Abstract:Sequences of parametrized Lyapunov equations can be encountered in many application settings. Moreover, solutions of such equations are often intermediate steps of an overall procedure whose main goal is the computation of $\text{trace}(EX)$ where $X$ denotes the solution of a Lyapunov equation and $E$ is a given matrix. We are interested in addressing problems where the parameter dependency of the coefficient matrix is encoded as a low-rank modification to a \emph{seed}, fixed matrix. We propose two novel numerical procedures that fully exploit such a common structure. The first one builds upon the Sherman-Morrison-Woodbury (SMW) formula and recycling Krylov techniques, and it is well-suited for small dimensional problems as it makes use of dense numerical linear algebra tools. The second algorithm can instead address large-scale problems by relying on state-of-the-art projection techniques based on the extended Krylov subspace. We test the new algorithms on several problems arising in the study of damped vibrational systems and the analyses of output synchronization problems for multi-agent systems. Our results show that the algorithms we propose are superior to state-of-the-art techniques as they are able to remarkably speed up the computation of accurate solutions.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the efficient solution of the parameterized Lyapunov equation sequence in multiple application backgrounds. Specifically, the form of such equations is as follows:
\[ A(v)X + XA(v)^T = -Q \]
where \( A(v) \) has the form:
\[ A(v) = A_0 - B_l D(v) B_r^T \]
Here, \( A_0, Q \in \mathbb{R}^{n \times n} \), \( B_l, B_r \in \mathbb{R}^{n \times k} \) are fixed matrices, and \( D(v) = \text{diag}(v_1, v_2, \ldots, v_k) \) contains the parameters \( v_i \in \mathbb{R} \setminus \{0\} \), which are encoded in the vector \( v = [v_1, v_2, \ldots, v_k]^T \in (\mathbb{R} \setminus \{0\})^k \). It is assumed that the number of parameters \( k \) is very small, that is, \( k = O(1) \).
The paper mentions that such parameterized Lyapunov equations are encountered in many application scenarios, such as the study of vibrating systems, parameterized model order reduction, low - gain feedback design, and the analysis of multi - agent systems. In particular, in optimizing the damped vibrating systems and the output synchronization problem of multi - agent systems, it is necessary to efficiently calculate \( \text{trace}(EX(v)) \), where \( E \) is a given matrix and \( X(v) \) is the solution of the Lyapunov equation.
To efficiently solve these problems, the paper proposes two new numerical methods:
1. **A method based on the Sherman - Morrison - Woodbury (SMW) formula and the cyclic Krylov technique**:
- This method is suitable for small - scale problems and utilizes dense linear algebra tools.
- By pre - computing the parameter - independent part and only computing the parameter - related part in the online stage, the efficiency is improved.
2. **A projection method based on the extended Krylov subspace**:
- This method is suitable for large - scale problems and utilizes state - of - the - art projection techniques.
- By constructing an extended Krylov subspace, it can effectively handle the situation of parameter changes without having to recompute the entire subspace every time.
Both methods make full use of the structure of the coefficient matrix \( A(v) \), that is, \( A(v) \) is an affine function of the parameter \( v \). In this way, the methods in the paper are superior to the existing state - of - the - art methods in terms of computational performance and memory requirements, especially in terms of computation time.