[2N6-GS-1] Fundamental AI, theory: Constraint satisfaction and optimization

[2N6-GS-1-03] DSSA+for3D

Expansion of DSSA+ into the 3D Space

〇Tomohiro Shiota¹, Katsutoshi Hirayama¹, Tenda Okimoto¹ (1. Kobe University)

Keywords:Multi-Agent System, Cooperative Problem Solving, Agent Simulation

Distributed Stochastic Search Algorithm(DSSA) is one of state-of-the-art distributed algorithms for the ship collision avoidance problems. In DSSA, when another ship approaches one’s own ship, she will select the optimal and safe course by exchanging intentions(headings) with each other. The extended version of DSSA+ can realize more efficient and general collision avoidance by considering not only the course change but also the speed change. However, since both them assume collision avoidance in two-dimensional space, further expansion and reconsideration of the existing framework are necessary to deal with collision avoidance problems of autonomous vehicles in three-dimensional space such as drones in the future. In this paper, we propose a new distributed collision avoidance algorithm with such an extension and evaluate its performance.