LSTM NEURAL NETWORK – ANALOG OF PROPORTIONAL INTEGRAL-DIFFERENTIAL CONTROLLER IN THE OPTICAL FEEDBACK CIRCUIT OF A SEMICONDUCTOR LASER

An effective method for narrowing the line width and stabilizing the generation of semiconductor lasers is the use of optical feedback. The main disadvantage of this method is its high sensitivity to fluctuations in the configuration parameters and the environment. To stabilize the laser operation, a proportional integro-differentiative (PID) algorithm is used. In this paper, an optimization option for the active optoelectronic feedback circuit is proposed, in which the PID algorithm is implemented based on the LSTM recurrent neural network.