Left-inversion soft-sensor of synchronous generator

Abstract

SUMMARY The online measurement of some crucial variables in synchronous generator (such as the power angle, the d-axis and q-axis transient electric and magnetic fields, and the d-axis and q-axis stator circuit currents) is very important in many applications, but they are difficult to be measured by ordinary sensors. In the paper, these variables are estimated (soft-sensed) on the basis of the online directly measurable variables of generator (such as the active and reactive powers and the amplitude of current). First, a new model of synchronous generator is proposed, which fully includes the online measurable variables of synchronous generator and is suitable for designing soft-sensors. Second, aiming at the specialty of the proposed model (a typical kind of nonlinear differential-algebraic subsystem), the original left-inversion soft-sensing algorithm proposed in a previous paper is expanded. Then a left-inversion soft-sensor is designed while the fourth-order transient practical model of synchronous generator is considered, and all of the to-be-estimated variables are soft-sensed. Finally, the simulation is performed on the basis of MATLAB/SimPowerSystems (MathWorks, Natick, MA, USA), in which the sixth-order subtransient practical model of generator is used to simulate the real generator, and the results demonstrate the validity of the proposed method. Copyright © 2011 John Wiley & Sons, Ltd.