The signalized intersection is the cardinal element of a freeway system. To improve the operational quality of the system it is necessary to develop traffic signalization to its highest level of efficiency. A state-of-the-art summary resulted in an extensive bibliography. A multiple purpose, microscopic, digital simulation model of traffic performance and control at an individual intersection was successfully formulated, programmed, tested, and refined. By applying the intersection simulation model, data were derived that led to an increased understanding of a variety of measures of effectiveness of performance, their dependency on the mode of traffic-signal control utilized, and their interdependency. The model was exercised comprehensively for fixed-time modes of intersection control in order to derive a controlled data base from which to judge objectively the relative effectiveness of experimental traffic-responsive philosophies of control. A new philosophy of intersection control, the basic queue-control scheme, appeared promising and was selected for pilot field implementation. One unique feature of this work was the elimination of considerable time and cost involved in designing and fabricating special- purpose traffic-control hardware, made possible by mocking up the new control logic in a real-time digital control computer. The experimental control technique could be developed into an operational system in a relatively short period of time. A special study of the relative effectiveness of various phasing schemes incorporating protected left turns was successfully completed. An inescapable conclusion that repeatedly emerged is the overriding influence of signal-cycle length on effectiveness of performance. The total delay equalization policy was found to be rational, virtually synonymous with delay minimization. The average delay equalization policy results in significant over-all delay excesses under high, unbalanced volume conditions.
Samenvatting