How many times do you need to train the robot arm to reach the target position well?

dvalenciar / robotic_arm_environment

Doosan robotic arm, simulation, control, visualization in Gazebo and ROS2 for Reinforcement Learning.

MIT License

172 stars 29 forks source link

How many times do you need to train the robot arm to reach the target position well? #5

Closed hyungtai-kim closed 1 year ago

hyungtai-kim commented 1 year ago

It is very helpful to experience reinforcement learning simulation using your examples. I'm running through your examples, but it's hard to see that reinforcement learning is working. How many times can you tell if you've reached 80% of your target?

dvalenciar commented 1 year ago

Currently, there is no way to reach the goal now since it is taking random action. Soon I will implement some RL algorithms as TD3 or SAC, to solve the task What you have now is just the environment