Force-Based Algorithm for Motion Planning of Large Agent.

Area of Science:

• Robotics
• Artificial Intelligence
• Computer Science

Background:

• Multi-agent systems require sophisticated motion planning for collision avoidance.
• Existing algorithms often have high computational overhead and depend on detailed state information.

Purpose of the Study:

• To present a distributed, efficient, scalable, and real-time motion planning algorithm for large groups of agents.
• To enable autonomous agents to generate individual trajectories independently using only relative positioning.

Main Methods:

• A novel force-based motion planning (FMP) algorithm is proposed.
• Each agent uses a control law with collision avoidance and navigational feedback terms.
• The algorithm relies solely on relative position information of neighboring agents, not velocity states.

Main Results:

• The FMP algorithm achieves collision-free motions with lower transition times compared to existing methods.
• It significantly reduces computational overhead by not requiring velocity state information.
• Simulations in dense and complex 2-D and 3-D scenarios demonstrate superior performance.

Conclusions:

• The proposed FMP algorithm offers an efficient and scalable solution for real-time motion planning in multi-agent systems.
• It effectively handles collision avoidance and navigation with reduced computational demands.
• The method outperforms existing approaches in complex simulated environments.