Real-Time Dynamic Transit Signal Priority Optimization for Coordinated Traffic Networks Using Genetic Algorithms and Artificial Neural Networks

Transit signal priority (TSP) has gained popularity in providing public transportation buses with preferential treatment at signalized intersections. Many studies have addressed its implementation in prompting enhanced public transportation service, such as reducing person delay and reducing transit travel time. However, most TSP implementations are done at the intersection level. Only a few studies have addressed the problem of integrating signal priority in coordinated real-time traffic signal control systems. A particular problem in this case is the uncertainty of predicting transit movements when considering the variability of dwell times at service stops. This study presents the development of a real-time traffic signal control integrating traffic signal timing optimization and TSP control using genetic algorithms (GA) and artificial neural networks (ANN) modeling. The GA is used to find near-optimal signal timings. Six different signal control systems were evaluated: fixed-time control with and without standard TSP, actuated signal control with and without standard TSP, real-time GA-based control without TSP, and real-time GA-based with advanced TSP logic. The standard TSP is implemented at the intersection level, by providing either early green (red truncation) or green extension strategies whenever a bus exists. A traffic signal control system that incorporates GA to optimize the fitness function and ANN for transit travel time prediction is developed. A microscopic simulation environment using VISSIM 4.3 simulation environment is used to test the previously mentioned six traffic control systems. The simulation results show that the proposed control system can reduce transit vehicle delay and improve schedule adherence. The reductions in delay and schedule adherence are statistically significant.