Eigensystem Realization Algorithm (ERA)

A data-driven, equation-free method that extracts spatio-temporal coherent structures from high-dimensional time-series data. Unlike ERA, which operates on impulse response data, DMD works directly on snapshot sequences from simulations or measurements.

• Computes eigenvalues and eigenvectors (dynamic modes) of a best-fit linear operator
• Each mode has an associated growth/decay rate and oscillation frequency
• Particularly effective for analyzing ambient synchrophasor data without requiring a distinct ringdown event
• Variants like Exact DMD and Sparsity-Promoting DMD improve noise robustness and mode selection

A classical signal processing method that fits a sum of exponentially damped sinusoids to a uniformly sampled signal. ERA can be viewed as a multi-input/multi-output generalization of Prony's method.

• Directly estimates frequency, damping, amplitude, and phase for each mode
• Works on a single output signal (e.g., one PMU channel)
• Sensitive to noise and model order selection
• Often used as a benchmark for validating ERA results in ringdown analysis
• Extended Prony methods incorporate least-squares fitting for improved accuracy

The process of extracting modal parameters from the transient oscillatory response of the grid following a sudden disturbance such as a line trip, generator outage, or fault clearing.

• ERA is one of the primary algorithms applied to ringdown data
• Captures inter-area oscillation modes with frequencies typically between 0.1–2.0 Hz
• Requires high-resolution PMU data sampled at 30–60 samples per second
• Key outputs: mode frequency, damping ratio, and mode shape
• Damping ratios below 3–5% indicate poorly damped modes requiring operator attention

A dimensionless parameter quantifying how rapidly an electromechanical oscillation decays, expressed as a percentage of critical damping.

• Calculated from ERA-identified eigenvalues: ζ = −σ / √(σ² + ω²)
• σ is the real part (decay rate) and ω is the imaginary part (angular frequency)
• Damping ratio > 5%: Generally acceptable stability margin
• Damping ratio < 3%: Poorly damped mode — requires corrective action
• Negative damping ratio: Unstable growing oscillation — emergency condition
• ERA provides confidence intervals for damping estimates through the Eigensystem Realization Algorithm with Data Correlation (ERA/DC) variant

A complex vector describing the relative amplitude and phase of oscillation participation across different generators or buses for a specific system mode.

• ERA extracts mode shapes from the identified state-space model's eigenvectors
• Mode shape magnitude indicates which generators dominate the oscillation
• Phase angles (typically 0° or 180°) reveal coherent groups swinging against each other
• Visualized on geographic maps to show inter-area oscillation patterns
• Critical for designing targeted damping controllers and validating dynamic models
• Mode shapes from ambient data (via ERA variants) enable continuous stability monitoring without waiting for disturbances

The ability of the power system to maintain synchronism under small disturbances, analyzed through linearization of the nonlinear differential-algebraic equations around an operating point.

• ERA provides a measurement-based alternative to model-based eigenvalue analysis
• The identified state-space model approximates the linearized system dynamics
• ERA eigenvalues correspond directly to the system's electromechanical modes
• Enables validation of dynamic simulation models against real grid behavior
• Discrepancies between ERA-identified modes and model-predicted modes indicate model calibration errors requiring correction