Universal Partial Entrainment Equation

• The Universal Partial Entrainment Equation is a reduced model that captures how a nonlinear system becomes partly locked to an external signal based on symmetry and coupling conditions.
• It distinguishes regimes of full locking, oscillatory entrainment, and drift by employing low-dimensional dynamics in settings like localized patterns and Kuramoto systems.
• The equation delineates entrainment boundaries using concepts such as Arnold tongues, invariant semi-balls, and effective phase response curves, providing practical insights into synchronization.

Searching arXiv for the specified papers and closely related work on entrainment and partial entrainment. The Universal Partial Entrainment Equation is a symmetry-reduced dynamical description of how a driven nonlinear system becomes only partly locked to an external signal. In the literature represented here, the phrase refers not to a single equation valid for all oscillatory media, but to a recurring reduced-form template that captures the slow dynamics of relative coordinates under weak or effectively weak forcing. In localized pattern-forming systems with broken translation and internal phase symmetry, the universal object is a two-dimensional locking system for relative position and relative phase (Mizrahi et al., 2022). In mean-field Kuramoto systems with i.i.d. natural frequencies, the corresponding universal relation is an implicit equation for the entrained fraction of oscillators in the thermodynamic limit (Bronski et al., 2020). In strongly amplitude-modulated high-frequency forcing of limit-cycle oscillators, the relevant reduced equation is a scalar phase-difference dynamics obtained after averaging and expressed through an effective phase response curve (Pyragas et al., 2015). These formulations differ in state variables and assumptions, but share a common role: they separate full locking from drift, delimit entrainment regions, and identify partially entrained regimes through reduced collective dynamics.

1. Conceptual scope and terminology

The term partial entrainment denotes regimes in which external forcing does not rigidly lock every relevant degree of freedom, yet some collective structure remains bounded or synchronized. In the Kuramoto setting, this means that only a subset of oscillators becomes mutually phase-locked or remains within a bounded phase spread, while others drift (Bronski et al., 2020). In the localized-pattern setting, the reduced dynamics has two soft coordinates, position and internal phase, so one may distinguish full entrainment, non-entrainment, and partially entrained regimes associated with bounded but non-stationary attractors such as limit cycles (Mizrahi et al., 2022). In the AMHF phase-reduction setting, partial entrainment corresponds to the absence of a stable fixed point in the reduced phase-difference equation and the consequent phase drift or phase slipping relative to the envelope frequency (Pyragas et al., 2015).

A useful unifying interpretation is that a “universal partial entrainment equation” is a low-dimensional normal-form-like description that emerges after eliminating fast or strongly contracting degrees of freedom. This suggests that universality is structural rather than literal: the reduced dynamics is determined by symmetry class, coupling architecture, and forcing protocol, whereas the detailed coupling functions depend on the underlying system.

2. Symmetry-based formulation for localized patterns

For localized patterns in one spatial dimension, the starting point is a pattern-forming PDE

$$\frac{\partial \Psi_u}{\partial t} = \mathcal{N}[\Psi_u],$$

with space-translation symmetry and a continuous internal symmetry (Mizrahi et al., 2022). The paper specializes to internal phase symmetry, represented by global phase rotations

$$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$

A localized stationary pattern $\psi_0(x)$ breaks both translation invariance and internal phase symmetry, which yields two zero modes of the linearized operator: a translation zero mode and a phase zero mode. Under the assumption that the rest of the spectrum lies strictly in the left half-plane, the long-time dynamics is dominated by the collective coordinates associated with these zero modes, namely the pattern position $\xi(t)$ and internal phase $\varphi(t)$ (Mizrahi et al., 2022).

Weak external forcing is introduced as a localized signal moving with velocity disparity $v$ and rotating with frequency detuning $\omega$. Projection onto the adjoint zero modes yields evolution equations for the collective coordinates, and in terms of the relative position and relative phase,

$y = \xi - v t, \quad \theta = \varphi - \omega t,$

the universal locking dynamics takes the form

$$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$

For phase symmetry, the coupling functions have the symmetry-constrained structure

$$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$

so that the reduced system becomes

$$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$0

This equation is explicitly identified in the reconstruction as the universal partial entrainment equation for localized patterns with broken internal phase symmetry (Mizrahi et al., 2022).

The universality claim is precise. Any one-dimensional pattern-forming system with translation invariance, internal phase symmetry, a localized pattern breaking both symmetries, and weak localized forcing periodic in internal phase reduces to a two-dimensional slow system of this structural form (Mizrahi et al., 2022). The specific PDE affects only the numerical profiles of $$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$1, $$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$2, $$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$3, and $$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$4, not the form of the reduction.

3. Full locking, oscillatory locking, and partial entrainment

In the symmetry-based two-degree-of-freedom system, a fully entrained state is a fixed point $$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$5 satisfying

$$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$6

Such a state locks both relative position and relative internal phase to constants (Mizrahi et al., 2022). Stability is determined by the Jacobian

$$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$7

with the usual conditions $$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$8 and $$G_\varphi \Psi = (e^{i\varphi}\psi, e^{-i\varphi}\psi^*), \quad \varphi \in \mathbb{R}.$$9 (Mizrahi et al., 2022).

Partial entrainment enters when no stable fixed point exists but the reduced system still supports bounded attractors. The paper states that “the occurrence of Hopf bifurcations implies that oscillatory ‘breather’ entrainment can be realized on limit cycles of the locking system” (Mizrahi et al., 2022). In that case, neither relative position nor relative phase is frozen at a constant value, but both can remain bounded in a periodic or more complicated orbit. This is the paper’s natural notion of partial entrainment in a multi-degree-of-freedom context.

A related but distinct picture appears in the AMHF reduction of a single forced limit-cycle oscillator. There, the averaged phase-difference equation is

$\psi_0(x)$0

where $\psi_0(x)$1 is the mismatch between envelope and natural frequency (Pyragas et al., 2015). Full entrainment corresponds to a stable fixed point of this scalar equation. Outside the corresponding Arnold tongue, the phase difference drifts monotonically, producing phase slips. The source text describes this as a natural notion of “partial entrainment” or incomplete locking in the AMHF setting (Pyragas et al., 2015). This suggests a broad dynamical interpretation: partial entrainment is the regime in which the reduced variables remain influenced by the drive but do not settle into complete frequency and phase locking.

4. Kuramoto formulation and the thermodynamic-limit equation

In the Kuramoto model

$\psi_0(x)$2

partial phase-locking is treated in two complementary ways (Bronski et al., 2020). In finite $\psi_0(x)$3, the strong notion is defined through a subset $\psi_0(x)$4 for which the internal phase spread becomes asymptotically small, measured by the semi-norm

$\psi_0(x)$5

The existence of an invariant semi-ball and a smaller attracting semi-ball around a phase-locked core yields a rigorous criterion for a partially phase-locked subset of size $\psi_0(x)$6 (Bronski et al., 2020).

The central finite-$\psi_0(x)$7 quantities are

$\psi_0(x)$8

and

$\psi_0(x)$9

If

$\xi(t)$0

then there exists a subset $\xi(t)$1 with $\xi(t)$2 and a vector $\xi(t)$3 such that a forward-invariant semi-ball of radius $\xi(t)$4 and a smaller attracting semi-ball of radius $\xi(t)$5 exist around $\xi(t)$6 in the $\xi(t)$7-coordinates (Bronski et al., 2020). Full phase-locking is the special case $\xi(t)$8.

In the large-$\xi(t)$9 random-frequency setting, the paper passes to a deterministic entrainment criterion. For i.i.d. natural frequencies with symmetric unimodal density $\varphi(t)$0, the minimal possible frequency spread for a fraction $\varphi(t)$1 of oscillators converges to a deterministic quantile function $\varphi(t)$2 defined by

$\varphi(t)$3

The De Smet–Aeyels-based entrainment criterion becomes

$\varphi(t)$4

and the equality

$\varphi(t)$5

is explicitly proposed as the best candidate for a universal partial entrainment equation for the Kuramoto model with i.i.d. frequencies (Bronski et al., 2020).

This equation is universal only within the specified class: sinusoidal coupling, mean-field scaling, complete graph, and symmetric unimodal frequency distributions (Bronski et al., 2020). Its function is not to evolve a state variable in time, but to determine the onset and size of partially entrained clusters in the thermodynamic limit.

5. Quantitative structure of locking and entrained fractions

The different formulations produce different observables, but all serve to delineate entrainment domains. In the localized-pattern equation, eliminating the phase variable from the fixed-point conditions yields an explicit relation between $\varphi(t)$6 and the relative offset $\varphi(t)$7,

$\varphi(t)$8

which describes the locking domain in the $\varphi(t)$9 plane (Mizrahi et al., 2022). Saddle-node and Hopf bifurcation curves delimit stable locking regions, and because $v$0 and $v$1 depend linearly on the forcing amplitude, the locking domains scale linearly with amplitude and form conical Arnold tongues in $v$2 space (Mizrahi et al., 2022).

In the Kuramoto thermodynamic limit, the main quantitative result is a pair of deterministic bounds for the largest partially entrained cluster. If $v$3 is the smallest solution of

$v$4

then, with high probability as $v$5,

$v$6

The lower bound comes from the rigorous partial-entrainment criterion, while the upper bound follows from the necessary condition that any entrained cluster must have natural frequencies lying within an interval of width $v$7 (Bronski et al., 2020). The paper reports that in numerical experiments the observed size of the largest entrained cluster is predicted extremely well by the upper bound (Bronski et al., 2020). A plausible implication is that the rigorous lower bound is conservative, whereas the interval-counting upper bound is close to asymptotically sharp for at least some frequency laws.

In the AMHF setting, the threshold for entrainment is determined by the extrema of the averaged coupling function $v$8. For $v$9, locking first appears when $\omega$0 can be balanced by $\omega$1; for $\omega$2, it is controlled by $\omega$3 (Pyragas et al., 2015). This produces Arnold tongues in the $\omega$4 plane. Although the reduced equation is scalar rather than multidimensional, its role is analogous: it partitions parameter space into complete locking and drift.

6. Model examples and explicit realizations

The localized-pattern framework is instantiated in the cubic–quintic complex Ginzburg–Landau equation

$\omega$5

driven by

$\omega$6

For Gaussian forcing $\omega$7, the reduced equations reproduce the existence and structure of locking domains in the $\omega$8 plane, multistability, and the presence of Hopf and Bogdanov–Takens bifurcations (Mizrahi et al., 2022). Broad forcing yields unimodal $\omega$9 and a single velocity-locking interval for each $y = \xi - v t, \quad \theta = \varphi - \omega t,$0, while narrow forcing can produce bimodal $y = \xi - v t, \quad \theta = \varphi - \omega t,$1 and more complex locking domains with overlapping regions (Mizrahi et al., 2022).

The Kuramoto analysis provides explicit examples for Gaussian and Cauchy frequency distributions. For $y = \xi - v t, \quad \theta = \varphi - \omega t,$2,

$y = \xi - v t, \quad \theta = \varphi - \omega t,$3

and numerically $y = \xi - v t, \quad \theta = \varphi - \omega t,$4, with $y = \xi - v t, \quad \theta = \varphi - \omega t,$5 at threshold (Bronski et al., 2020). For the Cauchy distribution with scale $y = \xi - v t, \quad \theta = \varphi - \omega t,$6,

$y = \xi - v t, \quad \theta = \varphi - \omega t,$7

and numerically $y = \xi - v t, \quad \theta = \varphi - \omega t,$8, with $y = \xi - v t, \quad \theta = \varphi - \omega t,$9 (Bronski et al., 2020). These values quantify the onset of partial entrainment in the specific sense formalized by the paper.

In the AMHF reduction, the Stuart–Landau oscillator yields

$$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$0

with $$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$1, and for the chosen square-wave envelope the extrema of the averaged coupling are

$$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$2

which gives the threshold

$$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$3

Because $$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$4 changes sign, both upward and downward frequency locking are possible (Pyragas et al., 2015). In the Morris–Lecar example, $$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$5 is positive almost everywhere, so entrainment is effective only for $$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$6, with threshold laws

$$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$7

for the harmonic envelope, and

$$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$8

for the square envelope (Pyragas et al., 2015).

7. Universality claims, limitations, and common misconceptions

A common misconception is that the Universal Partial Entrainment Equation is a single canonical formula applicable unchanged across oscillator networks, limit-cycle systems, and spatially localized patterns. The sources do not support that interpretation. Instead, they present three distinct universal structures, each tied to a different reduction principle.

In (Mizrahi et al., 2022), universality is symmetry-based. The equation is universal for one-dimensional localized patterns with translation invariance, internal phase symmetry $$\frac{dy}{dt} = c_\xi(y,\theta) - v, \qquad \frac{d\theta}{dt} = c_\varphi(y,\theta) - \omega.$$9, a localized pattern breaking both, and weak localized forcing periodic in phase. Its validity rests on weak forcing, a spectral gap, localized pattern and forcing, and the neglect of higher-order terms in the small parameter $$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$0. Strong forcing or loss of pattern stability lies outside the reduction (Mizrahi et al., 2022).

In (Bronski et al., 2020), universality is distributional and mean-field in character. The implicit equation

$$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$1

is universal only for the Kuramoto class with sinusoidal coupling, mean-field scaling, complete graph, and i.i.d. symmetric unimodal frequencies. Different coupling functions, graph topologies, or non-unimodal frequency laws would alter the algebraic factor $$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$2, the optimal interval in frequency space, or both (Bronski et al., 2020).

In (Pyragas et al., 2015), universality arises from averaging under strong amplitude-modulated high-frequency forcing. The reduced phase equation is universal for stable limit-cycle oscillators in the high-frequency regime $$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$3, with sufficiently small scaled coupling $$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$4 for the second-order Taylor expansion to remain valid even though $$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$5 can be large. Its central object is the effective PRC

$$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$6

or its multi-component analogue (Pyragas et al., 2015).

Another misconception is to equate partial entrainment with weak order parameter coherence alone. The cited works use technically sharper notions. In the Kuramoto paper, partial phase-locking is tied to invariant and attracting semi-balls for a subset $$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$7, and partial entrainment in the De Smet–Aeyels sense requires a bounded phase spread independent of $$c_\xi(y,\theta) = g(y)\cos\bigl[\theta - \alpha(y)\bigr], \qquad c_\varphi(y,\theta) = h(y)\cos\bigl[\theta - \beta(y)\bigr],$$8 (Bronski et al., 2020). In the localized-pattern paper, partial entrainment is associated with bounded attractors of the reduced two-dimensional dynamics rather than merely nonzero response to forcing (Mizrahi et al., 2022). In the AMHF phase-reduction paper, partial entrainment is inferred from nonlocked drift outside Arnold tongues rather than introduced as a formal theorem (Pyragas et al., 2015).

Taken together, these works indicate that the most robust meaning of a Universal Partial Entrainment Equation is a reduced collective equation whose solutions distinguish fixed-point locking from bounded nonstationary locking and from drift. In that sense, the equation is universal within a class defined by symmetry, coupling law, and asymptotic regime, but not across all entrainment phenomena without qualification.