Utilizing Genetic Algorithms for Autonomous RF FEM Simulation & Optimization

V. Gjokaj
NuPhotonics LLC,
United States

Keywords: genetic algorithim, antennas, FEM solver


The huge influx of interest has sprawled out for Artificial Intelligence, this has sprawled many new businesses and opportunities for businesses to become more efficient and more productive. The short coming of Artificial Intelligence is the need for large amounts of datasets and when custom solutions are needed for RF technologies it becomes hard for the AI to develop a solution. We have developed and offer Genetic Algorithm (GA) optimization for custom solutions needed for our clients. Genetic Algorithms follow Darwin's rule of evolution, the GA takes in a base design and utilizes Finite Element Method solvers to see the characteristics (performance) of the structure and only passes the desirable traits into the next generation and repeats these mating traits until the desirable performance is achieved. This us allows us to charge considerably less for custom designs of antennas and RF structures as we do not need an RF engineer to optimize the structure.