Volume 51 Issue 9
Sep.  2022
Turn off MathJax
Article Contents

Kang Ding, Wang Chunyang, Wang Zishuo, Wang Zeng, Zheng Qingquan. Optimization method of diffraction efficiency of liquid crystal spatial light modulator based on spline interpolation[J]. Infrared and Laser Engineering, 2022, 51(9): 20210827. doi: 10.3788/IRLA20210827
Citation: Kang Ding, Wang Chunyang, Wang Zishuo, Wang Zeng, Zheng Qingquan. Optimization method of diffraction efficiency of liquid crystal spatial light modulator based on spline interpolation[J]. Infrared and Laser Engineering, 2022, 51(9): 20210827. doi: 10.3788/IRLA20210827

Optimization method of diffraction efficiency of liquid crystal spatial light modulator based on spline interpolation

doi: 10.3788/IRLA20210827
Funds:  China Postdoctoral Science Foundation(2020 M673606 XB)
  • Received Date: 2021-11-05
  • Rev Recd Date: 2022-01-16
  • Accepted Date: 2022-02-09
  • Publish Date: 2022-09-28
  • To reduce the diffraction efficiency of liquid crystal spatial light modulator, an optimization method of diffraction efficiency of liquid crystal spatial light modulator based on spline interpolation is proposed. According to Tyman-Green interference principle, the phase modulation system is built. Loading on the modulator with a grayscale map of step change, the phase modulation curve of liquid crystal spatial light modulator is drawn by calculating the movement of interference fringes. The cubic spline inverse interpolation method is used to correct the phase modulation curve and to realize the compensation of the phase modulation amount. The diffraction efficiency test system of liquid crystal spatial light modulator is set up, and the experimental verification of the proposed optimization method is carried out and compared with the stochastic gradient descent method. The results show that when the beam deflection angle is 1.56°, 0.78°, 0.39° and 0.19°, the diffraction efficiency of the proposed method can increase by 30% to 40%. Compared with the stochastic parallel gradient descent method, the diffraction efficiency increases by 2%-8%. This method can effectively suppress the gate lobe energy, improve the diffraction efficiency of the main lobe beam, and overcome the disadvantages of the stochastic parallel gradient descent method, such as a number of iterations, slow optimization speed, and easy to fall into local optimal.
  • [1] Niu Qifeng, Wang Chunyang. High precision beam steering using a liquid crystal spatial light modulator [J]. Optical and Quantum Electronics, 2019, 51(6): 180. doi:  10.1007/s11082-019-1858-0
    [2] Huang Dajie, Fan Wei, Zhang Panzheng, et al. Development progress and performance test of optically addressed spatial light modulator [J]. Infrared and Laser Engineering, 2016, 45(9): 0920001. (in Chinese) doi:  10.3788/IRLA201645.0920001
    [3] Wei Chaojie, Yan Renpeng, Li Xudong, et al. Research progress of sub-nanosecond lasers for 3D imaging lidar [J]. Optics and Precision Engineering, 2021, 29(6): 1270-1280. (in Chinese) doi:  10.37188/OPE.20212906.1270
    [4] Wei Yu, Jiang Shilei, Sun Guobin, et al. Design of solid-state array laser radar receiving optical system [J]. Chinese Optics, 2020, 13(3): 517-526. (in Chinese) doi:  10.3788/CO.2019-0166
    [5] Qiu Jiawei, Zhang Zhen, Yu Saifen, et al. Development of 1.5 μm lidar for atmospheric detection (Invited) [J]. Infrared and Laser Engineering, 2021, 50(3): 20210079. (in Chinese) doi:  10.3788/IRLA20210079
    [6] Winker B, Mahajan M, Hunwardsen M. Liquid crystal beam directors for airborne free-space optical communi-cations[C]//IEEE Aerospace Conference Proceedings, Big Sky, 2004: 631-634.
    [7] Fu Qiang, Jiang Huilin, Wang Xiaoman, et al. Research status and development trend of space laser communication [J]. Chinese Optics, 2012, 5(2): 116-125. (in Chinese) doi:  10.3788/CO.20120502.0116
    [8] Tian Dapeng, Shao Xiaopeng. New emerging technologies in airborne optical imaging and measurement [J]. Optics and Precision Engineering, 2020, 28(6): 1221-1225. (in Chinese) doi:  10.3788/OPE.20202806.1221
    [9] Pan Guotao, Yan Yufeng, Yu Xin, et al. Design of optical system for quality evaluation of large rectangular aperture laser beam [J]. Chinese Optics, 2022, 15(2): 306-317. (in Chinese) doi:  10.37188/CO.2021-0130
    [10] Fang Zhou, Xu Xiangxiang, Li Xin, et al. SPGD algorithm with adaptive gain [J]. Infrared and Laser Engineering, 2020, 49(10): 20200274. (in Chinese) doi:  10.3788/IRLA20200274
    [11] Dayton D, Gonglewski J, Browne S. Control algorithms of liquid crystal phased arrays used as adaptive optic correctors[C]//Proceedings of SPIE, 2006, 6306: 630604.
    [12] Lin X, Jian Z, Li Y W. Influence of phase delay profile on diffraction efficiency of liquid crystal optical phased array [J]. Optics & Laser Technology, 2009, 41(4): 509-516.
    [13] Kong L, Lin Z, Feng X, et al. Modeling and analysis of liquid crystal optical phased array [J]. Chinese Optics Letters, 2012, 10(s2): S20501. doi:  10.3788/COL201210.S20501
    [14] Barrick G F, Bos P J, Titus C E, et al. Computing the liquid crystal director field in optical phased arrays [J]. Optical Engineering, 2004, 43(4): 924-932. doi:  10.1117/1.1666821
    [15] Harris S R, Gonglewski J D, Vorontsov M A, et al. Numerical optimization of the performance of nematic liquid crystal optical phased arrays[C]//Proc SPIE, 2003, 5162(11): 157-171.
    [16] Xiao F, Kong L, Chen J. Beam-steering efficiency optimization method based on a rapid-search algorithm for liquid crystal optical phased array [J]. Appl Opt, 2017, 56(16): 4585-4590. doi:  10.1364/AO.56.004585
    [17] Niu Qifeng. Research on technology of liquid crystal phased array beam deflection[D]. Changchun: Changchun University of Science and Technology, 2020. (in Chinese)
    [18] Huang Zhiwei. Study on beam control of liquid crystal optical phased array[D]. Changchun: Changchun University of Science and Technology, 2018. (in Chinese)
    [19] Rong Z Y, Han Y J, Wang S Z, et al. Generation of arbitrary vector beams with cascaded liquid crystal spatial light modulators [J]. Optics Express, 2014, 22(2): 1636-1644. doi:  10.1364/OE.22.001636
    [20] Yong Y, Zhou Q, Tian F, et al. Research on characteristics of reflective liquid crystal spatial light modulator [J]. Journal of Information & Computational Science, 2014, 11(7): 2331-2340.
    [21] Bogdanov V, Volkov Y. Shape-preservation conditions for cubic spline interpolation [J]. Siberian Advances in Mathematics, 2019, 29(1): 231-262.
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(12)  / Tables(2)

Article Metrics

Article views(114) PDF downloads(28) Cited by()

Related
Proportional views

Optimization method of diffraction efficiency of liquid crystal spatial light modulator based on spline interpolation

doi: 10.3788/IRLA20210827
  • 1. College of Ordnance Science and Technology, Xi’an Technological University, Xi’an 710021, China
  • 2. Xi'an Key Laboratory of Active Photoelectric Imaging Detection Technology1, Xi'an 710021, China
  • 3. College of Electronic and Information Engineering, Changchun University of Science and Technology, Changchun 130022, China
Fund Project:  China Postdoctoral Science Foundation(2020 M673606 XB)

Abstract: To reduce the diffraction efficiency of liquid crystal spatial light modulator, an optimization method of diffraction efficiency of liquid crystal spatial light modulator based on spline interpolation is proposed. According to Tyman-Green interference principle, the phase modulation system is built. Loading on the modulator with a grayscale map of step change, the phase modulation curve of liquid crystal spatial light modulator is drawn by calculating the movement of interference fringes. The cubic spline inverse interpolation method is used to correct the phase modulation curve and to realize the compensation of the phase modulation amount. The diffraction efficiency test system of liquid crystal spatial light modulator is set up, and the experimental verification of the proposed optimization method is carried out and compared with the stochastic gradient descent method. The results show that when the beam deflection angle is 1.56°, 0.78°, 0.39° and 0.19°, the diffraction efficiency of the proposed method can increase by 30% to 40%. Compared with the stochastic parallel gradient descent method, the diffraction efficiency increases by 2%-8%. This method can effectively suppress the gate lobe energy, improve the diffraction efficiency of the main lobe beam, and overcome the disadvantages of the stochastic parallel gradient descent method, such as a number of iterations, slow optimization speed, and easy to fall into local optimal.

    • 液晶空间光调制器是一种能够实现光束非机械式偏转的器件[1-2]。通过对空间光调制器的各个阵元施加不同的驱动电压,可以实现不同角度的光束偏转。由于液晶空间光调制器具有可随机偏转、功耗低、电控可编程、稳定性好等优点,在激光雷达[3-6]、空间光通信[7]、光学成像[8]和自适应光学[9-11]等领域有着广泛的应用。在激光相控阵雷达中,光束的衍射效率是一项最重要的系统性能指标,衍射效率过低会造成雷达探测距离下降和目标检测概率降低。受液晶盒厚度、相位凹陷、回程区以及电压量化等因素影响,液晶空间光调制器阵元间相位差不能保持一致,导致衍射效率低于理论值[12-13]

      近年来,国内外诸多学者已针对液晶空间光调制器的光束衍射效率优化方法开展了相关研究。Barrick G F等人结合时域有限差分(Finite Difference Time Domain, FDTD)法对液晶空间光调制器的衍射效率进行了优化,在入射光波长为1550 nm,光束偏转角度为15 mrad时,优化后衍射效率高于90%[14]。Harris S利用LabVIEW软件的单纯形法对衍射效率进行优化,优化后衍射效率相较优化前提高12%[15]。肖锋等人提出基于随机并行梯度下降法的衍射效率优化方法,通过实验验证该优化算法在液晶空间光调制器周期N=8,偏转角度为1.524°时,将衍射效率从优化前的37.76%提高到71.21%[16]。牛启凤等人采用Rosenbrock旋转坐标法进行了仿真优化,结果表明迭代2 000次左右满足优化终止条件,且通过实验验证经该算法优化后,入射激光为532 nm,光束偏转角度为0.052°时,衍射效率可以达到93.1%[17]。黄志伟等人分别对两种闪耀光栅模型下液晶空间光调制器衍射效率进行优化。在周期闪耀光栅模型下采用基于模式搜索法的衍射效率优化方法进行优化,当周期N=12时,优化后衍射效率达到82%。在非周期闪耀光栅模型下,采用基于蝙蝠算法的衍射效率优化方法,对阵元个数为360,偏转角度为2°时的驱动电压进行优化,仿真结果表明优化后衍射效率从优化前的53.84%提升到了64.76%[18]

      综上所述,基于时域有限差分法的优化方法需要建立不同角度下的电压模型,并采用时域有限差分法进行求解,优化过程复杂,很难应用于实际工程中。基于单纯形法、随机并行梯度下降法、Rosenbrock旋转坐标法、模式搜索法以及蝙蝠算法等优化方法都是通过迭代搜索最优驱动电压的方式对衍射效率进行优化,迭代次数多,优化速度慢,容易陷入局部最优 [14-19]。因此,文中针对工程应用中常用的迭代优化方法存在的问题提出了基于样条插值的衍射效率优化方法。依据泰曼-格林干涉原理,搭建相位调制测量系统,对液晶空间光调制器加载阶梯变化的灰度图,通过计算干涉条纹移动量,绘制液晶空间光调制器输入灰度与相位差的关系曲线,采用三次样条反插值法对所制相位调制曲线进行校正,使液晶空间光调制器输入灰度与对应相位差接近理想的线性关系,实现对相位调制量的相位补偿,生成新的灰度和电压的对应查找表(Look Up Table,LUT)。该方法不需要多次迭代寻找最优电压,只需将新生成的LUT表加载到液晶空间光调制器中,使输入灰度与相位调制量呈线性关系,保证液晶空间光调制器在加载不同灰度图时相邻阵元间相位差保持一致。文中所提优化方法解决了迭代优化算法中存在的迭代次数多,优化速度慢,易陷入局部最优的问题,有效抑制了栅瓣能量,提升了主瓣光束衍射效率。且所提优化方法易于在实际工程中应用,为液晶空间光调制器在激光雷达、空间光通信等领域的使用提供了有力保障。

    • 液晶空间光调制器中每个电极和它所控区域的液晶组成一个相控阵元,当激光入射到液晶空间光调制器时,通过对各个阵元施加不同的驱动电压,使液晶分子发生旋转,改变液晶的有效折射率,进而改变出射面上的光相位分布,实现2π相位调制,使光在指定方向上形成远场干涉加强,实现光束偏转[16-20]。液晶空间光调制器相位调制示意图如图1所示。

      Figure 1.  Schematic diagram of phase modulation of liquid crystal spatial light modulator

      计算机将一张包含相位信息的256级灰度图传送给驱动单元,驱动单元根据液晶空间光调制器加载的LUT表设置每个阵元的驱动电压,最终液晶空间光调制器精准地呈现出所需的相位分布,实现光束指向控制。液晶空间光调制器指定偏转角度远场光强分布可以表示为[17]

      式中:$\theta $为光束偏转角度;${U_1}(\theta )$表示远场光束复振幅分布;${U^*}_1(\theta )$$ {U_1}(\theta )$的共轭复数;a表示阵元宽度;$\Delta \phi $为相邻阵元间相位差;d为液晶盒厚度;$u = \sin (\theta )/\lambda $$\lambda $为激光波长。

      根据公式(1),液晶空间光调制器m级次衍射效率表示为[17]

      结合公式(1)、(2),当液晶空间光调制器阵元宽度a,液晶盒厚度d,激光波长与光束偏转角度确定时,m级次光强受液晶空间光调制器相邻阵元间相位差$ \Delta \phi $影响。由于液晶空间光调制器是通过离散电极进行驱动,驱动电压占空比不足100%,产生相位凹陷,再加之液晶分子的边缘效应以及制造工艺误差,综合导致$ \Delta \phi $产生误差,造成相邻阵元间相位差不能保持一致,该衍射级次上的光束被分散,降低光束衍射效率。

    • 依据泰曼-格林光学干涉原理搭建液晶空间光调制器相位调制测量系统,系统光路原理如图2(a)所示,实验装置图如图2(b)所示。

      Figure 2.  (a) Schematic diagram of the Tyman-Green light path; (b) Diagram of experimental device for phase modulation measurement system

      在相位调制量测量实验中,光源为1064 nm圆偏振光,扩束准直器束腰处光斑直径为8.0 mm,分光棱镜透反比为50∶50。CCD相机型号为FASTCAM Mini UX50,最大分辨率为1280×1024 pixel,像素尺寸10 μm。液晶空间光调制器为美国BNS公司的反射式相位型液晶空间光调制器,型号为HSP1920-850-1650-HSP8,响应波长为500~1650 nm,由1920×1152个电极阵元构成,其面阵大小为17.6 mm×10.7 mm,填充因数是95.7%,阵元宽度为9.2 μm,阵元间距为6.4 μm,液晶分子长轴方向为水平方向。实验具体步骤如下:

      (1) 调节偏振片,将圆偏光调节为线偏光,偏振方向为水平方向;

      (2) 调节分光棱镜使激光保持垂直入射;

      (3) 调整反射镜及液晶空间光调制器,使分光棱镜分出的两束光分别垂直入射到反光镜和液晶空间光调制器中;

      (4) 对液晶空间光调制器加载阶梯变化的灰度图;

      (5) 通过CCD相机采集反射镜和液晶空间光调制器反射光束的干涉条纹;

      (6) 根据干涉条纹计算液晶空间光调制器相位调制量,并绘制相位调制曲线。

      部分阶梯变化灰度图如图3所示,每幅图像上半部分的灰度值设为0,作为计算相位改变量的参考;图像的下半部分,灰度值从0递增到255,使液晶空间光调制器产生相应的相位延迟,光程差发生变化,产生有相对移动的干涉条纹。

      Figure 3.  (a) Gray scale of the lower half is 0; (b) Gray scale of the lower half is 125; (c) Gray scale of the lower half is 255

      通过CCD相机可以观察到如图4所示干涉条纹。为提高计算准确度,对条纹宽度进行10次测量取均值。由于实验所用液晶空间光调制器要求入射线偏光偏振方向与液晶分子保持严格一致,且整个实验需要保证激光光束始终垂直入射,CCD相机也要保持垂直角度采集,因此受光源偏振态、光束入射角度以及CCD相机采集角度的综合影响,干涉条纹右侧出现了错位和宽度不一致的现象,但如果保持始终测量同一条纹的移动量就不影响计算干涉条纹移动量与条纹间距的比例,所以对后续相位差的计算不造成影响。

      Figure 4.  (a) Interference fringe pattern when the gray level is 0; (b) Interference fringe pattern when the gray level is 125; (c) Interference fringe pattern when the gray level is 255

      图4所测干涉条纹可以观察到上半部分干涉条纹保持不变,下半部分干涉条纹发生偏移,且由图4(c)发现条纹偏移量已达到一个周期,即已完成2π的相位调制。

      光程差与干涉条纹移动量的关系为:

      式中:h是两光路的光程差;λ是激光波长;Δ是干涉条纹的移动量;Λ是干涉条纹的宽度。相位$ \delta $与干涉条纹的移动量的关系则为:

      式中:$ \delta $表示干涉条纹移动量对应的相位差。

      通过计算条纹间距与条纹错位长度,结合公式(3)、(4),计算不同灰度下产生的相位差。描点绘制原始相位调制曲线,如图5所示,从调制曲线可以看出,液晶空间光调制器加载的驱动灰度与相位调制量呈非线性变化趋势,灰度在0~120变化时调制相位起伏较大,在后半部分变化较为平缓。

      Figure 5.  Original phase modulation curve

      为实现2π相位调制,使周期内阵元构成较理想的相位面,需要保证相邻阵元间相位差相同。由公式(1)可知,当灰度与相位差为线性对应关系时,阵元间相位差$ {x_i} $保持一致,光束能量可以全部汇集在指定偏转角度上,此时光束衍射效率理论上达到最优。若灰度与相位差为非线性关系时,液晶空间光调制器相邻阵元间相位差$ {x_i} $不能保持一致,从而降低了光束在指定级次上的衍射效率。因此,对其开展基于样条插值的校正研究,补偿对应驱动灰度的相位调制量,使其呈线性关系,保证相邻阵元间达到相同的相位差,使液晶空间光调制器在加载灰度图后产生较为理想的相位面,实现衍射效率的优化。

    • 针对所测相位调制曲线的非线性关系,采用样条插值法对其进行校正。利用图5所示相位差与驱动灰度关系,反求满足线性关系的灰度,既已知输入灰度$ {x_i} $上的相位差$ y = f({x_i}) $,存在反函数$ x = {f^{ - 1}}({y_i}) $,以相位差$ {y_i} $为反插值节点,求解输入灰度$ x = g(y) $。为保证原始相位调制曲线的特性,选用具有保凸特性的三次样条插值法建立输入灰度与驱动灰度之间的关系。三次样条插值表达式为[21]

      式中:$ S(x) $为三次样条插值函数,表示输入灰度与驱动灰度关系;$ {f_i} $是灰度$ {x_i} $对应的相位差,既${f_i} = f({x_i}), (i = 0,1,2,\cdots,n)$${h_i} = {x_{i + 1}} - {x_i} (i = 0,1,2,\cdots,n)$$ {M_{\text{i}}} = S''({x_i}) $为待定参数。

      选择非扭结边界作为插值函数的边界条件,如公式(6)所示:

      式中:$ S_0^{'''}({x_0}) $$ S_1^{'''}({x_1}) $表示前两个灰度值的三阶导数;$ {\kern 1 pt} S_{n - 2}^{'''}({x_{n - 1}}) $$ S_{n - 2}^{'''}({x_n}) $表示最后两个灰度值的三阶导数。

      三次样条插值函数$ S(x) $在满足非扭结边界外,还需要满足以下条件:

      (1) 函数$ S(x) $在区间$ [{x_{i - 1}},{x_i}]{\kern 1 pt} {\kern 1 pt} {\kern 1 pt} ,{\kern 1 pt} {\kern 1 pt} {\kern 1 pt} i = 1,2,\cdots,n $上至多为三次幂;

      (2) $ S(x) $$ {S'}(x) $$ {S{''}}(x) $在插值节点区间内连续;

      (3) $ S({x_{\text{i}}}) = {y_i}{\kern 1 pt} {\kern 1 pt} {\kern 1 pt} ,{\kern 1 pt} {\kern 1 pt} {\kern 1 pt} i = 0,1,2,\cdots,n - 1 $

      在原始相位调制曲线中,相位完成一个变化周期对应的驱动灰度值在0~255的范围内,可将该段区间的曲线记为:

      式中:$ {\varphi _i} $表示测量相位;$ {g_i} $表示驱动灰度。

      为实现相位补偿,首先构造空间光调制器输入灰度与相位调制量的理想线性函数为:

      式中:$ \varphi ' $表示相位差,取值区间为$0 \sim 2\pi$$ g' $为输入灰度。

      利用所测灰度对应的相位差,对函数$ \varphi $进行三次样条反插值,可以得到驱动灰度$ g $与输入灰度$ g' $之间的关系,如图6所示。

      Figure 6.  Relationship between enter grayscale and drive grayscale

      根据图5中实验所测驱动灰度与相位差的关系和图6中输入灰度与驱动灰度的关系,求得液晶空间光调制器不同相位差对应校正后输入灰度如表1所示。

      Phase difference (π)Original grayscaleCorrected grayscale
      000
      0.21226
      0.42051
      0.64676
      0.876102
      192127
      1.2109153
      1.4132178
      1.6161204
      1.8205229
      2255255

      Table 1.  Phase difference of gray before and after correction

      根据表1绘制如图7所示的校正后相位调制曲线。从图6中可以看出:校正后相位调制曲线接近理想的线性关系,受测量误差影响,在输入灰度为30左右时存在较小误差,但整体液晶空间光调制器相位差与灰度有较好的线性关系。结合图5图6所示的经三次样条插值后所得的输入灰度与驱动灰度的关系以及校正后的相位调制关系生成新的LUT表,并加载到液晶空间光调制器中。

      Figure 7.  Contrast diagram of ideal phase and corrected phase modulation curve

    • 为验证所提方法对液晶空间光调制器衍射效率优化的有效性,文中搭建了如图8所示的衍射效率测试系统,图8(a)为系统原理图,图8(b)为系统实验装置图。并与随机并行梯度下降法进行对比分析,验证所提方法的优越性,随机并行梯度下降法性能评价函数值设为0.8,迭代次数设定为1000次。实验具体步骤如下:

      Figure 8.  (a) Diagram of experimental light path; (b) Graph of wxperimental scene

      (1) 调节偏振片,将圆偏光调整为线偏光,偏振方向为水平方向;

      (2) 为液晶空间光调制器加载不同相位差的灰度图;

      (3) 调节CCD相机,捕获远场光强分布;

      (4) 分析远场光强分布图,计算光束衍射效率。

      部分灰度图如图9所示。

      Figure 9.  (a) $ \Delta \phi $=$ \pi {\text{/2}} $, $ \theta $=1.56°; (b) $ \Delta \phi $=$ \pi {\text{/4}} $, $ \theta $=0.78°; (c) $ \Delta \phi $=$ \pi {\text{/8}} $, $ \theta $=0.39°

      在液晶空间光调制器衍射效率测试系统中分别测试光束偏转不同角度时光束衍射效率,测试结果如表2所示。

      Grating period/(°)Not optimizedStochastic parallel gradient descentThe proposed method
      1.5631.76%68.13%76.41%
      0.7838.80%81.61%83.89%
      0.5244.32%85.72%87.63%
      0.3949.91%87.08%90.19%
      0.3153.28%89.13%91.83%
      0.2655.91%90.64%92.37%
      0.2257.85%91.32%93.25%
      0.1958.93%91.63%93.84%

      Table 2.  Diffraction efficiency of beam at different deflection angles

      通过CCD相机观测液晶空间光调制器光束偏转时的远场光斑强度分布,并对远场光强进行归一化处理。光束偏转1.56°时,得到如图10的优化结果及光强分布图。

      Figure 10.  (a) Not optimized; (b) Stochastic parallel gradient descent method; (c) The proposed method

      图10(a)为未优化时光斑及光强分布结果,(b)为随机并行梯度下降法优化后结果,(c)为所提方法优化后的结果。由于未优化时液晶空间光调制器驱动灰度与相位差成非线性关系,相邻阵元间相位差不能保持一致,造成光束分散,光斑能量不集中,严重影响了指定级次光束的衍射效率。应用随机并行梯度下降法优化后,指定级次光束衍射效率有所提高,但优化后依旧存在旁瓣。所提优化方法由于直接校正液晶空间光调制器的LUT表,使得驱动灰度与相位差近似理想的线性关系,所以在加载灰度图后衍射效率有明显的提升,旁瓣数量减少,能量减弱。

      Figure 11.  (a) Not optimized; (b) Stochastic parallel gradient descent method; (c) The proposed method

      光束偏转角度为0.78°时实验结果如图11所示,当偏转角度为0.78°时,周期内增至8个阵元,受未优化时相位差与灰度的非线性关系影响,8个阵元间相位差不能一致保持为$ \pi {\text{/4}} $,造成旁瓣数量明显增多,指定衍射级次光束衍射效率较低。应用所提方法对原始相位调制量进行补偿优化后,相位差与灰度基本呈线性关系,并生成新的LUT表加载到液晶空间光调制器中,使液晶空间光调制器阵元间相位差基本保持在$ \pi {\text{/4}} $,有效抑制了栅瓣,衍射效率相较于未优化时提高了45.09%,比随机并行梯度下降法衍射效率提高了2.28%。

      图12为光束偏转0.39°时优化结果及远场光强分布图,由于偏转角度减小,周期内阵元数增至16,实验结果如图12所示,经所提方法优化后衍射效率达到90.19%,比未优化时提高了40.28%,相较随机并行梯度下降法提高了3.11%。光束偏转0.19°时,应用所提优化方法优化后,指定级次光束衍射效率如表2所示,所提方法优化后衍射效率为93.84%,分别比未优化和随机并行梯度下降法优化后衍射效率提高了34.91%和2.21%。

      Figure 12.  (a) Not optimized; (b) Stochastic parallel gradient descent method; (c) The proposed method

      对比分析表2中实验结果以及图10~12中远场光强分布图可以看出,未优化时由于驱动灰度与相位差的非线性关系,随着光束偏转角度减小,周期数增多,栅瓣数量也随之增多,极大地影响了光束衍射效率。基于随机并行梯度下降法的优化方法对一级光强有所提高,但对旁瓣强度抑制效果较差,栅瓣抑制效果不佳。所提优化方法有效抑制了一级光斑两侧旁瓣,栅瓣数量明显减少,光束能量基本汇聚在一级光斑处,光束衍射效率有明显的提高。实验结果显示所提优化方法能有效优化液晶空间光调制器光束衍射效率,优化效果优于随机并行梯度下降法等现有算法,且对不同偏转角度的光束衍射效率都有较好的优化效果。

      所提方法只需要对液晶空间光调制器实测相位调制曲线进行校正,更新调制器中的LUT表,使加载的灰度与形成相位差基本呈线性关系,保证光束偏转时相邻阵元间相位差相同,实现衍射效率的优化,无需通过迭代方式搜索最优驱动电压。在实际应用中,此方法可用于优化不同相位型液晶空间光调制器衍射效率,该方法为离线优化模式,需要在使用液晶空间光调制器之前应用所提方法对液晶空间光调制器相位调制曲线进行补偿校正。完成校正优化后,液晶空间光调制器中驱动灰度与相位基本呈线性关系,直接根据光束偏转角度加载对应阶梯变化灰度图,即可保证相邻阵元间相位差基本一致,有效提高光束衍射效率。

    • 文中采用三次样条反插值法对液晶空间光调制器原始相位调制曲线进行校正,实现对液晶空间光调制器相位调制量的相位补偿,使校正后输入灰度与相位调制量接近理想的线性关系,依据输入灰度与驱动灰度的关系以及校正后的相位调制关系生成新的LUT表,将其加载到液晶空间光调制器中,实现不同偏转角度光束衍射效率的优化。实验结果表明,所提方法无需多次迭代搜索最优驱动电压就可以有效提高液晶空间光调制器光束衍射效率,优化效果优于随机并行梯度下降法,且所提方法针对不同角度都有较好的优化效果。应用所提优化方法优化后,光束偏转0.78°时衍射效率达到83.89%,比优化前提高了45.09%,相较于随机并行梯度下降法提高了2.28%;偏转角度为0.19°时,衍射效率达到93.84%,比优化前提高了34.91%,相较于随机并行梯度下降法提高了2.21%。

Reference (21)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return