TY - JOUR
T1 - Artificial Neural Network PV Performance Prediction and Electric Power System Simulation of a Ship-tracking CubeSat
AU - Al Radi, Muaz
AU - Ghenai, Chaouki
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive licence to Shiraz University.
PY - 2023/6
Y1 - 2023/6
N2 - This paper presents results on the modeling and simulation of the electric power system (EPS) of a ship-tracking CubeSat. The main objective of the present study is to develop models for simulating the instantaneous and average performance of the CubeSat’s EPS components, namely the generation subsystem (solar PV cells), the storage subsystem (Li-ion batteries), and the CubeSat’s electric loads. The goal is to design a renewable power system to meet the CubeSat electric loads for marine ships tracking. Different simulation programs such as system tool kit and MATLAB software were used to simulate the power generation of the solar photovoltaic (PV) cells, the power consumption of the electric loads, and the instantaneous Li-ion battery performance parameters. First, a machine learning model based on Artificial Neural Networks was built and trained to predict the average power production of the CubeSat’s PV cells in terms of the orbit’s altitude and inclination and the PV efficiency. Different model architectures were compared and the optimal model was found. The optimal model showed substantially high accuracy as it achieved a regression coefficient (R) value of more than 0.9999 in the testing stage. Next, modeling and simulation analysis based on the algorithms developed in this study were used to investigate the short-term and long-term performance of the CubeSat battery system. Two simulations were carried out to investigate the performance of the proposed system, the first for 22 orbits and the second for 185 orbits. It was found that the chosen components, the proposed power budget, the sized solar PV and battery systems, and the chosen system parameters yielded a stable EPS as the batteries were always charged in the long run and their state of charge value kept oscillating between 0.91 and 0.8 until the end of the simulated mission.
AB - This paper presents results on the modeling and simulation of the electric power system (EPS) of a ship-tracking CubeSat. The main objective of the present study is to develop models for simulating the instantaneous and average performance of the CubeSat’s EPS components, namely the generation subsystem (solar PV cells), the storage subsystem (Li-ion batteries), and the CubeSat’s electric loads. The goal is to design a renewable power system to meet the CubeSat electric loads for marine ships tracking. Different simulation programs such as system tool kit and MATLAB software were used to simulate the power generation of the solar photovoltaic (PV) cells, the power consumption of the electric loads, and the instantaneous Li-ion battery performance parameters. First, a machine learning model based on Artificial Neural Networks was built and trained to predict the average power production of the CubeSat’s PV cells in terms of the orbit’s altitude and inclination and the PV efficiency. Different model architectures were compared and the optimal model was found. The optimal model showed substantially high accuracy as it achieved a regression coefficient (R) value of more than 0.9999 in the testing stage. Next, modeling and simulation analysis based on the algorithms developed in this study were used to investigate the short-term and long-term performance of the CubeSat battery system. Two simulations were carried out to investigate the performance of the proposed system, the first for 22 orbits and the second for 185 orbits. It was found that the chosen components, the proposed power budget, the sized solar PV and battery systems, and the chosen system parameters yielded a stable EPS as the batteries were always charged in the long run and their state of charge value kept oscillating between 0.91 and 0.8 until the end of the simulated mission.
KW - Artificial neural networks
KW - CubeSat
KW - Electric power system
KW - Li-ion battery
KW - Modeling and simulation
KW - Ship-tracking
KW - Solar PV
UR - http://www.scopus.com/inward/record.url?scp=85144161817&partnerID=8YFLogxK
U2 - 10.1007/s40998-022-00558-6
DO - 10.1007/s40998-022-00558-6
M3 - Article
AN - SCOPUS:85144161817
SN - 2228-6179
VL - 47
SP - 771
EP - 787
JO - Iranian Journal of Science and Technology - Transactions of Electrical Engineering
JF - Iranian Journal of Science and Technology - Transactions of Electrical Engineering
IS - 2
ER -