Reproduce “TD3 Based Collision Free Motion Planning for Robot Navigation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Hao Liu, Yi Shen, Chang Zhou, Yuelin Zou, Zijun Gao, Qi Wang
TD3-DWA combines DWA with TD3, using learned linear and angular velocities directly instead of sampled reference trajectories. In a 10 × 15 m Gazebo simulation over 100 tests, it recorded 2/100 collisions and an average route time of 11.6 s, versus DQN-DWA’s 3/100 and 14.5 s. Its average path was longer: 11.6 m versus 10.3 m. Evidence is limited to simulation, not real-world safety.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.