点击次数：

论文类型：期刊论文

发表时间：2018-10-01

发表刊物：International Journal of Innovative Computing, Information and Control

卷号：14

期号：5

页面范围：1755-1765

ISSN号：13494198

摘要：Stockouts are serious accidents at gas stations in China. However, it is hard to capture the abnormal consumption timely to avoid the occurrence of a stockout due to the stochastic arrivals of demands. In this paper, a real-time rolling approach to perceiving the time point of possibly abnormal consumption is developed by excluding the strictly secure periods. The proposed approach integrates dynamic contexts, current states and future contexts to support the real-time perceiving of possibly abnormal time points. A simulation system is implemented on the Arena simulation platform, through which we deeply analyze the simulation results to prove the effectiveness of the proposed real-time rolling approach. © 2018 ICIC International.