Run addpath_RRT.m to add repository to Matlab path.

Basic RRT implementations in 2D and 3D.

Uses RRT for swing-up of a simple pendulum. The RRT_pend.m matlab script will plot the RRT and output the control torques required. The RRT_pend_novisuals.m matlab script outputs the control torques required, without showing a plot. The RRT_pend_fromC.m matlab script saves the control torques from 'build-RRT_pend_C-Desktop_Qt_5_3_GCC_64bit-Debug/data.txt' into the matlab workspace. The data.txt file is generated from the C version of the RRT algorithm (source in 'RRT_pend_c'). The pend_sim_euler.slx is a simulink simulation of the simple pendulum. To use the simulink simulation, run either of the following first:

    >> control = RRT_pend;
            OR
    >> control = RRT_pend_novisuals;
            OR
    >> RRT_pend_fromC;      (after running the C code)

Uses RRT for swing-up of an acrobot. The RRT_acrobot.m matlab script outputs the control torques required, without showing a plot. The RRT_acrobot_fromC.m matlab script saves the control torques from 'build-RRT_acrobot_C-Desktop_Qt_5_3_GCC_64bit-Debug/data.txt' into the matlab workspace. The data.txt file is generated from the C version of the RRT algorithm (source in 'RRT_acrobot_c'). The acrobot_sim_v3.slx is a simulink simulation of the acrobot. To use the simulink simulation, run either of the following first:

    >> control = RRT_acrobot
            OR
    >> RRT_acrobot_fromC;      (after running the C code)

Not sure if PFL implementation is working properly (probably needs task space PFL).

• get basic RRT implementation for acrobot working
• implement in C (manually)
• test in CUDA
• partial feedback linearization (need to work on task space PFL)

• parallelize using multi-RRT (test with dual-RRT first)