Alignment Of Integrated Photonic Devices Based On Genetic Algorithm

More and more attention has been paid to the integrated photonic devices due to its capability of optical information processing for large optical system as well as compact structure, anti-interference, and good consistency. At present, the chip produce and packing restrict the development of integrated photonic devices, while alignment technology is the key in the integrated photonic devices packaging process and it is the basis of functional. It has great significance to find an alignment algorithm with high-precision, high speed, low movement at the coupling of integrated photonic devices.First of all, single-mode optical field distribution of fiber array model is established based on array waveguide and fiber array field distribution theory. The single-mode optical fiber array field distribution experiments are finished at the AA Modal Integrated photonic devices platform. Based on experiment results, the field distribution of single mode fiber array and waveguide array is analyzed, which is the foundation for alignment coupling technology. It is established that fiber array aligned with the waveguide array coupling model based on optical waveguide theory, and it is analyzed how offsets of horizontal, vertical and angle impact the planar waveguide devices alignment coupling, which will help the packing experiments.Based on single-mode optical fiber array alignment and array waveguide coupling model, it is completed that the simulation of hill climbing algorithm, simplex algorithm, pattern algorithm, genetic algorithm and its improved algorithm. Analysis the advantages and disadvantages of each algorithm in integrated photonic devices alignment is finished, and, according to the results of the analysis, the complex algorithm based on improved genetic algorithm is advanced.The experiments of fiber array aligned with planar waveguide are done at the AA Modal Integrated photonic devices platform. Based on the results of experiments and simulations, the complex algorithm based on improved genetic algorithm and pattern algorithm is advanced, which reduces 81.38% platform movement and 77.43% search time compared with hill climbing algorithm, and achieves alignment 100% successfully.