Traffic signal systems can be more effective if high-quality performance measures are available. However, these performance measurement methodologies require more detection than exists in standard signal systems. One useful performance measure would be the traffic flow state at the intersection (e.g., static queue, discharging queue, free flow). This paper develops and tests a method for detection of the state of real-time traffic flow by interpreting patterns of signal status and traffic flow conditions measured by stop bar detectors. This paper proposes a widely available algorithm to create a model in the form of a decision tree that is calibrated to translate these states from traffic operation and signal control patterns. The tree was calibrated with microscopic simulation data and validated with independent simulation and manually extracted field data sets. Test results indicated that the traffic flow states detection methodology performed well and could be efficiently developed on large data sets.
[1]
Darcy M. Bullock,et al.
Data-Driven Algorithms for Real-Time Adaptive Tuning of Offsets in Coordinated Traffic Signal Systems
,
2007
.
[2]
Peter J. Bickel,et al.
Day-to-Day Travel-Time Trends and Travel-Time Prediction from Loop-Detector Data
,
2000
.
[3]
Darcy M. Bullock,et al.
Event-Based Data Collection for Generating Actuated Controller Performance Measures
,
2007
.
[4]
Brian Lee Smith,et al.
Data-Mining Tools for the Support of Signal-Timing Plan Development
,
2001
.
[5]
Robert L. Bertini,et al.
Using Signal System Data and Buses as Probe Vehicles to Define the Congested Regime on Arterials
,
2008
.
[6]
Peter T. Martin,et al.
Monitoring Commuter Congestion on Surface Streets in Real Time
,
2002
.
[7]
Mark E Hallenbeck,et al.
Arterial Performance Monitoring Using Stop Bar Sensor Data
,
2008
.