Scenario 1 exhibited a decreasing trend from 157.1 ± 23.7 (1 obstacle) to 144.5 ± 40.2 (4 obstacles), whereas Scenario 2 exhibited a modest increase from 149.4 ± 30.5 to 166.9 ± 34.4 as the number of dynamic obstacles increased.
← all excerpts
SOAR-RL: Safe and Open-Space Aware Reinforcement Learning for Mobile Robot Navigation in Narrow Spaces.
1
—
—