Kuramoto Theory — Mathematical Foundations¶
The Kuramoto Model¶
Yoshiki Kuramoto (1975) proposed the simplest model of synchronization in a population of coupled oscillators:
where: - θ_i ∈ [0, 2π): phase of oscillator i - ω_i: natural frequency (drawn from distribution g(ω)) - K: global coupling strength - N: number of oscillators
Order Parameter¶
The complex order parameter z = R·exp(iΨ) summarizes the collective state:
- R = |z| ∈ [0, 1]: coherence (0 = incoherent, 1 = perfect sync)
- Ψ = arg(z): mean phase
The Kuramoto model can be rewritten using z:
Each oscillator is pulled toward the mean phase Ψ with force proportional to both K and R. This creates a positive feedback loop: partial sync (R > 0) creates a stronger pull, which increases sync further.
Critical Coupling¶
For a symmetric unimodal frequency distribution g(ω), there exists a critical coupling K_c below which the incoherent state (R = 0) is stable:
For Lorentzian g(ω) with half-width Δ centered at ω₀:
Above K_c, the order parameter grows as R ~ √(K - K_c).
Ott-Antonsen Reduction¶
For Lorentzian frequency distributions, the infinite-N Kuramoto model reduces exactly to a single complex ODE (Ott & Antonsen, 2008):
Steady-state: R_ss = √(1 - 2Δ/K) for K > K_c.
This is implemented in scpn_phase_orchestrator.upde.reduction and used
by the MPC supervisor for O(1) prediction.
Generalized Coupling¶
SPO extends the basic model with:
In these SCPN forms the coupling enters as a per-pair matrix K_ij (the K_nm
matrix), which absorbs the 1/N normalisation written explicitly as (K/N) in the
classic form above: to reproduce a uniform global coupling of strength K, set every
K_ij = K/N; a heterogeneous K_nm encodes the network topology directly. The
normalisation is therefore the caller's responsibility, not applied implicitly.
Sakaguchi-Kuramoto (Phase Lags)¶
The phase lag α_ij models transport delays or asymmetric coupling.Stuart-Landau (Amplitude Dynamics)¶
Adds amplitude r_i with Hopf bifurcation: μ > 0 → r → √μ (active), μ < 0 → r → 0 (quiescent).Simplicial (Higher-Order)¶
3-body interactions produce explosive (first-order) synchronization transitions (Gambuzza et al. 2023, Nature Physics).Inertial (Second-Order)¶
The swing equation for power grid stability. Inertia m_i represents rotating mass of generators.Spectral Alignment Function¶
The Synchrony Alignment Function (Skardal & Taylor, 2016) provides a closed-form approximation of R from the Laplacian eigenstructure of K:
where λ_j are eigenvalues and v^j eigenvectors of the graph Laplacian L = D - K. This enables coupling topology optimization without ODE integration (10x faster).
Inverse Problem¶
Given observed phases θ_i(t), infer K and ω by minimising:
The L1 penalty on K promotes sparsity (topology discovery). Gradients computed via JAX autodiff through the ODE solver.
How this maps to controls in the product¶
Each mathematical form links to a concrete engineering path:
- pairwise and lagged couplings feed the standard UPDE derivatives,
- amplitude terms map to Stuart-Landau and related imprint experiments,
- higher-order terms map to simplicial coupling studies,
- inertial terms map to swing-domain workflows where stability boundaries are evaluated against grid-like operating constraints.
This model page is used to keep mathematical assumptions visible next to control behavior. The same equations inform guard thresholds, action constraints, and the conditions under which a policy is allowed to adjust coupling or phase forcing.
References¶
- Kuramoto Y (1975). Self-entrainment of a population of coupled non-linear oscillators. Lecture Notes in Physics 39:420-422.
- Ott E, Antonsen TM (2008). Low dimensional behavior of large systems of globally coupled oscillators. Chaos 18:037113.
- Skardal PS, Taylor D (2016). Optimal synchronization of directed complex networks. Chaos 26:094807.
- Gambuzza LV et al. (2023). Stability of synchronization in simplicial complexes. Nature Physics 19:1427-1434.
- Filatrella G et al. (2008). Analysis of a power grid using a Kuramoto-like model. European Physical Journal B 61:485-491.
- O'Keeffe KP et al. (2017). Oscillators that sync and swarm. Nature Communications 8:1504.