State Estimation Extensive Criticality Analysis Performed on Measuring Units: A Comparative Study

• Type: Article in International Scientific Journal
• Date: 2024

Authors

Nishio, A (Author) Other; Do Coutto, MB (Author) Other; de Souza, JCS (Author) Other; Pereira, J (Author) FEP; Zanghi, E (Author) Other

Journal

Title: Journal of Control, Automation and Electrical Systems

Vol. 35

Pages: 1135-1146

ISSN: 2195-3880

Publisher: Springer Nature

Indexing

ISI Web of Knowledge - 0 Citations

Scopus - 0 Citations

Other information

Authenticus ID: P-017-AYD

DOI: 10.1007/s40313-024-01130-9

Abstract (EN): As one of the functions integrating energy management systems, state estimation (SE) is instrumental in monitoring power networks, allowing the best possible use of energy resources. It plays a decisive role in debugging if sufficient data are available, ruined if not. Criticality analysis (CA) integrates SE as a module in which elements of the estimation process-taken one-by-one or grouped (tuples of minimal multiple cardinality)-are designated essential. The combinatorial nature of extensive CA (ExtCA), derestricted from identifying only low-cardinality critical tuples, characterizes its computational complexity and imposes defiant limits in implementing it. This paper presents the methodology for ExtCA and compares algorithms to find an efficient solution for expanding the boundaries of this analysis problem. The algorithms used for comparison are one sequential Branch&Bound (a well-known paradigm for combinatorial optimization recently used in ExtCA) and two new parallels implemented on the central processing unit (CPU) and the graphics processing unit (GPU). The conceived parallel architecture favors evaluating massive combinations of diverse cardinality measuring unit (MU) tuples in ExtCA. The acronym MU refers to the aggregate of devices deployed at substations, such as a remote terminal unit, intelligent electronic device, and phasor measurement unit. The numerical results obtained in the paper show significant speed-ups with the novel parallel GPU algorithm, tested on different and real-scale power grids. Since, the visualization of the ExtCA results is still not a well-explored field, this work also presents a novel way of graphically depicting spots of weak observability using MU-oriented ExtCA.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 12