From Controllability to Information: A Unified Analysis via Gramian, Minimum Energy, Fisher Information Matrix and Entropy
Gabriel R. de Andrade Silva · 2025
Read From Controllability to Information: A Unified Analysis via Gramian, Minimum Energy, Fisher Information Matrix and Entropy by Gabriel R. de Andrade Silva (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-2507-08847.txt.
Introduction
Gabriel Silva ties steering power, steering cost, information, and entropy in one frame for linear flows. His test bed is the damped spring oscillator, a plain case with rich behavior. His 2025 work works out the steering Gramian set and its spread for each damping kind. That spread reads as least steering cost: how cheap a push can drive the system to a set state.
Damping grade and sway rate shift the cost in set ways. Information forms face the Gramian set and line up: keen readings and cheap steering prove two faces of one coin. Entropy stories join too: ease of steering tracks how much information the system holds and sheds. One frame thus binds cost and information for flow craft.
Catalog record
| CREATOR | Gabriel R. de Andrade Silva |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-08847
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
From Controllability to Information: A Unified Analysis via Gramian, Minimum Energy, Fisher Information Matrix and Entropy Gabriel R. de Andrade Silva Abstract This article explores the connections between controllability, control energy, information, and entropy in the context of linear dynamical systems, using the damped harmonic oscillator as a case study. We analytically derive the Controllability Gramian Matrix ( W c W_{c} ) and its determinant ( det ( W c ) \det(W_{c}) ) for different damping regimes. We analyze the physical interpretation of det ( W c ) \det(W_{c}) in terms of minimum control energy, highlighting the influence of system parameters (damping ζ \zeta and natural frequency ω n \omega_{n} ). We investigate the conceptual and formal relationships between the Gramian, the Fisher Information Matrix ( I I ), and Shannon ( H H ) and thermodynamic ( S S ) entropies, suggesting a duality between energetic control and estimation precision, as well as a link between ease of control and informational/thermodynamic entropy. The unified analysis offers insights into the fundamental principles governing the control of dynamical systems, integrating energy and information perspectives. 1 Introduction Understanding and controlling complex dynamical systems represent central challenges in various areas of science and engineering, from physics and biology to economics and social sciences. The ability to predict and influence the behavior of these systems is fundamental for the development of advanced technologies, process optimization, and solving global problems. At the core of control theory, the notion of controllability, introduced by Kalman in the 1960s, establishes the conditions under which it is possible to drive a system from an initial state to any desired final state in finite time, using a limited set of control inputs. Although fundamental, classical controllability is a qualitative property, providing no information about the cost or effort required to achieve such a transition. In practice, energy limitations are ubiquitous. Physical actuators have constraints on power, amplitude, and frequency, making purely qualitative analysis of controllability insufficient for many real applications. Thus arises the need for quantitative metrics that capture the control effort, allowing assessment of the ease or difficulty of maneuvering a system’s state. The Controllability Gramian Matrix ( W_{c} ), defined as the integral of the exponential of the system’s dynamics matrix multiplied by the input matrix and its transpose, has emerged as a powerful tool in this context. The spectral properties of the Gramian, such as its minimum and maximum eigenvalues, trace, and determinant, provide quantitative measures of the energy effort associated with control. In particular, the determinant of the Controllability Gramian ( \det(W_{c}) ) gained prominence after the seminal work of Joshi and Mitter (1977), which established its fundamental physical interpretation: the inverse of the determinant is related to the minimum energy required to control the system. The higher the determinant, the lower the required minimum energy, indicating a system that is ”easier” to control from an energetic perspective. This connection between an abstract mathematical property (the determinant of an integral matrix) and a measurable physical quantity (energy) opened new avenues for the analysis and design of efficient control systems. In parallel, concepts from information theory and statistical physics, such as the Fisher Information Matrix ( I ) and Entropy (both in Shannon’s formulation and thermodynamics), offer complementary perspectives for quantifying the uncertainty, information, and complexity inherent in dynamical systems. The Fisher Matrix, classically used in statistics to measure the amount of information that observable data carry about unknown parameters of a model, finds intriguing parallels in the control context, where it can be related to the curvature of the ”reachable state space” and the distribution of control capability. Entropy, in turn, quantifies disorder or lack of information about a system’s state. Recent investigations have explored the profound connections between these different metrics, suggesting that control energy, Fisher information, and entropy can be viewed as distinct facets of a more unified description of the dynamics and controllability of complex systems (Baggio et al., 2022). This article seeks to explore these connections in depth, using the damped harmonic oscillator as a canonical model system. Through detailed analysis of this system, we will derive analytical expressions for the Controllability Gramian Matrix and its determinant in different damping regimes (underdamped, critically damped, overdamped, and undamped). We will investigate the physical interpretation of the determinant in terms of minimum control energy, explicitly showing the influence of damping on the effort required to control the oscillator. Next, we will explore the relationships between the Gramian, the Fisher Information Matrix, and Shannon and thermodynamic entropies, seeking to elucidate how these different metrics complement each other in characterizing the controllability and complexity of the system. The ultimate goal is to provide a unified analysis that integrates the concepts of controllability, energy, and information, offering insights into the fundamental principles governing the control of dynamical systems. The work is organized as follows: Section 2 presents the mathematical modeling of the damped harmonic oscillator, including its state-space representation. Section 3 details the controllability analysis via the Gramian Matrix, with analytical calculation of the matrix and the matrix exponential for each regime. Section 4 focuses on the Gramian determinant, its analytical calculation, and its physical interpretation in terms of minimum control energy. Section 5 introduces the Fisher Information Matrix and explores its connection with the Gramian. Section 6 investigates the relationships with Shannon entropy and thermodynamic entropy. Section 7 presents a discussion of the results and their implications. Finally, Section 8 concludes the article with a recapitulation of the main contributions. 2 Mathematical Modeling of the Harmonic Oscillator The damped harmonic oscillator is a fundamental and ubiquitous physical system, serving as a model for a vast range of phenomena in physics, engineering, and other sciences. Its dynamics are described by a second-order linear differential equation that captures the interaction between restoring forces (proportional to displacement), dissipative forces (proportional to velocity), and potentially external forces. The general differential equation for the displacement x(t) of a damped and forced harmonic oscillator is given by: m\ddot{x}(t)+c\dot{x}(t)+kx(t)=F(t) (1) Where: • m is the mass of the oscillator. • c is the viscous damping coefficient. • k is the spring constant (stiffness). • x(t) is the displacement of the oscillator relative to its equilibrium position at time t . • \dot{x}(t) is the velocity of the oscillator at time t . • \ddot{x}(t) is the acceleration of the oscillator at time t . • F(t) is the external force applied to the system at time t . For systems and control analysis purposes, it is convenient to normalize this equation. Dividing by m and introducing the parameters of undamped natural frequency ( \omega_{n} ) and damping factor ( \zeta ), we have: \ddot{x}(t)+\frac{c}{m}\dot{x}(t)+\frac{k}{m}x(t)=\frac{F(t)}{m} (2) Defining: • Undamped natural frequency: \omega_{n}=\sqrt{k/m} • Damping factor: \zeta=\frac{c}{2\sqrt{mk}}=\frac{c}{2m\omega_{n}} • Normalized external force: f(t)=F(t)/m The normalized equation becomes: \ddot{x}(t)+2\zeta\omega_{n}\dot{x}(t)+\omega_{n}^{2}x(t)=f(t) (3) This is the standard form of the differential equation for the damped and forced harmonic oscillator. The behavior of the system crucially depends on the value of the damping factor \zeta : • \zeta=0 : Undamped oscillator. The system oscillates indefinitely with frequency \omega_{n} if there is no external force. • 0<\zeta<1 : Underdamped oscillator. The system oscillates with exponentially decreasing amplitude, with a damped frequency \omega_{d}=\omega_{n}\sqrt{1-\zeta^{2}}