Identifying Trips And Activities Using GPS Based Travel Survey Data

Personal travel survey is the foundation of urban traffic planning andmanagement. The data of travel and activity is an important reference for trafficproblem diagnosis, public transportation planning and traffic managementpolicy-making. With the rapid development of technology and urbanization process, the demand for personal travel information and travel characteristics are increasinggradually in urban planning, traffic management and residents’ behavior research. Andthe requirement on travel data is higher and higher. The artificial method includingpaper diaries or telephone interviews method is the most important way ofinvestigation on residents travel. Which need the respondents recall travel details andsubjective cognition. The survey data rely much on questionnaire and investigators.Generally respondents often forget or slip some trips, which leads to poor accuracy totravel models. At the same time, high burden is also responsible for a low overallresponse rate. So it could not meet the needs of current’s management and researchboth in accuracy and quantity.While with characteristics of quickly locating, recording space location andaccessing to the mass trajectory of travel, the application of GPS devices is becomingmore and more widely in resident travel survey. However the application of GPSdevices can solve many problems in traditional travel survey in a certain extent, at thesame time it raises huge challenges of how to process data and how to fetch featureinformation from the trajectory data. To solve these problems, not only haverevolutionized travel survey technology, and provide data support for the research ofbehavioral science, etc. Therefore it is of great utility value for us.This paper presents the techniques for processing data and identifying trips andactivities based on GPS travel survey data. By improving the method of stationarypoints, it proposes a new method which can be used to determine status based oncontinuous data segments. Moreover, it designs a trip-activity identifying algorithmcomposing of dividing status segments, identifying activities, recognizing trips andintermediate stop. The results indicate that the algorithm shows a high level ofprediction accuracy comparing with the travel diaries reported in paper forms. Thefindings can be used to identify trips, activities, and short-time intermediate stops inparticular, which are often omitted in paper-form travel survey. Among this, therecognition of intermediate stop is innovation point in this paper. The trip-activity identifying algorithm of this paper is a quite effective methodfor GPS data. The results can be used to process data and identify trips and activitiesbased on GPS travel survey data. Also it can significantly enhance the efficiency andaccuracy of travel survey and facilitate the development of transportation system.