Traditionally, electricity networks in rural areas are built using a centralised network infrastructure. Due to long distances between sparsely distributed consumers, such infrastructures can be cost intensive to build and maintain, and prone to several environmental risks such as bushfires or storms. Considering the uptake of distributed renewable energy sources, e.g. rooftop solar panels, this might not be an ideal network layout in rural areas. In this seminar talk, we discuss how evolutionary algorithms can assist in planning optimised electricity distribution networks.  Given a set of electricity consumers, their loads and locations, the aim is to find a network layout minimising the total problem-specific cost over a certain time frame. Such a network does not need to be interconnected but might include isolated microgrids or stand-alone power systems, which are no longer connected to the main grid. Additionally, we want to be able to consider further constraints such as restricted areas that might be harder to access, risk-prone, or do not allow for construction of network infrastructure. Due to the complexity of the problem, evolutionary algorithms are proposed to identify where to place isolated network parts and how to interconnect electricity consumers. We will discuss different evolutionary approaches, focussing on the problem representation and implementation of constrained areas represented as polygonal obstacles that could be generalised to other areas of application as well.

Manou Rosenberg is a Forrest Research Scholar and PhD candidate at the Department of Computer Science and Software Engineering at the University of Western Australia. She received her BSc and MSc in Mathematics at the RWTH Aachen University, Germany. During her time as a research assistant at the Institute of Energy Economics and Rational Energy Use in Stuttgart, she worked on modelling the effects of prosumer behaviour on the German electricity market. Her research interests include the application of evolutionary algorithms and AI to real-world problems, optimisation of network layouts and combinatorial optimisation.