Using an on-line BSE technique for wide-area oscillations monitoring

The dynamic behavior of large interconnected power systems can be provided by wide-area measurement systems. Motivated by this, the use of on-line techniques has been developed, as well as multivariate methods have been proposed to extract useful patterns from a large data set. In this paper, a multivariable technique to identify and extract dynamic patterns from simultaneously measured data is proposed. One dynamic pattern or a limited set of dynamic patterns can be sequentially extracted. The technique combines blind source extraction with a recursive least squares adaptive filter, which exploits the temporal structure of the measured signals. In order to show the applicability of the proposed methodology, a numerical simulation of a four-machine, two-area test system is carried out. Results indicate that the method has the ability to estimate modal responses and mode shapes.