Neural network-based predefined-time bipartite formation tracking control of uncertain heterogeneous Euler-Lagrange...

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Formation tracking is crucial for multi-agent systems.
• Uncertainties in Euler-Lagrange systems pose significant control challenges.
• Predefined-time control offers faster convergence than traditional methods.

Purpose of the Study:

• To address the task-space bipartite formation tracking problem for uncertain heterogeneous Euler-Lagrange systems.
• To develop a hierarchical control algorithm capable of achieving tracking in a predefined time.
• To enhance system robustness against dynamic uncertainties.

Main Methods:

• Design of a hierarchical predefined-time control algorithm.
• Utilization of a non-singular sliding surface for adjustable settling time.
• Integration of a radial basis function neural network to handle system uncertainties.
• Development of a leader state estimator and a formation-specific controller.

Main Results:

• The proposed algorithm successfully achieves task-space bipartite formation tracking.
• The non-singular sliding surface allows flexible adjustment of the settling time.
• The radial basis function neural network effectively mitigates dynamic uncertainties.
• Numerical simulations validate the algorithm's effectiveness and practical applicability.

Conclusions:

• The developed hierarchical predefined-time control algorithm is effective for uncertain Euler-Lagrange systems.
• The approach offers precise and adaptable formation tracking capabilities.
• The integration of neural networks enhances robustness in complex dynamic environments.