Research On Reconstruction Of Electromagnetic Radiation Sources Based On Planar Near-Field Scanning And Its Applications

The electromagnetic radiation sources localization and electromagnetic radiation evaluation in electromagnetic compatibility(EMC)are studied and explored in this dissertation.In order to localize electromagnetic radiation sources,a theoretical model characterizing the relationship between the current sources above the ground plane and the radiation near field,is constructed.Furthermore,a radiation source model based on numerical Green’s function is proposed to reconstruct the radiation sources in complex medium environments.To accelerate the reconstruction of the radiation sources,on the one hand,a sparse solver that only needs fewer near-field samples,is utilized to quickly solve the radiation source models.On the other hand,an adaptive sampling strategy is proposed to efficiently collect the near-field information of radiation sources.The current image obtained by solving the radiation source model accurately represents the real current distribution.Therefore,it can be used to localize the radiation sources and predict their radiation.Specifically,the research content includes the following four parts:Firstly,a new electromagnetic radiation source model enlightened by the traditional dipole-moment model,is proposed to characterize the relationship between current sources and their radiation field.In the proposed model,the physical currents are first decomposed into horizontal current and vertical current according to their directions.Source planes are selected on the planes where the horizontal and vertical currents exist.Then,some electric dipoles are placed on the source planes to replace all currents.Finally,the half-space Green’s function,which characterizes the relationship between the current sources above the ground plane and their radiation near field,is used to construct the radiation source model.Taking advantage of the sparse nature of the currents,the radiation source model can be solved using a sparse solver FOCUSS,which effectively reduces the number of required near-field samples.The results of numerical experiments demonstrate that the the proposed source model is effective and reliable.Secondly,according to the physical phenomenon that the field changes rapidly near the strong radiation sources,an adaptive sampling strategy is proposed.Firstly,a set of initial uniformly sampled data is used to solve the radiation source model,and a rough radiation source is obtained.Then,a region growing method is used to segment the source data based on the amplitude values’ similarity of the radiation source.After that,multiple strong source points are identified.New sampling points are adaptively added around the identified strong source points.Finally,the adaptively added samples and initial uniformly sampled data are used to solve radiation source model.The above adaptive sampling process continues until the coordinates of all strong source points converge.The experimental results show that compared with the traditional sampling method,the proposed sampling strategy is faster and has lower time cost.Thirdly,considering that the half-space Green’s function can not accurately characterize the relationship between source and field in complex medium,a radiation source model based on numerical Green’s function is proposed to reconstruct the radiation sources in medium environment.Different from the traditional layered medium Green’s function,the numerical Green’s function uses simulation tools to directly calculate the relationship between source and field in the complex medium.Firstly,model of the ground plane and the dielectric substrate(it has the same dimensions as the object that needs to be reconstructed)is constructed in the simulation software.Then,the source plane(the plane where radiation sources exist)is discretized into uniform grids,and a unit electric dipole is successively placed on each grid to simulate the radiation field information.Finally,the radiation source model based on numerical Green’s function is constructed by arranging all the radiation field data according to the location relationship between the field points and the electric dipole source points.Furthermore,an adaptive sampling method based on matrix approximation and volume sampling is used to reduce the number of simulations during the construction of numerical Green’s function and accelerate the establishment of the radiation source model.Finally,the planar near-field scanning which is the most widely used near-field measurement technology is studied in this dissertation.The radiation source models mentioned above are all solved by the data of planar near-field scanning.The study of planar near-field scanning lays a foundation for the practical application of these proposed radiation source models.Firstly,the classical technologies of near field to far field transformation and aperture field inversion based on plane wave spectrum expansion are studied.Based on these technologies,the tangential electric field components on any near-field plane of a strong directional radiator can be collected to calculate the far-field pattern and aperture field distribution of the radiator.Then,different from the traditional method of obtaining the receiving pattern of the probe using a large number of simulations,this dissertation proposes a fast method to calculate the receiving pattern of the probe.This proposed method only needs one simulation,which effectively improves the computational efficiency of probe’s receiving pattern.Finally,the probe error is corrected by using the computed probe receiving pattern,which effectively controls the error of the near field test system itself and ensures the accuracy of the near-field measurement data.This dissertation systematically studies the electromagnetic radiation source reconstruction and its applications in engineering.By using the adaptive sampling strategy and the numerical Green’s function,the solution of radiation source model is accelerated,and the practicability and application range of these radiation source models are extended.These research contents provide the guarantee of accuracy and efficiency for the extensive application of electromagnetic radiation source reconstruction in EMC.