What Is Harmonic Analysis?

Harmonic analysis is a branch of mathematics and applied signal processing concerned with representing functions or signals as superpositions of simpler oscillatory components, principally sinusoids. The field generalizes the classical theory of Fourier series, which decomposes periodic functions into sums of integer-multiple frequency components, to broader classes of functions and more abstract mathematical settings. In engineering practice, harmonic analysis provides the theoretical foundation for every technique that extracts frequency-domain information from time-domain measurements, from power systems monitoring to audio processing.

The discipline draws on mathematical analysis, linear algebra, and group theory, with roots in the work of Joseph Fourier in the early nineteenth century. Its practical reach extends into electrical engineering, acoustics, communications, and numerical methods, making it one of the most broadly applied areas of mathematics in engineering.

Fourier Analysis and the Fast Fourier Transform

Classical harmonic analysis is built on the Fourier series and the Fourier transform, which express a signal as a continuous integral or discrete sum of complex exponentials weighted by their spectral coefficients. For digital signals, the discrete Fourier transform (DFT) computes these coefficients exactly for a sampled sequence, while the fast Fourier transform (FFT) reduces the computation from O(N²) to O(N log N) operations, a reduction that made real-time spectral analysis computationally practical. The FFT is the workhorse algorithm in spectrum analyzers, audio equalizers, radar signal processors, and power-quality monitoring instruments. Data Compression and Harmonic Analysis by Daubechies and DeVore provides a rigorous treatment of how harmonic analysis concepts underpin modern compression algorithms, connecting pure theory to engineering practice.

Wavelet Transforms

Where the Fourier transform provides a purely global frequency representation with no time localization, wavelet transforms decompose a signal into components that are simultaneously localized in both time and frequency. A wavelet basis consists of scaled and shifted copies of a mother wavelet, and the wavelet coefficients reveal when frequency content appears and disappears in a non-stationary signal. This property makes wavelets particularly suited to transient detection, fault diagnosis in power systems, and image compression. The JPEG 2000 standard and certain radar pulse analysis systems rely on wavelet decompositions rather than FFTs. Discrete wavelet transform for harmonic analysis of electromagnetic transients in 500 kV transmission systems demonstrates the advantage of wavelet-based analysis over FFT when harmonics are embedded in non-stationary switching transients, which a standard DFT smears across frequency bins.

Harmonic Analysis in Power Systems

In electrical engineering, harmonic analysis refers specifically to computing the harmonic content of voltages and currents in power networks. A sinusoidal supply distorted by nonlinear loads, such as rectifiers, variable-frequency drives, and switch-mode power supplies, contains energy at integer multiples of the fundamental supply frequency (50 or 60 Hz). The IEEE 519-2014 standard defines voltage and current harmonic distortion limits at the point of common coupling and specifies measurement procedures grounded in DFT-based analysis. Power systems engineers use harmonic analysis to size filters, assess transformer derating, and diagnose resonance conditions that can amplify particular harmonic orders to damaging levels.

Applications

Harmonic analysis has applications in a wide range of fields, including:

• Power quality assessment and harmonic limit compliance in distribution and transmission networks
• Audio processing, music analysis, and room acoustics characterization
• Radar, sonar, and communications signal detection and modulation analysis
• Medical imaging, particularly MRI reconstruction and ultrasound signal processing
• Seismology and geophysical signal analysis for earthquake and subsurface characterization