Learning to Play Hard Exploration Games Using Graph-Guided Self-Navigation

Zhao, Enmin; Yan, Renye; Li, Kai; Li, Lijuan

doi:10.1109/ijcnn52387.2021.9534251

Cited by 1 publication

(1 citation statement)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ABM allows each vehicle to experience the traffic while its data are logged. As a result, the ABM simulation enables action planning and the action outcome estimation before the actual action has been performed [18]. Moreover, ABM can mimic the actual object without relying on statistical distributions, which facilitates the properties or attributes attachment of the modeled objects [19].…”

Section: Literature Review and Problem Statementmentioning

confidence: 99%

Identifying the influence of land logistic driver cognitive energy impact on supply chain performance through agent-based simulation

Tama,

Hardiningtyas,

Sudiarno

et al. 2024

EEJET

View full text Add to dashboard Cite

Logistic transports link demand generators, distributors, and producers in a supply chain network (SCN). The existence of logistic transports is critical to ensure whole nodes’ economic sustainability. This study explores the impact of human factors on SCN performance through cognitive energy expenditure (CEE) tracking. Agent-based model (ABM) simulation was used to analyze the impact of CEE from truck driver’s electroencephalography (EEG) data to obtain the postsynaptic potential values, which were then transformed to calorific energy. The fleet agents, retailers and distributor models were built based on the East Java, Indonesia, logistic transport route around Karanglo, Gempol, Bungurasih, and Gubeng. The frequency and the peak value of the EEG data, postsynaptic potential, and energy data indicate the same information. All data indicate that more challenging routes have higher frequency and higher peak values. The ABM simulation of the fleet agents shows balanced CEE throughout entire routes due to the precise rest period and eat scheduling. The average delivery success rate was 8 out of 30 or 26.7 % in each simulation time step. Hence, most goods delivery tasks can be completed by fleet agents in a balanced system. As a consequence, the SCN performance is also balanced due to the fluid inventory shift without overstock and stockouts. The rest and eat periods of a fleet agent were scheduled after the CEE has been peaked. The time lag between rest periods and transport operations has to be maintained to overcome fleet agents task buildup. Task buildup has a potential to decay both transport safety and inventory shift rates. Therefore, the upgrade in SCN performance is possible through proper fleet agents scheduling

show abstract

Section: Literature Review and Problem Statementmentioning

confidence: 99%