L. Wehenkel
Y. Jacquemart, B. Heilbronn, P. Pruvot
Paper published in The International Journal of Engineering
Intelligent Systems for Electrical Engineering and
Communications, Vol. 2 No 4 Dec. 1994, pp. 233-245
Keywords - Voltage security; statistical data analysis; machine learning; pattern
recognition; artificial neural networks.
,
T. Van Cutsem ,
M. Pavella
Abstract:
This paper provides a comparative study of various computer based
learning approaches to power system voltage security
assessment, on the basis of three systems : an academic 7-bus system,
a 320-bus EHV model of the Brittany subsystem of Electricité de
France, and a 1250-bus EHV+HV model of the latter system. The three
classes of learning methods, namely machine learning, artificial
neural networks and statistical pattern recognition, are found to have
complementary features. It is thus proposed to combine them in a tool
box, so as to exploit in an optimal way the large statistical data
bases of security information which may be generated thanks to the
rapid growth of computing power.
Wed Jan 18 18:48:41 MET 1995