TITLE:
Diamond Interchange Performance Measures Using Connected Vehicle Data
AUTHORS:
Enrique Saldivar-Carranza, Steven Rogers, Howell Li, Darcy Michael Bullock
KEYWORDS:
Performance Measures, Safety, Connected Vehicle, Trajectories, Diamond Interchange
JOURNAL NAME:
Journal of Transportation Technologies,
Vol.12 No.3,
July
28,
2022
ABSTRACT: Diamond interchanges are frequently used where a freeway intersects a two-way surface street. Most of the techniques to evaluate the
performance of diamond interchanges rely on the Highway Capacity Manual (HCM),
simulation, Automated Traffic Signal Performance Measures (ATSPMs), and
historical crash data. HCM and simulation techniques require on-site data
collection to obtain models’ inputs. ATSPMs need high-resolution controller
event data acquired from roadway sensing equipment. Safety studies typically
need 3 to 5 years of crash data to provide statistically significant results.
This study utilizes commercially available connected vehicle (CV) data to
assess the performance and operation of a three- and four-phase diamond
interchange located in Indianapolis, Indiana, and Dallas, Texas, respectively.
Over 92,000 trajectories and 1,400,000 GPS points are analyzed from August 2020
weekdays CV data. Trajectories are
linear-referenced to generate Purdue Probe Diagrams (PPDs) from which
arrivals on green (AOG), split failures, downstream blockage, and movement-based
control delay are estimated. In addition, an extension of the PPD is presented
that characterizes the complete journey of a vehicle travelling through both
signals of the diamond interchange. This enhanced PPD is a significant
contribution as it provides an analytical framework and graphical summary of
the operational characteristics of how the
external movements traverse the entire system. The four-phase control
showed high internal progression (99% AOG) compared to the moderate internal
progression of the three-phase operation (64% AOG). This is consistent with the
design objectives of three- and four-phase control models, but historically
these quantitative AOG measures were not possible to obtain with just detector
data. Additionally, a graphical summary that illustrates the spatial
distribution of hard-braking and hard-acceleration events is also provided. The
presented techniques can be used by any agency to evaluate the performance of
their diamond interchanges without on-site data collection or capital
investments in sensing infrastructure.