Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311443347> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W4311443347 endingPage "435" @default.
- W4311443347 startingPage "419" @default.
- W4311443347 abstract "AbstractReinforcement Learning (RL) and continuous nonlinear control have been successfully deployed in multiple domains of complicated sequential decision-making tasks. However, given the exploration nature of the learning process and the presence of model uncertainty, it is challenging to apply them to safety-critical control tasks due to the lack of safety guarantee. On the other hand, while combining control-theoretical approaches with learning algorithms has shown promise in safe RL applications, the sample efficiency of safe data collection process for control is not well addressed. In this paper, we propose a provably sample efficient episodic safe learning framework for online control tasks that leverages safe exploration and exploitation in an unknown, nonlinear dynamical system. In particular, the framework (1) extends control barrier functions (CBFs) in a stochastic setting to achieve provable high-probability safety under uncertainty during model learning and (2) integrates an optimism-based exploration strategy to efficiently guide the safe exploration process with learned dynamics for near optimal control performance. We provide formal analysis on the episodic regret bound against the optimal controller and probabilistic safety with theoretical guarantees. Simulation results are provided to demonstrate the effectiveness and efficiency of the proposed algorithm.KeywordsRobot safetySafe reinforcement learningSafe control" @default.
- W4311443347 created "2022-12-26" @default.
- W4311443347 creator A5032266950 @default.
- W4311443347 creator A5036126020 @default.
- W4311443347 creator A5078150431 @default.
- W4311443347 date "2022-12-15" @default.
- W4311443347 modified "2023-10-16" @default.
- W4311443347 title "Sample-Efficient Safe Learning for Online Nonlinear Control with Control Barrier Functions" @default.
- W4311443347 cites W2588802774 @default.
- W4311443347 cites W2735010720 @default.
- W4311443347 cites W2738778707 @default.
- W4311443347 cites W2798791976 @default.
- W4311443347 cites W2822752092 @default.
- W4311443347 cites W2889711700 @default.
- W4311443347 cites W2892521964 @default.
- W4311443347 cites W2899087494 @default.
- W4311443347 cites W2909324603 @default.
- W4311443347 cites W2963148914 @default.
- W4311443347 cites W2963525569 @default.
- W4311443347 cites W2964138223 @default.
- W4311443347 cites W2966735560 @default.
- W4311443347 cites W2968945909 @default.
- W4311443347 cites W2972429904 @default.
- W4311443347 cites W2977676795 @default.
- W4311443347 cites W3024987186 @default.
- W4311443347 cites W3038180127 @default.
- W4311443347 cites W3048735518 @default.
- W4311443347 cites W3098925401 @default.
- W4311443347 cites W3119686839 @default.
- W4311443347 cites W3186317065 @default.
- W4311443347 cites W3207570104 @default.
- W4311443347 cites W4205841652 @default.
- W4311443347 cites W4230431917 @default.
- W4311443347 doi "https://doi.org/10.1007/978-3-031-21090-7_25" @default.
- W4311443347 hasPublicationYear "2022" @default.
- W4311443347 type Work @default.
- W4311443347 citedByCount "0" @default.
- W4311443347 crossrefType "book-chapter" @default.
- W4311443347 hasAuthorship W4311443347A5032266950 @default.
- W4311443347 hasAuthorship W4311443347A5036126020 @default.
- W4311443347 hasAuthorship W4311443347A5078150431 @default.
- W4311443347 hasBestOaLocation W43114433472 @default.
- W4311443347 hasConcept C111919701 @default.
- W4311443347 hasConcept C119857082 @default.
- W4311443347 hasConcept C126255220 @default.
- W4311443347 hasConcept C154945302 @default.
- W4311443347 hasConcept C185592680 @default.
- W4311443347 hasConcept C198531522 @default.
- W4311443347 hasConcept C203479927 @default.
- W4311443347 hasConcept C2775924081 @default.
- W4311443347 hasConcept C33923547 @default.
- W4311443347 hasConcept C41008148 @default.
- W4311443347 hasConcept C43617362 @default.
- W4311443347 hasConcept C49937458 @default.
- W4311443347 hasConcept C50817715 @default.
- W4311443347 hasConcept C6557445 @default.
- W4311443347 hasConcept C86803240 @default.
- W4311443347 hasConcept C91575142 @default.
- W4311443347 hasConcept C97541855 @default.
- W4311443347 hasConcept C98045186 @default.
- W4311443347 hasConceptScore W4311443347C111919701 @default.
- W4311443347 hasConceptScore W4311443347C119857082 @default.
- W4311443347 hasConceptScore W4311443347C126255220 @default.
- W4311443347 hasConceptScore W4311443347C154945302 @default.
- W4311443347 hasConceptScore W4311443347C185592680 @default.
- W4311443347 hasConceptScore W4311443347C198531522 @default.
- W4311443347 hasConceptScore W4311443347C203479927 @default.
- W4311443347 hasConceptScore W4311443347C2775924081 @default.
- W4311443347 hasConceptScore W4311443347C33923547 @default.
- W4311443347 hasConceptScore W4311443347C41008148 @default.
- W4311443347 hasConceptScore W4311443347C43617362 @default.
- W4311443347 hasConceptScore W4311443347C49937458 @default.
- W4311443347 hasConceptScore W4311443347C50817715 @default.
- W4311443347 hasConceptScore W4311443347C6557445 @default.
- W4311443347 hasConceptScore W4311443347C86803240 @default.
- W4311443347 hasConceptScore W4311443347C91575142 @default.
- W4311443347 hasConceptScore W4311443347C97541855 @default.
- W4311443347 hasConceptScore W4311443347C98045186 @default.
- W4311443347 hasLocation W43114433471 @default.
- W4311443347 hasLocation W43114433472 @default.
- W4311443347 hasOpenAccess W4311443347 @default.
- W4311443347 hasPrimaryLocation W43114433471 @default.
- W4311443347 hasRelatedWork W1790524216 @default.
- W4311443347 hasRelatedWork W2923504512 @default.
- W4311443347 hasRelatedWork W3022038857 @default.
- W4311443347 hasRelatedWork W3036350490 @default.
- W4311443347 hasRelatedWork W3094206923 @default.
- W4311443347 hasRelatedWork W3158149540 @default.
- W4311443347 hasRelatedWork W3185561310 @default.
- W4311443347 hasRelatedWork W4287636452 @default.
- W4311443347 hasRelatedWork W4311969089 @default.
- W4311443347 hasRelatedWork W4319083788 @default.
- W4311443347 isParatext "false" @default.
- W4311443347 isRetracted "false" @default.
- W4311443347 workType "book-chapter" @default.