High-speed signalized intersections are hazardous from the standpoint of causing rear-end collisions between vehicles on the same approach. The signal funnel concept developed in germany is desirable at such locations since it substantially reduces the percent of vehicles stopping. However, the signal funnel has not been incorporated with the semiactuated traffic signal often used at intersections on major thoroughfares in the united states. The major objectives of this investigation were to design and evaluate a speed signal system capable of functioning effectively with semiactuated control. The study involved a traffic control system for a t-junction utilizing an advisory speed signal on the main approach. The evaluation was accomplished by computer stimulation models programmed in gpss/360. The figures of merit for each model were (1) total number of vehicles stopping on the main approach during 15 signal cycles, (2) percent of vehicles forced to stop against the red signal on the high-speed route, (3) average delay incurred per side road vehicle, and (4) average delay per side road vehicle stopped. The first simulation model described vehicle activity on a minor approach lane and a high-speed approach lane at a t-junction with a two-phase semiactuated controller. This model was validated by comparing simulation output to field data obtained at an intersection in a 45-mph speed zone. Field data were gathered for side flow ranging from 60 to 250 vph and main flow from 180 to 700 vph per lane. Linear regression equations involving the same variables were constructed. The corresponding equations from the field data and from the simulation were then statistically tested for equality of regression coefficients. The simulation model proved satisfactory for predicting the figures of merit for the traffic volumes involved. The second simulation model was similar to the first, but included a main route speed advisory signal and a more elaborate side route vehicle detection system. Data obtained from the speed signal simulation model were compared to the output from the first model, thus evaluating the proposed signal funnel. The traffic-actuated speed signal funnel stopped an average of only 2.0 percent of the main route traffic, while the conventional semiactuated controller stopped 20.9 percent. Furthermore, the speed signal system only 2.8 cars on the high-speed route during 15 typical signal cycles, compared to 25.8 cars stoffped in 15 cycles with the semiactuated control. The improvement in main road flow was obtained without causing excessive side road delay. (Author/publisher)
Abstract