Dynamic Event-Triggered Strategy-Based Optimal Control of Modular Robot Manipulator: A Multiplayer Nonzero-Sum Game...

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Modular robot manipulators (MRMs) face computational limitations in sensors and controllers, necessitating efficient communication paradigms.
• Traditional event-triggered mechanisms struggle with resource constraints, driving the need for advanced solutions like dynamic event-triggered control.
• Uncertain disturbances in MRM systems pose significant challenges to control accuracy and stability.

Purpose of the Study:

• To develop an optimal control scheme for MRM systems using a dynamic event-triggered mechanism.
• To address resource constraints and uncertain disturbances in MRM applications.
• To enhance resource utilization efficiency and system design flexibility in MRMs.

Main Methods:

• Established the dynamic model of the MRM using joint torque feedback and employed a data-driven neural network for uncertainty estimation.
• Transformed the MRM tracking control problem into a multiplayer nonzero-sum differential game framework.
• Utilized adaptive dynamic programming for static event-triggered control, introducing an internal dynamic variable to analyze dynamic trigger rules.

Main Results:

• Developed a dynamic event-triggered control strategy for MRMs with uncertain disturbances.
• Demonstrated improved resource utilization and system design flexibility compared to traditional methods.
• Excluded Zeno behavior by ensuring a positive minimum sampling interval via an exponential attenuation signal.

Conclusions:

• The proposed dynamic event-triggered control scheme ensures asymptotic stability for MRM systems, validated by Lyapunov theory.
• Experimental results confirm the effectiveness of the developed optimal control strategy.
• The method offers a robust and efficient solution for resource-constrained MRM applications.