AIRO Young Opportunity

Expires
30 Nov 2017
Institution
University of Erlangen
Country
Germany
Description
For complete information see the linked pdf file. A brief fummary from the file follows:

Transportation networks have an increasing share of border-crossing services. The conditions
to implement such services are often different in neighbouring countries. For resource planning,
this may make it necessary to change between resources at the border. In the case of railway
networks, for example, this might apply to different electricity systems which require a change
of locomotive, different track gauges which require suitable wagons, or labour agreements as
well as different technical skills which might require staff changes. The challenge is to provide
decision makers with suggestions on how to find optimal resource allocations to deal with these
differing regulations in the best possible way. Ideally, it is also possible to make suggestions
for additional investments into more flexible resources, such as locomotives that are suited for
different electricity systems or training of employees.
The PhD candidate shall develop mixed-integer optimization models for the above problem setting
of an optimal resource allocation and investment for border-crossing transport services, focussing
on a suitably chosen subset of resources to plan. The optimization models should form a hier-
archy in the sense that they represent the given problem in different levels of detail, for example
with respect to the planning horizon (strategic, tactical or operational view), or with respect to time
and space resolution. This enables the development of solution algorithms based on adaptive re-
finement or suitable decomposition approaches based on Benders decomposition, for example.
The mathematical work on these models will then focus on structural analysis, for example iden-
tifying combinatorial subproblems or drawing upon graph theory and polyhedral analysis. Data
uncertainties, for example in demand forecasts or in the form of delays, might require robust or
stochastic model extensions. The PhD candidate will spend secondments for technical and sci-
entific training at DB Cargo Polska in Zabrze, Poland. The PhD degree will be awarded by FAU
Erlangen-N├╝rnberg, Germany.