Vicsek-Kuramoto Model: Unified Collective Behavior

• The Vicsek-Kuramoto Model is a unified framework combining spatial alignment and phase synchronization to model swarming, vortex lattices, and emergent order.
• It employs a set of stochastic differential equations, incorporating noise and local interactions, to capture transitions from global synchrony to complex pattern formation.
• The model informs active matter research by linking microscopic dynamics to macroscopic phenomena such as crystalline swarming and phase-locked clusters.

The Vicsek-Kuramoto (VK) model constitutes a unification of two paradigms for collective behavior: the Vicsek model for self-propelled particle alignment and the Kuramoto model for phase synchronization. In the VK framework, each agent is endowed with both a spatial position and an internal phase, and interactions combine spatially local alignment (Vicsek mechanism) with coupling of the phase or natural frequency (Kuramoto mechanism). This generalization captures a broad range of spontaneous synchronization and pattern formation phenomena in active matter systems, including swarming, flocking, vortex lattices, and phase-locked clusters, and serves as a prototype for understanding the emergence of order due to alignment, frustration, and noise.

1. Mathematical Formulation and Model Classes

The canonical Vicsek-Kuramoto model involves $N$ self-propelled particles indexed by $i=1,\ldots,N$, each with position $\mathbf{r}_i(t)\in\mathbb{R}^2$, phase (heading or internal clock) $\theta_i(t)\in[0,2\pi)$, and angular velocity $\omega_i(t)\in\mathbb{R}$. The dynamics typically take the form: $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$ where $\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$, $v_0$ is speed, $K_\theta$, $K_\omega$ are alignment strengths for phase and angular velocity respectively, and $i=1,\ldots,N$0, $i=1,\ldots,N$1 are independent Wiener processes for angular and frequency noise. The averages $i=1,\ldots,N$2, $i=1,\ldots,N$3 are typically taken over neighbors within distance $i=1,\ldots,N$4 via suitably normalized kernels (Merino-Aceituno et al., 18 Dec 2025).

A critical extension introduces a frustration or phase-lag parameter $i=1,\ldots,N$5, yielding

$i=1,\ldots,N$6

with $i=1,\ldots,N$7 the set of neighboring particles. This generalizes the interaction to include ferromagnetic, frustrated, and anti-aligning regimes depending on $i=1,\ldots,N$8 (Lu et al., 12 Nov 2025).

Distinct variants exist reflecting differing choices for noise (scalar/intrinsic, vector/extrinsic), alignment neighborhoods, and inclusion of confining or tilt fields (Chepizhko et al., 2010, Bertoli et al., 18 Mar 2026).

2. Mean-Field Theory and Reduction to Phase Equations

Under mean-field assumptions, the VK model reduces, for both Vicsek and Kuramoto limits, to a stochastic phase equation: $i=1,\ldots,N$9 where $\mathbf{r}_i(t)\in\mathbb{R}^2$0 is the global order parameter. For Vicsek, typically $\mathbf{r}_i(t)\in\mathbb{R}^2$1; for Kuramoto, $\mathbf{r}_i(t)\in\mathbb{R}^2$2 is drawn from a distribution. The noise amplitude $\mathbf{r}_i(t)\in\mathbb{R}^2$3 encodes intrinsic or extrinsic noise (Chepizhko et al., 2010).

Continuous (second-order) and discontinuous (first-order) synchronization transitions emerge depending on noise realization: scalar (intrinsic) noise leads to a supercritical pitchfork bifurcation, while vector (extrinsic) noise induces a subcritical jump. Weighted mixtures ("mixed noise") interpolate between these, producing a tricritical point in the $\mathbf{r}_i(t)\in\mathbb{R}^2$4 plane ($\mathbf{r}_i(t)\in\mathbb{R}^2$5) (Chepizhko et al., 2010).

3. Macroscopic and Hydrodynamic Limits

The VK model admits rigorous kinetic and hydrodynamic reductions. Let $\mathbf{r}_i(t)\in\mathbb{R}^2$6 be the single-particle density. The kinetic (mean-field) equation is

$\mathbf{r}_i(t)\in\mathbb{R}^2$7

where $\mathbf{r}_i(t)\in\mathbb{R}^2$8, $\mathbf{r}_i(t)\in\mathbb{R}^2$9 are nonlocal mean alignments in $\theta_i(t)\in[0,2\pi)$0 and $\theta_i(t)\in[0,2\pi)$1 respectively (Merino-Aceituno et al., 18 Dec 2025).

In the hydrodynamic (macroscopic) limit, with strong local alignment (small correlation length), solutions rapidly equilibrate in $\theta_i(t)\in[0,2\pi)$2 to products of von Mises and Gaussian distributions, yielding closed PDEs for the fields: $\theta_i(t)\in[0,2\pi)$3 where $\theta_i(t)\in[0,2\pi)$4 is the density, $\theta_i(t)\in[0,2\pi)$5 the mean orientation, and $\theta_i(t)\in[0,2\pi)$6 the local mean angular velocity (Merino-Aceituno et al., 18 Dec 2025).

Different angular velocity regimes—small (SOHR-S) and large (SOHR-L) typical spin—produce distinct macroscopic behaviors, including pressure-like terms and nontrivial coupling between density, orientation, and rotational fluxes (Degond et al., 2013).

4. Pattern Formation: Lattices, Waves, and Synchronization

The VK model exhibits a wide diversity of emergent spatiotemporal structures. For unfrustrated interactions ($\theta_i(t)\in[0,2\pi)$7), the dominant long-time behavior is global synchronization or swarming, with flocking directions locked and density nearly uniform.

Introduction of alignment frustration parameter $\theta_i(t)\in[0,2\pi)$8 yields a bifurcation at $\theta_i(t)\in[0,2\pi)$9: for $\omega_i(t)\in\mathbb{R}$0 the instability is at $\omega_i(t)\in\mathbb{R}$1 (global synchrony); for $\omega_i(t)\in\mathbb{R}$2 a finite-wavelength Hopf-Turing bifurcation generates hexagonal resting lattices, vortex lattices (“respiratory” cells with oscillatory collective phase motion), dual-cluster lattices (unit-cell-level $\omega_i(t)\in\mathbb{R}$3 splitting with bimodal polarization), and anti-synchronized drifting lanes for $\omega_i(t)\in\mathbb{R}$4 near $\omega_i(t)\in\mathbb{R}$5 (Lu et al., 12 Nov 2025). The lattice spacing is set by the instability wavelength, $\omega_i(t)\in\mathbb{R}$6, with cluster-scale phase dynamics determined by the mean slip rate $\omega_i(t)\in\mathbb{R}$7.

Systematic scaling shows pattern boundaries in the $\omega_i(t)\in\mathbb{R}$8 plane are determined by equalizing the cluster diameter and lattice wavelength: $\omega_i(t)\in\mathbb{R}$9 This accurately predicts the region of robust lattice order (Lu et al., 12 Nov 2025).

Traveling orientation waves and globally rotating clusters are also reported, depending on the balance of alignment strengths and noise in the phase and angular velocity channels (Merino-Aceituno et al., 18 Dec 2025).

5. Synchronization Thresholds, Confinement, and Tilts

The onset of order is characterized by critical coupling thresholds. In the presence of a confining potential ($$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$0) and angular tilt $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$1, the Itô SDEs include: $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$2 Analysis in the mean-field limit yields for the fully normalized model a critical coupling

$$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$3

Thus, confinement raises the threshold quadratically with $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$4; the tilt $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$5 enters through the correction but does not affect the threshold at $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$6 (Bertoli et al., 18 Mar 2026).

Atypical normalization of the alignment kernel (in $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$7 or $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$8 only, or unnormalized) yields different scaling relations for $$\begin{aligned} \dot{\mathbf{r}}_i &= v_0\, \mathbf{p}(\theta_i),\ d\theta_i &= \omega_i\, dt + K_\theta\, \sin(\overline\theta_i - \theta_i)\,dt + \sqrt{2\alpha^2}\,dB^i_t,\ d\omega_i &= K_\omega\, (\overline\omega_i - \omega_i)\,dt + \sqrt{2\beta^2}\,d\tilde{B}^i_t, \end{aligned}$$9, as summarized below:

Normalization	$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$0	Scaling Dependence
Fully normalized	$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$1	Noise only
Unnormalized	$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$2	Noise, interaction range
$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$3-normalized	$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$4	Same as unnormalized
$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$5-normalized	$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$6	Noise only

(Bertoli et al., 18 Mar 2026)

6. Noise-Induced Bifurcations and Tricriticality

The transition to synchronization is classified by the nature of noise present:

• Scalar (intrinsic) noise:

The stationary order parameter satisfies

$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$7

yielding a continuous (second-order) transition at $\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$8.

• Vector (extrinsic) noise:

The self-consistency equation reads

$\mathbf{p}(\theta) = (\cos\theta, \sin\theta)$9

giving rise to a subcritical (first-order) jump in $v_0$0.

• Mixed noise:

A mixture $v_0$1 has a self-consistency $v_0$2, with a region of bistability and a tricritical point at $v_0$3 (Chepizhko et al., 2010).

This structural mapping between noise and the order of phase transitions directly links patterns in Vicsek and Kuramoto models.

7. Applications and Theoretical Implications

The VK model and its macroscopic reductions have been applied to analyze pattern formation in active matter including bacterial vortex arrays, sperm-cell rings, and rod-like swimmer collectives. The ability of frustrated orientational alignment to generate crystalline lattice order without explicit spatial forces challenges established views on pattern formation in non-equilibrium systems (Lu et al., 12 Nov 2025). Hydrodynamic limits capture global synchronization but fail to sustain traveling-wave and multi-cluster regimes observed in microscopic models, pointing to the importance of finite-range effects and higher-order closures (Merino-Aceituno et al., 18 Dec 2025).

A plausible implication is that the interplay of alignment strength, frustration, and noise provides a minimal tuning mechanism for engineering targeted spatiotemporal phases—ranging from classical flocking to crystalline swarming—within active matter and synchronization-based engineering systems.

---

References:

(Chepizhko et al., 2010, Lu et al., 12 Nov 2025, Degond et al., 2013, Merino-Aceituno et al., 18 Dec 2025, Bertoli et al., 18 Mar 2026)