Optimal partitioning of photovoltaic modules on a curved solar collector
Files
(Published version)
Date
2019
Authors
Kapsis, M.
Albrecht, A.R.
Pudney, P.J.
Editors
Elsawah, E.
S, S.
S, S.
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
23rd International Congress on Modelling and Simulation Supporting Evidence Based Decision Making the Role of Modelling and Simulation Modsim 2019, 2019 / Elsawah, E., S, S. (ed./s), pp.109-115
Statement of Responsibility
Conference Name
23rd International Congress on Modelling and Simulation (MODSIM2019) (1 Dec 2019 - 6 Dec 2019 : Canberra, Australia)
Abstract
The Australian Technology Network (ATN) group of universities designed and built a solar car to participate in the 2019 Bridgestone World Solar Challenge. The car is powered by 29 photovoltaic modules on its top surface. To get a useful voltage from the solar collector, modules are connected in series. However, the power generated by a group of modules in series is limited by the module with the lowest irradiance. Irradiance depends on the angle between the sun and the cell normal, which is influenced by the curvature of the collector and the position of the sun relative to the car. If all of the modules were facing the same direction then each module would receive the same irradiance and there would be no "series mismatch" losses, but the solar panel is curved for aerodynamic efficiency.Our challenge was to partition the modules into groups so that the energy generated by the solar collector is maximised during a six-day journey across Australia. We describe two methods for partitioning modules into groups. Our mixed-integer programming model provides optimal solutions for a single time instant, but cannot solve the problem for the entire journey. Using a Cross Entropy Optimisation method, we were able to find solutions that were within 0.03% of optimal for a single time instant, and could also find good solutions to the six-day problem.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
Copyright 2019 the Author(s) (https://creativecommons.org/licenses/by/4.0/)
Access Condition Notes: These proceedings are licensed under the terms of the Creative Commons Attribution 4.0 International CC BY License