Improving PMA and IO algorithms for annular Gaussian beam uniform illumination

Abstract

Beam shaping theory is naturally one of the inverse problems and it is unable to get a unique minimum resolution. In the case of high demands to target beam quality such as uniform illumination, or the complex style of incident beams, one still needs to improve the classical design algorithm to satisfy fabrications and applications. In this paper, new iteration algorithm based on phase mixture algorithm (PMA) and input-output (IO) algorithm is presented. By using random phase mixture factor instead of fixed phase mixture factor in PMA and random feedback factor in IO algorithm, and introducing a selection rule in each loop of the iteration, the degree of freedom of the iteration is increased and better target beam quality is obtained. A continuous diffractive optical element (DOE) for uniform illumination of annular Gaussian incident beam with diameter 240 mm is shown as a design example. Comparison of iteration results between the new algorithm and classical PMA or IO shows that the new algorithm provided better target beam quality and thinner DOE phase thickness. The design result of new algorithm only has 2.25% top profiled error (TPE) with a phase thickness of about 8 π while the best simulated result of PMA and IO algorithm has 3.42% TPE with a phase thickness of 12.4 π.