برگزاری دفاعیه دکتری مهندس محمد پور جعفری

Optimization of Traffic Signal Timing with Adaptive Transit Signal Priority

Abstract 

Nowadays, transit signal priority (TSP) plays a significant role in reducing delay and facilitating public transport (PT) vehicles through urban intersections. The development of adaptive transit signal priority strategies (ATSP) relying on the optimization of intersection performance measures has become the center of researchers’ attention in recent years. In this respect, the development of passenger-based strategies based on the optimization of performance measures for passengers compared to vehicle-based methods is considered as one of the novel fields in the respective studies. Generally, the optimization parameters in these studies consist of green time, cycle length, or offset, and no study has incorporated phase sequence in the optimization process of adaptive strategies so far. Therefore, a passenger-based kinematic wave-based adaptive transit signal priority strategy (KATC) is developed in this study, aiming to minimize the average passenger delay at the intersection level and based on the optimization of both green time and phase sequence. The analysis of the queuing process and the development of passenger vehicle delay models are carried out based on kinematic wave theory and for both undersaturated and oversaturated flow. Furthermore, the public transport delay models are developed based on mixed-use operation and various arrival cases of PT vehicles at the intersection. The optimization of average passenger delay is performed based on mixed-integer nonlinear programming (MINLP) and genetic algorithm (GA) is used as the solution method to solve the problem. The model is further applied to the PT operating condition with dedicated bus lanes as a passenger-based adaptive transit signal priority for BRT systems (PASPB). The PT vehicle delay models are developed in this respect based on the operational conditions with dedicated bus lanes and the existence of PT vehicle stops. The PT vehicle dwell time models are presented in this respect based on several cases of passenger service with one or two PT vehicles. The performance of the KATC strategy is evaluated in the VISSIM simulation environment and compared to the optimal solution of the SYNCHRO model and the KATC strategy without the capability of phase sequence optimization. Based on the acquired results, the KATC strategy can reduce the average PT passenger delay by 8.0% to 23.6% and average total passenger delay by 0.7% to 13.9% compared to the SYNCHRO model. Lower values of the saturation rate and increasing passenger occupancy have a positive effect on the delay improvements achieved by KATC. In addition, KATC improves the average speed and PT vehicle number of stops compared to other models. Regarding the PASPB strategy, the relevant method is deployed to a real intersection in Tehran with two conflicting BRT lines and evaluated against the current fixed-time signal timing, the SYNCHRO model, and the active strategy of phase insertion in the SUMO simulation environment. The results of the SUMO simulation experiments show that the PASPB strategy reduces the average total passenger delay by 12% and 5.9% (10.8 and 4.9 s/p) compared to the current fixed-time signal timing and SYNCHRO, respectively.  Compared to the phase insertion strategy, the results of the average passenger delay for the PASPB strategy are approximately equal to those of the relevant active strategy. Regarding PT passenger delay, the adoption of the PASPB strategy leads to a decrease in the referred performance measure by 24.8% and 11.8% (23.3 and 9.6 s/p) compared to the SYNCHRO model and phase insertion strategy, respectively. Besides, the results of the sensitivity analysis prove that in cases of increasing the PT vehicle or passenger vehicle demand, the application of the PASPB strategy leads to the best performance compared to other methods regarding the average passenger delay. Based on the results of the environmental performance measures in the current situation, the pollutant emissions and fuel consumption rates are higher in the PASPB strategy in comparison to the active strategy. However, PASPB delivers the best environmental results with increased general traffic demand. Furthermore, based on the results of the economic study, the benefit-to-cost ratio for implementing the PASPB strategy at a single signalized intersection is equal to 3.55.                  

Keywords: signal timing, adaptive transit signal priority, optimization, genetic algorithm, average passenger delay