A heuristic programming approach to arterial signal timing

Abstract

The objective of this research was to evaluate a heuristic programming approach to optimize arterial street signal timing variables. The approach used a maximum bandwidth progression procedure (PASSER II) in combination with a minimum delay solution procedure (TRANSYT). Arterial street scenarios based on a real-world street were used to test the heuristic procedure. Cycle length and intersection spacing were varied to determine their effect on the heuristic solution. Program runs were made for all phasing sequences, cycle lengths and intersection spacing combinations to evaluate the heuristic approach in a classical nonlinear programming optimization evaluation. Comparisons were made between the heuristic solution and the best solution from the complete enumeration of the TRANSYT program. A revised green split procedure was developed and evaluated for use in the PASSER II program. The NETSIM computer simulation program was used for an evaluative comparison of the minimum delay solutions. The comparison showed variability in results. However, overall it would appear that the heuristic approach does provide "good" results for this evaluation. The use of a minimum delay cycle length as part of the heuristic approach was shown. The heuristic programming approach of using the PASSER II optimal solution as the initial starting point for a minimum delay solution was shown to be a good optimization technique. The heuristic procedure provides a means of determining all the signal timing variables for an arterial street to produce an optimal progression solution or a good minimum delay solution.