Designing of broadband solar omnidirectional reflector with a chirped multilayer using an optimization procedure based on a genetic algorithm

We have proposed an innovative optimization tool to design a solar broadband omnidirectional reflector (BOR) that operates across the entire solar irradiation spectrum by using a two-step procedure based on genetic algorithm. The BOR consists of chirped multilayer forming a discrete array of sub-mirrors, each with different resonance wavelengths. The central wavelength distribution is based on the Padé approximation. Three different types of sub-mirrors has been tested: periodic, Thue-Morse, and hybrid multilayer. In the first step, an integer genetic algorithm optimizes the resonance wavelength distribution, the number of submirrors, and the number of layers in each submirror. In the second step, the thicknesses of the layers are randomized to further enhance the reflection of the solar irradiance spectrum. All numerical simulations of electromagnetic propagation were performed using the scattering matrix method. The best result was obtained with a periodic BOR consisting of 60 layers, achieving an overall reflectivity of 98 % for the solar irradiance spectral range between 400 and 2000 nm across all incident angles for the both polarizations. The optimization tool is tested by evaluating the reflectance efficiency of optimized BOR with experimental irradiance of light solar simulators.