Hill-climbing algorithm based on Zernike modes for wavefront sensorless adaptive optics

To solve the local optimum problem, a modified hill-climbing algorithm based on Zernike modes is presented for wavefront sensorless adaptive optics. This algorithm adopts the Zernike mode coefficients, instead of the actuators' voltages in a traditional hill-climbing algorithm, as the adjustable variables to optimize the object function. First, the prin- ciple of the algorithm is described. Then the efficiency of the modified and traditional hill-climbing algorithms are analyzed numerically and experi- mentally using a He-Ne laser beam shaping system with a 37-element unimorph deformable mirror. The results of both simulations and experi- ments demonstrate that the modified hill-climbing algorithm can eliminate the local optimum problem with a fast speed of about 100 iterations. © 2013 Society of Photo-Optical Instrumentation Engineers (SPIE). (DOI: 10.1117/1.OE.52.1.016601) Subject terms: wavefront sensorless; adaptive optics; hill-climbing algorithm; Zernike modes; deformable mirror.