Power system state recognition using data mining algorithms

Author

Alluri, Prem ; Solanki, Sarika Khushalani ; Solanki, Jitendra ; Menzies, T.

Author_Institution

Lane Dept. of Comput. Sci. & Electr. Eng., West Virginia Univ. (WVU), Morgantown, WV, USA

fYear

2013

fDate

22-24 Sept. 2013

Firstpage

1

Lastpage

6

Abstract

This paper describes an approach to identify the state of a power system using Data Mining Algorithms. With the advent of Phasor Measurement Units (PMU), a huge amount of power system data is being stored day-to-day in the Phasor Data Collectors (PDCs). Knowledge discovery and machine learning techniques can make use of this data to extract valuable information and interesting patterns in these databases. In this paper, Symbolic Aggregate ApproXimation (SAX) is used to convert the time stamped phasor data from the PMUs into symbolic strings and Data Mining (DM) algorithms are used to predict the current state of a power system. Along with the state, the location and cause of disturbance is also identified in minimal time. The effectiveness of different DM algorithms for determining state of a power system is also shown using the measures of accuracy, precision and recall.

Keywords

data mining; learning (artificial intelligence); phasor measurement; power engineering computing; power system faults; DM algorithms; PDC; PMU; SAX; current state prediction; data mining algorithms; knowledge discovery; machine learning techniques; phasor data collectors; phasor measurement units; power system data; power system state recognition; symbolic aggregate approximation; symbolic strings; time stamped phasor data; Accuracy; Data mining; Phasor measurement units; Power systems; Prediction algorithms; Time series analysis; Training;

fLanguage

English

Publisher

ieee

Conference_Titel

North American Power Symposium (NAPS), 2013

Conference_Location

Manhattan, KS

Type

conf

DOI

10.1109/NAPS.2013.6666949

Filename

6666949