WITS2015paper2152智能交通世界大会ITS智慧城市社区人工智能AI物联网IT报告课件教案.pdfVIP

22nd ITS World Congress, Bordeaux, France, 5–9 October 2015 Paper number ITS-2152 A Comparative Study of Two Alternative In-Vehicle Dynamic Route Guidance with Information Uncertainty Chee Chung Tong1*, Pei-Ying Lin2 1. Associate Professor, Tamkang University, Taiwan, 151 Ying-Chung Road, Tamsui, New Taipei City, TEL : +886-2 E-mail : cctong@mail.tku.edu.tw 2. Graduate Research Assistance, Tamkang University, Taiwan Abstract Current route guidance system provides general information with estimated travel time of recommended route for trip routing decision. However, such a general system considers neither the uncertainty of the estimated information, nor the potential differences of perceptions among drivers’ to information uncertainty. With a control experiment approach on node-to-node routing dynamics, this study was set to explore two variants of supplying additional travel time uncertainty based on different travel behavior models, one was a “personal match” index derived from Indifference Band model; the other was simply an “event strength” index applied as the associated occurrence probability of event outcomes from Cumulative Prospect Theory (CPT). Results from this study suggested that providing “Personal Match” index based on the Indifference Band model may have the advantage of stability in predictability and promoting higher rate of compliance over providing “Outcome Probability” index based on the CPT model. Keywords: route guidance, information uncertainty. Methodology Two major theoretical models are addressed in this session. Indifference Band Model [Times New Roman 12, italic, sentence case, aligned left, Heading 2] Probability of route switching to recommended guidance (compliance to guidance) is defined by the notion of indifferent band which is treate