Event-Triggered Memory Control for Interval Type-2 Fuzzy Heterogeneous Multi-Agent Systems

Sen Kong

2024-12-10

Abstract:This study explores the design of a memory-based dynamic event-triggered mechanisms (DETM) scheme for heterogeneous multi-agent systems (MASs) characterized by interval type-2 Takagi-Sugeno (IT2 T-S) fuzzy models. To address the complex nonlinear uncertainties inherent in such systems, discrete IT2 T-S fuzzy models are employed to accurately capture system dynamics. In response to the limitations on communication resources and computational capabilities within MASs, this research introduces a distributed DETM approach based on a dynamic threshold method. This mechanism effectively minimizes unnecessary communication while maintaining robust performance. The proposed memory-based control strategy not only reduces the conservatism associated with controller design conditions but also enhances overall controller performance. Furthermore, leveraging a non-parallel distributed compensation (non-PDC) strategy, a novel derivation method is developed for controller design conditions that significantly decreases conservatism. This leads to sufficient conditions for the asymptotic stability of the closed-loop system. The designed distributed event-triggered controllers improve the overall performance of MASs, as evidenced by numerical simulations that validate the effectiveness of the proposed approach. Overall, these findings advance the state-of-the-art in control strategies for heterogeneous MASs, offering practical solutions for real-world applications where resource constraints are critical.

Systems and Control

What problem does this paper attempt to address?

The problems that this paper attempts to solve mainly focus on the following aspects:

1. **Dealing with the complex nonlinear uncertainties in heterogeneous multi - agent systems**: Heterogeneous multi - agent systems (MASs) exhibit complex nonlinear uncertainties due to the diversity of their internal structures and behaviors, which makes accurate modeling difficult. For this reason, the paper adopts the interval type - 2 Takagi - Sugeno (IT2 T - S) fuzzy model to accurately capture the system dynamics.

2. **Coping with the limitations of communication resources and computing capabilities**: In multi - agent systems, the cooperative control among agents depends on data transmission, but each agent has limited communication, energy, and computing capabilities and cannot support high - frequency data transmission and control output signal calculation. By introducing the distributed dynamic event - triggering mechanism (DETM) based on the dynamic threshold method, the paper effectively reduces unnecessary communication while maintaining the robust performance of the system.

3. **Reducing the conservatism of controller design conditions and improving the overall performance**: Traditional controller design methods are often too conservative, which limits the flexibility and effectiveness of the controllers. The paper proposes a memory controller design method based on the non - parallel distributed compensation (non - PDC) strategy, which not only reduces the conservatism of controller design conditions but also improves the overall controller performance.

4. **Achieving the asymptotic stability of heterogeneous MAS**: By using appropriate mathematical tools and techniques (such as linear matrix inequalities LMIs), the paper solves the problems of new state deviation variables and multiple coupling terms in the controller stability analysis, thereby deriving sufficient conditions to ensure the asymptotic stability of heterogeneous MAS.

### Specific problem description

The main objective of the paper is to design a memory - based dynamic event - triggering controller for heterogeneous multi - agent systems to ensure the global asymptotic stability of the closed - loop system. Specifically, the paper attempts to solve the following problems:

- How to use the IT2 T - S fuzzy model to deal with the complex nonlinear uncertainties in heterogeneous MAS.
- How to reduce the communication burden by introducing the dynamic event - triggering mechanism while maintaining the robust performance of the system.
- How to design a memory controller based on the non - parallel distributed compensation (non - PDC) strategy to reduce the conservatism of controller design conditions and improve the overall performance.
- How to derive the controller design conditions to ensure the asymptotic stability of heterogeneous MAS by using appropriate mathematical tools and techniques (such as linear matrix inequalities LMIs).

### Solution overview

The solutions proposed in the paper include:

- **Modeling heterogeneous MAS based on the IT2 T - S fuzzy model**: Use the discrete IT2 T - S fuzzy model to accurately describe the system dynamics.
- **Distributed dynamic event - triggering mechanism (DETM) based on the dynamic threshold method**: By introducing dynamic threshold variables, reduce unnecessary communication while maintaining the robust performance of the system.
- **Memory controller design based on the non - parallel distributed compensation (non - PDC) strategy**: By introducing historical state information, reduce the conservatism of controller design conditions and improve the overall performance.
- **Deriving controller design conditions through linear matrix inequalities (LMIs)**: Use appropriate mathematical tools and techniques to derive sufficient conditions to ensure the asymptotic stability of heterogeneous MAS.

These solutions work together to make the control strategy proposed in the paper more efficient and have better performance in practical applications.