Control Algorithm Of The Model-based Wavefront Sensorless Adaptive Optics System

Due to the complex atmospheric structure and changeable motion state,it occurs the turbulence effect such as wavefront distortion and intensity fluctuation,as a result the imaging quality of the telescope is greatly reduced when observing the astronomical target.The conventional adaptive optics(AO)technology has always been an effective means to correct the turbulence in the atmosphere.However,the conventional AO technology is limited by the observation of the low elevation angle and the near surface of the astronomical target,which makes the wavefront detection have great error.Compared with the conventional AO technique,the wavefront sensorless AO technology eliminates a wavefront sensor,which makes the system a more simple structure and reduces the cost of the system,then avoids the distortion by scintillation effect in atmospheric turbulence,the system can still work.What's more,in the common used control algorithms based on the model,the control algorithm used in this paper is based on the linesr relationship between the masked detector signal(MDS)and the mean square gradient(MSG),which does not depend on the concrete basis function model and does not need to remove the system aberration,it has great potential in real time wavefront correction for its fast convergence and good correction effect.This paper carries out the experimental research and analysis of linear relationship control algorithm based on 37-element AO system,which chooses point target as experimental object.First,an AO system experiment platform is established with a 37-element piezoelectric deformable mirror(DM)?a CCD camera?a Shack Hartmann sensor,in which the DM generates wavefront aberration,the CCD camera grabs the distribution of far-field intensity and the Hartmann gets the information of wavefront.Through experiments and numerical analysis method,experiment results show MDS is linear to MSG and the slop is 0.018,which is very close to the theoretical value 1/(4?~2).Then,the software platform of AO system is built,and the correction ability of control algorithm is analyzed.The results show that the control algorithm can converge quickly and effectively reduce the aberration of wavefront.Finally,compared with the stochastic parallel gradient descent(SPGD)algorithm,it has the advantage of relatively faster convergence.