An iterative dynamic state estimation and bad data processing

Abstract This paper presents an iterative dynamic state estimation including detection and identification of bad data. The bad data are detected by tests of asymmetry on the updated information, and the identification of measurements with gross errors is made by finding the normalized updated information with the largest standard deviation. The iterative procedure avoids the recalculation of the Kalman gain as well as the covariance matrix of the estimation errors. In the estimation process the Riccati equation is solved by the double algorithm method, working only with its diagonal part. Two main types of test were carried out: on the iterative model itself; and on the bad data processing. The model has been tested on part of a large interconnected pool in the eastern United States.