Power System Tracking State Estimation and Bad Data Processing

In this paper the tracking of the time yarying power system static state is analysed. The relative importance of the small measurement noise and the eventual occurrence of gross error in the measurements is used as a design criterion for the development of fast tracking estimators. The basic approach uses a pre-estimation bad data detection/ elimination scheme based on the exponential smoothing of past estimations and logical checks followed by an estimation stage. Three different estimation algorithms were tested: plain weighted least squares, quadratic square-root and linear criterion. Numerical results showing the performance of the three estimators in a simulated test case are also presented.