Feature Adaptive Generator Model Calibration

This paper addresses a key challenge in practical applicability of automatic model calibration, namely determining a reasonable set of parameters from grid events. The proposed approach extracts features from the event data and model validation results based on a change point detection algorithm. A similarity-based parameter screening approach together with weighted nonlinear least square optimization are designed to allow for easy integration of the extracted features. The proposed approach shows superior performance in terms of reasonable parameter choice, response accuracy, and computational speed, based on tests with NERC synthetic data and WECC field data.