Genetic Algorithm Methodology to Optimize Signal Timing for Vehicle and Pedestrian Delays

There are two objectives for this research. One is to develop an effective procedure to optimize intersection signal timing by minimizing total delay for both vehicles and pedestrians. The second objective is to establish guidance for the pedestrian crossing phase pattern (two-way or scramble) and for the length of WALK phase when a scramble crossing is used. An optimization procedure for signal plans in an isolated intersection was developed. The procedure yielded up to four phases for vehicles with either the two-way or scramble pedestrian crossing phase. A simple Genetic Algorithm (GA) was used in finding suitable signal plans because of the existence of a very large solution searching space. Furthermore, a look-up table was generated to guide the decision between two-way and scramble phases. The guidance considered different combinations of vehicle and pedestrian volumes. Compared with the Highway Capacity Software (HCS) GA function, the proposed procedure has the same accuracy and slightly better capability when pedestrian delay is not considered. However, when pedestrian delay is considered in signal timing, such a comparison could not be made because HCS does not compute delay for pedestrians while the proposed GA procedure does.