Echo-state network based adaptive fuzzy sliding-mode consensus control scheme for nonlinear multi-agent systems with...

Area of Science:

• Control Systems Engineering
• Artificial Intelligence
• Robotics

Background:

• Nonlinear multi-agent systems (MASs) face challenges from external disturbances and uncertainties.
• Conventional sliding-mode control (SMC) offers robustness but suffers from chattering.
• Existing methods often struggle with accurate state estimation and disturbance rejection in complex MASs.

Purpose of the Study:

• To develop a novel hybrid control strategy for nonlinear MASs.
• To address limitations of conventional controllers, including chattering and disturbance effects.
• To enhance robustness, tracking performance, and control efficiency in MASs.

Main Methods:

• Integration of an adaptive fuzzy sliding-mode (AFSM) controller to mitigate chattering.
• Utilization of an echo state network (ESN) optimized by the chaotic whale optimization algorithm (CWOA) for improved state estimation.
• Incorporation of a nonlinear disturbance observer unit (NDOU) for estimating and compensating disturbances and unmeasured states.
• Lyapunov stability analysis to ensure system stability.

Main Results:

• The proposed hybrid controller effectively suppresses chattering and enhances disturbance rejection.
• The optimized ESN improves state estimation accuracy and dynamic adaptability.
• Synergistic integration leads to fast, robust, and reliable consensus among agents.
• Simulation results demonstrate superior tracking performance and control efficiency compared to conventional methods.

Conclusions:

• The novel hybrid control strategy offers a significant advancement for nonlinear MASs.
• The integrated approach overcomes limitations of single-technique controllers, providing practical effectiveness.
• This framework enhances the reliability and performance of MASs in uncertain and disturbed environments.