论文部分内容阅读
高能激光系统中通常包含光束净化装置以对激光器出射光束的波前畸变进行连续校正。为研究随机并行梯度下降自适应光学方法在光束净化问题上的可行性,按算法运行时序连续改变37单元变形镜的面形,以在实验光路中引入高能激光器输出光束的常见动态波前畸变,同时采用随机并行梯度下降算法控制同一变形镜对此动态畸变进行校正。实验结果显示,在事先消除系统初始像差的情况下,系统分别以1 kHz和2 kHz的模拟迭代速率工作时,激光束的动态像差都得到了充分抑制,光束质量在整个校正时段内始终维持在较好水平。这表明随机并行梯度下降自适应光学方法应用在光束净化系统中是可行的。
High-energy laser systems usually include a beam cleaning device to continuously correct the wavefront distortion of the laser exit beam. In order to study the feasibility of stochastic parallel gradient descent adaptive optics method on the beam cleaning problem, the surface shape of 37-unit deformable mirror was changed continuously according to the algorithm running sequence to introduce the common dynamic wavefront distortion of the output beam of high-energy laser in the experimental optical path, At the same time, a random parallel gradient descent algorithm is used to control the same anamorphic mirror to correct the dynamic distortion. The experimental results show that the dynamic aberrations of the laser beam are fully suppressed when the system is operated at simulated iteration rates of 1 kHz and 2 kHz, respectively, without prior system initial aberrations, and the beam quality is constant throughout the correction period Maintain at a good level. This shows that the stochastic parallel gradient descent adaptive optics method is feasible in beam cleaning system.