Caputo Barrier Functions and Their Applications to the Safety, Safety-and-Stability, and Input-to-State Safety of a Class of Fractional-Order Systems

Zheren Zhu, Bingbing Shen, Le Yao, Yi Chai, Zhihuan SongMathematics 2025 · DOI

TL;DR

This paper extends the powerful concept of barrier functions, crucial for ensuring safety in dynamical systems, to the realm of fractional-order systems. The authors introduce novel Caputo barrier functions and establish theoretical frameworks for guaranteeing safety, stability, and input-to-state safety in these complex systems, paving the way for safer fractional-order control design.

What Problem Does This Paper Address?

Ensuring the safety of dynamical systems is paramount, especially for complex and large-scale systems. Barrier functions have emerged as a highly effective tool for this purpose, allowing systems to remain within predefined safe regions. While extensively studied for traditional integer-order systems, the application of barrier functions to fractional-order systems (FOSs) has been largely unexplored. FOSs, characterized by their memory and hereditary properties, present unique challenges that make direct application of integer-order control theories difficult. This paper aims to bridge this gap by adapting and extending barrier function theory to FOSs, specifically those described by Caputo fractional derivatives.

Methods

The core contribution of this paper lies in the development and adaptation of barrier functions for Caputo fractional-order systems (CFOSs). The authors introduce two novel types of barrier functions:

  • Caputo Reciprocal Barrier Function (CRBF): This is an adaptation of the well-established reciprocal barrier function (RBF) for integer-order systems. It's designed to ensure that the system's state remains within a safe set.
  • Caputo Zeroing Barrier Function (CZBF): This is an adaptation of the zeroing barrier function (ZBF), which is particularly useful for input-to-state safety analysis.

Using these novel barrier functions, the paper establishes several key theoretical results:

  • Safety Criteria: Theorems are presented that guarantee the forward invariance of the safe set for CFOSs, ensuring that the system remains safe.
  • Synchronous Safety and Stability: The paper extends previous work to propose theorems for uniformly asymptotic stability and exponential stability with guaranteed safety for CFOSs. This is crucial for industrial applications where both safety and stability are critical.
  • Caputo Input-to-State Safety (ISSf): A new definition for ISSf is proposed for CFOSs, inspired by the proof of the CRBF. This definition unifies safety and ISSf and facilitates controller design. Two criteria for Caputo ISSf are then derived using CRBF and CZBF.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Bridging the Fractional-Order Gap: The paper successfully demonstrates that the powerful safety guarantees provided by barrier functions for integer-order systems can be effectively extended to fractional-order systems, a previously under-researched area.
  • Novel Caputo Barrier Functions: The introduction of Caputo RBF and Caputo ZBF provides the essential tools for analyzing and controlling the safety of CFOSs.
  • Unified Safety and Stability: The development of theorems that guarantee both safety and stability simultaneously is a significant step towards practical control design for FOSs, as safety alone is often insufficient.
  • Redefining Input-to-State Safety: The new definition of Caputo ISSf offers a more unified and practical approach to analyzing the robustness of safety in the presence of external disturbances for FOSs.
  • Computational Trade-offs: The paper highlights that while CRBFs offer theoretical advantages, CZBFs are more computationally tractable for practical implementation in fractional-order systems due to the complexity of fractional derivatives.

Key Results

The paper establishes theoretical frameworks and criteria for:

  • Forward Invariance of Safe Sets: Theorems 1 and 2 guarantee that the defined safe sets remain forward-invariant for CFOSs using Caputo RBF and CZBF, respectively.
  • Uniform Asymptotic Stability with Safety: Theorem 4 shows that CFOSs can achieve uniform asymptotic stability while remaining within the safe set.
  • Exponential Stability with Safety: Theorem 5 demonstrates that CFOSs can achieve exponential stability while maintaining safety.
  • Caputo Input-to-State Safety: Theorems 6 and 7 provide criteria for Caputo ISSf using both Caputo RBF and Caputo ZBF, respectively, confirming that the system remains safe even under bounded external disturbances.

Limitations and Future Work

  • Finite-Time Stability: The authors acknowledge that their current work focuses on infinite-time stability and that proving finite-time stability for fractional-order systems is significantly more challenging and was not explored in depth.
  • Computational Complexity of CRBF: The paper notes that the implementation of CRBFs is computationally complex due to the nature of fractional derivatives, making CZBFs more practical for real-world applications.
  • Application to Practical Cases: The authors state that future work will focus on applying the developed theoretical framework to practical fractional-order control problems. This suggests that while the theory is established, its real-world implementation and validation are still ongoing.

Summary updated

Abstract

Safety control based on barrier functions has gradually become one of the emerging and more important directions in the field of safety. Scholars are attempting to apply barrier functions to integer-order dynamical systems, such as general nonlinear systems, hybrid systems, linear systems, etc. Moreover, the introduction of barrier functions has even expanded the research approaches on safe reinforcement learning. However, there is very little research on the safety control problem of fractional-order dynamical systems. Based on our previous work, this article further explores, in depth, the problem of the transfer and adaptability of barrier functions for integer-order systems in fractional-order systems, and it also proposes the Caputo reciprocal barrier function and Caputo zeroing barrier function. And we established two theorems, which proved that we can also achieve uniform asymptotic stability or exponential stability with guaranteed safety. In the end, we created a new description for the definition of input-to-state safety under Caputo’s fractional-order systems, and we used this description and the above two Caputo barrier functions to construct two criteria of the Caputo input-to-state safety. Thus, we, finally, established the embryonic form of the theoretical framework of safety control based on barrier functions for fractional-order systems.
BibTeX
@inproceedings{zhu2025caputo,
  author = {Zheren Zhu and Bingbing Shen and Le Yao and Yi Chai and Zhihuan Song},
  title = {{Caputo Barrier Functions and Their Applications to the Safety, Safety-and-Stability, and Input-to-State Safety of a Class of Fractional-Order Systems}},
  booktitle = {Mathematics},
  year = {2025},
  doi = {10.3390/math13081215},
}
Figure