000097378 001__ 97378 000097378 005__ 20210902121930.0 000097378 0247_ $$2doi$$a10.3390/app10238386 000097378 0248_ $$2sideral$$a121764 000097378 037__ $$aART-2020-121764 000097378 041__ $$aeng 000097378 100__ $$0(orcid)0000-0002-1361-9529$$aPlaced, J.A.$$uUniversidad de Zaragoza 000097378 245__ $$aA deep reinforcement learning approach for active SLAM 000097378 260__ $$c2020 000097378 5060_ $$aAccess copy available to the general public$$fUnrestricted 000097378 5203_ $$aIn this paper, we formulate the active SLAM paradigm in terms of model-free Deep Reinforcement Learning, embedding the traditional utility functions based on the Theory of Optimal Experimental Design in rewards, and therefore relaxing the intensive computations of classical approaches. We validate such formulation in a complex simulation environment, using a state-of-the-art deep Q-learning architecture with laser measurements as network inputs. Trained agents become capable not only to learn a policy to navigate and explore in the absence of an environment model but also to transfer their knowledge to previously unseen maps, which is a key requirement in robotic exploration. 000097378 536__ $$9info:eu-repo/grantAgreement/ES/DGA/T04-FSE$$9info:eu-repo/grantAgreement/ES/MINECO/DPI2015-68905-P 000097378 540__ $$9info:eu-repo/semantics/openAccess$$aby$$uhttp://creativecommons.org/licenses/by/3.0/es/ 000097378 590__ $$a2.679$$b2020 000097378 591__ $$aPHYSICS, APPLIED$$b73 / 160 = 0.456$$c2020$$dQ2$$eT2 000097378 591__ $$aENGINEERING, MULTIDISCIPLINARY$$b38 / 91 = 0.418$$c2020$$dQ2$$eT2 000097378 591__ $$aCHEMISTRY, MULTIDISCIPLINARY$$b101 / 178 = 0.567$$c2020$$dQ3$$eT2 000097378 591__ $$aMATERIALS SCIENCE, MULTIDISCIPLINARY$$b201 / 333 = 0.604$$c2020$$dQ3$$eT2 000097378 592__ $$a0.435$$b2020 000097378 593__ $$aComputer Science Applications$$c2020$$dQ2 000097378 593__ $$aEngineering (miscellaneous)$$c2020$$dQ2 000097378 593__ $$aProcess Chemistry and Technology$$c2020$$dQ2 000097378 593__ $$aInstrumentation$$c2020$$dQ2 000097378 593__ $$aMaterials Science (miscellaneous)$$c2020$$dQ2 000097378 593__ $$aFluid Flow and Transfer Processes$$c2020$$dQ2 000097378 655_4 $$ainfo:eu-repo/semantics/article$$vinfo:eu-repo/semantics/publishedVersion 000097378 700__ $$0(orcid)0000-0001-5977-8720$$aCastellanos, J.A.$$uUniversidad de Zaragoza 000097378 7102_ $$15007$$2520$$aUniversidad de Zaragoza$$bDpto. Informát.Ingenie.Sistms.$$cÁrea Ingen.Sistemas y Automát. 000097378 773__ $$g10, 23 (2020), 8386 [1-21]$$pAppl. sci.$$tAPPLIED SCIENCES-BASEL$$x2076-3417 000097378 8564_ $$s507152$$uhttps://zaguan.unizar.es/record/97378/files/texto_completo.pdf$$yVersión publicada 000097378 8564_ $$s426328$$uhttps://zaguan.unizar.es/record/97378/files/texto_completo.jpg?subformat=icon$$xicon$$yVersión publicada 000097378 909CO $$ooai:zaguan.unizar.es:97378$$particulos$$pdriver 000097378 951__ $$a2021-09-02-10:54:27 000097378 980__ $$aARTICLE