Simultaneous Identification and Correction of Multiple Network Parameter Errors by Mixed-Effects Models

Abstract

Accurate parameter values are essential for the secure and stable operation of power systems. Parameters may be erroneous in real power systems, while existing methods may provide false results in the case of multiple interacting parameter errors. This paper proposes an LME-HTI method for the simultaneous identification of multiple parameter errors based on a linear mixed-effects (LME) model and hypothesis testing identification (HTI). The residual equations from multiple-snapshot state estimation are used to formulate the LME model in which the parameter errors are considered as the fixed effects and the measurement errors are considered as the random effects. By solving the LME model, all the parameter errors and the variances of the measurement errors can be simultaneously estimated. Then, hypothesis testing is performed to infer whether each parameter error is zero. An extended state augmentation (ESA) method is proposed to further remove the misidentified but correct parameters from the suspicious parameter set while correcting the left erroneous parameters by constrained nonlinear optimization. In this way, only a limited number of parameters with strong evid