Application of sequential Response Surface Methodology to the power system state estimation

Problems of estimation in statistical inference are mainly generalized by the samples. Assigned numerical values to a population parameter will expect a close estimation based on the available sample. In this paper, a use of sequential response surface methodology (sequential-RSM) is proposed for the power system state estimator solution. Offline Data (measurements and the state variables) are collected for building regression model (RM). Since each available measurement takes consideration of dynamic nature of bus load in the respective bus, each state variable is given its regression model for the solution. Finally, the model validation based on reasonableness of the regression coefficients is analyzed by checking a candidate model against independent data. The minimum number of measurements is made available for each trial. Results obtained when the IEEE 14-bus testing system was analyzed are presented.