Genetic Algorithm-Based Fast Real-Time Automatic Mode-Locked Fiber Laser

The nonlinear polarization evolution-based passively mode-locked fiber laser is attracting increasing attention that can be attributed to its variety of operating regimes. However, the precise polarization tuning required to achieve these different regimes and the extreme vulnerability of these lasers to environmental disturbances have substantially hindered their widespread applications. Here, we experimentally demonstrate the first genetic algorithm-based real-time automatic mode-locked fiber laser, in which the fitness functions for the different regimes are based on temporal information only and where a modified genetic algorithm is proposed to accelerate the mode-locking time. The laser demonstrates an outstanding time-consumption performance, particularly when searching the second-order harmonic mode-locking regime and the Q-switching regime.