Bearing-Only Formation Control With Prespecified Convergence Time.

Area of Science:

• Robotics
• Control Theory
• Multi-Agent Systems

Background:

• Bearing-only formation control is crucial for autonomous systems where agents rely solely on relative directional information.
• Existing methods often lack finite-time convergence guarantees or continuous control inputs, posing challenges for practical applications.

Purpose of the Study:

• To develop a new bearing-only formation control law for multi-agent systems.
• To achieve finite-time convergence to target formations with guaranteed collision avoidance.
• To enable user-defined convergence times and ensure smooth control inputs.

Main Methods:

• A novel control law utilizing a time-varying scaling gain for bearing-only formation control.
• Development of a leader-follower control structure for global convergence.
• Analysis of the bearing Laplacian matrix properties for collision avoidance and smooth control.

Main Results:

• The proposed control law achieves finite-time convergence to target formations.
• Sufficient conditions for almost global convergence and inter-agent collision avoidance are established.
• The leader-follower structure ensures global convergence while preserving collision avoidance and smooth control inputs.
• Validation through simulations and experiments on a multi-robot hardware platform.

Conclusions:

• The novel bearing-only formation control strategy effectively achieves target formations in finite time.
• The use of a time-varying gain allows for user-specified convergence rates and continuous control derivatives.
• The proposed methods ensure robust performance, including collision avoidance and smooth control, validated by empirical evidence.