MORFAE

Funding Program: BMBF, Mathematics for Innovations

Duration: 11/2022 until 10/2025

Project partners: University of Heidelberg, Robert Bosch GmbH

Subproject of the WG Continuous Optimization: Multilevel Iterations with Sequential Homotopy Methods

The joint project MORFAE with the company BOSCH aims at realizing potentials of mathematical MSO to promote "sustainability of mobility in urban and rural areas". The aim is to develop real-time methods for calculating optimal driving modes for partially or fully autonomous, networked individual vehicles and interconnected entire vehicle fleets in road traffic. In this way, the total energy consumption of all vehicles is to be minimized while guaranteeing safety and conserving resources. Resource-constrained vehicle computers and more powerful external infrastructure ("cloud") that communicate with each other and with other vehicles are available as computing resources. A focus is on optimal vehicle operating strategies that minimize CO2 emissions or energy consumption under uncertainties and strict constraints. Complex models for electric powertrains, internal combustion engines, hybrid powertrains with battery to combinations with fuel cells are to be used that have integer decisions and implicit discontinuities in the dynamics. Another focus is the much more complex - simultaneous - optimization of entire vehicle fleets with heterogeneous propulsion technology in a spatial segment. This leads to large distributed but specially structured problems of optimization.

In the subproject of TU Clausthal, building on preliminary work on novel sequential homotopy methods, new structure-exploiting numerical methods are developed and their suitability for the distributed solution of the underlying optimization problems in real time is investigated, especially with respect to resource-constrained onboard hardware systems