Generalized Chirikov Map

Updated 11 July 2026

Generalized Chirikov map is a family of extensions to the standard map that models the transition between regular and stochastic behavior using additional kick mechanisms.
The approach employs phase-shifted secondary kicks, incommensurate spatial harmonics, and fractional memory effects to represent complex transport and stability phenomena.
These models reveal practical insights into stickiness, anomalous diffusion, and the persistence of stable islands even in highly chaotic regimes.

Searching arXiv for recent and foundational papers on generalized Chirikov maps and related extensions. arxiv_search(query="generalized Chirikov map standard map extension phase-shifted secondary kicks", max_results=10) arxiv_search(query="Sub-diffusive behavior in the Standard Map (Palmero et al., 2020)", max_results=5) arxiv_search(query="A map for systems with resonant trappings and scatterings (Artemyev et al., 2019)", max_results=5) arxiv_search(query="Fractional Equations of Kicked Systems and Discrete Maps (Tarasov et al., 2011)", max_results=5) arxiv_search(query="Determining the boundary of dynamical chaos in the generalized Chirikov map via machine learning (Chernyshov et al., 15 Sep 2025)", max_results=5) The generalized Chirikov map denotes a family of extensions of the Chirikov–Taylor standard map, the area-preserving kicked map

$p_{n+1}=p_n+K\sin x_n,\qquad x_{n+1}=x_n+p_{n+1}\quad (\mathrm{mod}\ 2\pi),$

used to model the transition between regular and stochastic dynamics. In the supplied literature, generalization is achieved by adding phase-shifted secondary kicks, incommensurate spatial harmonics, non-diffusive scattering and trapping rules, discontinuous or fractional kicks, or power-law phase advances. The notion is also used more broadly for similar area-preserving, periodically kicked Hamiltonian systems in which stickiness, islands, cantori, and resonance overlap organize transport, mixing, and the onset of chaos (Artemyev et al., 2019, Palmero et al., 2020, Chernyshov et al., 15 Sep 2025).

1. Standard-map reference framework

The standard map is the baseline against which generalized constructions are defined. It is a two-dimensional area-preserving map for phase and action variables, and it exhibits the coexistence of regular and chaotic motion. For small perturbation strength, the phase space is dominated by Kolmogorov–Arnold–Moser invariant structures; for larger perturbation strength, global stochasticity emerges, although the literature emphasized here also shows that chaos never fully eliminates all stable structures (Ermann et al., 2018, Nieto et al., 2024).

A central numerical benchmark is the known critical value for the onset of global chaos in the standard map,

$K_c \approx 0.971635,$

which is recovered in the machine-learning study of the generalized phase-shifted model (Chernyshov et al., 15 Sep 2025). At the same time, the standard map does not exhibit a boundary to chaos in the sense of complete disappearance of regular islands. A systematic numerical search at large parameter values confirms Chirikov’s prediction that islets of stability persist for arbitrarily large $K$ , with the parameter interval decaying as $\Delta K \propto K^{-1}$ , the maximal area decaying as $A \propto K^{-2}$ , and the three-dimensional volume in $(\theta,p,K)$ space decaying as $V \propto K^{-3}$ (Nieto et al., 2024).

Mixed phase space is therefore not a perturbative residue but a structural feature. In the standard map, sticky trapping near the boundaries of KAM islands and resonant satellite islands produces anomalous transport rather than simple Gaussian diffusion. This observation underlies several later generalizations that explicitly encode trapping, memory, or broad jump distributions (Palmero et al., 2020).

2. Principal constructions of generalized Chirikov maps

One direct generalization adds a second sequence of kicks with amplitude $K_\alpha$ and phase shift $\tau$ . The Hamiltonian is

$H = \frac{I^2}{2} + \cos \theta \left[ K \sum_{n=-\infty}^{\infty} \delta(t - nT) - K_\alpha \sum_{n=-\infty}^{\infty} \delta(t - T(n + \tau)) \right],$

and the corresponding discrete map is

$K_c \approx 0.971635,$ 0

It reduces to the standard map for $K_c \approx 0.971635,$ 1, and to a phase-shifted Chirikov map for $K_c \approx 0.971635,$ 2 (Chernyshov et al., 15 Sep 2025).

Another generalization is the incommensurate standard map, obtained by replacing the single kick harmonic with two or three incommensurate spatial harmonics. For two harmonics,

$K_c \approx 0.971635,$ 3

and for three harmonics,

$K_c \approx 0.971635,$ 4

Here $K_c \approx 0.971635,$ 5 and $K_c \approx 0.971635,$ 6 are irrational frequencies. This extension preserves the kicked-Hamiltonian structure while increasing the effective complexity of the potential landscape (Ermann et al., 2018).

A third construction arises from slow-fast dynamics. The slow-fast map

$K_c \approx 0.971635,$ 7

is area-preserving and is conjugate to the standard map with large parameter $K_c \approx 0.971635,$ 8 after the rescaling $K_c \approx 0.971635,$ 9. Its significance lies in converting problems of slow transport into a large- $K$ 0 standard-map regime (Blumenthal et al., 2018).

Fractional generalizations depart from the Markovian update rule altogether. For $K$ 1, the kicked equation with the fractional Riemann–Liouville derivative,

$K$ 2

induces a discrete map in which the present state depends on all previous states through the kernel

$K$ 3

For the fractional Chirikov map,

$K$ 4

so the evolution becomes non-Markovian with power-law memory (Tarasov et al., 2011).

3. Memory, stickiness, and anomalous diffusion

In the standard map, stickiness refers to the tendency for chaotic orbits to spend extended times trapped near the boundaries of regular islands before escaping into the chaotic sea. This mechanism yields anomalous, non-Gaussian transport and motivates a continuous-time random walk description with fat-tailed waiting times

$K$ 5

The associated fractional diffusion equation is

$K$ 6

with periodic-boundary solutions expressed through the Mittag–Leffler function $K$ 7 (Palmero et al., 2020).

A numerical route to the anomalous exponent uses a variant of the Ulam method to approximate the Perron–Frobenius operator. Phase space is discretized into cells, a transfer matrix $K$ 8 is built, and the decay of the leading nontrivial eigenvalue is fitted by

$K$ 9

For the standard map at $\Delta K \propto K^{-1}$ 0, the fit gives $\Delta K \propto K^{-1}$ 1, indicating strong sub-diffusive dynamics due to stickiness (Palmero et al., 2020).

Fractional discrete maps provide an explicit dynamical generalization of this transport picture. Their defining property is long-term memory: the present state evolution depends on all past states with weights represented by combinations of power-law functions. Standard maps are recovered as the special case $\Delta K \propto K^{-1}$ 2, whereas fractional maps admit long-range memory, pseudochaotic attractors, and modified relaxation and bifurcation properties (Tarasov et al., 2011).

This suggests a useful distinction within generalized Chirikov dynamics. Some extensions alter the kick geometry or resonance content while remaining Markovian; others encode anomalous transport by inserting memory kernels directly into the map. Both routes preserve the role of the map as a reduced description of nontrivial Hamiltonian transport.

4. Stability boundaries, islets, and escape

In the phase-shifted generalized Chirikov map, regular and chaotic regimes were mapped through maximal Lyapunov exponents on $\Delta K \propto K^{-1}$ 3 grids in $\Delta K \propto K^{-1}$ 4, and these Lyapunov maps were then classified by a ResNet18 architecture. The model reaches $\Delta K \propto K^{-1}$ 5 classification accuracy, reproduces the standard-map threshold $\Delta K \propto K^{-1}$ 6, and identifies two-dimensional boundaries in $\Delta K \propto K^{-1}$ 7 for fixed $\Delta K \propto K^{-1}$ 8. For $\Delta K \propto K^{-1}$ 9, the critical boundary is approximately linear, $A \propto K^{-2}$ 0; for $A \propto K^{-2}$ 1, the boundary becomes a smooth nonlinear function, the symmetry under $A \propto K^{-2}$ 2 is lost, and the chaotic wedge grows with $A \propto K^{-2}$ 3 (Chernyshov et al., 15 Sep 2025).

At large perturbation strength, stability is organized into increasingly small islets rather than removed entirely. High-resolution exit-basin computations verify Chirikov’s scaling predictions for islet area and parameter range, and extend them to a three-dimensional volume law. The reported numerical exponents are close to $A \propto K^{-2}$ 4, $A \propto K^{-2}$ 5, and $A \propto K^{-2}$ 6 for $A \propto K^{-2}$ 7, $A \propto K^{-2}$ 8, and $A \propto K^{-2}$ 9, respectively, for both period-4 and period-2 families (Nieto et al., 2024).

A qualitatively different limit is the discontinuous, “very steep” analog of the standard map,

$(\theta,p,K)$ 0

which replaces the smooth sinusoidal kick by a step function. For rational $(\theta,p,K)$ 1 and rational initial conditions $(\theta,p,K)$ 2, the arithmetic condition

$(\theta,p,K)$ 3

determines the global alternatives: if $(\theta,p,K)$ 4 is even, all orbits at $(\theta,p,K)$ 5 are bounded; if $(\theta,p,K)$ 6 is odd, there is a unique escaping trajectory and all other orbits are bounded. The escaping orbit is exceptionally slow, with numerical evidence suggesting $(\theta,p,K)$ 7 for the period length of the unbounded trajectory (Arnold et al., 2016).

These results show that generalized Chirikov dynamics need not approach chaos monotonically with increasing nonlinearity. Depending on the modification, one encounters shrinking islets, deformed stability boundaries, or sparse arithmetic escape channels rather than a single universal route.

5. Transport-oriented generalizations and separatrix models

For resonant systems with strong scatterings and trappings, a new iterative map was introduced to generalize the standard Chirikov description. The standard map is said to describe weak, diffusive scatterings, whereas the new map incorporates two additional effects: a nonzero mean drift in the slow action-like variable $(\theta,p,K)$ 8 and rare, strong trapping-induced jumps. For $(\theta,p,K)$ 9,

$V \propto K^{-3}$ 0

with $V \propto K^{-3}$ 1, $V \propto K^{-3}$ 2, and $V \propto K^{-3}$ 3. The transition between stochastic and regular dynamics is controlled by the ratio $V \propto K^{-3}$ 4, with critical scaling

$V \propto K^{-3}$ 5

Above this threshold the dynamics is more stochastic; below it, correlations grow and diffusion approximations break down (Artemyev et al., 2019).

The Lévy map $V \propto K^{-3}$ 6 is another transport-centered generalization. It is a two-dimensional nonlinear map designed to generate step lengths with tail

$V \propto K^{-3}$ 7

through

$V \propto K^{-3}$ 8

For $V \propto K^{-3}$ 9 and $K_\alpha$ 0, the map recovers the rippled-billiard map and connects to the standard map near fixed points. The onset of global chaos is determined via Chirikov’s overlap criterion, with stochasticity parameter

$K_\alpha$ 1

and global chaos when $K_\alpha$ 2 (Mendez-Bermudez et al., 2014).

The Kepler map is a power-law separatrix generalization relevant to nearly parabolic cometary motion. It updates orbital energy and phase as

$K_\alpha$ 3

with the exponent $K_\alpha$ 4 placing it in the general separatrix-map family

$K_\alpha$ 5

In this sense, the Kepler map is a power-law generalization of the Chirikov separatrix construction rather than a simple modification of the sinusoidal kick (Shevchenko, 2013).

In celestial mechanics, Chirikov diffusion has also been implemented through a Hadjidemetriou-type symplectic map for the asteroidal three-body resonance $K_\alpha$ 6. There, diffusion across the resonance is associated with the stochastic layer generated by the layer resonance, whereas diffusion along the resonance requires driving resonances and is much larger in the eccentricity-like directions (1009.3558).

The incommensurate standard map has a quantum version in which the unitary kick operator contains two or three incommensurate harmonics. For two harmonics,

$K_\alpha$ 7

and for three harmonics an additional incommensurate term is included. The two-harmonic case behaves like a 2D Anderson model in momentum space and shows an Aubry–André-type metal-insulator transition in real space at small kicks; the three-harmonic case exhibits an Anderson transition analogous to the 3D case, with numerical transition region $K_\alpha$ 8 for $K_\alpha$ 9 (Ermann et al., 2018).

A distinct quantum extension is the quantum Chirikov criterion for resonance overlap. In the two-particle box model with a small mass defect, classical resonances are translated into quantum resonance bands, but quantum mechanics excludes resonances that occupy less than one Hamiltonian eigenstate. The resonance width in quantum numbers is

$\tau$ 0

and the existence condition is $\tau$ 1. This yields a finite quantum chaos threshold,

$\tau$ 2

with undestroyed groups of eigenstates serving as quantum analogues of classical KAM tori (Yampolsky et al., 2021).

Beyond direct map generalizations, the Fibonacci trace map has been proposed as another candidate for the simplest conservative system with highly non-trivial dynamics. Restricted to the compact component $\tau$ 3 of the invariant surface $\tau$ 4 for $\tau$ 5, it is conservative and exhibits persistent homoclinic tangencies, a stochastic sea of full Hausdorff dimension as $\tau$ 6, and infinitely many elliptic islands. The literature summarized here states that it has all the essential properties previously obtained for the Taylor–Chirikov standard map (Yessen, 2015).

Taken together, these developments define the generalized Chirikov map not as a single formula but as a research program: a class of standard-map extensions, separatrix analogues, and related conservative models used to study resonance overlap, anomalous transport, mixed phase space, and quantum-classical correspondences in nonlinear Hamiltonian dynamics.