Prof. Lilin Yi’s Work on Intelligent Mode Locked Fiber Laser Selected for OPN’s Optics in 2019

Date: 29 November 2019

Human-like algorithm for passive MLFL, the work by Prof. Lilin Yi in the Department of Electronics Engineering is selected in  Optics in 2019 , a special issue of Optics & Photonics News (OPN), the news magazine of the Optical Society of America (OSA). For the past 37 years, the OSA has been inviting world-renowed experts to select the breakthroughs in the field of optics published over the past year. The selected researches are featured as a special issue in OPN every December. The work on intelligent mode locked fiber laser by Prof. Yi’s research group, published in Optica (IF: 9.263) in March 2019, is among the selected researches in this year’s Optics in 2019.

 

 

OSA-OPN Optics in 2019

Ultrafast lasers are one of the most advanced and active research directions in the field of optoelectronic technology, and are also the basis and starting point of ultrafast optical system researches. Ultrashort pulses generated by ultrafast lasers are used as basic tools in studies of physics, chemistry, biology, and information science. Mode-locking is the main method of generating ultrashort pulses including active mode-locking, passive mode-locking, and hybrid mode-locking. Among them, passive mode-locking based on nonlinear polarization evolution (NPE) is favored by academia because of its simple structure and excellent pulse performance. In NPE-based mode locked lasers, mode-locking can be achieved by adjusting the polarization controllers in the laser cavity. However, manual polarization control takes a long time to achieve mode locking and has a poor repeatability. Besides, the mode-locking regime in the NPE-based mode locked laser is extremely vulnerable to environmental disturbances. Once the laser detaches from the mode-locking regime, it is difficult to recover back to the mode-locking regime. Therefore, the industrial applications of the NPE-based mode locked lasers are very limited.

Prof Yi’s team has been working on the above problem for five years, and finally proposed and implemented an intelligent mode locked laser based on a human-like algorithm (HLA) (Optica [6(3): 362, 2019]). The paper was ranked #1 in the top download list on the March issue of Optica.

 

The configuration of intelligent mode-locked fiber laser

Inspired from the manual mode-locking tuning process, HLA aims to combine the speed, resolution and precision of machines with the human’s logic. Benefited from the direct and high-efficient human’s logic, HLA is less complex, which is more in line with the real-time requirement in mode-locking applications.

Under the HLA approach, the automatic mode-locking process from the turning on to the fundamental mode-locking regime took merely 0.22s and the recovering from the detachment took only 14.8ms, which substantially refresh  their previous records respectively. Through the different discrimination criteria established via extracting and analyzing the characteristics of different pulsation regimes, the HLA-based intelligent mode locked fiber laser is capable of automatically locking onto various regimes containing the fundamental mode-locking regime, the harmonic mode-locking regimes, the Q-switching regimes and the Q-switched mode-locking regimes. Further, the laser can switch among these pulsation regimes as required.

Discriminating different pulsation regimes requires no expensive devices including oscilloscopes and optical spectrum analyzers as in traditional methods. It can be achieved by a control circuit with an FPGA, a low-speed ADC and low-speed DACs. Therefore, the intelligent mode locked fiber laser can be designed as a portable module with low cost which meets the requirements of industrial applications.

The research is partly supported by National Natural Science Foundation of China (61575122).

 

Reference:

https://www.osa-opn.org/home/articles/volume_30/december_2019/features/optics_in_2019/

 

https://www.osa-opn.org/home/articles/volume_30/december_2019/extras/human-like_algorithm_for_passive_mlfl/