Planing path algo

---

This project is about optimizing path planing path algorithms for agent in 3D space.

---

Usage Example:

python3 main.py -c diffusion_3D.json -m denoise_model_002 -a rrt test
press 1 to show path found by RRT algorithm

press 2 to show noised path

press 3 to show denoised path by pre-trained model

Create own map

• How to add own models: Just add file to models3D folder, it must be .obj file and model must be composed from triangles.
• Set custom map: It must be writen in json file, and located in "configurations folder" example:

//example.json
{
  //name_of_uour_model.obj, [[possition_xyz], [orientation_rotX_rotY_rotZ]]
  "robot": ["robot.obj", [[0.0,0.0,0.0], [-1.5707963267948966,0.0,0.0]]],
  //name_of_obstacles.obj [[[possition_1], [orientation_1]], ... [[possition_i], [[orientation_i]]]
  "obstacles": ["cube.obj",
    [[[10.0, 0.0, 0.0], [0.0,0.0,0.0]],
    [[10.0, 0.0, 10.0], [0.0,0.0,0.0]]]],
  //name_of_goal.obj, [[possition_xyz], [orientation_rotX_rotY_rotZ]]
  "goal": ["goal.obj",[[20.0, 0.0, 20.0], [0.0,0.0,0.0]]]
}

Create own model for denoising

• You can use crated datasets in folder ./dataset

example: ./denoise_simple

• You have to add your pre-trained TF model to ./pretrained_models

There are implemented

• 3D engine and parser for .obj files
• SO3 space
• Rapidly Exploring Random Tree algorithm
• TF model for denoising path