基于模糊控制的波前校正技术

Wavefront Correction Technology Based on Fuzzy Control

模糊控制由于不依赖变形镜的响应模型,用于波前校正时具有实用性,其可行性已被证实。对自适应光学系统中模糊比例积分微分(PID)控制的波前校正效果进行了评估,包括对模糊校正带宽的测试、对模糊输入和输出论域的选取以及对模糊规则库以及波面加权模板的优化。实验结果表明,论域的选取合理,采用3×3的波前加权模板效果更好。当PID控制器3个参数矢量K_P、K_I和K_D的初始值都为0时,模糊输出论域采用隶属度正分布的规则库实用性更强。当哈特曼-夏克传感器刷新频率为120Hz时,系统校正带宽为5~6Hz,与常规PID一致。

Fuzzy control is used in wavefront correction and its feasibility has been confirmed because it does not depend on the response model of deformable mirror. The wavefront correction effect of fuzzy proportion integration differentiation （PID） control in adaptive optical system is assessed, including vesting fuzzy correction bandwidth, selecting fuzzy input and output domains, and optimizing fuzzy rule base and wavefront weighted template. Experimental results show that the selection of fuzzy domain is reasonable, and the better results are obtained with a 3 X 3 wavefront weighted template. When the initial values of Kp, KI and KD of PID controller are 0, the rule base with membership degree positive distribution is more suitable for the fuzzy output domain. When the refresh rate of the Hartmann-Shack sensor is 120 Hz, the system correction bandwidth is 5-6 Hz, which is consistent with the conventional PID.