Optimization of traffic flow splits.

The objective of this study is the development of a method for predicting changes in a specified measure of the level of service provided by a street system under various system controls. The procedure consists of the development of a maximization model, an analysis of model sensitivity, and the comparison of this model output to the output from models presently in use. The model is an application of the technique of maximization under constraints. The objective function selected for maximization is the number of vehicle- miles per hour produced within the designated street system. The variables studied include street capacities, flow characteristics, one-way streets, freeway capacities, and ramp capabilities. The model is used to illustrate the effectiveness of various possible control systems when evaluated by increases in the objective function. The results of this analysis indicated that increasing the capacity of a freeway has the greatest effect on the total system output. Although the development of the technique for measuring system output is the primary objective of this study, the application of these results in the decision- making process is the ultimate goal. There are two major areas where use of this model type would add significantly to the field of traffic assignment and control. First, it provides an objective measure for evaluating control methods, and second, it provides a method for determining the system effect of changes in an individual element. The results of the model were used to illustrate the decision- making process for three selected criteria: (1) a minimum acceptable average velocity, (2) a maximum acceptable number of stored vehicles, and (3) a maximum output per dollar expenditure. /author/.