Recent Advances in Wide Range Parametric Modeling

The development of an electromagnetic (EM) parametric model can be challenging when the geometrical parameters of the model differ significantly. To overcome this issue, a paper reviewed an automatic system algorithm that utilizes second-order derivative information. This method involves dividing the entire geometrical region based on its degree of smoothness, creating sub-regions for both flat and uneven areas. Each sub-region's geometrical parameters are treated as variables, and a neural network model is established to represent the sub-region, which is called the sub-model. However, moving from one sub-model to an adjacent one results in a different EM response at the boundary, causing multidimensional discontinuity. The paper proposes a solution to this problem, allowing for a continuous overall model.