Preliminary Study of a Hybrid Genetic Algorithm/Expert System for Modeling Complex Radar Signatures

摘要

I made an initial study of a hybrid genetic algorithm/expert system (HGAES) to model targets with nonlinear radar imaging effects caused by features such as cavities and canopies. The model for the nonlinear parameters was relatively simple, so it should be suitable for incorporation into hardware-in-the-loop and software-in-the-loop simulations that currently use point scatter models. I demonstrated the algorithm on simulated two-dimensional (2-D) inverse synthetic aperture radar (ISAR) images using a simple technique to determine the initial scattering centers. Many of the ideas used in developing the algorithm can be extended to more complex targets and from 2-D to 3-D images. A major issue in the development of an HGAES is knowledge representation. My conclusions are that models determined using this technique have the potential to enhance the accuracy of weapon systems simulations; thus, this technique is worth further investigation.