Transitions of bifurcation diagrams of a forced heteroclinic cycle

Basic Information

• Paper ID: 2312.00729
• Title: Transitions of bifurcation diagrams of a forced heteroclinic cycle
• Authors: Isabel S. Labouriau, Alexandre A. P. Rodrigues
• Classification: math.DS (Dynamical Systems)
• Publication Date: January 3, 2025
• Paper Link: https://arxiv.org/abs/2312.00729

Abstract

This paper investigates periodic perturbations of attracting robust heteroclinic cycles defined on a two-dimensional sphere by reducing the analysis to the study of single-parameter families of maps on the circle. The zero set of this family forms bifurcation diagrams on a cylindrical surface. As the attracting strength of the cycle and the amplitude of periodic perturbations vary, different bifurcation diagrams and transitions between them are obtained. The paper identifies a threshold for the cycle's attracting strength, above which frequency-locked periodic solutions with arbitrarily long periods bifurcate from the cycle as the perturbation period decreases. Below this threshold, further transitions are discovered, producing frequency-locked invariant tori and frequency-locked suspended horseshoes, all arising from heteroclinic tangencies in the map family.

Research Background and Motivation

1. Research Problem

This paper studies the dynamical behavior of robust heteroclinic cycles under periodic forcing, with particular focus on structural changes in bifurcation diagrams and dynamical transition phenomena. Heteroclinic cycles are cyclic structures formed by heteroclinic orbits connecting saddle-type invariant sets. While structurally unstable in general dynamical systems without symmetry or other constraints, they can become robust in equivariant vector fields.

2. Problem Significance

• Biological Relevance: Heteroclinic cycles represent the "no-winner competition" concept, which is more biologically relevant in evolutionary game theory and other biological scenarios than "winner-takes-all" assumptions
• Intermittency Dynamics: Heteroclinic cycles and networks serve as skeletons for understanding complex dynamics, capable of generating intermittent and irregular behavior
• Practical Applications: Important applications in modeling experimental data such as C. elegans neural activity

3. Limitations of Existing Methods

Although periodic perturbations are natural in modeling many biological effects, systematic studies of time-periodic perturbations and their bifurcations are relatively scarce. Existing research primarily focuses on specific parameter ranges or special cases, lacking complete qualitative descriptions.

4. Research Motivation

The paper aims to provide a complete qualitative description of dynamics and bifurcations in the neighborhood of periodically forced heteroclinic cycles, particularly systematic analysis of different bifurcation diagram types and their transitions.

Core Contributions

1. Complete Bifurcation Analysis: Provides complete qualitative descriptions of all possible persistent bifurcation diagrams under different forcing amplitudes and saddle values
2. Threshold Determination: Identifies the golden ratio Φ = (1+√5)/2 as a critical threshold for the cycle's attracting strength
3. Frequency-Locking Phenomena: Systematically analyzes the generation mechanisms of frequency-locked periodic solutions, invariant tori, and suspended horseshoes
4. Codimension-2 Bifurcations: Identifies discrete-time Bogdanov-Takens bifurcations as organizing centers for horseshoe dynamics and singular attractors
5. Transition Mechanisms: Provides detailed descriptions of transition processes and conditions between different bifurcation diagrams

Methodology Details

Problem Setup

Study of periodically perturbed differential equations:

ẋ = x(1-r²) - αxz + βxz² + γ(1-x)sin(2ωt)
ẏ = y(1-r²) + αyz + βyz²  
ż = z(1-r²) - α(y²-x²) - βz(x²+y²)

where r² = x² + y² + z², β < 0 < α, |β| < α, γ,ω ∈ ℝ₊₀.

Model Architecture

1. Dimension Reduction Method

• Reduces the three-dimensional problem to a two-dimensional map on the cylindrical surface {(y,s) : s ∈ ℝ (mod 2π), y > 0}
• First return map: G(y,s) = (y^(δ²) + γy^(δ²-δ)(1+k sin s), s - ln y/K)
• Further simplification to a zero-point problem for single-parameter families of circle maps

2. Auxiliary Parameter Method

Introduces auxiliary parameter τ to study fixed points of the map G_τ(y,s) = G(y,s) + (0, ln τ/K), equivalent to solving the circle map family:

g(τ,s) = τ^(δ²) + γτ^(δ²-δ)(1+k sin s) - τ = 0

3. Bifurcation Diagram Classification

Systematically analyzes bifurcation diagram structures for different values of parameter k:

• k > 1: Open curve structure
• 0 < k < 1: Closed curve structure
• k = 1: Transition case
• k = 0: Limiting case

Technical Innovations

1. Systematic Classification Method

Uses contact equivalence relations to classify bifurcation diagrams, providing a complete topological classification framework.

2. Threshold Theory

Proves the mathematical rigor of the golden ratio Φ as the boundary between "weak attraction" and "strong attraction" cases.

3. Stability Analysis

Provides detailed analysis of bifurcation branch stability near fold points, identifying attracting, saddle, and completely repelling fixed points.

Experimental Setup

Theoretical Analysis Framework

This paper is primarily based on theoretical analysis, with results verified through the following methods:

1. Parameter Space Partitioning

• Partitions different parameter regions in the (δ,γ) plane
• Each region corresponds to a specific bifurcation diagram type
• Analyzes transition behavior at region boundaries

2. Geometric Analysis

• Utilizes geometric properties of the function F_δ(τ) = τ^p(δ) - τ^δ
• Determines bifurcation diagram structure through intersection points of horizontal lines with the graph
• Analyzes fold point locations and types

3. Stability Criteria

• Computes eigenvalues of Jacobian matrices
• Analyzes Hopf bifurcation and saddle-node bifurcation conditions
• Determines tangency conditions of invariant manifolds

Experimental Results

Main Results

1. Bifurcation Diagram Classification (k > 1)

• Region A: δ < Φ, γ > γ₊(δ) - Two open curves spanning the cylinder, no fold points
• Region B: δ < Φ, γ < γ₊(δ) - Two non-spanning open curves, each with one fold point
• Region C: δ > Φ - One non-spanning open curve with one upper critical fold point

2. Bifurcation Diagram Classification (0 < k < 1)

• Region W: δ < Φ, γ > γ₋(δ) - No zeros
• Region X: δ < Φ, γ₊(δ) < γ < γ₋(δ) - One contractible closed curve with two fold points
• Region Y: δ < Φ, γ < γ₊(δ) - Two non-contractible closed curves, each with two fold points
• Region Z: δ > Φ - One non-contractible closed curve with two fold points

3. Existence of Frequency-Locked Solutions

For δ > Φ and arbitrary γ > 0, there exists ω₁ > 0 such that for any ω > ω₁, pairs of frequency-locked periodic solutions appear at saddle-node points.

Ablation Studies

1. Effects of Parameter k

• k = 0: Bifurcation diagram simplifies to a circle of constant τ
• k = 1: Transition case with fold points moving to boundaries
• Variations in k lead to fundamental topological changes in bifurcation diagrams

2. Critical Role of Threshold δ = Φ

• δ < Φ: Weak attraction case, may exhibit Hopf bifurcations and invariant tori
• δ > Φ: Strong attraction case, primarily exhibits frequency-locked periodic solutions

Important Findings

1. Bogdanov-Takens Bifurcation

A codimension-2 discrete Bogdanov-Takens bifurcation point is discovered at parameters (τₘ(δ), δ, MF(δ)/(1±k)), serving as an organizing center for horseshoe dynamics.

2. Horseshoe Dynamics

Transverse intersections exist between heteroclinic tangency surfaces in certain parameter regions, producing dynamics equivalent to Smale horseshoes.

Related Work

Main Research Directions

1. May-Leonard Systems: Work by Dawes and Tsai provides foundations for the weak attraction case
2. Afraimovich et al.: Discovered invariant closed curves and symbolic dynamics under small-amplitude perturbations
3. Wang-Young Theory: Studies of singular attractors in the strong attraction case

Advantages of This Paper

• Provides complete parameter space analysis
• Determines precise threshold conditions
• Unifies theoretical frameworks for weak and strong attraction cases

Conclusions and Discussion

Main Conclusions

1. Threshold Theory: The golden ratio Φ is the precise threshold distinguishing weak and strong attraction dynamics
2. Complete Classification: Provides complete classification of all possible bifurcation diagram types and transition conditions
3. Dynamical Richness: The system exhibits diversity ranging from frequency-locking to chaotic dynamics

Limitations

1. Model-Specific: Based on a specific three-dimensional system; generalizability requires verification
2. Numerical Verification: Primarily based on theoretical analysis; lacks extensive numerical experiments
3. Stability: Stability of certain invariant tori remains incompletely determined

Future Directions

1. Generalization: Extension to more general heteroclinic networks
2. Numerical Methods: Development of effective numerical methods to verify theoretical predictions
3. Applied Research: Specific applications in biological and physical systems

In-Depth Evaluation

Strengths

1. Theoretical Rigor: Rigorous mathematical analysis with complete proofs
2. Systematicity: Provides a complete bifurcation theoretical framework
3. Novelty: Identifies new threshold phenomena and transition mechanisms
4. Clarity: Extensive figures make complex theory accessible

Weaknesses

1. Computational Complexity: Explicit computation of certain parameters is difficult
2. Experimental Verification: Lacks direct comparison with numerical simulations
3. Application Scope: Applicability to real systems requires further verification

Impact

1. Theoretical Contribution: Provides important advances in periodic perturbation theory of heteroclinic cycles
2. Methodology: Dimension reduction and bifurcation analysis methods have universal significance
3. Interdisciplinary: Significant application value in dynamical systems, biomathematics, and related fields

Applicable Scenarios

• Modeling competition dynamics in biological systems
• Study of switching dynamics in neural networks
• Analysis of convection patterns in fluid mechanics
• General research on periodically forced dynamical systems

References

This paper cites 27 important references covering classical works in heteroclinic cycle theory, bifurcation theory, May-Leonard systems, and related fields, providing a solid theoretical foundation for the research.