Comparison of alternative real-time performance measures for measuring signal phase utilization and identifying oversaturation

Edward J Smaglik, Darcy M. Bullock, Douglas Gettman, Christopher M. Day, Hiromal Premachandra

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Various control algorithms process vehicle detector on and off times with the objective of quantifying split performance. Two such metrics are green occupancy ratio (GOR) and volume-to-capacity (v/c) ratio, both of which can be used to estimate the efficiency of a split as well as estimate oversaturated conditions. The GOR metric is attractive because of its simplicity of implementation; the v/c ratio is based on well-established traffic flow concepts and is independent of detector geometry although additional processing is required to develop the metric. Characterizing the impact of detection zone length and vehicle speed on the GOR as well as comparing its performance with another metric is important in understanding the robustness of GOR as a surrogate performance metric. This paper analyzes the difference between calculated GOR and v/c ratios over different detection zone lengths and differing vehicle speeds, compares these values with a calculated delay metric, and observes the effectiveness of GOR as an indicator of oversaturation. On the basis of the analysis, the paper documents the influence of approach speed and detection zone length on the calculated GOR metric. The paper concludes that the GOR values tend to reach a saturated value of 1.0 more rapidly than v/c, making it challenging for the analyst to identify oversaturated conditions by the metric alone. For practitioners using the GOR metric, it is critical that the system be calibrated properly because of the sensitivity of GOR to the detection zone length and vehicle speed. However, because direct measurement of v/c is difficult to obtain, GOR may still be a reasonable surrogate in real-time performance assessment of traffic signal systems.

Original languageEnglish (US)
Pages (from-to)123-131
Number of pages9
JournalTransportation Research Record
Issue number2259
DOIs
StatePublished - Dec 1 2011

Fingerprint

Detectors
Traffic signals
Signal systems
Geometry
Processing

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Mechanical Engineering

Cite this

Comparison of alternative real-time performance measures for measuring signal phase utilization and identifying oversaturation. / Smaglik, Edward J; Bullock, Darcy M.; Gettman, Douglas; Day, Christopher M.; Premachandra, Hiromal.

In: Transportation Research Record, No. 2259, 01.12.2011, p. 123-131.

Research output: Contribution to journalArticle

Smaglik, Edward J ; Bullock, Darcy M. ; Gettman, Douglas ; Day, Christopher M. ; Premachandra, Hiromal. / Comparison of alternative real-time performance measures for measuring signal phase utilization and identifying oversaturation. In: Transportation Research Record. 2011 ; No. 2259. pp. 123-131.
@article{ee23ed56e93248b6b944fe7cb5824fac,
title = "Comparison of alternative real-time performance measures for measuring signal phase utilization and identifying oversaturation",
abstract = "Various control algorithms process vehicle detector on and off times with the objective of quantifying split performance. Two such metrics are green occupancy ratio (GOR) and volume-to-capacity (v/c) ratio, both of which can be used to estimate the efficiency of a split as well as estimate oversaturated conditions. The GOR metric is attractive because of its simplicity of implementation; the v/c ratio is based on well-established traffic flow concepts and is independent of detector geometry although additional processing is required to develop the metric. Characterizing the impact of detection zone length and vehicle speed on the GOR as well as comparing its performance with another metric is important in understanding the robustness of GOR as a surrogate performance metric. This paper analyzes the difference between calculated GOR and v/c ratios over different detection zone lengths and differing vehicle speeds, compares these values with a calculated delay metric, and observes the effectiveness of GOR as an indicator of oversaturation. On the basis of the analysis, the paper documents the influence of approach speed and detection zone length on the calculated GOR metric. The paper concludes that the GOR values tend to reach a saturated value of 1.0 more rapidly than v/c, making it challenging for the analyst to identify oversaturated conditions by the metric alone. For practitioners using the GOR metric, it is critical that the system be calibrated properly because of the sensitivity of GOR to the detection zone length and vehicle speed. However, because direct measurement of v/c is difficult to obtain, GOR may still be a reasonable surrogate in real-time performance assessment of traffic signal systems.",
author = "Smaglik, {Edward J} and Bullock, {Darcy M.} and Douglas Gettman and Day, {Christopher M.} and Hiromal Premachandra",
year = "2011",
month = "12",
day = "1",
doi = "10.3141/2259-11",
language = "English (US)",
pages = "123--131",
journal = "Transportation Research Record",
issn = "0361-1981",
publisher = "US National Research Council",
number = "2259",

}

TY - JOUR

T1 - Comparison of alternative real-time performance measures for measuring signal phase utilization and identifying oversaturation

AU - Smaglik, Edward J

AU - Bullock, Darcy M.

AU - Gettman, Douglas

AU - Day, Christopher M.

AU - Premachandra, Hiromal

PY - 2011/12/1

Y1 - 2011/12/1

N2 - Various control algorithms process vehicle detector on and off times with the objective of quantifying split performance. Two such metrics are green occupancy ratio (GOR) and volume-to-capacity (v/c) ratio, both of which can be used to estimate the efficiency of a split as well as estimate oversaturated conditions. The GOR metric is attractive because of its simplicity of implementation; the v/c ratio is based on well-established traffic flow concepts and is independent of detector geometry although additional processing is required to develop the metric. Characterizing the impact of detection zone length and vehicle speed on the GOR as well as comparing its performance with another metric is important in understanding the robustness of GOR as a surrogate performance metric. This paper analyzes the difference between calculated GOR and v/c ratios over different detection zone lengths and differing vehicle speeds, compares these values with a calculated delay metric, and observes the effectiveness of GOR as an indicator of oversaturation. On the basis of the analysis, the paper documents the influence of approach speed and detection zone length on the calculated GOR metric. The paper concludes that the GOR values tend to reach a saturated value of 1.0 more rapidly than v/c, making it challenging for the analyst to identify oversaturated conditions by the metric alone. For practitioners using the GOR metric, it is critical that the system be calibrated properly because of the sensitivity of GOR to the detection zone length and vehicle speed. However, because direct measurement of v/c is difficult to obtain, GOR may still be a reasonable surrogate in real-time performance assessment of traffic signal systems.

AB - Various control algorithms process vehicle detector on and off times with the objective of quantifying split performance. Two such metrics are green occupancy ratio (GOR) and volume-to-capacity (v/c) ratio, both of which can be used to estimate the efficiency of a split as well as estimate oversaturated conditions. The GOR metric is attractive because of its simplicity of implementation; the v/c ratio is based on well-established traffic flow concepts and is independent of detector geometry although additional processing is required to develop the metric. Characterizing the impact of detection zone length and vehicle speed on the GOR as well as comparing its performance with another metric is important in understanding the robustness of GOR as a surrogate performance metric. This paper analyzes the difference between calculated GOR and v/c ratios over different detection zone lengths and differing vehicle speeds, compares these values with a calculated delay metric, and observes the effectiveness of GOR as an indicator of oversaturation. On the basis of the analysis, the paper documents the influence of approach speed and detection zone length on the calculated GOR metric. The paper concludes that the GOR values tend to reach a saturated value of 1.0 more rapidly than v/c, making it challenging for the analyst to identify oversaturated conditions by the metric alone. For practitioners using the GOR metric, it is critical that the system be calibrated properly because of the sensitivity of GOR to the detection zone length and vehicle speed. However, because direct measurement of v/c is difficult to obtain, GOR may still be a reasonable surrogate in real-time performance assessment of traffic signal systems.

UR - http://www.scopus.com/inward/record.url?scp=84857867866&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84857867866&partnerID=8YFLogxK

U2 - 10.3141/2259-11

DO - 10.3141/2259-11

M3 - Article

AN - SCOPUS:84857867866

SP - 123

EP - 131

JO - Transportation Research Record

JF - Transportation Research Record

SN - 0361-1981

IS - 2259

ER -