A Neural Network Embeddable Algorithm Based on Inherent Characteristic Parameters for RCS Calculation of 3D Objects

The rapid development of intelligent computational electromagnetics (CEM) has the potential to overcome the difficulties of the great cost in calculation and memory for traditional CEM methods. However, most of the methods appeared in recent years perform poor in generalization and can only be used in specific applications. In this paper, a neural network (NN) embeddable algorithm base on inherent characteristic parameters (ICP) for the radar cross section (RCS) calculation of three-dimensional objects is proposed. By deploying the fact that ICP is only related to a perfectly electrically conducting (PEC) target's geometric properties (TGP), a NN can be designed to map the relationship between TGP and ICP. In this way, the ICP of any specific PEC target can be directly extracted using the well-trained NN and the monostatic or bistatic RCS can be calculated with high efficiency. Numerical results are presented to demonstrate the usefulness of the proposed method.