An Approach to GaN HEMT Modeling Based on Genetic Algorithm

In this paper, we used Angelov model to fit large signal drain-source current I
ds
of six GaN HEMT devices with different gate widths and gate lengths, and used Matlab platform to build Genetic Algorithm (GA) to extract parameters. The results showed that the fitting accuracy was as high as 99.6% after using GA. The algorithm has the characteristics of fast convergence speed, high accuracy and strong reusability. At the same time, the parameter extraction can be completed in batches with the assistance of Matlab platform, which greatly improves the modeling efficiency.