A ridesplitting market equilibrium model with utility-based compensation pricing

Qing Long Lu, Moeid Qurashi, Constantinos Antoniou

Research output: Contribution to journalArticlepeer-review

Abstract

The paper develops a theoretic equilibrium model for ridesplitting markets with specific considerations of origin-destination demand patterns, competition with other transport modes, characteristics of en route matching, and spatial allocation of ridesplitting vehicles, to adequately portray the intertwined relationships between the endogenous variables and decisions. The operation property of the market under distance-based unified pricing is analyzed through the response of system performance indicators to the decisions. Moreover, a gradient descent algorithm is derived to find optimal operating strategies in the monopoly scenario and social optimum scenario. Leveraging the tight connection between trip’s utility and level of service (LoS), the paper then proposes a utility-based compensation pricing method to alleviate the inequity issue in ridesplitting, which results from the variance in waiting time and detour time and the implementation of unified pricing. Specifically, the trip fare of those with an initial utility smaller than a threshold will be compensated following a predefined compensation function. We compare its effectiveness and influence in different scenarios through numerical experiments at Munich. The results show that the proposed pricing method can improve the LoS and equity without losing any profit and welfare, and can even achieve increments in maximum profit and social welfare under certain conditions.

Original languageEnglish
Pages (from-to)439-474
Number of pages36
JournalTransportation
Volume51
Issue number2
DOIs
StatePublished - Apr 2024

Keywords

  • Equity
  • Level of services
  • Market equilibrium
  • Pricing
  • Ridesplitting

Fingerprint

Dive into the research topics of 'A ridesplitting market equilibrium model with utility-based compensation pricing'. Together they form a unique fingerprint.

Cite this