Practical Application of Hilbert Transform Techniques in Identifying Inter-area Oscillations

Abstract

Disturbances in large power systems can exhibit nonlinear, time-varying behavior. Traditional modal identification from field data is via techniques, such as Prony analysis, which assume data stationarity. The Hilbert transform and analytic function can be used to analyze inter-area oscillatory behavior of power systems with the stationarity assumption relaxed. However, reducing the data to simple numerical results can be achieved more effectively when stationarity is assumed. The application process is not straightforward and subtle changes can yield considerable variation in the results observed. An example is the effect of discrete time calculations of the Hilbert transform over a window of finite length. Application of the newer modal identification technique, Hilbert analysis, is examined relative to the more established Prony analysis, with particular reference to the considerable structural differences which exist between the two methods. Prony analysis yields modes which are directly expressed as exponentially modulated sinusoids, whereas the Hilbert method provides a more general solution. Synthetic and measured signals are used in the comparison. Some general conclusions are drawn from the analysis of several signals, including sets of measured field data.