Parameter estimation under non-stationary circumstances using extended signal model

[摘要] In the context of transient processes in power system measurements, a sampledsignal is analyzed with respect to electro-magnetic influences. These aremost likely superposed sinusoids that can be found next to the fundamentalsystem sinusoid. However, such signal is not stationary, i.e. has a varyingmodel order which means that temporary and exponentially damped sinusoidsappear eventually. In our previous work, the exponential damping was onlyconsidered indirectly but not explicitly incorporated in the signal modelused. Regarding the amplitude estimation, this led to inaccuracies since theamplitude was incorrectly assumed constant within a window of samples. Inthis paper, we use ESPRIT's ability to also estimate a parameter ofexponential damping, improve the corresponding signal model, advance theamplitude estimation algorithm and back projection algorithm. Especiallyshort-term signal components can be tracked far more precise.

[学科分类] 电子、光学、磁材料