An implementation of DancingPRMstar. Current version requires a lot of effort to refine the overall performance / readability also might have lots of bugs. Please enjoy it as an underdeveoped implementation.
How to run Installation tips:
- OMPL(Open Motion Planning Library, https://ompl.kavrakilab.org/, tested on 1.1.0)
- V-REP(http://www.coppeliarobotics.com/, tested on 3.2.2 EDU)
- It is based on Lazy PRM*(Lazy Collision Checking in Asymptotically-Optimal Motion Planning, Kris Hauser, 2015) with my own implemenetation of DSPT(Fully dynamic algorithms for maintaining shortest paths trees, D. Frigioni, Journal of Algoritms, 2000).
- Trajectory optimizer is not provided in this version. Please refer to the corresponding source code for the implementation tip. You can use uBLAS or Eigen library for linear algebra computation; this version is Eigen-friendly though.
- Routines for experiment and debug might remain incomplete/unused.
- Copy "DancingPRMstar.cpp" and "DancingPRMstar.h" into $(OMPL_PATH)/src/ompl/geometric/planners/prm/
Successful compilation would generate a library file(e.g., libompl.*). Copy it into /usr/local/lib so that V-REP can load the ompl library file successfully.
-
Replace v_repExtOMPL directory in V-REP directory($(VREP_PATH)/programming/v_repExtOMPL) with downloaded v_repExtOMPL.
-
Compile v_repExtOMPL and copy libv.repExtOMPL.* into V-REP directory(vrep.app/Contents/MacOS/)
-
Run V-REP and check the console ouptut to make sure that all the library files are successfully loaded.
-
Open example_scene.ttt in V-REP and Run!
If you have any problem with installation or following the above steps, feel free to mail me or leave a comment.