Weather-related traffic signal timing along a Minneapolis/St. Paul corridor reduced vehicle delay nearly eight percent and vehicle stops by over five percent.
Traffic and weather data were collected during a "normal" PM peak period and during three "adverse weather" PM peaks. For this study, "adverse weather" was defined as a snowstorm with three or more inches of snow. Simulation software, which models and optimizes traffic signal timing, was used to develop a network model of the corridor for analysis of weather impacts during the PM peak hour (5:00 PM to 6:00 PM). The network model was designed using signal spacing and intersection geometry data; existing split cycle and offset data; as well as traffic volume and occupancy data. Travel time run data were utilized to calculate average corridor speed and to adjust model offsets to represent existing conditions. "Normal" conditions (i.e., existing signal timing in good weather) were simulated as a basis for comparison.
Traffic statistics were measured with the existing timing under adverse weather conditions. Traffic volumes were found to be 15 to 30 percent lower during the PM peak hour. Average speed dropped 40 percent from 44 mi/h to 26 mi/h. The saturation flow rate decreased by 11 percent from 1,800 vehicles per hour per lane (vphpl) to 1,600 vphpl. Start-up delay increased by 50 percent from two seconds to three seconds. Percentile signal delay per vehicle actually decreased (because there were fewer vehicles traveling the corridor). Stops per vehicle remained the same. These statistics were used to modify the network model for simulation of existing system operation in "adverse" conditions.
To determine if a new timing plan could improve operations under "adverse" conditions, optimization was performed on the network model using cycle lengths between 100 and 180 seconds. A cycle length of 125 seconds improved measures of effectiveness. A comparison of existing and optimized timing under adverse conditions is shown in the table below. Optimization reduced vehicle delay by nearly eight percent and decreased vehicle stops by over five percent.
Measures of Effectiveness Table
Cycle Length (sec)
Signal Delay per Vehicle (sec)
Average Stops per Vehicle
Average Speed (mi/h)
Based on the simulation, it appeared that corridor operation was not radically affected by the adverse weather. Even though average speed decreased, vehicle delay did not increase. This is primarily due to the fact that there were fewer vehicles on the corridor during adverse weather. Further, the existing timing plan bandwidths were large enough to accommodate lower speeds.
Author: Maki, Pamela J.
Published By: Minnesota DOT
Prepared by Short Elliott Hendrickson, Inc. for the Minnesota DOT
Source Date: 1999URL: http://trafficware.infopop.cc/downloads/00005.pdf
Average User Rating
Typical Deployment Locations
coordinated signals, signal coordination, centralized signal control, signal synchronization, traffic signals, advanced signal control, signal timing optimization, coordinated signal control, advanced signal controller, traffic signal retiming, retiming