Research On The Compound Optimization Algorithm Of The Phase Hologram Of Liquid Crystal On Silicon

Spatial light modulator of liquid crystal on silicon plays an important role in many fields such as optical information processing and adaptive optics.Phase holograms are the key to liquid crystals on silicon and their optimal generation is usually realized by the classical iterative Fourier transform algorithm.However,when the phase hologram obtained by traditional algorithm is used for diffraction calculation,the obtained results are not completely satisfactory,and there is still large room for improvement in diffraction efficiency and similarity of light field.In this paper,the generation of phase holograms of liquid crystals on silicon is studied theoretically and experimentally by combining traditional algorithm with intelligent optimization algorithm.The main tasks of the thesis are as follows1.Based on the analysis of the basic structure and working principle of liquid crystal on silicon,the existing optimization algorithms are analyzed in detail for the 1920x1080 pixel liquid crystal on silicon chip,and the trait of each algorithm are summarized.The iterative Fourier transform algorithm is selected as the preposition algorithm of compound optimization and the intelligent optimization algorithm as the compound method of postposition algorithm2.On the premise of improving the design of the composite algorithm,three application scenarios with different focuses are designed,namely,equal proportion beam splitting,non-equal proportion beam splitting and super-gaussian beam shaping,and the simulation research of each composite algorithm is carried out respectively.It is concluded that the compound effect of the differential evolution algorithm is the best for the traditional algorithm,and the compound effect of the iterative Fourier transform algorithm and the differential evolution algorithm can often reduce the root mean square error of the preoptimization results by more than 2 orders of magnitude.3.Designed and completed the experiment of non-equal proportion beam splitting,and verified the above research conclusion.When the GS algorithm is combined with the differential evolution algorithm,the RMSE is reduced by 2 orders of magnitude.By combining SIFTA algorithm with differential evolution algorithm,the RMSE is reduced by 3 orders of magnitude.It is shown that the exper:imental results are basically consistent with the simulation results,and the compound algorithm improves the traditional algorithm to some extent.In this paper,the phase recovery algorithm for liquid crystal on silicon applications is simulated and experimentally studied,and the original algorithm is optimized and improved by combining the traditional iterative Fourier transform algorithm with the intelligent optimization algorithm,and the compound effect of the iterative Fourier transform algorithm and the differential evolution algorithm is obtained.