On Tensor-based Polynomial Hamiltonian Systems
Shaoxuan Cui, Guofeng Zhang, Hildeberto Jardón-Kojakhmetov, Ming Cao · 2025
Read On Tensor-based Polynomial Hamiltonian Systems by Shaoxuan Cui, Guofeng Zhang, Hildeberto Jardón-Kojakhmetov, Ming Cao (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-21487.txt.
Introduction
Shaoxuan Cui and three co-authors extend a classic test in systems theory in a 2025 paper. For simple linear systems, a clean matrix test tells whether the system conserves energy in the Hamiltonian sense. Many real systems are polynomial, not linear, and had no such tidy test. Their work fills that gap using tensor forms.
They define a new kind of tensor that plays the role the special matrices play for linear systems. A polynomial system then passes the test exactly when all its tensors have that property. They also give a practical stability test for such systems. Worked numeric examples confirm the theory, and a closing discussion builds intuition for the results.
Catalog record
| CREATOR | Shaoxuan Cui, Guofeng Zhang, Hildeberto Jardón-Kojakhmetov, Ming Cao |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-21487
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On Tensor-based Polynomial Hamiltonian Systems Shaoxuan Cui, Guofeng Zhang, Hildeberto Jardón-Kojakhmetov, Ming Cao Abstract It is known that a linear system with a system matrix A A constitutes a Hamiltonian system with a quadratic Hamiltonian if and only if A A is a Hamiltonian matrix. This provides a straightforward method to verify whether a linear system is Hamiltonian or whether a given Hamiltonian function corresponds to a linear system. These techniques fundamentally rely on the properties of Hamiltonian matrices. Building on recent advances in tensor algebra, this paper generalizes such results to a broad class of polynomial systems. As the systems of interest can be naturally represented in tensor forms, we name them ”tensor-based polynomial systems”. Our main contribution is that we formally define Hamiltonian cubical tensors and characterize their properties. Crucially, we demonstrate that a tensor-based polynomial system is a Hamiltonian system with a polynomial Hamiltonian if and only if all associated system tensors are Hamiltonian cubical tensors—a direct parallel to the linear case. Additionally, we establish a computationally tractable stability criterion for tensor-based polynomial Hamiltonian systems. Finally, we validate all theoretical results through numerical examples and provide a further intuitive discussion. I Introduction A Hamiltonian system constitutes a class of dynamical systems whose equations of motion can be expressed via a scalar function known as the Hamiltonian [ 1 ] . Originating from the analysis of physical systems, this mathematical framework provides a natural formalism for studying conservative mechanical systems [ 2 ] . Canonical examples range from elementary systems like the harmonic oscillator (where the total mechanical energy remains conserved) to complex phenomena exemplified by the N -body problem [ 2 ] . Modern theoretical developments have extended the Hamiltonian paradigm beyond classical mechanics, finding applications in diverse domains including ecological modeling [ 3 , 4 ] and economic systems analysis [ 5 , 6 ] . Notably, Hamiltonian systems exhibit profound connections with optimal control theory, as solutions to optimal control problems inherently follow Hamiltonian trajectories [ 7 ] . The control-theoretic perspective further considers Port-Hamiltonian systems, incorporating external inputs [ 8 , 9 , 10 ] – a formulation particularly relevant in modern energy-based control methodologies. The linear Hamiltonian system is the most fundamental type of Hamiltonian systems. It is well-established that a linear Hamiltonian system corresponds to a quadratic form of the Hamiltonian function [ 11 , 2 ] . By leveraging the properties of Hamiltonian matrices and the linear systems theory, the behavior of such systems has been extensively studied [ 2 ] . Moreover, the properties of Hamiltonian matrices are widely utilized in the field of optimal control [ 12 ] , including H_{2} control [ 13 ] and H_{\infty} control [ 14 ] . These properties also play a significant role in quantum linear systems [ 15 ] and quantum Hamiltonian systems [ 16 ] . In contrast, the behavior of general nonlinear Hamiltonian systems has not been fully explored. A natural question arises: how can the well-understood case of linear Hamiltonian systems be generalized to uncover certain classes of nonlinear Hamiltonian systems? The behavior of a linear system is closely related to its associated system matrix. With advancements in tensor algebra, it has recently been demonstrated that polynomial systems, where the power is a positive integer, can be represented in tensor forms. This representation allows for further analysis of the system using tensor properties [ 17 , 18 , 19 , 20 , 21 ] . However, most of these results [ 17 , 18 , 19 , 20 , 21 ] are limited to the case of asymptotic stability. For autonomous Hamiltonian systems, an equilibrium is at most Lyapunov stable [ 2 ] . In this work, we aim to utilize tensor properties to study polynomial Hamiltonian systems. The main contributions of this paper are as follows. First, we establish a necessary and sufficient condition for a polynomial system to be Hamiltonian with a polynomial Hamiltonian, using tensor algebra. Both the theoretical framework and the tensor techniques introduced are novel to the field. Second, we provide a stability criterion for each equilibrium of a tensor-based polynomial Hamiltonian system, which can be easily verified through tensor operations. Finally, we illustrate our theoretical results with application examples. The paper is organized as follows: In Section II , we introduce preliminary concepts related to tensors, which are essential for deriving our main results. In Section III , we review the theory of linear Hamiltonian systems. In Section IV , we present our main results on tensor-based polynomial Hamiltonian systems. In Section V , we demonstrate our findings with examples and discuss potential applications. Finally, in the last section, we conclude the paper and provide further discussion. Notation: \mathbb{R} denotes the set of real numbers. For a matrix M\in\mathbb{R}^{n\times r} and a vector a\in\mathbb{R}^{n} , M_{ij} and a_{i} denote the element in the i th row and j th column and the i th entry, respectively. The symbol \textbf{det}(R) denotes the determinant of a matrix or a tensor R [ 22 ] . The symbol \textbf{tr}(R) denotes the trace of a matrix or a tensor R [ 22 ] . In the following section, we introduce the preliminaries on tensors and some further notations regarding tensors. II Tensor preliminaries A tensor T\in\mathbb{R}^{n_{1}\times n_{2}\times\cdots\times n_{k}} is a multidimensional array, where the order is defined as the number of its dimension k and each dimension n_{i} , i=1,\cdots,k, is called a mode of the tensor. A tensor is a cubical tensor if every mode has the same size, i.e., T\in\mathbb{R}^{n\times n\times\cdots\times n} . We further write a k -th order n dimensional cubical tensor as T\in\mathbb{R}^{n\times n\times\cdots\times n}=\mathbb{R}^{[k,n]} . A cubical tensor T is a supersymmetric tensor if T_{j_{1}j_{2}\ldots j_{k}} is invariant under any permutation of the indices. For the rest of the paper, a tensor always refers to a cubical tensor. For a tensor A\in\mathbb{R}^{[k,n]} and a vector x\in\mathbb{R}^{n} , the tensor-vector product Ax is a (k-1) -th order n dimensional tensor (i.e., Ax\in\mathbb{R}^{[k-1,n]} ), whose component is given by (Ax)_{i_{1}i_{2}\ldots i_{k-1}}=\sum_{i_{k}=1}^{n}A_{i_{1}\ldots i_{k-1}i_{k}}x_{i_{k}}. Similarly, by repeatedly using the above operation, the component of the tensor-vector product to the power k-2 , Ax^{k-2}\in\mathbb{R}^{[2,n]} is defined as (Ax^{k-2})_{i_{1}i_{2}}=\sum_{i_{3},\ldots,i_{k}=1}^{n}A_{i_{1}i_{2}i_{3}\ldots i_{k}}x_{i_{3}}\ldots x_{i_{k}}. The tensor-vector product to the power k-1 , Ax^{k-1}\in\mathbb{R}^{[1,n]} yields a vector, whose i -th component is given by (Ax^{k-1})_{i_{1}}=\sum_{i_{2},\ldots,i_{k}=1}^{n}A_{i_{1}i_{2}\ldots i_{k}}x_{i_{2}}\ldots x_{i_{k}}. The tensor-vector product to the power k:Ax^{k} is defined as a scalar Ax^{k}=x^{\top}(Ax^{k-1})=\sum_{i_{1},\ldots,i_{k}=1}^{n}A_{i_{1}i_{2}\ldots i_{k}}x_{i_{1}}\ldots x_{i_{k}}. For a matrix R\in\mathbb{R}^{n\times n} and a tensor A\in\mathbb{R}^{[k,n]} , the matrix-tensor product RA\in\mathbb{R}^{[k,n]} is as follows: (RA)_{i_{1}i_{2}\ldots i_{k}}=\sum_{j=1}^{n}R_{i_{1}j}A_{ji_{2}\ldots i_{k}}. In addition, we have the following result. Lemma 1 [ 19 ] Given a one dimensional homogenous polynomial function f(x) : \mathbb{R}^{n}\rightarrow\mathbb{R}. The function f(x) can be uniquely represented by the tensor form Ax^{m} , m is the order of A , A is a super-symmetric cubical tensor. In the following, we prove the following associative law for matrices and tensors. Lemma 2 If A\in\mathbb{R}^{[k,n]} is a tensor and B,C\in\mathbb{R}^{n\times n} are matrices, BCA=B(CA) . Proof: We have that (CA)_{i_{1}i_{2}\ldots i_{k}}=\sum_{j=1}^{n}C_{i_{1}j}A_{ji_{2}\ldots i_{k}}. It follows that \left(B(CA)\right)_{i_{1}i_{2}\ldots i_{k}}=\sum_{j=1}^{n}B_{i_{1}j}(CA)_{ji_{2}\ldots i_{k}}=\sum_{j=1}^{n}B_{i_{1}j}\left(\sum_{t=1}^{n}C_{jt}A_{ti_{2}\ldots i_{k}}\right). On the other hand, (BC)_{i_{1}j}=\sum_{j}B_{i_{1}t}C_{tj} . Then, (BCA)_{i_{1}\cdots i_{k}}=\sum_{j=1}^{n}(BC)_{i_{1}j}A_{ji_{2}\ldots i_{k}}=\sum_{j}\left(\sum_{t}B_{i_{1}t}C_{tj}\right)A_{ji_{2}\ldots i_{k}}. Thus, BCA=B(CA) . ∎ In addition, we have the