Implementation of Genetic Algorithm in Polarization Control System

In this study, a polarization control system with an embedded genetic algorithm has been developed. The objective is to implement endless and accurate control with adaptive characteristic. To achieve this, Genetic Algorithm is used which consumes low size of memory besides providing high speed. To have an economic system, the PIC32MX microcontroller with 72 MHz speed is chosen. Throughout this study different genetic parameters have been examined to determine best genetic parameters for achieving the fastest execution time. The best genetic parameters which are population size, selection method, crossover probability and mutation probability are defined as 8×16 matrix, tournament method, 0.7 and 0.01 respectively. Referring to the earlier designs , this structures performs better with good flexibility and speed due to the genetic algorithm. Truly endless and adaptive operation is confirmed and this system is able to stabilize the polarization changes in 17 µs and 300 µs in the best and the worst case respectively. Apart from adaptive operation, this solution is also more economic in comparison to the FPGA solution with Look-Up-Table.

Autorentext

Ghazaleh Mamdoohi is a Postdoctoral Research Fellow at Shahid Bahonar University of Iran. Her research interest focuses on Third-order Optical Non-linearities, Optical Amplifier, Artificial Intelligent Algorithm and Polarization Controller.