{"id":"https://openalex.org/W4388692115","doi":"https://doi.org/10.1109/lra.2023.3333248","title":"Safe Reinforcement Learning With Dead-Ends Avoidance and Recovery","display_name":"Safe Reinforcement Learning With Dead-Ends Avoidance and Recovery","publication_year":2023,"publication_date":"2023-11-15","ids":{"openalex":"https://openalex.org/W4388692115","doi":"https://doi.org/10.1109/lra.2023.3333248"},"language":"en","primary_location":{"id":"doi:10.1109/lra.2023.3333248","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lra.2023.3333248","pdf_url":null,"source":{"id":"https://openalex.org/S4210169774","display_name":"IEEE Robotics and Automation Letters","issn_l":"2377-3766","issn":["2377-3766"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Robotics and Automation Letters","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102967827","display_name":"Xiao Zhang","orcid":"https://orcid.org/0009-0009-8081-2079"},"institutions":[{"id":"https://openalex.org/I116953780","display_name":"Tongji University","ror":"https://ror.org/03rc6as71","country_code":"CN","type":"education","lineage":["https://openalex.org/I116953780"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiao Zhang","raw_affiliation_strings":["Department of Computer Science and Technology, Tongji University, Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0009-8081-2079","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Technology, Tongji University, Shanghai, China","institution_ids":["https://openalex.org/I116953780"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100646312","display_name":"Hai Zhang","orcid":"https://orcid.org/0009-0003-1301-8605"},"institutions":[{"id":"https://openalex.org/I116953780","display_name":"Tongji University","ror":"https://ror.org/03rc6as71","country_code":"CN","type":"education","lineage":["https://openalex.org/I116953780"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hai Zhang","raw_affiliation_strings":["Department of Computer Science and Technology, Tongji University, Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0003-1301-8605","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Technology, Tongji University, Shanghai, China","institution_ids":["https://openalex.org/I116953780"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050285917","display_name":"Hongtu Zhou","orcid":"https://orcid.org/0000-0002-0609-4226"},"institutions":[{"id":"https://openalex.org/I116953780","display_name":"Tongji University","ror":"https://ror.org/03rc6as71","country_code":"CN","type":"education","lineage":["https://openalex.org/I116953780"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongtu Zhou","raw_affiliation_strings":["Department of Computer Science and Technology, Tongji University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-0609-4226","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Technology, Tongji University, Shanghai, China","institution_ids":["https://openalex.org/I116953780"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101425674","display_name":"Chang Shan Huang","orcid":"https://orcid.org/0009-0000-4815-8761"},"institutions":[{"id":"https://openalex.org/I116953780","display_name":"Tongji University","ror":"https://ror.org/03rc6as71","country_code":"CN","type":"education","lineage":["https://openalex.org/I116953780"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chang Huang","raw_affiliation_strings":["Department of Computer Science and Technology, Tongji University, Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0000-4815-8761","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Technology, Tongji University, Shanghai, China","institution_ids":["https://openalex.org/I116953780"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103209609","display_name":"Di Zhang","orcid":"https://orcid.org/0009-0001-7135-272X"},"institutions":[{"id":"https://openalex.org/I116953780","display_name":"Tongji University","ror":"https://ror.org/03rc6as71","country_code":"CN","type":"education","lineage":["https://openalex.org/I116953780"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Di Zhang","raw_affiliation_strings":["Department of Computer Science and Technology, Tongji University, Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0001-7135-272X","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Technology, Tongji University, Shanghai, China","institution_ids":["https://openalex.org/I116953780"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100396642","display_name":"Chen Ye","orcid":"https://orcid.org/0000-0002-0048-6974"},"institutions":[{"id":"https://openalex.org/I116953780","display_name":"Tongji University","ror":"https://ror.org/03rc6as71","country_code":"CN","type":"education","lineage":["https://openalex.org/I116953780"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chen Ye","raw_affiliation_strings":["Department of Computer Science and Technology, Tongji University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-0048-6974","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Technology, Tongji University, Shanghai, China","institution_ids":["https://openalex.org/I116953780"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5007977858","display_name":"Junqiao Zhao","orcid":"https://orcid.org/0000-0002-7864-3255"},"institutions":[{"id":"https://openalex.org/I116953780","display_name":"Tongji University","ror":"https://ror.org/03rc6as71","country_code":"CN","type":"education","lineage":["https://openalex.org/I116953780"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Junqiao Zhao","raw_affiliation_strings":["Department of Computer Science and Technology, Tongji University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-7864-3255","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Technology, Tongji University, Shanghai, China","institution_ids":["https://openalex.org/I116953780"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I116953780"],"apc_list":null,"apc_paid":null,"fwci":1.269,"has_fulltext":false,"cited_by_count":8,"citation_normalized_percentile":{"value":0.8341098,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":98},"biblio":{"volume":"9","issue":"1","first_page":"491","last_page":"498"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9945999979972839,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10653","display_name":"Robot Manipulation and Learning","score":0.9944000244140625,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.9091026782989502},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.7397966980934143},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.727607250213623},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.631801962852478},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.583428680896759},{"id":"https://openalex.org/keywords/reinforcement","display_name":"Reinforcement","score":0.5553568005561829},{"id":"https://openalex.org/keywords/boundary","display_name":"Boundary (topology)","score":0.5115900635719299},{"id":"https://openalex.org/keywords/state","display_name":"State (computer science)","score":0.4585326910018921},{"id":"https://openalex.org/keywords/risk-analysis","display_name":"Risk analysis (engineering)","score":0.44524285197257996},{"id":"https://openalex.org/keywords/computer-security","display_name":"Computer security","score":0.38772067427635193},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.33424684405326843},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.18875399231910706},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.11466005444526672},{"id":"https://openalex.org/keywords/business","display_name":"Business","score":0.10056307911872864},{"id":"https://openalex.org/keywords/social-psychology","display_name":"Social psychology","score":0.08519917726516724},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.0811048150062561}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.9091026782989502},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.7397966980934143},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.727607250213623},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.631801962852478},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.583428680896759},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.5553568005561829},{"id":"https://openalex.org/C62354387","wikidata":"https://www.wikidata.org/wiki/Q875399","display_name":"Boundary (topology)","level":2,"score":0.5115900635719299},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.4585326910018921},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.44524285197257996},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.38772067427635193},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33424684405326843},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.18875399231910706},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.11466005444526672},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.10056307911872864},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.08519917726516724},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.0811048150062561},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/lra.2023.3333248","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lra.2023.3333248","pdf_url":null,"source":{"id":"https://openalex.org/S4210169774","display_name":"IEEE Robotics and Automation Letters","issn_l":"2377-3766","issn":["2377-3766"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Robotics and Automation Letters","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","display_name":"Reduced inequalities","score":0.47999998927116394}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":33,"referenced_works":["https://openalex.org/W2145339207","https://openalex.org/W2966735560","https://openalex.org/W2968983352","https://openalex.org/W2978134924","https://openalex.org/W3100789280","https://openalex.org/W3135301907","https://openalex.org/W3150718622","https://openalex.org/W3176452384","https://openalex.org/W3184308941","https://openalex.org/W3205794883","https://openalex.org/W4206497039","https://openalex.org/W4226220005","https://openalex.org/W4287370827","https://openalex.org/W4287629996","https://openalex.org/W4312347470","https://openalex.org/W4313005236","https://openalex.org/W4316843729","https://openalex.org/W4385767976","https://openalex.org/W6736057607","https://openalex.org/W6737893269","https://openalex.org/W6747473740","https://openalex.org/W6762823601","https://openalex.org/W6764053384","https://openalex.org/W6771280675","https://openalex.org/W6774126978","https://openalex.org/W6779812412","https://openalex.org/W6784630725","https://openalex.org/W6797010411","https://openalex.org/W6802659552","https://openalex.org/W6802880167","https://openalex.org/W6803800456","https://openalex.org/W6810543476","https://openalex.org/W6848903118"],"related_works":["https://openalex.org/W2366107444","https://openalex.org/W4388145910","https://openalex.org/W2920061524","https://openalex.org/W4310083477","https://openalex.org/W2328553770","https://openalex.org/W2381570729","https://openalex.org/W1976205134","https://openalex.org/W4248336175","https://openalex.org/W2031260042","https://openalex.org/W2391445434"],"abstract_inverted_index":{"Safety":[0],"is":[1,67,81,168],"one":[2],"of":[3,148,155,183],"the":[4,22,74,114,141,145,159,171,175],"main":[5],"challenges":[6],"in":[7,13,139,150],"applying":[8],"reinforcement":[9],"learning":[10],"to":[11,28,33,52,69,77,92,102],"tasks":[12,186],"realistic":[14],"environments.":[15],"To":[16],"ensure":[17],"safety":[18,131,142,149,157,197],"during":[19],"and":[20,60,83,117,128],"after":[21],"training":[23],"process,":[24],"existing":[25],"methods":[26],"tend":[27],"adopt":[29],"overly":[30,39],"conservative":[31,40],"policies":[32],"avoid":[34],"unsafe":[35,61],"situations.":[36],"However,":[37],"an":[38,136],"policy":[41,109,121,127],"severely":[42],"hinders":[43],"exploration.":[44,90],"In":[45],"this":[46],"letter,":[47],"we":[48,65,96],"propose":[49],"a":[50,54,86,98,107,119,129,164],"method":[51],"construct":[53,66],"boundary":[55,64],"that":[56,110,122,188],"discriminates":[57],"between":[58],"safe":[59,79,178],"states.":[62],"The":[63,125],"equivalent":[68],"distinguishing":[70],"dead-end":[71],"states,":[72],"indicating":[73],"maximum":[75],"extent":[76],"which":[78,140],"exploration":[80],"guaranteed,":[82],"thus":[84],"has":[85,191],"minimum":[87],"limitation":[88],"on":[89,135],"Similar":[91],"Recovery":[93],"Reinforcement":[94],"Learning,":[95],"utilize":[97],"decoupled":[99],"RL":[100],"framework":[101],"learn":[103],"two":[104],"policies,":[105],"(1)":[106],"task":[108,115,193],"only":[111],"considers":[112],"improving":[113],"performance,":[116],"(2)":[118],"recovery":[120,126],"maximizes":[123],"safety.":[124],"corresponding":[130],"critic":[132,143],"are":[133],"pre-trained":[134],"offline":[137],"dataset,":[138],"evaluates":[144],"upper":[146],"bound":[147],"each":[151],"state":[152],"as":[153],"awareness":[154],"environmental":[156],"for":[158],"agent.":[160],"During":[161],"online":[162],"training,":[163],"behavior":[165],"correction":[166],"mechanism":[167],"adopted,":[169],"ensuring":[170],"agent":[172],"interacts":[173],"with":[174,195],"environment":[176],"using":[177],"actions":[179],"only.":[180],"Finally,":[181],"experiments":[182],"continuous":[184],"control":[185],"demonstrate":[187],"our":[189],"approach":[190],"better":[192],"performance":[194],"fewer":[196],"violations":[198],"than":[199],"state-of-the-art":[200],"algorithms.":[201]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":2}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}