Microwave-Circuit Optimization with Parallel Enhanced Fast Messy Genetic Algorithm (pefmGA)

Abstract

Fast messy genetic optimization is found suitable for complex microwave circuit design. Increase in computation speed is achieved using several ordinary computers connected to a network. Calculations are running on background so that computers can be used for other purposes at the same time. Dynamic change of bounds, search space segmentation and gradient incorporation have significantly improved convergence rate. The new method has found global minimum in each run, while classic methods failed for some starting points.