Tool: Optimisation of combined bike sharing - public transport system

This tool assumes that fleet sizes for New Shared Modes are also subject to optimization. In this case, a full social welfare function is presented, in which users’ preferences are taken into account to optimize fleet sizes, with an application to the case of bike-sharing and public transport. This is possible with a full social welfare function that incorporates not only users’ benefits, but also operators’ profit (total operator revenue minus total operator cost) and two additional components related to positive and negative externalities of mobility, namely the health benefits from active mobility and climate change cost, measured as the social cost of carbon from motorised mobility.

The framework, developed by the University of Twente, is solved a bi-level optimization problem. The upper-level problem aims to optimize fleet sizing for both bike-sharing and public transport. The lower-level problem is a modified user equilibrium model, used to evaluate both route choice and mode choice by users within the network. The decision variables at this level are demand and traffic/passenger flows across different transportation modes. The upper-level problem is solved using a genetic algorithm, while the lower-level problem is addressed using a modified Method of Successive Averages (MSA). The lower-level model is embedded as a function within the upper-level algorithm, allowing both levels to be solved iteratively.