Adaptive Optics Correction Using Recurrent Neural Networks for Wavefront Prediction.

Adaptive Optics is a field aimed to improve the quality of the images received by terrestrial telescopes by the use of optical instrumentation, although it heavily relies on different techniques to control it. Neural networks have proven to be versatile in many different situations, therefore, implementing them to Adaptive Optics is the next reasonable step. Similar to weather forecasting, the present paper focuses on the implementation of neural networks in prediction of next stages of the atmosphere. Presented results are comparable with those with traditional systems but with neural networks being cheaper and easier to implement.