A new weighted BPR model for travel time estimation sensitive to “on the road elements”

Document Type : Original Article



2 1 Geomatics Engineering Faculty, K.N. Toosi University of Technology, Valiasr St., Tehran, Iran

3 Department of Civil Engineering, K. N. Toosi University of Technology, Tehran 1996715433, Iran;

4 Remote Sensing Department, Kerman Graduate University of Technology, Mahan, Iran



Numerous variables, including vehicle volume, road width, on-street parking, and border parking, can all affect the amount of time a vehicle travels on a given route. While these factors have varying effects, they are typically considered invariant across all segments of a route. None of the previous studies considered an area where these factors could exert influence. In this paper, a new model named "weighted BPR" is proposed that considers an effective area in for each factor where the vehicle’s speed changes. These areas are incorporated in the model using some weights. True travel times across many streets are then used to estimate the model weights. Provided that we know which factors are present in a route, we can, then, estimate how long it would take for a vehicle to travel through the route. In addition to estimating the travel time, such model can be used to guide city planners how new structures affect the travel time in cities. Due to the large number of such factors, this research examines only three: bus stops, crosswalks, and speed bumps. To estimate the weight The prototype was evaluated on 37 distinct sections of Tehran's streets. The results indicated that the estimated travel time using the developed model is more precise than that computed using conventional methods. Experiments demonstrated that the technique developed in this paper can estimate the travel time with an accuracy of 2.93 to 3.27 minutes while this value computed by the model developed by national traffic experts ranges between 7.8 to 23.09 minutes.


Main Subjects