Description
The PhD research will focus on three complementary challenges in low voltage grid decision support, each of them representing a fundamental hurdle to the realistic-scale roll-out of smart-grid decision systems in the field, and each requiring a considerable progression of the current state-of-the-art in data science technology applied to distribution grids.
The first focus is on optimizing the extraction and stochastic simulation of relevant information from the limited number of available grid data sources. Standard techniques for data-wrangling and data cleaning should be automatically combined in an optimal way and enhanced and/or customized to overcome the domain-specific flaws and deficiencies in the available data. Based on the resulting enhanced data, tools can be developed that stochastically generate relevant synthetic energy profiles.

The second focus is on forecasting the appearance of grid congestions such as under-voltage, over-voltage, and over-current. The main challenge here is in the exceptional nature and therefore the sparsity of occurrences of such congestions. Regression techniques that can make accurate extrapolations from the sparsely available grid data should be investigated, combined, enhanced and if necessary developed from scratch to effectively tackle this challenge.

The third focus is on overcoming the difficulties related to the correct interpretation of the collected, enhanced, processed, generated and forecasted data. Domain-specific and versatile representation and analysis techniques will be developed. For this purpose, AI-based clustering and visualization methods will be combined, enhanced and complemented with domain-specific insights and practical experiences in the field.

Collaboration with University of Leuven
Registration deadline: 11/01/2021


Qualification
Masters Diploma in Electrical Engineering with experience in data science. 
or
Masters Diploma in Computer Science with thorough expertise in data science techniques and affinity with smart grids.
Offer
The PhD student receives a PhD grant from the University of Leuven. 
VITO concludes a financing agreement with the University of Leuven, with VITO undertaking to provide an annual allowance matching the net remuneration of an assistant, plus management costs. The University of Leuven will pay the selected PhD student a PhD grant matching the before mentioned net amount.

Requisition

Location: 
Genk
Jobfield: 
PhD
ID: 
31400