Font Size: a A A

Reconstruction Algorithm Of Electromagnetic Interference Source Based On Near-Field Scanning

Posted on:2020-11-30Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y F ShuFull Text:PDF
GTID:1360330578973954Subject:Electronic Science and Technology
Abstract/Summary:PDF Full Text Request
With modern electronic products continuely developing toward high speed,high density,low power consumption,miniaturization,etc.,their electromagnetic compatibility problems have become increasingly serious.In the face of such complex electronic products,conventional full-wave simulation methods have been awkward to analyze their electromagnetic compatibility problems accurately and efficiently,because full-wave methods are very inefficient in dealing with complex multi-scale problems.On the other hand,the detailed structure of some devices can not been obtained due to the intellectual property protection.Near-field scanning technique has shown great advantages in such awkward situations.This scanning technique can initially diagnose and locate electromagnetic interference problems without the details of the device structure.In addition,the electromagnetic interference source reconstruction algorithm based on near-field scanning can quickly quantify and analyze the electromagnetic interference problem in complex systems by constructing the equivalent source model of noisy devices.This technique provides a powerful theoretical tool for electromagnetic interference risk assessment layout design in the early stage of electronic products,and electromagnetic interference modification,greatly reducing the research and development costs of electronic products,shortening the product development cycle,and becoming research hot point in industrial and academic community.Our work is closely related to the real engineering,and carries out related technical research on how to construct the equivalent radiation source model of electromagnetic interference source based on near-field scanning technology accurately and efficiently.The main contents of this paper include the following three points:1.Aiming at the low efficiency of near-field scanning in practical engineering,this paper proposes an iterative algorithm based on single-plane phaseless scanning,which inspire by the convensional double-plane iterative algorithm.The algonthm successfully iterates out a reliable equivalent radiation source model by using interpolated amplitude and the original amplitude of the scanning plane.Compared to the conventional double-plane iterative algorithm,this method reduces the scan time by half while maintaining accuracy.Both simulation and experiment examples demonstrate the effectiveness of the method In addition,we also verified the robustness of the algorithm with noise-added data.This algorithm compromises accuracy and efficiency and is suitable for electromagnetic interference problem analysis in the early stage of architecture design.2.Aiming at the problem of modeling the equivalent radiation source with multi-reflection or diffraction effects,we propose a reconstruction algorithm based on artificial neural network.The algorithm takes multiple reflection or diffraction effects into account,which effectively improves the reconstruction accuracy of the equivalent radiation source,providing a new solution for the analysis and diagnosis of electromagnetic interference problems in complex environments.The simulation and experiment examples not only prove that the method can predict the field distribution above the scanning surface,but also have some abilities to predict the field distribution below the scanning surface.It is meaningful for locating the electromagneti interference source.3.Aiming at the efficiency problem of multi-frequency modeling of equivalent radiation source in engineering,we propose a differential evolution reconstruction algorithm with fixed equivalent dipole position based on prior information.In this case,the equivalent dipoles with different dipole moments can reuse the mesh in full-wave simulation,which is the basis of broadband modeling.It can be used to quickly estimate coupling power at multiple frequencies.We use the algorithm to perform equivalent radiation source reconstruction on an actual high-speed chip,and estimate the coupling power between the chip and the surrounding transmission line,which provides quantitative guidance for the next chip layout.
Keywords/Search Tags:Near-field scanning, equivalent radiation source reconstruction, electromagnetic interference, phase retrieval, deep learning, artificial neural network, differential evolution algorithm
PDF Full Text Request
Related items