Abstract:Using physics-informed neural networks to solve physical inverse problems is becoming a trend. However, it is difficult to solve the scheme that only introduces physical knowledge through the loss function. Constructing a reasonable loss function to make the results converge becomes a challenge. To address the challenge of physics-informed neural network models for inverse design of electromagnetic devices, a deep physics-informed neural network is introduced by using the mode matching method. The physical equations have been integrated into the network structure when the network is constructed. This feature makes the deep physics-informed neural network have a more concise loss function and higher computational efficiency when solving physical inverse problems. In addition, the training parameters of deep physics-informed neural networks are physically meaningful compared with those of traditional physics-informed neural networks. Users can control the network by parameters more easily. Taking the scattering parameter design of a two-port waveguide for example, we present a new metal topology inverse design scheme and give a detailed explanation. In numerical experiments, we target a set of physically realizable scattering parameters and inversely design the metallic septum by using a deep physics-informed neural network. The results show that the method can not only achieve the design target but also obtain solutions with different topologies. The establishment of multiple solutions is extremely valuable in implementing the inverse design. It can allow the designer to determine the size and location of the design area more freely while achieving the performance requirements. This scheme is expected to promote the application and development of the inverse design of electromagnetic devices.

Design of Plane Wave Generator Based on Strong Robust Inverse Radiation Problem Solving Method

A New Hybrid Equivalent Modeling Method of Low-Frequency Radiation Source Based on GS and JADE Algorithms and Phaseless Near-Field Data

Conjugate Adjoint Gradient-Based Inverse Design Method for Aperiodic Frequency-Selective Surface with Weighted Loss

Plane-Wave Synthesis: A Sparse Representation Perspective

Neural Network-Based Surrogate Model for Inverse Design of Metasurfaces

Modeling And Optimization Of Standing Wave Oscillator Via Geometric Programming

A Novel Pseudospectral Time Domain Method for Complex EM Simulations

Effective Plane Wave Generator Based on Reference Field Shaping using Super-Gaussian Distribution

Investigation of Array-based Plane Wave Generator for Compact Antenna Test Range Application

Design of an Array-based Plane Wave Generator for Compact Field Antenna Testing

Determining Aperture Field for Arbitrary Phaseless Far-Field Utilizing Inverse Design Method Based on Spectral Analysis

Inverse Design Method of Metal Structures in Multiport Waveguide Based on Numerical Green’s Function

A Hybrid Schatten P -Norm and Lp -Norm with Plane Wave Expansion Method for Near–Near Field Transformation

An Adaptive High-Order Transient Algorithm to Solve Large-Scale Anisotropic Maxwell’s Equations

An Effective Plane Wave Synthesis Method

Inverse Design of Microwave Waveguide Devices Based on Deep Physics-Informed Neural Networks

Phase-Informed Discontinuous Galerkin Method for Extremely High-Frequency Wave Modeling

Determining the Minimum Number of Plane Wave Generator Radiating Elements Based on SVD of the Transfer Matrix

Design of Anti-Phase Feeding Network for W8JK Array Based on In-Phase Power Divider

Near-Field Beam Focusing and Steering Generator Based on 3D Curved Substrate Integrated Waveguide

Waveform Optimization for Large-Scale Multi-Antenna Multi-Sine Wireless Power Transfer