Advanced Signal Control Strategies and Analysis Methodologies for Diverging Diamond Interchanges
StatisticsView Usage Statistics
This dissertation introduces three new methodologies to improve traffic signal operations of Diverging Diamond Interchanges (DDIs). Methodology one applies to a DDI without signals for left-turns from the freeway off-ramp. This methodology combines Webster's method and the specific characteristics of a DDI to determine traffic signal operation parameters such as cycle length, phasing splits, and phasing sequence. Comparing to methodology one, methodology two can handle more general and complex cases. Both methodologies can be implemented at a DDI by one traffic controller and operate successfully for a variety of controller types including pre-timed, fully actuated, and coordinated actuated control. Methodology three, also called proposed operation 3, combines Genetic Algorithm and a professional simulation tool such as VISSIM to search for the optimal operations for DDIs based on the phasing scheme of methodology 1 or 2. As a case study, methodology two is comprehensively studied based on a proposed DDI located at Moana Lane and U.S. 395, in Reno, Nevada. Through testing in a hardware-in-the-loop platform, this methodology can operate successfully for pre-timed, fully actuated, and coordinated actuated traffic signal controls. Microscopic simulation models were developed to evaluate the traffic signal operation of each scenario. The simulation results revealed that proposed methodology 2 reduces average delay by 17% in the morning (AM) peak hour and 28% in the afternoon (PM) peak hour at the Moana DDI, when compared to the methodology presented by staff from the City of Reno, NV. The average total delays of different cycle lengths show that the optimal cycle length changes with the variation of saturation flow ratios at this DDI. The simulations illustrated that the performances of the same traffic signal operation varied when it was applied to a variety of traffic volume distributions among routes. Therefore, developing a traffic signal operation for a DDI based on its traffic volume distributions on routes, instead of turning movement volumes, is necessary. The results also indicate that the range of signal operation performance on a variety of traffic volume distributions on routes reduces when the space between the two crossovers intersections of a DDI increases.