Energy Management System for Energy Hubs Using Artificial Intelligence

Project description

Heavy-duty/ medium-duty electric vehicles (HD/MDEVs) compared to light-duty EVs have very big batteries to provide for greater power requirements over a long distance. The only way to guarantee a reasonable charging time is to use fast, high-power charging stations (> 1 MW for each charging plug). Therefore, uncontrolled charging can have adverse impacts on the electricity network, such as:

• overloading of distribution network assets like lines and transformers
• undervoltage issues
• increased EV charging costs as the charging often coincides with a period of high energy price
• substantial use of fossil fuel-based sources of electricity esp. during the evening peak
• the unused potential of EV battery as an asset for ancillary services
• increased battery degradation due to charging at full power for shorter durations

The aforementioned difficulties can be solved by intelligently controlling the charging systems, energy storage systems, and fuel cells in the energy hub. Heuristic or mathematical programming methods have frequently been employed in the development of energy management systems for energy hubs. However, when nonlinearities of the energy hub's subsystems, such as distribution grids, thermal storage, and battery degradation, are taken into consideration, the performance of these algorithms is constrained and highly time-consuming. With the advancement of approaches such as machine learning (ML), the potential to utilize these techniques for energy hubs is huge.

The PhD candidate will develop a ML-based algorithm for the power and energy management of the energy hubs that is fast, robust, and scalable. The developed algorithm would next be evaluated in a hardware-in-loop setup and possibly at one of the project partners' test sites.

PhD candidate:

Leila Shams Ashkezari: l.shamsashkezari@tudelft.nl 

Supervisor:

Gautham Ram Chandra Mouli: g.r.chandramouli@tudelft.nl 

Neil Yorke-Smith: N.Yorke-Smith@tudelft.nl

Promotor:

Pavol Bauer: P.Bauer@tudelft.nl

Funded by:

Dutch government