Manipulation task planning for heterogeneous object stacking based on reinforcement learning
The paper propose a new way of solving the Pallet Packing Problem by modelling it as a Markov decision process. This allows the program to make decisions step-by-step based only on the current state and adapt to any error in execution. By applying reinforcement learning techniques, an agent can be t...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
2019
|
Subjects: | |
Online Access: | http://hdl.handle.net/10356/76393 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-76393 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-763932023-03-04T18:59:38Z Manipulation task planning for heterogeneous object stacking based on reinforcement learning Pham, Minh Khang Domenico Campolo School of Mechanical and Aerospace Engineering DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Robotics DRNTU::Engineering::Mechanical engineering::Robots The paper propose a new way of solving the Pallet Packing Problem by modelling it as a Markov decision process. This allows the program to make decisions step-by-step based only on the current state and adapt to any error in execution. By applying reinforcement learning techniques, an agent can be trained from simulation to learn a model-free near-optimal policy that maximize the discounted cumulative rewards, which is proportional to the original objective function. Experiments show positive results on simulations involving packing up to 12 boxes into a grid-based pallet of size 8*8*6. Bachelor of Engineering (Mechanical Engineering) 2019-01-07T00:32:41Z 2019-01-07T00:32:41Z 2018 Final Year Project (FYP) http://hdl.handle.net/10356/76393 en Nanyang Technological University 38 p. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Robotics DRNTU::Engineering::Mechanical engineering::Robots |
spellingShingle |
DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Robotics DRNTU::Engineering::Mechanical engineering::Robots Pham, Minh Khang Manipulation task planning for heterogeneous object stacking based on reinforcement learning |
description |
The paper propose a new way of solving the Pallet Packing Problem by modelling it as a Markov decision process. This allows the program to make decisions step-by-step based only on the current state and adapt to any error in execution. By applying reinforcement learning techniques, an agent can be trained from simulation to learn a model-free near-optimal policy that maximize the discounted cumulative rewards, which is proportional to the original objective function. Experiments show positive results on simulations involving packing up to 12 boxes into a grid-based pallet of size 8*8*6. |
author2 |
Domenico Campolo |
author_facet |
Domenico Campolo Pham, Minh Khang |
format |
Final Year Project |
author |
Pham, Minh Khang |
author_sort |
Pham, Minh Khang |
title |
Manipulation task planning for heterogeneous object stacking based on reinforcement learning |
title_short |
Manipulation task planning for heterogeneous object stacking based on reinforcement learning |
title_full |
Manipulation task planning for heterogeneous object stacking based on reinforcement learning |
title_fullStr |
Manipulation task planning for heterogeneous object stacking based on reinforcement learning |
title_full_unstemmed |
Manipulation task planning for heterogeneous object stacking based on reinforcement learning |
title_sort |
manipulation task planning for heterogeneous object stacking based on reinforcement learning |
publishDate |
2019 |
url |
http://hdl.handle.net/10356/76393 |
_version_ |
1759855441959976960 |