Optimization Fitting of VCSEL L-I Characteristic Curve Based on Least Squares Method

This paper presents a novel curve fitting optimization model for modeling and optimizing the key parameters of the light intensity-current (L-I) characteristic curve of vertical cavity surface emitting lasers (VCSELs). The study constructs a progressive optimization model based on the application of optimization algorithms such as Adagrad, Adam, AdamW, and RMSProp, on the foundation of least squares fitting. By incorporating the temperature-dependent characteristics of thermal impedance and slope efficiency, as well as regularization methods such as L1, L2, and elastic net regularization, the fitting performance and generalization capability of the model are significantly enhanced. The experimental results demonstrate that the optimized model is capable of more accurately simulating the operating characteristics of VCSELs under different temperature conditions, thereby validating the effectiveness of the proposed method. The innovation of this study lies in the proposal of a nonlinear fitting framework that integrates optimization algorithms, and the enhancement of VCSEL modeling accuracy through temperature characteristic correction and regularization techniques, thereby more effectively simulating the light intensity-current operating characteristics of VCSELs.