the main aim is to address the multiple POD problems by exploiting the compressive system identification - a time-efficient approach in a complex network analysis. A typical power network is considered as a single graph, and the mathematical formulation of the POD problem is initialized using the dc power-flow model and graph theory concepts. Next, a sparse representation-based formulation for this problem (POD-SRP) is reported and further improved and generalized in case of multiple large-scale outages. Practical and technical challenges associated with this sparse recovery problem are partially addressed by developing new SRP solvers. Furthermore, a new sparse-based mathematical formulation for POD is introduced and termed as 'Binary-POD-SRP,' which specifically deals with two particular issues, namely, the high coherence and the signal dynamic outrange. Finally, the identification performance of the proposed framework is evaluated by a variety of case studies, which are modeled using IEEE standard test-beds. We specifically discuss how the inherent challenges within large-scale multiple-outages can be solved by applying these new techniques and formulations.

Original languageBritish English
Article number8731707
Pages (from-to)4462-4473
Number of pages12
JournalIEEE Transactions on Industry Applications
Issue number5
StatePublished - 1 Sep 2019


  • Compressive system identification (CSI) sparse recovery
  • power outage identification
  • power system monitoring
  • smart grid (SG)


Dive into the research topics of 'Compressive System Identification for Multiple Line Outage Detection in Smart Grids'. Together they form a unique fingerprint.

Cite this