Centre dedocumentation
scientifique

Multi-agent reinforcement learning to unify order-matching and vehicle-repositioning in ride-hailing services / Mingyue Xu in International journal of geographical information science IJGIS, vol 37 n° 2 (February 2023)  

Public [article]  inInternational journal of geographical information science IJGIS > vol 37 n° 2 (February 2023) . - pp 380 - 402 Titre : Multi-agent reinforcement learning to unify order-matching and vehicle-repositioning in ride-hailing services Type de document : Article/Communication Auteurs : Mingyue Xu, Auteur ; Peng Yue, Auteur ; Fan Yu, Auteur ; et al., Auteur Année de publication : 2023 Article en page(s) : pp 380 - 402 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Analyse spatiale [Termes IGN] appariement de données localisées [Termes IGN] apprentissage profond [Termes IGN] autopartage [Termes IGN] Chine [Termes IGN] distribution spatiale [Termes IGN] interaction humain-espace [Termes IGN] modèle de Markov [Termes IGN] système d'information urbain [Termes IGN] système multi-agents [Termes IGN] taxi [Termes IGN] transmission de données [Termes IGN] zone d'activité économique Résumé : (auteur) The popularity of ride-hailing platforms has significantly improved travel efficiency by providing convenient and personalized transportation services. Designing an effective ride-hailing service generally needs to address two tasks: order matching that assigns orders to available vehicles and proactive vehicle repositioning that deploys idle vehicles to potentially high-demand regions. Recent studies have intensively utilized deep reinforcement learning to solve the two tasks by learning an optimal dispatching strategy. However, most of them generate actions for the two tasks independently, neglecting the interactions between the two tasks and the communications among multiple drivers. To this end, this paper provides an approach based on multi-agent deep reinforcement learning where the two tasks are modeled as a unified Markov decision process, and the colossal state space and competition among drivers are addressed. Additionally, a modifiable agent-specific state representation is proposed to facilitate knowledge transferring and improve computing efficiency. We evaluate our approach on a public taxi order dataset collected in Chengdu, China, where a variable number of simulated vehicles are tested. Experimental results show that our approach outperforms seven existing baselines, reducing passenger rejection rate, driver idle time and improving total driver income. Numéro de notice : A2023-058 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1080/13658816.2022.2119477 Date de publication en ligne : 07/09/2022 En ligne : https://doi.org/10.1080/13658816.2022.2119477 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102396 [article]

Machine learning for the distributed and dynamic management of a fleet of taxis and autonomous shuttles / Tatiana Babicheva (2021)  

Public Titre : Machine learning for the distributed and dynamic management of a fleet of taxis and autonomous shuttles Titre original : Machine Learning pour la gestion distribuée et dynamique d’une flotte de taxis et navettes autonomes Type de document : Thèse/HDR Auteurs : Tatiana Babicheva, Auteur ; Leïla Kloul, Directeur de thèse ; Dominique Barth, Directeur de thèse Editeur : Bures-sur-Yvette : Université Paris-Saclay Année de publication : 2021 Importance : 190 p. Format : 21 x 30 cm Note générale : bibliographie Thèse de Doctorat de l'Université Paris-Saclay, Informatique Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique [Termes IGN] apprentissage par renforcement [Termes IGN] autopartage [Termes IGN] calcul d'itinéraire [Termes IGN] méthode heuristique [Termes IGN] navigation autonome [Termes IGN] OpenStreetMap [Termes IGN] optimisation (mathématiques) [Termes IGN] réseau neuronal artificiel [Termes IGN] réseau routier [Termes IGN] taxi [Termes IGN] trafic routier [Termes IGN] trafic urbain [Termes IGN] véhicule électrique [Termes IGN] ville intelligente Index. décimale : THESE Thèses et HDR Résumé : (auteur) In this thesis are investigated methods to manage shared electric autonomous taxi urban systems under online context in which customer demands occur over time, and where vehicles are available for ride-sharing and require electric recharging management. We propose the heuristics based on problem decomposition which include road network repartition and highlighting of subproblems such as charging management, empty vehicle redistribution and dynamic ride-sharing.The set of new methods for empty vehicle redistribution is proposed, such as proactive, meaning to take into account both current demand and anticipated future demand, in contrast to reactive methods, which act based on current demand only.We provide the reinforcement learning in different levels depending on granularity of the system.We propose station-based RL model for small networks and zone-based RL model, where the agents are zones of the city obtained by partitioning, for huge ones. The complete information optimisation is provided in order to analyse the system performance a-posteriori in offline context.The evaluation of the performance of proposed methods is provided in set of road networks of different nature and size. The proposed method provides promising results outperforming the other tested methods and the real data on the taxi system performance in terms of number of satisfied passengers under fixed fleet size. Note de contenu : 1- Introduction 2- State-of-the-art 3- Modelling the electrical aTaxisystem 4- Functional architecture of aTaxi system management 5- Reinforcement learning for aTaxi system optimisation 6- Evaluation scenarii 7- Numerical evaluation of aTaxi systems 8- Conclusion and discussion Numéro de notice : 28591 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/INFORMATIQUE Nature : Thèse française Note de thèse : thèse de Doctorat : Informatique : Paris-Saclay : 2021 Organisme de stage : Données et Algorithmes pour une ville intelligente et durable (UVSQ) DOI : sans En ligne : https://tel.archives-ouvertes.fr/tel-03230845/document Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=97968

Road network structure and ride-sharing accessibility: A network science perspective / Mingshu Wang in Computers, Environment and Urban Systems, vol 80 (March 2020)  

Public [article]  inComputers, Environment and Urban Systems > vol 80 (March 2020) Titre : Road network structure and ride-sharing accessibility: A network science perspective Type de document : Article/Communication Auteurs : Mingshu Wang, Auteur ; Zheyan Chen, Auteur ; Lan Mu, Auteur ; et al., Auteur Année de publication : 2020 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Analyse spatiale [Termes IGN] Atlanta (Géorgie) [Termes IGN] autopartage [Termes IGN] densité de population [Termes IGN] gestion urbaine [Termes IGN] migration pendulaire [Termes IGN] modèle conceptuel de données localisées [Termes IGN] OpenStreetMap [Termes IGN] réseau routier [Termes IGN] système d'information géographique Résumé : (auteur) The prosperity of ride-sharing services has rippled in the communities of GIScience, transportation, and urban planning. Meanwhile, road network structure has been analyzed from a network science perspective that focuses on nodes and relational links and aims to predictive models. However, limited empirical studies have explored the relationship between road network structure and ride-sharing accessibility through such perspective. This paper utilizes the spatial Durbin model to understand the relationship between road network structure and ride-sharing accessibility, proxied by Uber accessibility, through classical network measures of degree, closeness, and betweenness centrality. Taking the city of Atlanta as a case study, we have found in addition to population density and road network density, larger values of degree centrality and smaller values of closeness centrality of the road network are associated with better accessibility of Uber services. However, the effects of betweenness centrality are not significant. Furthermore, we have revealed heterogeneous effects of degree centrality and closeness centrality on the accessibility of Uber services, as the magnitudes of their effects vary by different time windows (i.e., weekday vs. weekend, rush hour in the morning vs. evening). Network science provides us both conceptual and methodological measures to understand the association between road network structure and ride-sharing accessibility. In this study, we constructed road network structure measures with OpenStreetMap, which is reproducible, replicable, and scalable because of its global coverage and public availability. The study resonates with the notion of cities as the set of interactions across networks, as we have observed time-sensitive heterogeneous effects of road network structure on ride-sharing accessibility. Numéro de notice : A2020-190 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article DOI : 10.1016/j.compenvurbsys.2019.101430 Date de publication en ligne : 12/11/2019 En ligne : https://doi.org/10.1016/j.compenvurbsys.2019.101430 Format de la ressource électronique : url article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=94852 [article]

Usability of an opportunistic interface concept for ad hoc ride-sharing / Michael Rigby in International journal of cartography, vol 2 n° 2 (December 2016)  

Public [article]  inInternational journal of cartography > vol 2 n° 2 (December 2016) . - pp 115 - 147 Titre : Usability of an opportunistic interface concept for ad hoc ride-sharing Type de document : Article/Communication Auteurs : Michael Rigby, Auteur ; Stephan Winter, Auteur Année de publication : 2016 Article en page(s) : pp 115 - 147 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique web [Termes IGN] autopartage [Termes IGN] convivialité [Termes IGN] ingénierie des connaissances [Termes IGN] interface web [Termes IGN] recherche du chemin optimal, algorithme de [Termes IGN] site web [Termes IGN] système de transport intelligent Résumé : (Auteur) Interacting with ride-sharing systems for ad hoc travel is a complex spatio-temporal task. The dynamics of service supply and demand challenge the rigidity of traditional human–computer interfaces, introducing service uncertainty and creating a knowledge gap which hinders a client's travel planning. Such interface constraints may mean that a client user is unable to find any ride matching their intentions. To overcome this, a novel visual interface concept, launch pads, has been suggested to replace the traditional interface within a two-step negotiation. To close the proposed approach's feedback loop, this paper investigates human understanding and use of the launch pad metaphor. Usability testing of launch pads is performed using a spatial cognitive engineering approach in directed wayfinding scenarios using various alternative representations. Results highlight that the variances of user interaction times depend on the representation used and reveal potential information overload issues. Using these findings, a minimum decision-making time is defined to tune the system's architecture. Numéro de notice : A2016--060 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1080/23729333.2016.1145040 En ligne : http://dx.doi.org/10.1080/23729333.2016.1145040 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=84215 [article]