Swift integration of distributed energy generation, especially with renewable energy sources, asks for a smarter home and (micro)grid. From conversion of sun light to electricity injected to the grid there are numerous continuously fluctuating generation and consumption steps, and all of them currently suffers from numerous unknowns/imprecision in forecasting. Simultaneously the number of generation sites of different size is increasing and their behaviour is also changing as storage, electrical vehicles, and smart home are rapidly integrating our life.
Through the case of photovoltaic energy integration in the future smart grid the PhD candidates will work on PV power generation forecasting. Local weather measurement is one the key for precise forecasting and this could be gathered by dedicated sensors (FPGA, microcontroller boards) potentially integrated in the Pv module, system. Complementary, the Pv system should leverage the large arrays sensors in the smart home and taping into the Internet of things. Imec in-house sensor and Pv module technology expertise will be key assets in the demonstration and testing of these concepts.
To forecast the PV generation the PhD candidate will expand imec’s state-of-the-art and patented forecasting simulation framework using a combination of deep neural networks and physics based models. Therefore we look for candidates with previous experience with large simulation frameworks, big data analysis and use of neural network based model development. In the model development the fast ramp-rate and precision under varying weather conditions will be essential for balancing the local micro-grid. Further performance metrics and their specific targets will be defined in collaboration with colleagues working on the smart-grid, load management. The final aim of PhD topic will be an accurate and fast PV forecasting simulation framework coupled to sensors. Real-life validation will be critical element of the project for which imec has existing tests sites, and access large dataset via its industrial partners.
Important to note that the here developed tools and approaches are aimed to be generalized to other renewable energy sources, and in general to demand response management of the smart grid.
The student will be part imec’s PV module team and also benefit from close collaboration with various experts in EnergyVille as well as its state-of-the-art measurement facilities.
Required background: computer science major with, physics/electrical engineering minor
Type of work: 30% experimental, 60% computing, 10% literature and interaction w other experts. specify percentage dedicated to literature, technology study, experimental work, other
Supervisor: Jef Poortmans
Daily advisors: Hans Goverde. Dimitrios Anagnostos
The reference code for this PhD position is SE1712-17. Mention this reference code on your application form.