openai_ros example turtlebot2 does only circular movements
Hi! I followed the OpenAI tutorial with the turtlebot2 example and used the start_qlearning.py file, but when I run the launchfiles the turtlebot only moves in a circle then it seems to restart (some kind of collision?) and this happens all the time..
This is the output from the terminal:
NODES / turtlebot2_maze (turtle2_openai_ros_example/start_qlearning.py) ROS_MASTER_URI=http://localhost:11311 process[turtlebot2_maze-1]: started with pid [8204] [ERROR] [1598386894.407863, 24.150000]: NOT Initialising Simulation Physics Parameters [WARN] [1598386894.414771, 24.157000]: Start Init ControllersConnection [WARN] [1598386894.416358, 24.157000]: END Init ControllersConnection /home/henry/.local/lib/python2.7/site-packages/gym/logger.py:30: UserWarning: WARN: Box bound precision lowered by casting to float32 warnings.warn(colorize('%s: %s'%('WARN', msg % args), 'yellow')) [WARN] [1598386896.373830, 24.964000]: mod=120 [WARN] [1598386896.376594, 24.964000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.379235, 24.964000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.382009, 24.964000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.384714, 24.964000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.386477, 24.964000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.387933, 24.964000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.389596, 24.964000]: ############### Start Step=>0 [WARN] [1598386896.391132, 24.964000]: Next action is:2 [WARN] [1598386896.603553, 25.170000]: mod=120 [WARN] [1598386896.605311, 25.170000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.606913, 25.170000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.609013, 25.170000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.610608, 25.170000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.611779, 25.170000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.612802, 25.170000]: NOT done Validation >>> item=-inf< 0.5 [ERROR] [1598386896.613971, 25.170000]: TurtleBot2 is Ok ==> [WARN] [1598386896.615367, 25.170000]: [30.0, 30.0, 30.0, 30.0, 30.0, 30.0] 1 [WARN] [1598386896.616569, 25.170000]: # state we were=>30.030.030.030.030.030.0 [WARN] [1598386896.617597, 25.170000]: # action that we took=>2 [WARN] [1598386896.619158, 25.170000]: # reward that action gave=>1 [WARN] [1598386896.620522, 25.170000]: # episode cumulated_reward=>1 [WARN] [1598386896.622563, 25.170000]: # State in which we will start next step=>30.030.030.030.030.030.0 [WARN] [1598386896.624011, 25.170000]: NOT DONE [WARN] [1598386896.625498, 25.170000]: ############### END Step=>0 [WARN] [1598386896.626990, 25.170000]: ############### Start Step=>1 [WARN] [1598386896.628236, 25.170000]: Next action is:2 [WARN] [1598386896.844001, 25.377000]: mod=120 [WARN] [1598386896.845982, 25.377000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.847937, 25.377000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.849289, 25.377000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.850256, 25.377000]: NOT done Validation >>> item=-inf< 0.5 [WARN] [1598386896.851219, 25.377000]: NOT done Validation >>> item ...
Hey,
I think I solved this one.
I'll just link the thread I opened, but in short summary, there is probably no backwards or forwards potential actions set for your turtlebot.
ROS THREAD