A complete classification of control sets for singular linear control systems on the Heisenberg group

Basic Information

• Paper ID: 2510.08751
• Title: A complete classification of control sets for singular linear control systems on the Heisenberg group
• Authors: Adriano Da Silva (Universidad de Tarapacá), Okan Duman (Yildiz Technical University), Anderson Felipe Penagos Rojas (Universidade Estadual de Campinas)
• Classification: math.OC (Optimization and Control)
• Publication Date: October 13, 2025
• Paper Link: https://arxiv.org/abs/2510.08751

Abstract

This paper investigates control sets for linear control systems associated with singular derivatives on the Heisenberg group. Under Lie algebra rank conditions, a complete characterization of these control sets is provided through analysis of the trace and determinant of relevant 2×2 submatrices.

Research Background and Motivation

Problem Background

1. Limitations of Traditional Research: Traditional studies of linear control systems have primarily focused on Euclidean spaces. However, when the underlying space possesses non-trivial geometric structure (such as Lie groups), the dynamical behavior exhibits characteristics without counterparts in Euclidean settings.
2. Importance of Control Systems on Lie Groups: Linear control systems on Lie groups have deep connections with differential geometry, representation theory, and nonlinear dynamics. Jouan proved that every control-affine system with complete vector fields is equivalent to a linear control system on a Lie group or homogeneous space.
3. Specificity of Nilpotent Lie Groups: For linear control systems on nilpotent Lie groups, the properties of control sets strongly depend on the eigenvalues of the derivative associated with the drift vector field.
4. Research Gap: While the regular case (where the associated derivative is invertible) has been studied, the singular case (where the associated derivative is non-invertible) remains a critical gap, presenting fundamentally different and more complex phenomena.

Research Motivation

The Heisenberg group, as a three-dimensional nilpotent but non-abelian Lie group, combines simple structure with highly non-trivial geometric properties, making it an ideal test platform for exploring how group structure influences controllability. This paper specifically addresses the unexplored and challenging problem in the singular case.

Core Contributions

1. Complete Classification Framework: Provides a complete classification of control sets for singular linear control systems on the Heisenberg group, where the associated derivative is non-invertible.
2. Invariant-Based Analysis: Through conjugation of systems via group automorphisms to reduced canonical forms, detailed analysis is performed based on fundamental invariants of the derivative (trace trA and determinant detA of the 2×2 submatrix A).
3. Discovery of Rich Behavioral Patterns: Reveals diverse behavioral patterns:
   • Dichotomy when detA = trA = 0: either global controllability or complete uncontrollability
   • Characterization based on spectrum and parameters when detA ≠ 0 and trA = 0
   • Connection with affine systems when detA = 0 and trA ≠ 0
4. Theoretical Foundation: Establishes concrete foundations for extending the study of singular linear control systems to higher-dimensional nilpotent Lie groups.

Detailed Methodology

Problem Formulation

Study linear control systems on the Heisenberg group H = (ℝ² × ℝ, ∗):

ΣH: ẋ(τ) = X(x(τ)) + Σᵢ uᵢ(τ)Zᵢ(x(τ))

where X is a linear vector field, Zᵢ are left-invariant vector fields, with the goal of completely characterizing control sets.

Theoretical Framework

Heisenberg Group Structure

The Heisenberg group is defined as H := (ℝ² × ℝ, ∗), where:

(v₁, z₁) ∗ (v₂, z₂) := (v₁ + v₂, z₁ + z₂ + ½ω(v₁,v₂))

Its Lie algebra is h := (ℝ² × ℝ, ·, ·), satisfying:

[(ζ₁, α₁), (ζ₂, α₂)] := (0, ω(ζ₁, ζ₂))

Derivative and Automorphism Structure

Proposition 3.4: The explicit forms of the derivative D and automorphism P are:

D = (A  0 )  ∈ Der(h),  P = (P  0   ) ∈ Aut(H)
    (η^T trA)              (ξ^T detP)

Linear Control System Representation

Single-input linear control systems in coordinates are represented as:

v̇ = Av + uζ
ż = z·trA + uα + ω(v, θη + u·½ζ)

Technical Innovations

1. Conjugation Simplification Strategy: Utilizes automorphisms P to selectively reduce matrix A to Jordan canonical form while preserving spectral properties.
2. Classification Criteria: Establishes classification criteria based on LARC and ad-rank conditions:
   • LARC: ω(Aζ, ζ) ≠ 0
   • ad-rank condition: ω(Aζ, ζ)(α detA + ω(Aζ, θη)) ≠ 0
3. Case-by-Case Analysis: Systematic analysis according to different combinations of detA and trA.

Main Results

Case 1: detA = trA = 0

Theorem 4.1:

1. If ω(Aζ, θη) ≠ 0, then H is the unique control set of ΣH
2. If ω(Aζ, θη) = 0, then the plane kerD is a continuum of singleton control sets of ΣH

Proof Strategy: Through construction of conjugate systems and Lyapunov-type functions F(x,y,z) = 3zσ + y(y² - 2xσ), orbit separation properties are established.

Case 2: detA ≠ 0 and trA = 0

Theorem 4.3:

1. If the ad-rank condition is satisfied, then C_{ℝ²} × ℝ is the unique control set of ΣH
2. If the ad-rank condition is not satisfied but A has purely imaginary eigenvalues, conclusion (1) holds
3. If the ad-rank condition is not satisfied and A has real eigenvalues, then kerD is a continuum of singleton control sets

Key Techniques:

• For purely imaginary eigenvalues, periodic orbits are constructed using rotation matrix properties
• For real eigenvalues, function G(x,y,z) = z + σy + σ²ln(y-σ) is constructed to prove orbit monotonicity

Case 3: detA = 0 and trA ≠ 0

Theorem 4.5: The unique control set is CH = π⁻¹(C^A_{ℝ²}), where π is the canonical projection onto the last two components.

Proof Method: Through conjugacy relations with affine systems, control set correspondence is established using affine control system theory.

Technical Details Analysis

Application of Conjugation Technique

The paper skillfully employs the automorphism group Aut(H) to simplify complex systems:

PDP⁻¹ = (PAP⁻¹  0  )
         (η̂^T    trA)

where η̂ can be adjusted through appropriate choice of P.

Fiber Controllability Analysis

Proposition 4.2 demonstrates the controllability of the fiber {0} × ℝ under different parameter conditions:

• α ≠ 0: Periodic orbits exist connecting any two points
• α = 0: Positive or negative time reachability is determined by the sign of (detA)⁻¹ω(Aζ,ζ)

Lyapunov Function Construction

In proving uniqueness of control sets, the paper constructs sophisticated Lyapunov-type functions, such as:

F(x,y,z) = 3zσ + y(y² - 2xσ)
G(x,y,z) = z + σy + σ²ln(y-σ)

The monotonicity of these functions along trajectories ensures orbit separation.

Related Work

Historical Development

1. Markus (1962): First extended the linear control system framework to matrix groups
2. Ayala & Tirao: Generalized to arbitrary Lie groups, establishing a unified geometric perspective
3. Jouan: Proved equivalence between control-affine systems and linear control systems on Lie groups

Existing Research

• Existence of control sets in the regular case has been established
• Properties of control sets on low-dimensional groups
• Relationships between derivative eigenvalues and control sets on nilpotent Lie groups

Contribution of This Paper

This paper fills the theoretical gap in the singular case, providing a complete classification framework.

Conclusions and Discussion

Main Conclusions

1. Complete Classification: Based on the trace and determinant of matrix A, completely characterizes the control set structure of singular linear control systems on the Heisenberg group.
2. Rich Phenomena: Discovers a rich spectrum of behaviors ranging from global controllability to complete uncontrollability.
3. Geometric Insights: Reveals deep connections between group geometric structure and control-theoretic properties.

Limitations

1. Dimensional Restriction: Results are specifically for the three-dimensional Heisenberg group; generalization to higher dimensions requires further research
2. Singular Assumption: Only considers the singular derivative case
3. Single-Input Systems: Primarily analyzes single-input systems

Future Directions

1. Generalization to higher-dimensional nilpotent Lie groups
2. Analysis of multi-input systems
3. Study of singular linear control systems on non-nilpotent Lie groups

In-Depth Evaluation

Strengths

1. Theoretical Completeness: Provides a complete solution to the problem with exhaustive classification
2. Methodological Innovation: Cleverly combines Lie group theory, control theory, and dynamical systems methods
3. Technical Depth: Sophisticated proof techniques, particularly in Lyapunov function construction
4. Clear Structure: Well-organized paper with clear logical flow

Weaknesses

1. Limited Applicability: Practical application value of theoretical results requires further exploration
2. Computational Complexity: Discrimination conditions in certain cases are computationally complex
3. Lack of Numerical Verification: Absence of numerical experiments to verify theoretical results

Impact

1. Theoretical Contribution: Makes important contributions to control system theory on Lie groups
2. Methodological Value: The analytical framework provided can be generalized to other Lie groups
3. Foundation for Future Research: Establishes solid theoretical foundations for subsequent research

Application Scenarios

1. Non-holonomic constraint systems in robotics
2. Quantum control systems
3. Geometric control theory research
4. Sub-Riemannian geometry applications

References

The paper cites important literature in the field, including the classical textbook "The Dynamics of Control" by Colonius & Kliemann, as well as a series of works by Ayala, Da Silva, and others on control systems on Lie groups, reflecting good academic heritage and development trajectory.

Overall Assessment: This is a high-quality theoretical research paper that makes substantial contributions to the important field of control systems on Lie groups. The paper demonstrates high technical depth and strong theoretical completeness, providing important foundations for further development in this field.