Probabilistic Model Predictive Safety Certification for Learning-Based Control
Reinforcement learning (RL) methods have demonstrated their efficiency in simulation. However, many of the applications for which RL offers great potential, such as autonomous driving, are also safety critical and require a certified closed-loop behavior in order to meet the safety specifications in...
Saved in:
Published in | IEEE transactions on automatic control Vol. 67; no. 1; pp. 176 - 188 |
---|---|
Main Authors | , , , |
Format | Journal Article |
Language | English |
Published |
New York
IEEE
01.01.2022
The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Reinforcement learning (RL) methods have demonstrated their efficiency in simulation. However, many of the applications for which RL offers great potential, such as autonomous driving, are also safety critical and require a certified closed-loop behavior in order to meet the safety specifications in the presence of physical constraints. This article introduces a concept called probabilistic model predictive safety certification (PMPSC), which can be combined with any RL algorithm and provides provable safety certificates in terms of state and input chance constraints for potentially large-scale systems. The certificate is realized through a stochastic tube that safely connects the current system state with a terminal set of states that is known to be safe. A novel formulation allows a recursively feasible real-time computation of such probabilistic tubes, despite the presence of possibly unbounded disturbances. A design procedure for PMPSC relying on Bayesian inference and recent advances in probabilistic set invariance is presented. Using a numerical car simulation, the method and its design procedure are illustrated by enhancing an RL algorithm with safety certificates. |
---|---|
AbstractList | Reinforcement learning (RL) methods have demonstrated their efficiency in simulation. However, many of the applications for which RL offers great potential, such as autonomous driving, are also safety critical and require a certified closed-loop behavior in order to meet the safety specifications in the presence of physical constraints. This article introduces a concept called probabilistic model predictive safety certification (PMPSC), which can be combined with any RL algorithm and provides provable safety certificates in terms of state and input chance constraints for potentially large-scale systems. The certificate is realized through a stochastic tube that safely connects the current system state with a terminal set of states that is known to be safe. A novel formulation allows a recursively feasible real-time computation of such probabilistic tubes, despite the presence of possibly unbounded disturbances. A design procedure for PMPSC relying on Bayesian inference and recent advances in probabilistic set invariance is presented. Using a numerical car simulation, the method and its design procedure are illustrated by enhancing an RL algorithm with safety certificates. |
Author | Wabersich, Henk J. Hewing, Lukas Zeilinger, Melanie N. Carron, Andrea |
Author_xml | – sequence: 1 givenname: Henk J. orcidid: 0000-0002-8554-1484 surname: Wabersich fullname: Wabersich, Henk J. email: wabersich@kimpeter.de organization: Institute for Dynamic Systems and Control, ETH Zürich, Zürich, Switzerland – sequence: 2 givenname: Lukas orcidid: 0000-0002-2770-4687 surname: Hewing fullname: Hewing, Lukas email: lhewing@ethz.ch organization: Institute for Dynamic Systems and Control, ETH Zürich, Zürich, Switzerland – sequence: 3 givenname: Andrea orcidid: 0000-0002-3219-4132 surname: Carron fullname: Carron, Andrea email: carrona@ethz.ch organization: Institute for Dynamic Systems and Control, ETH Zürich, Zürich, Switzerland – sequence: 4 givenname: Melanie N. orcidid: 0000-0003-4570-7571 surname: Zeilinger fullname: Zeilinger, Melanie N. email: mzeilinger@ethz.ch organization: Institute for Dynamic Systems and Control, ETH Zürich, Zürich, Switzerland |
BookMark | eNo9kEtLAzEUhYNUsK3uBTcDrqfmPZllHXxB1YJ1HTLJHUmpk5qkQv-9U1pcXQ5851z4JmjUhx4QuiZ4Rgiu71bzZkYxJTOGec2YOENjIoQqqaBshMYYE1XWVMkLNElpPUTJORmjt2UMrWn9xqfsbfEaHGyKZQTnbfa_UHyYDvK-aCBm33lrsg990YVYLMDE3vdf5b1J4Iom9DmGzSU678wmwdXpTtHn48OqeS4X708vzXxRWiaqXApWc1w7iwWx0FknKmcdw7hqGVZKWkKr2rUVM2CJVB1rW2VJWzspJEghOJui2-PuNoafHaSs12EX--GlppIIUnEu1EDhI2VjSClCp7fRf5u41wTrgzU9WNMHa_pkbajcHCseAP7xmhFOhWR_31FpzQ |
CODEN | IETAA9 |
CitedBy_id | crossref_primary_10_1109_LRA_2021_3097073 crossref_primary_10_3390_app12041995 crossref_primary_10_1016_j_advengsoft_2022_103245 crossref_primary_10_1016_j_amc_2024_128860 crossref_primary_10_1038_s41586_021_04301_9 crossref_primary_10_1002_rnc_5955 crossref_primary_10_1016_j_automatica_2022_110399 crossref_primary_10_1016_j_engappai_2023_105878 crossref_primary_10_1109_TCST_2023_3243993 crossref_primary_10_1109_LCSYS_2021_3089641 crossref_primary_10_1016_j_jfranklin_2022_12_024 crossref_primary_10_1109_TIV_2023_3336964 crossref_primary_10_1109_ACCESS_2022_3228922 crossref_primary_10_1109_TRO_2022_3197072 crossref_primary_10_1109_TAC_2022_3175628 crossref_primary_10_1109_ACCESS_2022_3188857 crossref_primary_10_1109_TCST_2022_3142629 crossref_primary_10_1016_j_arcontrol_2024_100945 crossref_primary_10_1109_TAC_2023_3241282 crossref_primary_10_1109_LRA_2023_3322644 crossref_primary_10_1109_TAC_2023_3294868 crossref_primary_10_1016_j_compchemeng_2024_108601 crossref_primary_10_1109_TRO_2024_3400838 crossref_primary_10_1109_TAC_2022_3227907 crossref_primary_10_1002_rnc_6560 crossref_primary_10_1109_TAC_2023_3291952 crossref_primary_10_1177_09596518231153445 |
Cites_doi | 10.3182/20070822-3-za-2920.00076 10.1038/nature14236 10.1142/9789813143654_0010 10.1016/s0005-1098(99)00113-2 10.1109/CDC.2012.6426855 10.1109/TCST.2008.2012116 10.1016/j.automatica.2013.02.003 10.1137/1.9781611970777 10.1109/CDC.2018.8619829 10.1109/IROS40897.2019.8967736 10.23919/ECC.2013.6669436 10.1016/j.ifacol.2017.08.324 10.3182/20080706-5-KR-1001.02593 10.1109/CDC.2014.7039601 10.1016/j.automatica.2020.109095 10.1007/s12555-011-0300-6 10.23919/ECC.2018.8550288 10.1007/978-3-0348-8407-5_21 10.1109/IROS.2011.6095101 10.1016/j.jprocont.2016.03.005 10.1016/j.automatica.2008.03.027 10.1109/ICRA.2012.6225035 10.1109/TCST.2009.2017934 10.1007/b94608 10.1109/TCST.2018.2886159 10.1109/ICRA40945.2020.9196867 10.1109/IROS.2015.7354161 10.1109/CDC.2018.8619554 10.1080/00207179.2017.1323351 10.1109/TNNLS.2017.2654539 10.1109/TCST.2005.847331 10.1002/oca.939 10.1109/CDC.2017.8263812 10.15607/RSS.2007.III.041 10.1109/CDC.2018.8619139 10.1109/CDC.2013.6761117 10.1109/CDC.2018.8619572 10.1109/ECC.2015.7330913 10.1109/CDC.2016.7798979 10.23919/ECC.2018.8550160 10.1007/978-3-319-23461-8_9 10.1177/0278364916645661 10.1109/TAC.2018.2876389 |
ContentType | Journal Article |
Copyright | Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022 |
Copyright_xml | – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022 |
DBID | 97E RIA RIE AAYXX CITATION 7SC 7SP 7TB 8FD FR3 JQ2 L7M L~C L~D |
DOI | 10.1109/TAC.2021.3049335 |
DatabaseName | IEEE All-Society Periodicals Package (ASPP) 2005-present IEEE All-Society Periodicals Package (ASPP) 1998-Present IEL CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Mechanical & Transportation Engineering Abstracts Technology Research Database Engineering Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional |
DatabaseTitle | CrossRef Technology Research Database Computer and Information Systems Abstracts – Academic Mechanical & Transportation Engineering Abstracts Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Engineering Research Database Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional |
DatabaseTitleList | Technology Research Database |
Database_xml | – sequence: 1 dbid: RIE name: IEL url: https://proxy.k.utb.cz/login?url=https://ieeexplore.ieee.org/ sourceTypes: Publisher |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Engineering |
EISSN | 1558-2523 |
EndPage | 188 |
ExternalDocumentID | 10_1109_TAC_2021_3049335 9314256 |
Genre | orig-research |
GrantInformation_xml | – fundername: Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung; Swiss National Science Foundation grantid: PP00P2 157 601/1 funderid: 10.13039/501100001711 – fundername: Swiss National Centre of Competence in Research NCCR Digital Fabrication grantid: #51NF40_141853 |
GroupedDBID | -~X .DC 0R~ 29I 3EH 4.4 5GY 5VS 6IK 97E AAJGR AASAJ AAYOK ABQJQ ABVLG ACGFO ACGFS ACIWK ACNCT AENEX AETIX AI. AIBXA AKJIK ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD F5P HZ~ H~9 IAAWW IBMZZ ICLAB IDIHD IFIPE IFJZH IPLJI JAVBF LAI M43 MS~ O9- OCL P2P RIA RIE RIG RNS TAE TN5 VH1 VJK XFK ~02 AAYXX CITATION 7SC 7SP 7TB 8FD FR3 JQ2 L7M L~C L~D |
ID | FETCH-LOGICAL-c357t-539409dc051cefcd57dcd3007b30886c1279db73aec168f3bb8c1b9d656e65543 |
IEDL.DBID | RIE |
ISSN | 0018-9286 |
IngestDate | Fri Sep 13 04:38:56 EDT 2024 Fri Aug 23 02:38:08 EDT 2024 Wed Jun 26 19:25:43 EDT 2024 |
IsPeerReviewed | true |
IsScholarly | true |
Issue | 1 |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-c357t-539409dc051cefcd57dcd3007b30886c1279db73aec168f3bb8c1b9d656e65543 |
ORCID | 0000-0002-3219-4132 0000-0003-4570-7571 0000-0002-2770-4687 0000-0002-8554-1484 |
PQID | 2615174458 |
PQPubID | 85475 |
PageCount | 13 |
ParticipantIDs | proquest_journals_2615174458 crossref_primary_10_1109_TAC_2021_3049335 ieee_primary_9314256 |
PublicationCentury | 2000 |
PublicationDate | 2022-Jan. 2022-1-00 20220101 |
PublicationDateYYYYMMDD | 2022-01-01 |
PublicationDate_xml | – month: 01 year: 2022 text: 2022-Jan. |
PublicationDecade | 2020 |
PublicationPlace | New York |
PublicationPlace_xml | – name: New York |
PublicationTitle | IEEE transactions on automatic control |
PublicationTitleAbbrev | TAC |
PublicationYear | 2022 |
Publisher | IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
Publisher_xml | – name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
References | ref13 ref57 ref12 ref56 ref15 ref59 ref14 ref58 ref11 Ljung (ref34) 1998 ref55 ref10 ref54 ref19 Rawlings (ref37) 2009 Lfberg (ref64) 2004 Berkenkamp (ref23) 2017 Chowdhury (ref35) 2017 ref45 ref48 ref42 ref41 Schulman (ref18) 2015 ref44 Rasmussen (ref52) 2006 ref43 Garca (ref16) 2015; 16 ref49 Skogestad (ref4) 2007; 2 ref8 ref9 ref3 Larsen (ref32) 2017; 50 (ref65) 2019 ref6 ref5 Deisenroth (ref29) 2011 ref40 Amodei (ref7) 2016 Silver (ref47) 395; 32 Bishop (ref53) 2006 ref31 Borrelli (ref50) 2003; 290 Agrawal (ref51) 2019 ref30 ref33 ref1 ref38 Schulman (ref46) 2015; 3 ref24 ref26 ref25 ref20 Merel (ref2) 2017 ref63 ref22 ref21 Lederer (ref36) 2019; 32 ref28 ref27 Pola (ref39) 2006 Achiam (ref17) 2017 ref60 ref62 ref61 |
References_xml | – start-page: 163 year: 1998 ident: ref34 article-title: System identification, in Signal Analysis Prediction contributor: fullname: Ljung – volume: 16 start-page: 1437 year: 2015 ident: ref16 article-title: A comprehensive survey on safe reinforcement learning publication-title: J. Mach. Learn. Res. contributor: fullname: Garca – start-page: 1889 volume-title: Int. Conf. Mach. Learn. year: 2015 ident: ref18 article-title: Trust region policy optimization contributor: fullname: Schulman – volume-title: Proc. Int. Symp. Math. Theory Netw. Syst. year: 2006 ident: ref39 article-title: Invariance in stochastic dynamical control systems contributor: fullname: Pola – start-page: 9562 volume-title: Proc. Adv. Neural Inf. Process. Syst. year: 2019 ident: ref51 article-title: Differentiable convex optimization layers contributor: fullname: Agrawal – ident: ref33 doi: 10.3182/20070822-3-za-2920.00076 – ident: ref1 doi: 10.1038/nature14236 – ident: ref55 doi: 10.1142/9789813143654_0010 – ident: ref59 doi: 10.1016/s0005-1098(99)00113-2 – ident: ref9 doi: 10.1109/CDC.2012.6426855 – ident: ref62 doi: 10.1109/TCST.2008.2012116 – year: 2009 ident: ref37 article-title: Model Predictive Control: Theory and Design contributor: fullname: Rawlings – ident: ref24 doi: 10.1016/j.automatica.2013.02.003 – year: 2006 ident: ref53 article-title: Pattern Recognition and Machine Learning (Information Science and Statistics) contributor: fullname: Bishop – ident: ref56 doi: 10.1137/1.9781611970777 – ident: ref12 doi: 10.1109/CDC.2018.8619829 – start-page: 908 year: 2017 ident: ref23 article-title: Safe model-based reinforcement learning with stability guarantees contributor: fullname: Berkenkamp – ident: ref49 doi: 10.1109/IROS40897.2019.8967736 – ident: ref60 doi: 10.23919/ECC.2013.6669436 – ident: ref61 doi: 10.1016/j.ifacol.2017.08.324 – start-page: 22 volume-title: Proc. Int. Conf. Mach. Learn. year: 2017 ident: ref17 article-title: Constrained policy optimization contributor: fullname: Achiam – year: 2016, ident: ref7 article-title: Concrete problems in AI safety contributor: fullname: Amodei – year: 2019 ident: ref65 article-title: MOSEK Optimization Toolbox for MATLAB Manual. Version 9.0 – volume: 32 start-page: 659 volume-title: Proc. Adv. Neural Inf. Process. Syst. year: 2019 ident: ref36 article-title: Uniform error bounds for Gaussian process regression with application to safe control contributor: fullname: Lederer – start-page: 465 volume-title: Proc. 28th Int. Conf. Mach. Learn. year: 2011 ident: ref29 article-title: Pilco: A. model-based and data-efficient approach to policy search contributor: fullname: Deisenroth – ident: ref58 doi: 10.3182/20080706-5-KR-1001.02593 – ident: ref45 doi: 10.1109/CDC.2014.7039601 – ident: ref44 doi: 10.1016/j.automatica.2020.109095 – year: 2017 ident: ref2 article-title: Learning human behaviors from motion capture by adversarial imitation contributor: fullname: Merel – ident: ref6 doi: 10.1007/s12555-011-0300-6 – volume: 50 start-page: 2536 issue: 1 volume-title: 20th IFAC World Congr. year: 2017 ident: ref32 article-title: Safe learning for distributed systems with bounded uncertainties contributor: fullname: Larsen – year: 2006 ident: ref52 article-title: Gaussian Processes for Machine Learning (Adaptive Computation and Machine Learning) contributor: fullname: Rasmussen – ident: ref31 doi: 10.23919/ECC.2018.8550288 – ident: ref5 doi: 10.1007/978-3-0348-8407-5_21 – ident: ref30 doi: 10.1109/IROS.2011.6095101 – ident: ref42 doi: 10.1016/j.jprocont.2016.03.005 – ident: ref40 doi: 10.1016/j.automatica.2008.03.027 – ident: ref25 doi: 10.1109/ICRA.2012.6225035 – volume: 32 start-page: 387 volume-title: Proc. 31st Int. Conf. Int. Conf. Mach. Learn year: 395 ident: ref47 article-title: Deterministic policy gradient algorithms contributor: fullname: Silver – ident: ref8 doi: 10.1109/TCST.2009.2017934 – ident: ref54 doi: 10.1007/b94608 – ident: ref63 doi: 10.1109/TCST.2018.2886159 – ident: ref15 doi: 10.1109/ICRA40945.2020.9196867 – ident: ref19 doi: 10.1109/IROS.2015.7354161 – ident: ref38 doi: 10.1109/CDC.2018.8619554 – ident: ref43 doi: 10.1080/00207179.2017.1323351 – ident: ref14 doi: 10.1109/TNNLS.2017.2654539 – volume: 290 year: 2003 ident: ref50 article-title: Constrained Optimal Control of Linear and Hybrid Systems contributor: fullname: Borrelli – volume: 3 start-page: 1889 year: 2015 ident: ref46 article-title: Trust region policy optimization contributor: fullname: Schulman – ident: ref3 doi: 10.1109/TCST.2005.847331 – ident: ref10 doi: 10.1002/oca.939 – ident: ref28 doi: 10.1109/CDC.2017.8263812 – ident: ref48 doi: 10.15607/RSS.2007.III.041 – ident: ref13 doi: 10.1109/CDC.2018.8619139 – ident: ref41 doi: 10.1109/CDC.2013.6761117 – ident: ref27 doi: 10.1109/CDC.2018.8619572 – start-page: 844 volume-title: Proc. Int. Conf. Mach. Learn. year: 2017 ident: ref35 article-title: On kernelized multi-armed bandits contributor: fullname: Chowdhury – volume: 2 year: 2007 ident: ref4 article-title: Multivariable Feedback Control: Analysis and Design contributor: fullname: Skogestad – ident: ref20 doi: 10.1109/ECC.2015.7330913 – ident: ref22 doi: 10.1109/CDC.2016.7798979 – ident: ref57 doi: 10.23919/ECC.2018.8550160 – ident: ref21 doi: 10.1007/978-3-319-23461-8_9 – ident: ref26 doi: 10.1177/0278364916645661 – ident: ref11 doi: 10.1109/TAC.2018.2876389 – start-page: 284 volume-title: Proc. CACSD Conf. year: 2004 ident: ref64 article-title: Yalmip: A toolbox for modeling and optimization in MATLAB contributor: fullname: Lfberg |
SSID | ssj0016441 |
Score | 2.5806212 |
Snippet | Reinforcement learning (RL) methods have demonstrated their efficiency in simulation. However, many of the applications for which RL offers great potential,... |
SourceID | proquest crossref ieee |
SourceType | Aggregation Database Publisher |
StartPage | 176 |
SubjectTerms | Algorithms Bayesian analysis Certificates Certification Learning Optimization Prediction algorithms Predictive control Predictive models Probabilistic logic Probabilistic models Probability theory reinforcement learning (RL) Safety Safety critical Statistical inference stochastic systems Tubes Uncertainty |
Title | Probabilistic Model Predictive Safety Certification for Learning-Based Control |
URI | https://ieeexplore.ieee.org/document/9314256 https://www.proquest.com/docview/2615174458/abstract/ |
Volume | 67 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV1LS8NAEB5qT3rwVcVqlT14EUyaZLOb5FiDpQgtBVvoLWQf8aC0UtOD_npnN2kU9eAthywsMzsz37wBrpUQAQ0ZdbRgzAkFOqyCFdxB31l5Skmp7bim8YSP5uHDgi1acNv0wmitbfGZds2nzeWrldyYUFk_oT4-Mb4DO7EXVL1aTcbA2PVK66IAB3GTkvSS_myQoiMY-K5JKVG72O3LBNmdKr8UsbUuwwMYb-9VFZU8u5tSuPLjx8jG_178EPZrmEkG1bs4gpZeHsPet-GDHZhM1yjLpjbWjGomZinaC5muTeLGqEDymBe6fCepKbwu6sgeQYhL6pGsT84dWkBF0qrY_QTmw_tZOnLq7QqOpCwqHWZ2oidKolRKXUjFIiUVRcggKGoeLv0gSpSIaK6lz-OCChFLXyQKAaDmCELoKbSXq6U-A6IREyFSoD5VCbIhzwNOc4HHEexIwZMu3GwJnr1WQzQy63x4SYbMyQxzspo5XegY-jX_1aTrQm_LoayWsrcsMHAsCkMWn_996gJ2A9OuYEMmPWiX642-RBBRiiv7ej4BQqbDCA |
link.rule.ids | 315,786,790,802,27957,27958,55109 |
linkProvider | IEEE |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV1LT8JAEJ4gHtSDLzSiqHvwYmKRdrste8RGggqEREi4Nd1HPWjAYDnor3d2W6pRD9566CabmZ2Zb94AF0oIj_qMOlow5vgCHVbB0sBB31m1lJJS23FNg2HQm_j3UzatwFXZC6O1tsVnumk-bS5fzeXShMquOXXxiQVrsI52vsXzbq0yZ2Ase653UYS9dpmUbPHrcSdCV9BzmyapRO1qty8jZLeq_FLF1r50d2CwulleVvLcXGaiKT9-DG3879V3YbsAmqSTv4w9qOjZPmx9Gz9Yg-FogdJsqmPNsGZi1qK9kNHCpG6MEiSPSaqzdxKZ0uu0iO0RBLmkGMr65NygDVQkysvdD2DSvR1HPafYr-BIysLMYWYrOlcS5VLqVCoWKqkoggZBUfcE0vVCrkRIEy3doJ1SIdrSFVwhBNQBwhB6CNXZfKaPgGhERYgVqEsVRzYkiRfQROBxhDtSBLwOlyuCx6_5GI3Yuh8tHiNzYsOcuGBOHWqGfuV_Benq0FhxKC7k7C32DCALfZ-1j_8-dQ4bvfGgH_fvhg8nsOmZ5gUbQGlANVss9SlCikyc2Zf0Cbztxl4 |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Probabilistic+Model+Predictive+Safety+Certification+for+Learning-Based+Control&rft.jtitle=IEEE+transactions+on+automatic+control&rft.au=Wabersich%2C+Henk+J.&rft.au=Hewing%2C+Lukas&rft.au=Carron%2C+Andrea&rft.au=Zeilinger%2C+Melanie+N.&rft.date=2022-01-01&rft.pub=IEEE&rft.issn=0018-9286&rft.eissn=1558-2523&rft.volume=67&rft.issue=1&rft.spage=176&rft.epage=188&rft_id=info:doi/10.1109%2FTAC.2021.3049335&rft.externalDocID=9314256 |
thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0018-9286&client=summon |
thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0018-9286&client=summon |
thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0018-9286&client=summon |