Python Motion Planning v2.0.dev2 Release!

Release v2.0.dev2 - 2025/12/21

Update

1. New planners:
   • GBFS (Greedy Best-First Search)
   • JPS (Jump Point Search)
   • RRT-Connect
   • Voronoi Planner
2. New controller:
   • RPP (Regulated Pure-Pursuit)
3. Improve graph search planners:
   • Add diagonal argument, which determines whether expand nodes diagonally or not.
4. Improve sample search planners:
   • Now you can plan in continuous space.
   • Use Faiss to accelerate the node search.
   • Improve asymptotically optimal RRT* planner. Add stop_until_sample_num to provide an option to wait for a better result.
5. Improve Grid:
   • Some interfaces changes (such as type_map.array -> type_map.data)
   • New function point_float_to_int, etc.
6. Improve visualizer:
   • Switching to PyVista, a 3D visualization library that renders faster than Matplotlib
   • Some interface changes (such as dividing the original Visualizer class into three classes BaseVisualizer, Visualizer2D, and Visualizer3D)
   • Add NE (Navigation Error), DTW (Dynamic Time Warping), nDTW (normalized Dynamic Time Warping) metrics to get_traj_info function.

Python Motion Planning v2.0.dev1 Release!

Release v2.0.dev1 - 2025/10/3

Update

1. Brand new and more standardized interfaces, but it is not compatible with previous versions.
2. Provide a toy simulator which provides simple physics simulation and supports multi-agent and is aligned with OpenAI gymnasium library.
3. Provide different robot models including circular omnidirectional robots and differential-drive robots.
4. Support N-D path planning. But only 2D and 3D path planning can be visualized.
5. More powerful visualizer.

However, due to the complete reconstruction of the library, many algorithms require time to be migrated. Currently, only some algorithms have been migrated, so this version is an unstable development version.

Python Motion Planning v1.1.1 Release!

Release v1.1.1 - 2025/4/11

Update

1. Clarify the annotated data type of env argument in planners.
2. Discard the preset obstacles in the init() function of Grid and Map, only preserving boundaries.
3. Supplement examples of adding obstacles in the document.

Python Motion Planning v1.1 Release!

Release v1.1 - 2025/3/6

Update

1. You can pip install python-motion-planning now !
2. Modify the Mkdocs documentation and provide an online version.
3. Delete learning-base algorithms DQN and DDPG temporarily for stability.
4. Delete some redundant files.

Python Motion Planning v1.0 Release!

Release v1.0 - 2024/11/22

V1.0 is our first stable version.

You can find 17 global planners, 8 local planners and 7 curve generators in Python Motion Planning in v1.0. Below are the algorithms:

python_motion_planning
├── curve_generation
│   ├── bezier_curve
│   ├── bspline_curve
│   ├── cubic_spline
│   ├── dubins_curve
│   ├── fem_pos_smooth
│   ├── polynomial_curve
│   └── reeds_shepp
├── global_planner
│   ├── evolutionary_search
│   │   ├── aco
│   │   └── pso
│   ├── graph_search
│   │   ├── a_star
│   │   ├── d_star
│   │   ├── d_star_lite
│   │   ├── dijkstra
│   │   ├── gbfs
│   │   ├── jps
│   │   ├── lazy_theta_star
│   │   ├── lpa_star
│   │   ├── s_theta_star
│   │   ├── theta_star
│   │   └── voronoi
│   └── sample_search
│       ├── informed_rrt
│       ├── rrt
│       ├── rrt_connect
│       └── rrt_star
└── local_planner
    ├── apf
    ├── ddpg
    ├── dqn
    ├── dwa
    ├── lqr
    ├── mpc
    ├── pid
    └── rpp