The statistical interval estimation of the mean and the hypothesis testing of population proportions for transformer tap position estimation

Abstract Frequent changes in the modern electrical power systems necessitate network state estimating for energy management system. Efficient network state estimation depends on deriving a correct model of the power system. To best model, the precise values of system parameters should be known. Therefore, many studies have paid attention to estimate unknown network parameters. Among them, transformer tap position (TTP) is one of the most important parameters ignored in the power system. This paper estimates TTP based on the statistical interval estimation of the mean and the hypothesis testing of population proportions. We exploit statistical rules jointly with weighted least squares (WLS) method to efficiently obtain unknown TTPs. Because of the non-recursive form, this method does not suffer the calculation divergence problem and is also simpler than recursive methods such as Bayesian approach. The proposed approach is also compared with Seidel type Recursive Bayesian Approach (SRBA) in the IEEE 14-bus system to show its advantages. Furthermore, it is applied on a real power system.