You are here: Home Chairs Chair for Resilience … Open positions

Open positions

Bachelor- und Masterarbeiten

Master Thesis - Efficiency and redundancy of power flow patterns in the European electricity network

Background:

Measures from information theory allow the quantification of efficiency as well as redundancy in networks with directed transport. The application of these measures to ecosystems, for example, shows that above a certain system size the ratio of efficiency to redundancy seems to concentrate in a small range of values, which raises the question of whether there is a connection between the classification in this range of values and the stability of the system.

In the proposed Master Thesis this relation between efficiency and redundancy is studied for power flow patterns in the European electricity system. In the interconnected European power grid, nodal imports and exports result in power flows connecting different regions both on a national and international scale. The basis for the thesis work are time series for imports, exports, and power flows, which are analyzed using measures from information theory to assess efficiency and redundancy of flows as a transport mechanism in this system. The analysis considers direct physical flows between neighboring regions, as well as distributed bilateral import-export relationships derived from flow allocation methods. The research studies how efficiency and redundancy depend on different system characteristics for current and future electricity systems and discusses the applicability of these measures for energy system analysis.

Tasks:

The following tasks are subject of the proposed topic:

  • Familiarization with data format of power flow patterns in the European electricity system
  • Application of information theoretic measures to the available data, characterizing efficiency and redundancy
  • Comparison of measures for neighboring as well as distributed import-export relationships
  • Comparison of current and future electricity system scenarios in view of efficiency and redundancy
  • Discussion of the informative value of these measures for energy system analysis, as a basis for optimization of the latter

Your profile:

  • You are a student of SSE, general engineering, physics, mathematics, or computer science
  • Highly interested in theoretical models and techniques to study energy systems
  • Programming skills (e.g. in Python) welcome (but not mandatory)

What you can expect:

  • Novel theoretical approaches in an interdisciplinary research environment
  • Data from cutting edge energy system models
  • Creative freedom in combination with intense support

Dates:

  • Start date: as soon as possible
  • Duration: According to examination regulations (typically 6 months)

Contact:

  • Prof. Dr. Alexander Stolz, Resilience Engineering for Technical Systems

alexander.stolz@inatech.uni-freiburg.de

 

  • Dr. Mirko Schäfer, Control and Integration of Grids

mirko.schaefer@inatech.uni-freiburg.de

 



SAFETY ANALYSIS OF AUTONOMOUS DRIVING BY MARKOV MODELING

Experimental Student Thesis (Master) 

Background
Autonomous driving functions up to SAE Level 3 and even more so in the future up to Level 5 place enormous demands on the safety case, since numerous possible scenarios must be evaluated in terms of safety.

Task

In the context of a larger national research project, the master thesis will use advanced vehicle system-level Markov models and simulations to identify critical scenarios, overall system state transitions and sequences. The objective is to provide a scalable and adaptable modeling and simulation to control state explosions. At the same time, the goal is to go far beyond the classical Markov assumption that overall system state transitions depend only on the current state and are time-independent.

Planned Tasks

  • Literature research on Markov modeling previously used for safety assessment of autonomous driving
  • Systems Modelling Language (SysML) model development of a generic vehicle architecture including driver, environment and other road users
  • Development of a scaling Markov diagram modeling based on the vehicle architecture
  • Transition, failure and recovery modeling and their parameterization
  • Markov simulation using different numerical methods in comparison, also combined and nested, to determine safety parameters
  • Evaluation of the feasibility, scalability and quality of the results with regard to possible key statements

 

What you bring with you

  • You are perusing a scientific university degree (Bachelor, Diploma or Master) in applied engineering or natural sciences, or in a comparable field (MINT subjects)
  • You have elementary knowledge in solving matrix equations (inversion of a matrix) and partial differential equations 
  • You have first experiences in a programming language, e.g. Python, C++
  • Knowledge of system modeling in a sub-discipline or of technologies relevant to autonomous driving would be an advantage 
  • With your independent and structured way of working you enrich the interdisciplinary team of our department

 

What you can expect

  • We are committed to excellent supervision and comprehensive training, while at the same time giving you a great deal of freedom to shape contents and implementations
  • You be part of an interdisciplinary team of scientists from a wide range of disciplines 
  • We have a very good office atmosphere in a modern working environment with excellent IT infrastructure and technical equipment of laptops and workstations

 

Starts: As soon as possible

Timeframe: According to examination regulations

More topics on request

 

Contact

Prof. Dr. Alexander Stolz

alexander.stolz@inatech.uni-freiburg.de

Resilience Engineering for Technical Systems| INATECH RTS

Faculty of Engineering | University of Freiburg

 

Dr. Ivo Häring

ivo.haering@emi.fraunhofer.de 

 


COOPERATIVE MASTERTHESIS:

INVESTIGATION OF THE POTENTIAL FOR AUTOMATED RESILIENCE ANALYSIS OF IT INFRASTRUCTURES                                                                                                                      

Professorship: Prof. Alexander Stolz (Universität Freiburg, D) and Client: CuriX AG

Initial situation:  

The existing software solution CuriX collects automatically measurement data (e.g. CPU utilization, free memory on hard disks, failed login attempts) and structural data (e.g. operating systems used, persons responsible, services provided) from IT infrastructures. The data is analyzed using various methods (e.g., statistical methods, machine learning, static analysis methods) to alert users on impending problems (e.g., system failure). Currently, these predictions are communicated as raw alerts (e.g., email, alert icon in the user interface) without any relation to resilience. 

Cooperative Masterthesis: Investigation of the potential for automated resilience analysis of IT infrastructures

In the future, the alerting functionality of CuriX should enhanced in the way that problems should be placed in the context of resilience, e.g. by grouping similar alerts to resilience incidents or issues. This should enable the user (i.e., typically the IT administrator) to derive and implement measures that can increase the resilience of the IT infrastructure.

Goal of the Master:

The existing and planned functionalities for CuriX-software should be analyzed and evaluated with regard to the existing taxonomy of resilience research. 

For example, it is planned to classify upcoming bottlenecks in the free memory on a hard disk into the category "resource bottleneck"; here, among other things, existing categorizations are to be assigned or elaborated. Likewise, anomalies in the system are analyzed, which are to be brought into the context of resilience. CuriX AG would like to follow existing glossaries of resilience research and use terms from or the vocabulary of resilience research for the software.

 

Key task is to investigate and evaluate the extent to which automated statements can be made about the resilience of IT infrastructures. It is planned to investigate two dimensions:  

  • Generic or generally valid statements: Assuming that the structure of IT infrastructures is known and sufficient measurement data is available, the aim is to investigate which and how generally valid statements about the resilience of IT infrastructures can be derived automatically. E.g., an indication of whether a higher homogeneity/heterogeneity of the operating systems used is more favorable or less favorable. 
  • Regarding given resilience goals: it should be investigated which statements about resilience (e.g. a resilience score) and resilience-increasing measures can be made and how.  

Within the scope of the work, statements are to be verified experimentally (by means of existing or to be developed software components and existing IT infrastructure). 

For both dimensions, preconditions are to be worked out that must be fulfilled in order to be able to make statements, e.g. which information on technology, organization, training, budget must be available.

Organisation:

The work takes place as a cooperative work with CuriX AG. 

The student will have a temporary contract for the duration of the Master work.

 

Starts: As soon as possible

Timeframe: According to examination regulations

More topics on request

 

 

Contact

Prof. Dr. Alexander Stolz

alexander.stolz@inatech.uni-freiburg.de

Resilience Engineering for Technical Systems| INATECH RTS

Faculty of Engineering | University of Freiburg

 

CuriX AG: 

Dr. Uli Siebold

uli.siebold@curix.ai

Head of Research | CuriX AG (Pratteln, Switzerland)

 


Experimental Student Thesis (Master) zum Thema RESILIENTES PATH-FINDING FÜR AGENTENBASIERTES MODELLIEREN

Agentenbasierte Modelle (ABM) mit individuell agierenden Agenten sind vor allem auf zwei Algorithmen angewiesen: das Path-Finding und die Kollisionsvermeidung. Besonders der Path-Finding Algorithmus hat einen großen Einfluss auf die Simulation und vor allem die benötigten Ressourcen, da er bei jeder Berechnung einer Route für jeden Agenten aufgerufen werden muss. D.h. die Dauer der Simulation hängt von der Anzahl an Agenten ab und der Effizienz des Path-Findings. Der bevorzugte Weg vom Standort eines Agenten zu einem bestimmten Ziel ist eine gerade Linie. Liegen allerdings Hindernisse auf diesem Weg, gibt es viele verschiedene Möglichkeiten einen geeigneten Weg zu finden. Ein Vergleich existierender Algorithmen wird angestrebt und die Idee eines resilienten Path-Findings wird untersucht. Das Ziel dieser Masterarbeit ist es zu bestimmen was einen Path-Finding Algorithmus resilient macht gegenüber diversen Hindernissen, die sich z.B. in Größe, Form, Position und Zeit verändern. Außerdem soll untersucht werden wie die Leistung des Algorithmus quantifiziert werden kann abgesehen von seiner Schnelligkeit. Szenarien werden entwickelt (z.B. mehrere aufeinandertreffende Menschenmassen) um den Algorithmus zu testen, welche anhand von frei verfügbaren Daten validiert werden sollen. Die generelle Ausrichtung der Arbeit kann individuell gestaltet werden. Eine mögliche Ausrichtung könnte sein, dass selbstlernende Ansätze untersucht werden um Path-Finding Algorithmen zu verbessern.

Voraussetzungen:

  • Studium in einem technischen oder naturwissenschaftlichen Fach, z.B. Informatik, Mathematik, Ingenieurwissenschaften, Physik, Mediendesign
  • Gute Programmierkenntnisse (z.B. C++, Java, Python)
  • Interesse an Programmierung und Softwareentwicklung für relevante wissenschaftliche Fragestellungen

 

Starts: As soon as possible

Timeframe: According to examination regulations More topics on request

Contact: Prof. Dr. Alexander Stolz, alexander.stolz@inatech.uni-freiburg.de