Bicycle route choice model developed using revealed preference GPS data

摘要

To better understand bicyclists' preferences, bicycle-mounted GPS units were used toobserve the behavior of 164 cyclists in Portland, Oregon, U.S.A. for several days each.Trip purpose and several other trip-level variables were recorded by the cyclists, and theresulting trips were coded to a highly detailed bicycle network. The authors used the1,449 valid non-exercise trips to estimate a bicycle route choice model. As part of thisresearch, a choice set generation algorithm was developed based on multiplepermutations of labeled path attributes, which seemed to out-perform comparableimplementations of other route choice set generation algorithms. The choice model wasformulated as a Path-Size Logit model to account for overlapping route alternatives.Estimation results are intuitive and suggest that cyclists are sensitive to the effects ofdistance, turn frequency, slope, intersection control, and traffic volumes. In addition,cyclists appear to place relatively high value on off-street bike paths, enhancedneighborhood bikeways (bicycle boulevards), and bridge facilities. Finally, estimationresults support segmentation by commute versus non-commute trip types. The routechoice model presented in this paper is currently being implemented as part of thePortland region's travel forecasting system.