Development, Testing and Evaluation of a Nodal Restraint Assignment Procedure (1153-4)

This research proposed a traffic assignment procedure in which capacity restraints are applied to nodes instead of links. The development is based on the concept that the capacity of an urban street system is constrained by nodes (intersections) instead of links. The nodal restraint assignment procedure was developed by utilizing the concept of the intersection sum of critical lane volumes in the Highway Capacity Manual 1985. A nodal impedance adjustment subroutine was incorporated in the assignment process to account for intersection delays where link impedances were held constant and nodal impedances were updated from iteration to iteration. The impedance for each turning movement at a node is determined by the association of all the movements encountered at the node.|The proposed procedure then was applied to a test network (Preston Road in North Dallas). In the application, various assignment procedures and different impedance adjustment function parameters were used to test the robustness of the procedure.|The results from the nodal restraint assignment procedure were compared to the selected "best" of the available conventional capacity restraint assignments, based upon traffic counts at major intersections along Preston Road. The evaluation was based on various micro-level analyses including mean difference, root mean square errors, turning movements as a percentage of approach volumes, and a series of paired t-tests. These analyses show that the nodal restraint assignment generally produced better turning movement replications than the available capacity restraint assignments.