Multiple Leader Candidate and Competitive Position Allocation for Robust Formation against Member Robot Faults

Area of Science:

• Robotics
• Distributed Systems
• Artificial Intelligence

Background:

• Traditional multi-robot formation structures often lack robustness against member robot failures.
• Existing position allocation methods can be rigid or computationally intensive.
• There is a need for adaptable and resilient multi-robot coordination strategies.

Purpose of the Study:

• To propose a novel Multiple Leader Candidate (MLC) structure for decentralized robot systems.
• To introduce a Competitive Position Allocation (CPA) algorithm for robust formation control.
• To demonstrate the applicability and performance of the MLC structure and CPA algorithm in multi-robot systems, including environmental sensing.

Main Methods:

• Development of a Multiple Leader Candidate (MLC) structure where a system leader is dynamically chosen.
• Implementation of a decentralized Competitive Position Allocation (CPA) algorithm for robot position assignment through inter-robot competition.
• Validation through numerical simulations and experimental results.

Main Results:

• The proposed MLC structure and CPA algorithm demonstrate robustness against the disappearance of member robots.
• The system achieves reduced overall cost compared to existing formation control methods.
• Feasibility and performance of the proposed approach are confirmed via simulations and experiments.

Conclusions:

• The MLC structure and CPA algorithm offer a resilient and cost-effective solution for multi-robot formations.
• This decentralized approach enhances fault tolerance in dynamic robotic systems.
• The findings are applicable to various fields, including environmental sensing and coordinated robotics.