From THD to Causality: Epistemology of Artificial Intelligence-Based Harmonic Analysis in Hybrid Microgrids

Ana Nuril Achadiyah (Department of Electrical Engineering and Informatics, State University of Malang, 65114, Indonesia)
Arif Nur Afandi (Department of Electrical Engineering and Informatics, State University of Malang, 65145, Indonesia)
Syaad Patmanthara (Department of Electrical Engineering and Informatics, State University of Malang, 65145, Indonesia)

Energy: Jurnal Ilmiah Ilmu-ilmu Teknik
ENERGY: JURNAL ILMIAH ILMU-ILMU TEKNIK (Special Issue on Engineering Paradigm 2025 Edition)

The increasing penetration of PV, wind turbines, battery storage (BESS), and electric vehicle charging stations (EVCS) in hybrid microgrids complicates the harmonic landscape. Common practices rely on FFT-based measurements and THD/TDD indices, but source attribution and causality assignment are often uncertain. We map how epistemological positions shape how we measure, explain, and justify technical claims about harmonics. We then propose an Epistemically-Informed Harmonic AI (EPI-HAI) framework that combines standardized measurements (IEC/IEEE), physics-constrained AI modeling (KCL/KVL, impedance), XAI (SHAP/Grad-CAM), and uncertainty management to strengthen epistemic trust. A vignette of a PV–BESS–EVCS microgrid demonstrates that triangulation of evidence (n-order patterns, operating logs, line impedance) is more valid than mere spectral correlation. The three main contributions of this article are, the compilation of a map of the relationship between epistemology and methodology in harmonic analysis, the formulation of transparent and accountable physics-based artificial intelligence (AI) design principles and a discussion of pedagogical implications that can be applied in the development of power engineering curricula.