Intelligent Laser Emitting and Mode Locking of Solid‐State Lasers Using Human‐Like Algorithms

Abstract

Solid‐state lasers based on mode‐locked technology are widely studied for their strong ability of average or peak power scaling and wide wavelength coverage. However, it usually takes a long time and a lot of effort to manually align an ultrafast solid‐state laser to achieve laser emission and stable mode‐locking. Here, an approach based on intelligent human‐like algorithms is proposed for aligning a solid‐state ultrafast laser system. The intelligent system is based on multi‐algorithm fusion and can fully simulate the process of observation, analysis, decision, and action of an experienced experimenter in the adjustment of the laser alignment and mode‐locking. The intelligent adjustment starts from the state of no laser emission, adopts the neural network, the modified augmented random search (ARS) algorithm, random search, and sliding window strategy, and takes the fluorescence and speckle patterns as indications to realize the laser emission and stable mode‐locking automatically. Several validation experiments are conducted using this intelligent system, and the stable mode‐locked pulses can be achieved within 40 s. This technology provides an efficient solution to the ultrafast solid‐state laser that requires full automatic laser emitting and stable mode‐locking for the first time.