US 7,498,821 B2 · Non-linear observers in electric power networks

Abstract

In order to make accurate predictions on whether or not a power grid is stable, it is desirable to know the frequency and phase of at least each “control area” in the power grid. Each control area will ideally have power generation matched with its load. Rather than measure thousands of data points at low resolution in order to calculate potential grid instabilities, one may make few highly accurate measurements at low voltage to predict frequency and phase of high voltage transmission lines. Predicted data at the transmission points can then be used to calculate the relative stability of the power grid. The method includes the use of non-linear Luenberger observers and Kalman filters. The method can also be used to compute “Virtual measurements” for any component in the connected grid based on lower cost or easier to obtain measurement at other points on network. This method applies to any connected network of components.

Claims13

Dependent claims

A method of detecting instability in a power grid, the method comprising:

measuring a first frequency at a first location on the power grid;
measuring a first phase angle at the first location on the power grid;
comparing the first frequency and the first phase angle to a model of the power grid;
estimating a second frequency and a second phase angle at a second location on the power grid based on the first frequency, the first phase angle, and the power grid model;
computing moving averages of absolute phase difference between pairs of phase angle estimates;
comparing the moving averages with predetermined limits;
and
alerting operators to a potential power grid instability based on the moving averages.

5 more elements

Dependencies are inferred from the claim text.

Non-linear observers in electric power networks

Life of this patent

Non-linear observers in electric power networks

Life of this patent

Abstract

Claims13