延迟随机并行梯度下降算法用于相干合成中倾斜控制的数值模拟  

Numerical Simulation of Tilt-Tip Control in Coherent Beam Combining Using Delayed-Stochastic Parallel Gradient Descent Algorithm

作　　者：王雄[1] 王小林[1] 周朴[1] 许晓军[1] 舒柏宏[1] 

机构地区：[1]国防科学技术大学光电科学与工程学院,湖南长沙410073

出　　处：《中国激光》2013年第B12期25-32,共8页Chinese Journal of Lasers

摘　　要：对延迟随机并行梯度下降（D-SPGD）算法在远距离相干合成中倾斜控制进行系统研究，并与传统的随机并行梯度下降（SPGD）算法进行对比。理论研究结果表明，该算法可以明显降低传输延时对控制带宽的限制，显著提高远距离系统倾斜控制的效果，有望在远距离相干合成系统中得到实际应用。The systematic research on tilt-tip control in coherent beam combining using delayed-stochastic parallel gradient descent （D-SPGD）algorithm is demonstrated. The control effects are compared with those of the traditional stochastic parallel gradient descent （SPGD） algorithm. The theoretical research results reveal that the D-SPGD algorithm can weaken the limitation of time-delay on control bandwidth, and improve the tilt-tip control effect of long-distance system obviously. D-SPGD algorithm is a promising method to be applied in practical system of coherent beam combining over long distance.

关 键 词：激光光学 相干合成 倾斜波前 延迟随机并行梯度下降算法 主动控制 

分 类 号：TN248[电子电信—物理电子学]