Over a billion of people all over the world, mostly in rural areas, do not have access to electricity. The main obstacle to an affordable and sustainable rural electrification is the lack of power electronics technology compatible with unforeseeable changes in either energy production or load profile. SPEARHEAD’s work focuses on studying the use of modular power electronics systems as a means to solve this problem. During their TA research, the SPEARHEAD user team tested two pilot elementary power modules in controlled conditions to validate their hardware, firmware and software design.
SPEARHEAD’s TA stay allowed the user team to get extensive feedback on their design choices. “Both of our elementary power modules had flaws that we would have never been able to detect if not for the tests we performed at the ICCS-NTUA laboratory in Athens,” says Luiz Fernando Lavado Villa, leader of the TA user group SPEARHEAD. In terms of hardware, we determined the best compromise between simplicity and robustness of different parts of both elementary boards. In terms of firmware, the user team understood the importance of looking into a more advanced mode of task scheduling. Finally, in terms of software, the SPEARHEAD team have decided to look further into Cython in order to embed future applications in low-cost SOCs such as a raspberry pi.
The most significant achievements for the SPEARHEAD TA group were the emergence of the notion of “Mathware” and the recognition that modular power electronics systems may require embedded mathematical applications for machine learning. These applications can provide clues to the state-of-health of a certain power converter and can help local practitioners to better manage their converter park as the microgrid changes over time. According to the SPEARHEAD approach, this is formalised by a “mathware” abstraction layer.