An Integrated Wireless Power Transmission and Multi-Physics Framework for PINN-Based Optimization of Space-Based Solar Power Systems
- Paper ID
114188
- author
- company
Turkish-German University
- country
Türkiye
- year
2026
- abstract
Space-Based Solar Power (SBSP) has been revisited due to its potential to generate electricity by harnessing high-intensity solar radiation outside the atmosphere and transmitting this energy to Earth. However, in large-scale architectures such as NASA’s RD1 (Heliostat Swarm) and RD2 (Mature Planar Array), several challenges must be addressed collectively: thermal cycles created by illumination-eclipse transitions, temperature-dependent efficiency variations of photovoltaic (PV) cells, and radiation-induced degradation. This multi-physics approach increases computational costs when using numerical solvers across a wide parameter space, while data-driven approaches alone may be limited in physical consistency and generalizability. This study proposes a Physics-Informed Neural Network (PINN)-based modeling pipeline to enhance system performance analysis of the SBSP energy chain during the early design phase. In the proposed approach, the electrical behavior of PV cells is represented by the Single Diode Model (SDM), while thermal behavior is modeled through radiative heat balance equations applicable to space environments. The goal is to predict panel temperature and power output with physical consistency under orbit-dependent conditions. NASA TSIS-1 and NAIF SPICE datasets will be used for input generation. The derived panel outputs will be integrated with parametric representations of the energy transfer chain to determine optimal design parameters and investigate optimal design regions. This study aims to provide a feasible analysis and optimization framework that can quickly and consistently evaluate SBSP panel performance during the concept development phase.