Flat broadband frequency upconversion within a thin-film lithium niobate waveguide achieved by multi-objective genetic algorithm particle swarm optimization

In the field of nonlinear infrared frequency upconversion within a poled thin film lithium niobate (TFLN) waveguide for spectroscopy, there is a persistent demand for achieving a flat broadband response, characterized by the minimal variation in output intensity across the desired wavelength range. We propose a design method that significantly broadens the spectral bandwidth and enhances the response flatness through multi-objective genetic algorithm particle swarm optimization (GAPSO). This approach minimizes human intervention in the optimization process, thereby enhancing efficiency and accuracy compared to traditional methods that depend on preset parameters. Compared to the traditional chirped periodically poled TFLN waveguide-based infrared frequency upconversion scheme, a remarkable spectral bandwidth expansion from 180 nm to 312 nm (a 73% increase) and an improved flatness from 1.71 dB to 0.56 dB (a reduction of over 67%) is achieved. This work not only paves the way for a more efficient flat broadband infrared frequency upconversion scheme but also opens new avenues for advancements in nonlinear optical applications, such as telecommunications and sensing technologies.