tony23545/MotionPlanning

Course project for CSE571 (Probabilistic Robotics). Implement Differentiable Particles Filter, Motion Planning Networks and DDPG for a complete navigation pipeline.

MotionPlanning

This is our final project for Probabilistic Robotics. We implemented 3 components for a complete navigation pipeline:

• Differentiable Particles Filter for localization.
• Motion Planning Networks for global path planning.
• Deep Deterministic Policy Gradient for local control.

run the whole pipeline

python main.py

run MPNet and DDPG with ground truth position

python MPNet_DDPG.py

run the individual module, go to the corresponding folder.

link to the dataset:

rrt* dataset
rrt dataset