Questions tagged [reinforcement-learning]
Reinforcement learning is a technique wherein an agent improves its performance via interaction with its environment. For this reason it is a commonly used machine learning technique in robotics.
32
questions
16
votes
2
answers
5k
views
Programming a line following robot with reinforcement learning
I am considering programming a line following robot using reinforcement learning algorithms. The question I am pondering over is how can I get the algorithm to learn navigating through any arbitrary ...
- 575
8
votes
2
answers
15k
views
What is the definition of `rollout' in neural network or OpenAI gym
I'm relatively new to the area. I run into several time the term ``rollout'' in training neural networks. I have been searching for a while but still not sure what it means.
For example, the number ...
- 115
8
votes
2
answers
388
views
How do I model a robot?
The answers I received to the question on training a line following robot using reinforcement learning techniques, got me to think on how to train a robot. I believe there are essentially two ways -
...
- 575
7
votes
2
answers
1k
views
Are inverse kinematics and reinforcement learning competitive techniques?
Are inverse kinematics and reinforcement learning techniques contending techniques to solve the same problem viz. movement of robotic manipulators or arm?
By a glance through the wikipedia article, ...
- 575
7
votes
2
answers
838
views
Measuring speed of movement in Webots
I have been experimenting with different fitness functions for my Webots robot simulation (in short: I'm using genetic algorithm to evolve interesting behaviour).
The idea I have now is to reward/...
- 203
5
votes
1
answer
572
views
Tuning Line follower PID constants with Q-learning
I am currently working on a line follower buggy and have managed to tune the PID constants manually. The buggy follows the line at a moderate speed.
I will now like to take things further and learn ...
- 51
5
votes
1
answer
72
views
Has hierarchical learning been embodied in a robot before?
I've been reading about hierarchical reinforcement learning (HRL) and it's applications. A well-written literature review on the subject can be found here. However, I was wondering if research has ...
- 161
4
votes
3
answers
3k
views
PID tuning with (Deep) Reinforcement Learning
I am trying to implement a RL algorithm for an adaptive PID in a robot system. My doubt consists in the creation of the possible states in the problem.
I mean, I understand quite well the problem ...
- 470
4
votes
2
answers
262
views
Building robots with high reliability, durability, and battery life
I'm involved in research on psychologically plausible models of reinforcement learning, and as such I thought it'd be nice to try and see how well some to the models out there perform in the real ...
- 143
3
votes
1
answer
535
views
Ornstein Uhlenbeck vs Epsilon Greedy [closed]
There are many methods of exploring in a Reinforcement Learning setting but two of the most used ones are Ornstein Uhlenbeck (OU) processes and epsilon-greedy approaches. Could anyone elucidate the ...
- 51
2
votes
0
answers
51
views
Difference between CHOMP and DRL
in terms of motion planning, what are the difference between gradient-based motion planning (for instance, CHOMP http://www.nathanratliff.com/thesis-research/chomp) and deep reinforcement learning?
...
- 324
1
vote
1
answer
119
views
Applications of Reinforcement Learning
Many global-control applications of robotics involve incomplete world information and at best can be represented as a POMDP.
Given this, can we really apply RL to most robotics applications? From ...
- 393
1
vote
2
answers
3k
views
What are myopic and non-myopic policies?
I am reading this paper: A survey on Policy Search for Robotics and I came across these terms. Can someone give an answer, please?
- 11
1
vote
2
answers
129
views
What type of rigid body rotation can best be learned by neural networks?
I am training a reinforcement learning network in simulation for a robot which at the current stage learns Euler Angles to rotate the end-effector based on the actual state. The performance is overall ...
- 11
1
vote
1
answer
748
views
Having Issues Importing and Using RLGlue locally with Python For Reinforcement Learning
I'm going through an online course on Coursera for Reinforcement Learning that makes use of RLGlue. I want to try to run and adapt the code locally, but am having issues using RLGlue, and not sure why....
- 111
1
vote
2
answers
2k
views
Mujoco and MoveIt!
I want to implement deep reinforcement learning using a UR5 robot. A little research told me nowadays researchers are using openAI gym, Mujoco, rllab as their frameworks. The thing is, I want to train ...
- 11
1
vote
1
answer
117
views
Training Multiple Robots for different tasks at the same time using Deep Reinforcement Learning
I'm wondering if a single agent can train multiple robots to perform different tasks simultaneously. If possible, can you please recommend me some research papers and implementations that I can take a ...
1
vote
0
answers
126
views
Real-time python reinforcement learning library
I'm looking for a reinforcement learning library that can be used for real-time robot control. What I first had in mind was ROS to describe the robot, Mujoco to simulate physics, and OpenAI gym to ...
- 111
0
votes
2
answers
101
views
Reinforcement Learning in global and local path planning for mobile robots and self-driving car
Most RL courses start with grid world problem like this where robot has to navigate from start to end and RL helps in generating the optimum policy. (State-action pairing).
I am not able to relate ...
- 145
0
votes
1
answer
127
views
what are the main steps to build a swarm robots system and train it to achieve foraging task using deep Q network
I studied reinforcement learning deeper and prepared myself to use Webots, and when I decided to build a swarm robots system and drive it by deep_Q_networks I feel too confused
how can I begin and is ...
- 1
0
votes
1
answer
48
views
Reinforcement train butterfly robot in virtual reality?
Suppose I want to train butterfly robot with reinforcement learning. So I need physically correct simulation of aerodynamics and material physics (because butterfly wings should bend etc). Also I ...
- 115
0
votes
0
answers
10
views
I want to apply reinforcement learning to a manipulator. Seeking advice!
I am a university student currently studying reinforcement learning. I am trying to apply RL to a manipulator for the first time and feel a bit overwhelmed, so I would greatly appreciate any advice ...
- 1
0
votes
0
answers
10
views
How to import gazebo world into gym gazebo?
I've created my gazebo world included .world file, urdf files for models and bot, some plugin packages (realsense for gazebo) and .launch file. How can I register it as an environment in gym gazebo ...
- 119
0
votes
0
answers
28
views
how do I replace built-in algorithms on my PixHawk/ardupilot with custom reinforcement learning algorithms m?
I’ve been working on a drone project for 2 months now. Originally, it was supposed to take 2 only weeks. Everyday, there’s a new, unexpected problem that ends of being more difficult than it seems.
My ...
0
votes
0
answers
23
views
End2End learning for robotics tasks such as grasping, manipulation
I am in RL but new to robotics. I am trying to know what the best to train an RL policy in an end2end fashion for grasping or manipulations tasks using images. I can think about three ways. Can ...
- 1
0
votes
0
answers
58
views
reset model position in ros2
I want to reset the position of robot under certain circumstances. I am trying to do it with python node. I know we call '/reset_world" service to to this but i am unable to call it with python. ...
- 45
0
votes
0
answers
38
views
Configuration is inaccurate with more Mujoco mj_steps
I tried to construct a simulation env following fetch_pick_and_place. I noticed that the following code is used to initialize the env:
...
- 52
0
votes
1
answer
642
views
Gazebo ROS2 how to set Entity State when calling Set Model State via Node client
new ROS2/Gazebo user here.
I am trying to implement set up a script for reinforcement learning, currently making a gazebo environment class where upon reset the robot is teleported to a random place ...
- 1
0
votes
0
answers
50
views
Relation between control and reinforcement learning
Has the relation between Control and reinforcement learning been studied by the scientific community? I would like to read credible studies about each ones pros and cons when it comes to control ...
0
votes
1
answer
44
views
Setting up rewards to account for UAV crashes (Reinforcement Learning)
I am working on a project to implement a collision avoidance algorithm on a real UAV. I'm interested in understanding the process to set up a negative reward to account for scenarios wherein there is ...
- 105
0
votes
0
answers
30
views
Can supervised learning be used to solve the inverted pendulum problem?
I know that reinforcement learning has been used to solve the inverted pendulum problem.
Can supervised learning be used to solve the inverted pendulum problem?
For example, there could be an ...
- 101
0
votes
1
answer
58
views
Add failsafe to reinforcement learning algorithm
I'm working on a hexapod that uses A3C to learn how to walk. Ideally I would test it all in a simulator for some structure to the weights/policy but I don't have enough time for that. Obviously there ...
- 25