Bi-level Hybrid Algorithm for Greener Environment via Vehicular Networks in a Single Intersection

Bi-level Hybrid Algorithm for Greener Environment via Vehicular Networks in a Single Intersection

In this paper, we propose a bi-level optimization model (BLOM) with improved hybrid metaheuristics. The hybrid GA and PSO algorithm is applied in both upperlevel and lower-level model. This improved hybrid algorithm differs from the general hybrid algorithm which GA or PSO is applied in the single upper-level or lower-level model. BLOM is intended to schedule the phases of each isolated traffic signal and eco-driving environmentally. The upper-level optimization model (ULOM) considers the real-time traffic characteristics of the traffic flows near the signalized road intersection. At the same time, vehicles in the lower-level optimization model (LLOM) retrieve the real-time traffic signals using vehicular networks. Then, the traffic signals update the schedule and the vehicles are optimized motion states for greener environment factor respectively. We evaluate the performance of BLOM in a single road intersection using OMNET++ and SUMO. From the simulation results, we conclude that the BLOM with improved hybrid algorithm reduce fuel consumption and CO2 emissions compared with Maximize Throughput Model (MaxTM). Moreover, compared with the ordinary single algorithm, the proposed improved hybrid algorithm decreases the average operation cycle.