Towards the creation of a backcasting-oriented model of mobility systems: identification of surrogate models from detailed traffic simulations

An emerging approach is expected to change the way of designing and assessing future mobility of people and goods. Whether it will be more electrified, connected, automated, digitalized or not, it will depend on which targets we aim to reach at given time horizons: environmental, economic, societal, etc. In the new paradigm, these targets will be set first, then the most appropriate roadmap of policies and technologies needed will be established (“backcasting”), as opposed to the current approach where potential actions are assumed, then their impacts are “forecasted” and assessed. Backcasting can be seen as a dynamic optimisation process (optimal control). It will be enabled by the availability of a system dynamic model that describes mobility systems in very macroscopic terms, yet capable of representing all causality loops linking the policies sought to the impacts expected. 
This doctoral topic represents the first building block of such an ambitious research program, as it targets a system dynamic model of the transportation network subsystem. A precise definition of manipulable and exogenous input, internal dynamic states, as well as of significant output variables, shall be the result of a preliminary, System Dynamics analysis. To identify and quantify the causal dependencies between these variables, a multimodal transportation network simulator (e.g., MatSim) will be used. Simulation experiments will be designed and run. The corresponding inputs and outputs variables will be used to construct surrogate models (the most suitable technique is to be identified) and tune their parameters, to provide the quantitative dependencies sought. The analysis will focus on different scenarios, to anticipate the effects of exogenous variables. Different networks or territorial scales will be attempted as well, to investigate the variability of the transfer functions identified and their scalability.

Keywords: Systems Dynamics, Backcasting, Transport Networks, Surrogate Models

• Academic supervisor    Dr. SCIARRETTA Antonio, IFP Energies nouvelles, ORCID 0000-0002-4643-0706
• Doctoral School    ED580 – STIC Sciences et Technologies de l’Information et de la Communication, https://www.universite-paris-saclay.fr/en/doctoral-schools/sciences-and-technologies-information-and-communication 
• PhD location    IFP Energies nouvelles, Rueil-Malmaison, France 
• Duration and start date    3 years, starting in fourth quarter 2023
• Employer    IFP Energies Nouvelles, Rueil-Malmaison, France
• Academic requirements    University Master degree in Transportation Engineering or Information Engineering or Artificial Intelligence
• Language requirements    Fluency in French or English, willingness to learn French