Design and Optimization of Graphene-Based Ultra-Wideband Surface Plasmon Resonance Solar Absorber Using Machine Learning

The new generation of the surface plasmon resonance broadband type solar absorber for daily activities is produced in this work. To use the recent absorber efficiently, the perfect amount of absorption can be created with the unique design (combination of one cylinder and two squares) in proper materials of titanium (Ti), titanium nitride (TiN), and tungsten (W) performed from the top to bottom of the structure. As an ideal performance, the current design can work in multilayers of the atmosphere, from the ultraviolet part to the middle infrared layer. Machine learning is also applied in this work to verify the detail absorption points in each layer change and design optimization is achieved. With the numerical absorption analyzed, the wavelength of 1700 nm identifies 95.45% light absorption and 97.3% at 700 nm. The longest rate was performed by 0.2 and 3 mu m with the resulting rate of 93.22%. The current work can be used in some solar applications of solar ventilation, solar lighting, and solar heating processes.