Study on estimation methods for travel data using GPS and context data

2014 Fiscal Year Final Research Report

• Project/Area Number: 25630215
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Civil engineering project/Traffic engineering
• Research Institution: Nagoya University
• Principal Investigator: MORIKAWA Takayuki 名古屋大学, 環境学研究科, 教授 (30166392)
• Co-Investigator(Kenkyū-buntansha): YAMAMOTO Toshiyuki 名古屋大学, エコトピア科学研究所, 教授 (80273465) ; MIWA Tomio 名古屋大学, エコトピア科学研究所, 准教授 (60422763) ; USUI Tomotaka 名古屋大学, 経済学研究科, 特任准教授 (20549448)
• Co-Investigator(Renkei-kenkyūsha): SATO Hitomi 名古屋大学, 未来社会創造機構, 特任講師 (00509193)
• Project Period (FY): 2013-04-01 – 2015-03-31
• Keywords: スマートフォン / 位置情報 / 活動推定 / 交通調査

Outline of Final Research Achievements

Traditional Personal Trip (PT) survey has problems about low accuracy and disability of continuity because of significant high cost and low respondent rate. In order to avoid these disadvantages, GPS data collected by smart phone were utilized in this study to infer trip purpose data. Based on the method of Decision Tree and Random Forest, model for inferring trip purpose was developed and some analyses proceeded subsequently. It is found that the accuracy of trip purpose inference calculated by the model developed in this study is more than 70%. Furthermore, it can be concluded that 1 month continuous GPS data is minimum requirement to get the satisfied results. Also, the results show that the accuracy of trip purpose reference can be improved by increasing the number of respondents.

Free Research Field

交通計画