Open AccessLED Solar Spectrum Computer Simulation Based on Non-dominated Sorting Genetic Algorithm
Author(s) -
Lin Liu
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1578/1/012103
Subject(s) - sorting , genetic algorithm , computer science , spectrum (functional analysis) , sorting algorithm , algorithm , physics , machine learning , quantum mechanics
The solar computer simulation technology has been extensively applied in the fields of polymer curing test, solar cell detection and calibration, and satellite heat balance test, etc. The non-negative least-squares solution of the overdetermined equations based on the non-dominated sorting genetic algorithm (NSGA) to optimize the monochromatic light-emitting diode (LED) matching light source combination, replace some monochromatic LED with white LED to simulate the solar spectrum, and discuss several LED lights with different peak wavelengths that can be replaced by white LED. The simulation results suggest that in the range of 300 ∼ 1100 nm, as the monochromatic LED types replaced by white LED increase, the total number of LEDs used is reduced, and the spectral match decreases. When LEDs with three different peak wavelengths are replaced, the correlation index in the fitting based on the algorithm is 0.9035, where the number of LEDs can be reduced by 15.6%, and the simulation spectrum is basically consistent with the target one. The proposed method has a small spectral mismatch and can be used to distinguish two absorption valleys of the standard solar spectrum AM1.5 accurately.