Parameter identifiability analysis of power system transient models based on profile likelihood

The reliability of power system dynamic simulation and transient stability assessment depends on the accuracy of parameters. Parameter identification is a crucial way of obtaining accurate parameters. However, the assessment of identifiability should be carried out a proiri. This paper demonstrates a new approach to evaluate the identifiability of power system transient model parameters, which considers both the model structure and the input/output data. By exploiting the profile likelihood, confidence intervals of each parameter can be established, based on which, the identifiability indices are calculated. Numerical tests are conducted accordingly to demonstrate the performance of the proposed approach.