Modelling the Effects of E-Scooters in Urban Traffic Using Artificial Neural Networks

Abstract

E-scooters have become a popular transportation alternative, especially for short distances. Compared to motor vehicles, e-scooters are operated at lower speeds on urban highways, therefore; changing traffic flow characteristics and reducing road traffic safety. Within the scope of this study, the effects of different demand levels of e-scooters on Istanbul Bagdat Avenue traffic were examined and modeled using artificial neural networks. Trained artificial neural network can estimate the hourly traffic volume and average speed values with a R2 value of 0.9699 and 0.9853, respectively. The effects of e-scooters on traffic can also be examined for different traffic conditions with the trained network structure. © 2022 IEEE.