OPTIMAL DESIGN OF GRADED REFRACTIVE INDEX PROFILE FOR BROADBAND OMNIDIRECTIONAL ANTIREFLECTION COATINGS USING GENETIC PROGRAMMING

To eliminate the average reflectance of antireflection coatings to the greatest extent, a Genetic Programming (GP) algorithm is proposed to design and optimize the graded refractive index distribution profile for broadband omnidirectional antireflection coatings. The proposed GP-index profile in this paper can obtain an extremely low average reflectance of 4.61×10-7% over a wide range of incident angles and wavelengths which is obviously superior to the average reflectance of 8.09×10-3%, 3.29×10-4% and 4.35×10-5% for linear profile, cubic profile and quintic profile. That means, Fresnel reflection almost can be eliminated by the optimal GP-index profile for omnidirectional incidence over a broad wavelength range. Moreover, it is demonstrated the proposed GP-index profile has better robustness, and it still has the best broadband and omnidirectional antireflection characteristics for the TiO2/SiO2 graded-index AR coating. Therefore, the proposed GP-index profile is obviously superior to the conventional linear profile, cubic profile and quintic profile, and the design methodology presented in this paper that uses a genetic programming technique is a quite convenient means to pursue an optimal nonlinear refractive index profile with broadband and omnidirectional antireflection characteristics.