On the dynamic pricing of pool ride-hailing services in bus lanes

Abstract

Ride-hailing vehicles contribute to traffic congestion in urban areas, where spatial constraints and uneven multi-modal distribution of infrastructure are a constant problem. In fact, roaming empty vehicles cause additional delays to other concurrent network users without delivering passengers to their destinations. Ride-splitting is one potential solution to counteract the negative impact of ride-hailing on traffic. In this work, we provide a dynamic non-equilibrium modelling framework for ride-splitting, where pool passengers are allowed to use dedicated bus lanes and can potentially travel faster than solo users. The objective is to develop a ride-splitting pricing policy between solo and pool options to encourage trip sharing with the goal of minimizing overall delays in multi-modal networks with bus lanes. Therefore, a Model Predictive Control (MPC) framework is set forward to investigate the price difference between the two ride-hailing alternatives with the objective of reconsidering the space allocation between the available modes. The results show that the proposed strategy is able to adjust the time-dependent fare changes based on the multi-modal demand and speeds in different parts of the network.