Increasing the accuracy of trip rate information from passive multi-day GPS travel datasets: Automatic trip end identification issues

摘要: With the availability of Global Positioning System (GPS) receivers to capture vehicle location, it is now feasible easily collect multiple days travel data automatically. However, GPS-collected are not ready for direct use in trip rate or route choice research until ends identified within large GPS streams. One common parameter used divide trips dwell time, time a stationary. Identifying particularly challenging when there chaining with brief stops, such as picking up and dropping off passengers. It hard distinguish these stops from those caused by traffic controls congestion. Although method effective many cases, foolproof recent indicates additional logic improves dividing. While some studies incorporating more than identify having been conducted, including actual evaluate success dividing methods have limited. In this research, 12 ten-day real-world datasets were develop, calibrate compare three start points stream. The true end each advance stream using supplemental log completed participants so that accuracy automated division could be measured compared. A heuristic model, which combines heading change, distance between road network, performs best, correctly identifying 94% ends.