Coupling And Alignment Algorithm Research On Array Waveguide Devices

Integrated photonic devices are key component of the next generation of all optical network communication system, which mainly used for the conversion and transmission of the optical signal. The demand for integrated photonic devices is growing rapidly with the development of optical communication system, and the requirement for the quality of the device is higher and higher. Currently, the manufacturing cost of the integrated photonic devices is mainly concentrated in the packaging process, as the alignment and coupling of the photonic devices is the key technologies of the packaging process, so as to reduces the cost and improve the quality of integrated photonic devices the proposed of new efficient alignment craft and search algorithm is of great significance.Firstly, a set of simple and reasonable manual packaging process is putted forward based on the distribution and coupling theory of single-mode fiber and waveguide; an automate packaging equipment is designed to solve the inefficiencies and high cost of manual packaging process and the automated packaging process is analyzed in accordance with the automated device.The paper studies the Particle Swarm algorithm in the alignment process application, the result of simulation analysis show that it is an excellent search algorithm; the distribution and coupling theory of single-mode fiber and waveguide is applied to the coupling and alignment process of array waveguide devices, and some alignment algorithm are proposed by this, include the alignment algorithm based on the analytical solution of the distribution model, the alignment algorithm based on the interpolation fitting distribution model, the alignment algorithm based on the least-squares fitting distribution model, then analysis the performance of each algorithm in the search process by the simulation result.Completed the experiments of the particle swarm algorithm and the alignment algorithm based on the distribution model, The experiment results show that the model-based algorithm can search to the target location with high coupling efficiency within20steps of platform move, but the stability of search results, the algorithm combine with least-squares and hill-climbing can overcome this shortcoming with good engineering application value.