Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration
The time-optimal control problem (TOCP) has faced new practical challenges, such as those from the deployment of agile autonomous vehicles in diverse uncertain operating conditions without accurate system calibration. In this study to meet a need to generate feasible speed profiles in the face of un...
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Capítulo de libro |
| Lenguaje: | inglés |
| Publicado: |
IntechOpen
2022
|
| Materias: | |
| Acceso en línea: | https://www.intechopen.com/chapters/81321 |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1879164822509060096 |
|---|---|
| author | Liao, Hsuan-Cheng Chou, Han-Jung Liu, Jing-Sin |
| author_browse | Chou, Han-Jung Liao, Hsuan-Cheng Liu, Jing-Sin |
| author_facet | Liao, Hsuan-Cheng Chou, Han-Jung Liu, Jing-Sin |
| author_sort | Liao, Hsuan-Cheng |
| building | IntechOpen |
| collection | Acceso abierto |
| description | The time-optimal control problem (TOCP) has faced new practical challenges, such as those from the deployment of agile autonomous vehicles in diverse uncertain operating conditions without accurate system calibration. In this study to meet a need to generate feasible speed profiles in the face of uncertainty, we exploit and implement probabilistic inference for learning control (PILCO), an existing sample-efficient model-based reinforcement learning (MBRL) framework for policy search, to a case study of TOCP for a vehicle that was modeled as a constant input-constrained double integrator with uncertain inertia subject to uncertain viscous friction. Our approach integrates learning, planning, and control to construct a generalizable approach that requires minimal assumptions (especially regarding external disturbances and the parametric dynamics model of the system) for solving TOCP approximately as the perturbed solutions close to time-optimality. Within PILCO, a Gaussian Radial basis functions is implemented to generate control-constrained rest-to-rest near time-optimal vehicle motion on a linear track from scratch with data-efficiency in a direct way. We briefly introduce the importance of the applications of PILCO and discuss the learning results that PILCO would actually converge to the analytical solution in this TOCP. Furthermore, we execute a simulation and a sim2real experiment to validate the suitability of PILCO for TOCP by comparing with the analytical solution. |
| doi_str_mv | 10.5772/intechopen.103690 |
| first_indexed | 2026-10-15T21:40:27Z |
| format | Book Chapter |
| fullrecord | <oai_dc:dc xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd"><identifier>intech-81321</identifier><datestamp>2022-04-16</datestamp>
<dc:title>Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration</dc:title>
<dc:creator>Hsuan-Cheng Liao</dc:creator>
<dc:creator>Han-Jung Chou</dc:creator>
<dc:creator>Jing-Sin Liu</dc:creator>
<dc:subject>Physical Sciences, Engineering and Technology</dc:subject>
<dc:description>The time-optimal control problem (TOCP) has faced new practical challenges, such as those from the deployment of agile autonomous vehicles in diverse uncertain operating conditions without accurate system calibration. In this study to meet a need to generate feasible speed profiles in the face of uncertainty, we exploit and implement probabilistic inference for learning control (PILCO), an existing sample-efficient model-based reinforcement learning (MBRL) framework for policy search, to a case study of TOCP for a vehicle that was modeled as a constant input-constrained double integrator with uncertain inertia subject to uncertain viscous friction. Our approach integrates learning, planning, and control to construct a generalizable approach that requires minimal assumptions (especially regarding external disturbances and the parametric dynamics model of the system) for solving TOCP approximately as the perturbed solutions close to time-optimality. Within PILCO, a Gaussian Radial basis functions is implemented to generate control-constrained rest-to-rest near time-optimal vehicle motion on a linear track from scratch with data-efficiency in a direct way. We briefly introduce the importance of the applications of PILCO and discuss the learning results that PILCO would actually converge to the analytical solution in this TOCP. Furthermore, we execute a simulation and a sim2real experiment to validate the suitability of PILCO for TOCP by comparing with the analytical solution.</dc:description>
<dc:publisher>IntechOpen</dc:publisher>
<dc:date>2022-04-16</dc:date>
<dc:type>Chapter, Part Of Book</dc:type>
<dc:identifier>https://www.intechopen.com/chapters/81321</dc:identifier>
<dc:identifier>doi:10.5772/intechopen.103690</dc:identifier>
<dc:language>en</dc:language>
<dc:relation>ISBN:978-1-83768-946-0</dc:relation>
<dc:rights>https://creativecommons.org/licenses/by/3.0/</dc:rights>
<dc:source>https://www.intechopen.com/books/12879 ; Artificial Intelligence Annual Volume 2022</dc:source>
</oai_dc:dc> |
| id | intech-81321 |
| institution | IntechOpen |
| isbn | 978-1-83768-946-0 |
| language | English |
| last_indexed | 2026-10-15T21:40:27Z |
| publishDate | 2022 |
| publishDateRange | 2022 |
| publishDateSort | 2022 |
| publisher | IntechOpen |
| publisherStr | IntechOpen |
| record_format | oai_dc |
| spelling | intech-813212022-04-16 Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration Hsuan-Cheng Liao Han-Jung Chou Jing-Sin Liu Physical Sciences, Engineering and Technology The time-optimal control problem (TOCP) has faced new practical challenges, such as those from the deployment of agile autonomous vehicles in diverse uncertain operating conditions without accurate system calibration. In this study to meet a need to generate feasible speed profiles in the face of uncertainty, we exploit and implement probabilistic inference for learning control (PILCO), an existing sample-efficient model-based reinforcement learning (MBRL) framework for policy search, to a case study of TOCP for a vehicle that was modeled as a constant input-constrained double integrator with uncertain inertia subject to uncertain viscous friction. Our approach integrates learning, planning, and control to construct a generalizable approach that requires minimal assumptions (especially regarding external disturbances and the parametric dynamics model of the system) for solving TOCP approximately as the perturbed solutions close to time-optimality. Within PILCO, a Gaussian Radial basis functions is implemented to generate control-constrained rest-to-rest near time-optimal vehicle motion on a linear track from scratch with data-efficiency in a direct way. We briefly introduce the importance of the applications of PILCO and discuss the learning results that PILCO would actually converge to the analytical solution in this TOCP. Furthermore, we execute a simulation and a sim2real experiment to validate the suitability of PILCO for TOCP by comparing with the analytical solution. IntechOpen 2022-04-16 Chapter, Part Of Book https://www.intechopen.com/chapters/81321 doi:10.5772/intechopen.103690 en ISBN:978-1-83768-946-0 https://creativecommons.org/licenses/by/3.0/ https://www.intechopen.com/books/12879 ; Artificial Intelligence Annual Volume 2022 |
| spellingShingle | Physical Sciences, Engineering and Technology Liao, Hsuan-Cheng Chou, Han-Jung Liu, Jing-Sin Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration |
| title | Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration |
| title_full | Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration |
| title_fullStr | Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration |
| title_full_unstemmed | Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration |
| title_short | Velocity Planning via Model-Based Reinforcement Learning: Demonstrating Results on PILCO for One-Dimensional Linear Motion with Bounded Acceleration |
| title_sort | velocity planning via model based reinforcement learning demonstrating results on pilco for one dimensional linear motion with bounded acceleration |
| topic | Physical Sciences, Engineering and Technology |
| url | https://www.intechopen.com/chapters/81321 |
| work_keys_str_mv | AT liaohsuancheng velocityplanningviamodelbasedreinforcementlearningdemonstratingresultsonpilcoforonedimensionallinearmotionwithboundedacceleration AT chouhanjung velocityplanningviamodelbasedreinforcementlearningdemonstratingresultsonpilcoforonedimensionallinearmotionwithboundedacceleration AT liujingsin velocityplanningviamodelbasedreinforcementlearningdemonstratingresultsonpilcoforonedimensionallinearmotionwithboundedacceleration |