摘要:
自适应光学中普遍采用的比例-积分-微分(PID)控制严格依赖于变形镜的响应模型,响应模型的标定非常繁琐且性能容易随时间和温度而漂移。采用模糊控制摆脱了对变形镜响应模型的依赖,通过波面的快速重构,从波面中提取波面评价指标,将该评价指标及其微分作为模糊PID控制器的输入。选择模糊输入和输出论域的隶属度函数,设计出模糊规则库,通过模糊推理、解模糊等过程,对PID控制器的3个参数即比例kp、积分k i和微分k d进行模糊自整定,实现了自适应控制。采用高速数字信号处理器(DSP)设计了硬件验证测试平台,通过该平台的测试,证明该方法有效可行,在不需要变形镜响应矩阵的前提下,能够将光斑的衍射极限倍增因子β值从10—12校正到3—4,与传统的PID控制基本一致,但稳定性更好。因无需标定,故降低了变形镜的安装要求。
Abstract:
In an adaptive optics system, proportion-integration-differentiation (PID) controller is widely used for correcting wave front, but the controller is strictly dependent on the response model of deformable mirror. In this paper, a novel wave front correction method is proposed. The method, combining fuzzy control and PID control, does not depend on the response model of the deformable mirror. Based on rapid wave front reconstruction, the wave front evaluation indexes, extracted from the reconstructed wave front, are employed for the input of fuzzy controller and PID controller. Thus, the model response matrix of deformable mirror is not required. Each actuator of deformable mirror corresponds to an independent fuzzy PID controller. By designing the fuzzy controller, including fuzzy rule base selection and fuzzy reasoning, the three parameters of PID controller, the proportional kp, the integral ki and the differential kd, are adjusted automatically. A high rapid DSP hardware platform is constructed to verify the method. Test results show that the method can be used to correct the diffraction limit multiplication factorβ of the light spot from 10–12 to 3–4, which is basically the same as the traditional PID control, but its stability is better. Because the model does not need to calibrate the deformable mirror, the installation of the deformable mirror is easier.
首页
登录
注册
下载: