MoM Research On Electromagnetic Scattering Characteristics Of Periodic Structure And Design Of Metasurface Element

Due to its unique electromagnetic properties,the periodic structure is widely used in antenna design,microwave engineering,and optical engineering.The well-known applications include: phased array antenna,frequency selective surface,wave-guide filter,organic solar cell,photon Crystals and various meta-material structures.With the increasing application of periodic structures,the increasing complexity of structural requirements and the emergence of various array elements and composite structures,the analysis and design of electromagnetic scattering of periodic structures is becoming more and more complex,and the computational workload is increasing.Therefore,it is of great theoretical and practical significance to carry out more in-depth theoretical analysis of the periodic structure and to develop accurate and reliable periodic structure calculation methods to simulate the rapid and accurate design of periodic structures.The Moment method has been widely used as a high-precision method in computational electromagnetic simulation.Firstly,based on the physical meaning of the periodic Green’s function,the spatial and spectral domain expressions of the periodic Green’s function are derived.And the diagonal block prepossessing based on region decomposition is proposed.The method was applied to the periodic volume fraction equation(VIE)to effectively analyze the electromagnetic properties of thin film organic solar cells.However,the Ewald method has a disadvantage:each additive summation term must be written in an exponential closed form,which is difficult to directly apply to the calculation of the periodic Green’s function in a layered medium.In order to overcome this difficulty,this paper applies the Ewald method combined with the Shank transform method to calculate the electromagnetic properties of the metasurface,and for the slow calculation of pix-elated metasurface elements,this paper uses the impedance matrix and the right vector pumping row.Innovative methods quickly and accurately calculate their electromagnetic properties.The correctness and applicability of the method are verified by an example.Aiming at the shortcomings of traditional optimization algorithm combined with full-wave simulation to optimize the periodic structure’s low efficiency and slow calculation,this paper proposes two artificial intelligence methods for the design optimization of 1-bit metasurface unit structure and finally achieves the goal of reducing RCS in the frequency band.For the optimized design of pix-elated metasurface elements,we use the Convolutional Neural Network(CNN)classification method to establish the relationship between the surface pattern of the element and the phase of the electromagnetic reflection.Finally,we combine the binary bat optimization algorithm to find a 1-bit unit with a constant phase difference.And the diffuse reflection front is formed by means of arrays,and finally the RCS reduction in the3.8GHz-10.7GHz wide frequency band is more than 10 dB.For the CNN,the training sample capacity is large,and it takes time and classification accuracy to construct the network model.With lower disadvantages,this paper proposes a convex structure metasurface unit,which is combined with Support Vector Machine(SVM)regression method and genetic optimization algorithm to finally optimize the 1-bit super-surface unit.Finally,through the array,the index of RCS reduction of 10 dB in the 3.5GHz-10.6GHz band is achieved.