针对自适应光学系统对波前处理计算量和实时性要求的提高, 提出了一种基于现场可编程门阵列(FPGA)的自适应光学系统波前处理算法。该算法利用核心处理模块重复利用的方式完成波前斜率计算, 利用矩阵与向量相乘的可分解性完成波前复原计算。在像素时钟的同步下, 完成整个波前处理, 给出促动器所需的促动量。以一片Virtex-4 LX80 FPGA作为主核心处理芯片进行了实验验证, 结果表明: 该算法可降低50%的硬件资源, 提高了系统波前处理能力; 另外, 算法可实现在当前帧结束前完成整个波前处理运算, 提高了系统的波前处理速度和整个自适应光学系统的控制带宽。在室内的Shack-Hartmann波前传感器的的自适应光学系统上进行了激光光源的校正实验, 结果显示光源能力集中度有了明显的提高。

A wave-front processing algorithm was proposed based on Field Programming Gate Array(FPGA) for improving computational complexity and real-time ability of adaptive optical systems. The reuse of core processing module was used to calculate a wave-front slope, and the decomposition of matrix and vector was used to calculate the wave-front restoration. Under the synchronization of pixel clock, the whole wave front was processed and the momentum required by an actuator was given. A piece of Virtex-4 LX80 FPGA was taken as a main chip to perform experimental verification, and the experiment results show that the algorithm reduces the hardware resources by 50%, and improves wave-front processing ability of the system. Furthermore, all of the wave-front processing operations can be implemented in the current frame before the processing end, which improves processing speeds and control bandwidths of the adaptive optic system. The new algorithm has carried on the correction experiment of the laser light source for a Shack-Hartmann in the room, and it shows a higher concentration ratio of light source.