CusADi: A GPU Parallelization Framework for Symbolic Expressions and Optimal Control

The parallelism afforded by GPUs presents significant advantages in training controllers through reinforcement learning (RL). However, integrating model-based optimization into this process remains challenging due to the complexity of formulating and solving optimization problems across thousands of...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE robotics and automation letters Ročník 10; číslo 2; s. 899 - 906
Hlavní autoři: Jeon, Se Hwan, Hong, Seungwoo, Lee, Ho Jae, Khazoom, Charles, Kim, Sangbae
Médium: Journal Article
Jazyk:angličtina
Vydáno: Piscataway IEEE 01.02.2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
Closed form solutions
Codes
Controllers
Exact solutions
Graphics processing units
Optimal control
Optimization
Optimization and optimal control
Reinforcement learning
Robot programming
Software tools
software tools for robot programming
Trajectory
ISSN:2377-3766, 2377-3766
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The parallelism afforded by GPUs presents significant advantages in training controllers through reinforcement learning (RL). However, integrating model-based optimization into this process remains challenging due to the complexity of formulating and solving optimization problems across thousands of instances. In this work, we present CusADi , an extension of the casadi symbolic framework to support the parallelization of arbitrary closed-form expressions on GPUs with CUDA . We also formulate a closed-form approximation for solving general optimal control problems, enabling large-scale parallelization and evaluation of MPC controllers. Our results show a ten-fold speedup relative to similar MPC implementation on the CPU, and we demonstrate the use of CusADi for various applications, including parallel simulation, parameter sweeps, and policy training.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2377-3766
2377-3766
DOI:10.1109/LRA.2024.3512254