Enhancing Empirical Modal Extraction by Logarithmic Scaling and Normalization of Multi-Area Signal Measurements

This paper presents a hierarchical scaling and normalization method to multi-area signal measurements that dynamically relates both the angular phase domain and its amplitude in empirical analysis of power system oscillations. The proposed approach combines logarithmic relations and the Hilbert transform to derive an effective multi-area data scaling and normalization method, improving objectively the numerical performance and reliability of data-based modal extraction algorithms. This method is developed in order to numerically minimize multiscale angular phase and amplitude effects in decomposition and identification processes of inter-area oscillation modes. It employs conventional data-based analysis algorithms on interconnected power systems to achieve this objective. Results show that the presented method guarantees the most effective description of interscale interaction effects and fluctuations among detected modal oscillation patterns, enhancing empirical modal extraction for inter-area electromechanical modes.