A Holistic Assessment Method For Deep Reinforcement Learning
Open a Terminal
git clone git@github.com:YD-19/HAM4DRL.git --recurse-submodules
if you want to train in our small environment, you need to modify the description file in turtlebot3 package.
roscd turtlebot3_description/urdf && gedit turtlebot3_waffle_pi.gazebo.xacro
find the base_scan model and modify as follow:
<scan>
<horizontal>
<samples>24</samples>
<resolution>1</resolution>
<min_angle>-1.57</min_angle>
<max_angle>1.57</max_angle>
</horizontal>
</scan>
After modification, make the files
cd ~/HAM4DRL && catkin_make
Open a new terminal:
source ~/HAM4DRL/devel/setup.bash
roslaunch turtlebot_DDPG acps_ddpg.launch
Open another terminal:
source ~/HAM4DRL/devel/setup.bash
rosrun turtlebot_DDPG ddpg_main_1804.py