Optimization of Flexible Transit Routes under Mixed Demand Considering Carbon Emissions

This paper concentrates on optimizing the routing for flexible transit with time windows, in which vehicles are tasked with both picking up and dropping off passengers concurrently. Besides introducing the time penalty cost to describe the passengers’ satisfaction on early or late arrival of vehicles, we include carbon emissions cost to investigate the impact of carbon emissions. An adaptive genetic algorithm (AGA) with adaptive mutation operator is proposed to solve this problem. Also, we introduce the elite retention into the algorithm in order to speed up the convergence. The case study uses the ride-hailing services data with its origin or destination located in Nanjing Railway Station to study the performance of the algorithm under mixed demand. Finally, we perform a sensitivity analysis to investigate how operational factors (vehicle speed, vehicle capacity, weighting ratios between operation, penalty, and carbon emissions costs) influence the performance of the result of routing optimization.