Extracting the Inconsistencies in the Urban Highways Considering the Crash Occurrences

The present study aims to identify the inconsistencies resulting from the interferences between the geometric features of the highway, access points and traffic control devices by extracting the frequent association rules and discovering the patterns that contribute to the recurrent crashes. A case study was conducted on a 117-kilometer urban highway around Mashhad. The data for five years of crashes were procured from the corresponding authorities and placed on the network under a geo spatial software environment. Then, the all features of the highway were collected. The trip times between the nodes and existing features were calculated via coding based on the data obtained from one of the route planning software. The operating speed was subsequently estimated. According to the obtained databank and considering the geometric features of the road (horizontal and vertical), the access points, road width, operating speed and position of the traffic control devices (signs and speed control camera), efforts were made to divide the road to the sections. Allocating the crash data to the sections and using the FP-Growth Algorithm, the frequent rules affecting the recurrent crash occurrences were extracted. The results showed that there are different combinations of the geometric road-related factors such as flat-straight alignment and horizontal curve along with cases like right-hand exit, consecutive entrance and exit, and reduced road width in addition to the presence of speed control camera and advanced direction signs on the highway sections together cause inconsistencies like the recurrent crash occurrences.

Extracting the Inconsistencies in the Urban Highways Considering the Crash Occurrences Keywords:

Association rules , crash analysis , Data-mining , Geo spatial system , inconsistency