Navigation and service

Advertising division: IEK-3 - Electrochemical Process Engineering
Reference number: 2019M-023, Electrical engineering, mechanical engineering, energy engineering, physics, mathematics, computer science

Master thesis: Handling Big Data in Renewable Energy Systems: Finding the Right Data Qualities and Clustering Indicators for Time Series Aggregation

Background:
Facing the turnaround in energy policy and the growing share of intermittent energy sources such as photovoltaic and wind, the Institute of Energy and Climate Research (IEK-3) develops energy system models. These focus on designing cost-optimal and feasible energy systems based on existing input time series such as electrical or chemical energy demand and energy generating potentials like wind speeds or irradiance at specific places for a large number of discrete time steps. In order to keep the models computationally tractable, the input time series are usually clustered and, in this way, aggregated to a small number of typical days or weeks, with the goal to reduce the complexity and number of variables for large systems. By choosing a few representative days for modeling instead of a large number of days, a number of basic dataset qualities usually differ from the original dataset, e.g. mean value, variance, gradients, and correlation between different data sets. Depending on the clustering algorithm for choosing the representative days, these qualities can be taken into account. However, there are a large number of indicators and dataset qualities that can be examined. This topic is highly relevant for models of the future energy systems and data reduction is a fundamental method for managing big data.

Your Job:
The offered thesis focuses on the influence of different dataset qualities on the final output of energy system models. In a further step, a clustering method should be developed, which takes this information into account within the clustering algorithm. Moreover, a set of indicators should be suggested, that can be used to estimate the quality of the chosen representative periods a priori, i.e. without running the energy system model itself.

  • Literature research on the most important qualities of time series for energy system models
  • Induction into our existing energy system model
  • Running test cases for a small energy system model and with the same time series which was clustered with respect to different dataset qualities
  • Analysis of the output deviation
  • Literature research on good indicators for evaluating the clustering process a priori
  • Running test cases with different clustering parameters and searching for good clustering indicators that mirror good energy system model outputs
  • Conclusion and general advise which dataset qualities should be clustered and what indicators stand for a sufficient clustering process

Your Profile:
Very good academic marks in electrical engineering, mechanical engineering, energy engineering, physics, mathematics, computer science or related fields of study. Ability to work autonomously and analytically within a project team. Ideally you already have experience in modelling, programming (preferred in Python) and a high affinity for complex mathematical problems.

Our Offer:

  • A pleasant working environment within a highly competent, international team in one of the most prestigious research facilities in Europe
  • You will be supported by top-end scientific and technical infrastructure as well as close guidance by experts
  • You will have the opportunity to work with excited researchers from various scientific fields and take part in the design of a future European energy system
  • Your work is remunerated
  • Depending on your performance, the small work packages can be adapted

Contact:
Maximilian Hoffmann
Institute of Energy and Climate Research (IEK)
IEK-3: Electrochemical Process Engineering
Process and Systems Analysis
52425 Jülich

http://www.fz-juelich.de/iek/iek-3
Phone.: +49 2461 61-85402
e-mail: max.hoffmann@fz-juelich.de

Advertisement as pdf file:  Master thesis: Handling Big Data in Renewable Energy Systems: Finding the Right Data Qualities and Clustering Indicators for Time Series Aggregation (PDF, 245 kB)