A set of signal-processing and diagnostic methods that decompose electrical waveforms into frequency components (harmonics and interharmonics), quantify their amplitudes and phases (e.g., via Fourier or time–frequency techniques), and use those measurements to assess power quality impacts and to design mitigation measures.

Principle

Non‑sinusoidal currents or voltages can be represented as superpositions of sinusoidal components at integer or non-integer multiples of the fundamental frequency; because equipment response and heating are frequency-dependent, identifying spectral components permits targeted mitigation and impact assessment.

Demonstration

Illustrative measurement — Situation: A distribution node exhibits elevated neutral currents and customer complaints of flicker. Recognition: A captured voltage/current waveform is processed with an FFT and STFT to reveal strong 3rd and 5th harmonic current components and transient interharmonics. Action: Engineers attribute peaks to identified nonlinear loads, specify a tuned filter and adjust load scheduling. Consequence: Harmonic amplitudes at the node fall below thresholds and neutral current is reduced.

Misapplication

Treating total harmonic distortion (THD) alone as a complete power‑quality descriptor and ignoring harmonic phase, individual order amplitudes, interharmonics, or non‑stationary behavior; the semantic error is reducing a frequency-domain spectrum to a single scalar without preserving actionable detail.

Consequence

Appropriate harmonic analysis enables design of filters, selection of equipment rated for harmonic heating, and mitigation strategies; misapplied analysis can lead to mis-specified filters, unresolved resonances, or overlooked intermittent distortions.

Reversal

When signals are strongly non‑stationary or contain transient, wideband disturbances, standard steady‑state Fourier analysis misrepresents temporal behavior; time‑frequency or wavelet methods are then required to capture evolving spectral content.

Boundary

Clearly within: periodic or quasi‑periodic voltage/current distortion in power systems where frequency-domain decomposition reveals sources and impacts. Boundary case: rapidly varying converter control interactions that require combined electromagnetic transient and spectral analysis. Clearly outside: pure random noise characterization where statistical time-domain metrics are primary.

Semantic Tension

Spectral Resolution ↔ Temporal Resolution — higher frequency resolution requires longer observation windows, degrading time localization and potentially hiding transient events important for diagnosis.

Synthesis

Harmonic analysis converts time‑domain distortion into a frequency‑domain signature that supports targeted mitigation, but its usefulness depends on choosing methods and windows appropriate to the stationarity and timescale of the disturbances.