Minimum information loss based state estimation for power systems

Author

Sun, Hongbin ; Gao, Feng ; Zhang, Boming

Author_Institution

Dept. of Electr. Eng., Tsinghua Univ., Beijing

fYear

0

fDate

0-0 0

Abstract

The mathematical base of state estimation (SE) method for power systems is studied based on information theory. A more general SE method based on the minimum information loss (MIL) principle is proposed. The new method is suitable for different probability distributions of measurement errors, it can utilize different types of information including that of analog and digital measurements, and a unified estimation of power flow and network topology can be performed by it. It is proved that weighted least square (WLS) and weighted least absolute value (WLAV) estimation methods are both special cases of the MIL estimation. Based on the MIL method, it is concluded that large magnitude currents are the approximate condition of WLS estimations for current magnitude measurements commonly used in SE´s for sub-transmission and distribution networks. A statistical study of a small test case and a numerical experiment in a real power network are carried out

Keywords

electric current measurement; information theory; least squares approximations; load flow; power system state estimation; statistical distributions; MIL estimation; current magnitude measurements; information loss minimization; information theory; network topology estimation; power flow estimation; power systems state estimation; probability distributions; weighted least absolute value estimation methods; weighted least square estimation methods; Fluid flow measurement; Information theory; Least squares approximation; Load flow; Measurement errors; Power measurement; Power system measurements; Power systems; Probability distribution; State estimation; Power systems; SE; current measurement; information theory; topology error;

fLanguage

English

Publisher

ieee

Conference_Titel

Power Engineering Society General Meeting, 2006. IEEE

Conference_Location

Montreal, Que.

Print_ISBN

1-4244-0493-2

Type

conf

DOI

10.1109/PES.2006.1708918

Filename

1708918