A two steps procedure in state estimation gross error detection, identification, and correction

• We propose a two-step state estimation (SE) formulation.
• The gross error analysis step uses the CME as objective function.
• The weight matrix (WM) is a percentage of the measurement magnitude.
• After the errors are corrected SE uses as WM the inverse of the measurement SD.
• The two-step state estimation is valid for multiple gross error analysis.
• The gross error detection test accuracy is a function of the ball radio magnitude.

In this paper a two steps approach for the power system state estimation is proposed. The first step is the gross error detection test when all the measurements are assumed as possible of having errors. With that assumption, a rule for the measurement’s weights as being the inverse of a constant percentage of the measurement’s magnitudes is proposed. Then, using the error as the objective function of the state estimation (SE) process to be minimized, the gross error analysis is performed. In case a gross error is detected, the Composed Measurement Error, in its normalized form (CMEN), is used to identify the measurement(s) with error(s). The measurement(s) with error(s) is corrected using the Composed Normalized Error (CNE). In the second step, the state estimation is again performed, but using as the weight for each measurement the inverse of the measurement’s standard deviation as proposed in the classical estimators. In this step, the set of measures is the original set, but with the measures flagged with gross errors replaced by their estimated values. The IEEE-14, IEEE-57 bus systems and the 45-bus equivalent of the Southern part of Brazil are used to perform the tests.