Dynamic event-driven neural network-based adaptive fault-attack-tolerant control for wheeled mobile robot system.

Area of Science:

• Robotics and Control Systems
• Cyber-Physical Systems Security
• Fault-Tolerant Control

Background:

• Wheeled mobile robots (WMRs) are susceptible to actuator faults, external disturbances, and sophisticated communication attacks.
• Ensuring reliable operation and performance recovery in WMRs under these combined threats is a significant challenge.
• Limited communication resources further complicate the design of robust control strategies.

Purpose of the Study:

• To investigate and address the fault-attack control problem for WMRs.
• To develop an event-observer based compensation controller for enhanced reliability and performance recoverability.
• To achieve communication-efficient control through dynamic event conditions and adaptive triggering.

Main Methods:

• Establishment of a WMR dynamic model incorporating actuator faults, disturbances, and communication attacks.
• Development of an event-based proportional-integral observer (PIO) with embedded state, actuator fault, and disturbance estimators.
• Design of a second-order adaptive sliding mode fault-compensation reliable controller utilizing observer outputs and attack information.
• Implementation of a dynamic event condition and adaptive trigger scheme for communication efficiency.

Main Results:

• The proposed controller effectively compensates for actuator faults and attenuates disturbances.
• The system demonstrates reliable tracking control performance even under communication attacks.
• The event-observer approach, combined with neural network approximation, ensures performance recoverability.
• Communication efficiency is achieved via adaptive triggering mechanisms in sensor and actuator channels.

Conclusions:

• The presented event-observer based compensation control strategy significantly enhances the reliability and recoverability of WMRs.
• The method successfully addresses complex scenarios involving actuator faults, disturbances, and communication attacks.
• The proposed approach offers a promising solution for robust and communication-efficient control of WMRs in adversarial environments.