基于NLPCA和JADE算法的EMI辐射源噪声诊断方法研究
发布时间：2018-09-11 20:29
【摘要】：随着大规模集成电路、高频开关元件以及无线通讯模块在电子设备中的广泛使用,大量电子辐射噪声随之产生,并以空间电磁场的形式向外传播,不仅影响设备自身的正常工作,而且还会通过电磁场辐射的形式干扰外部设备的工作。目前的电磁兼容问题分析方法大多是从理论和建模方面分析辐射问题,并不能对噪声源特征进行判定和诊断,且由于模型的理想化,导致了结果与实际有着一定的误差,因此复杂电子设备电磁辐射噪声源的识别和诊断成为了电磁兼容辐射分析中的一大挑战。针对上述问题,本文研究了“基于NLPCA和JADE算法的EMI辐射源噪声诊断方法研究”,具体包含以下内容:(1)从天线辐射理论出发,通过对复杂电子设备辐射机理的分析,针对实际电路中辐射线缆长度为电大尺寸时信号传播存在时延以及线路阻抗失配时电路存在反向行波这两种情况建立了适用性广泛的辐射卷积噪声模型。(2)研究复杂电子设备多辐射源的定位方法。利用NLPCA和JADE算法分别从信号的时域和频域角度对多辐射源辐射噪声进行分析,获得若干个独立的噪声信号,作为辐射噪声超标源头的备选信号。利用功率谱对分离后的独立辐射电磁干扰噪声信号进行分析,提取出各信号的频率特征,通过与PCB板的疑似辐射源进行比对,从而诊断出引起辐射超标的主要原因。(3)进行三组验证性实验,通过对双共模模型、双差模模型和共差模混合模型的噪声信号盲分离,并对分离后的独立信号进行能量频谱分析,找出辐射超标的原因,验证了该定位方法的可行性。最后通过两个实际的复杂电子产品进行案例分析,验证了该方法实际应用的有效性,成功找出了辐射超标的原因,并根据超标原因采取对应的整改,使电子设备通过国标测试。通过理论的分析研究和实验结果的验证,本文提出的定位方法对复杂电子设备的电磁兼容多辐射源信号分离研究有着重要的借鉴价值,而且对电磁兼容的整改工作有着积极的指导意义。
[Abstract]:With the widespread use of large scale integrated circuits, high-frequency switching elements and wireless communication modules in electronic devices, a large amount of electronic radiation noise is generated and propagates outward in the form of space electromagnetic fields. It not only affects the normal operation of the equipment itself, but also interferes with the work of the external equipment through electromagnetic radiation. At present, most of the analysis methods of EMC are based on theory and modeling, which can not judge and diagnose the characteristics of noise source, and because of the idealization of the model, there is a certain error between the results and the reality. Therefore, the identification and diagnosis of electromagnetic radiation noise sources in complex electronic equipment has become a major challenge in EMC radiation analysis. In order to solve the above problems, this paper studies the method of noise diagnosis of EMI radiation source based on NLPCA and JADE algorithm, which includes the following contents: (1) based on the antenna radiation theory, the radiation mechanism of complex electronic equipment is analyzed. In this paper, a widely applicable radiation convolution noise model is established for the real circuit in which the signal propagation delay exists when the length of the radiation cable is electrically large and the circuit has reverse traveling wave when the line impedance mismatches. (2) the complex convolution noise model is studied. Location method of multiple radiation sources for miscellaneous electronic equipment. NLPCA and JADE algorithms are used to analyze the radiated noise of multiple radiation sources in the time domain and frequency domain respectively, and several independent noise signals are obtained, which are used as alternative signals for the radiation noise to exceed the standard source. The independent radiated electromagnetic interference (EMI) noise signal after separation is analyzed by power spectrum, and the frequency characteristics of each signal are extracted, and compared with the suspected radiation source of the PCB board. Thus, the main causes of radiation exceeding the standard are diagnosed. (3) three groups of confirmatory experiments are carried out. The noise signals of the two common mode models, the double differential modes model and the common-difference mode mixed model are separated blindly, and the energy spectrum of the separated independent signals is analyzed. The reason of radiation exceeding the standard is found, and the feasibility of this method is verified. Finally, through the case analysis of two real complex electronic products, the validity of the method is verified, the reason of radiation exceeding the standard is found out successfully, and the corresponding rectification is taken according to the reason of exceeding the standard, so that the electronic equipment can pass the national standard test. Based on the theoretical analysis and experimental results, the localization method proposed in this paper has important reference value for the study of EMC multi-emitter signal separation in complex electronic equipment. Moreover, it has positive guiding significance to the rectification and reform of EMC.
【学位授予单位】：南京师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN03
本文编号：2237768
[Abstract]:With the widespread use of large scale integrated circuits, high-frequency switching elements and wireless communication modules in electronic devices, a large amount of electronic radiation noise is generated and propagates outward in the form of space electromagnetic fields. It not only affects the normal operation of the equipment itself, but also interferes with the work of the external equipment through electromagnetic radiation. At present, most of the analysis methods of EMC are based on theory and modeling, which can not judge and diagnose the characteristics of noise source, and because of the idealization of the model, there is a certain error between the results and the reality. Therefore, the identification and diagnosis of electromagnetic radiation noise sources in complex electronic equipment has become a major challenge in EMC radiation analysis. In order to solve the above problems, this paper studies the method of noise diagnosis of EMI radiation source based on NLPCA and JADE algorithm, which includes the following contents: (1) based on the antenna radiation theory, the radiation mechanism of complex electronic equipment is analyzed. In this paper, a widely applicable radiation convolution noise model is established for the real circuit in which the signal propagation delay exists when the length of the radiation cable is electrically large and the circuit has reverse traveling wave when the line impedance mismatches. (2) the complex convolution noise model is studied. Location method of multiple radiation sources for miscellaneous electronic equipment. NLPCA and JADE algorithms are used to analyze the radiated noise of multiple radiation sources in the time domain and frequency domain respectively, and several independent noise signals are obtained, which are used as alternative signals for the radiation noise to exceed the standard source. The independent radiated electromagnetic interference (EMI) noise signal after separation is analyzed by power spectrum, and the frequency characteristics of each signal are extracted, and compared with the suspected radiation source of the PCB board. Thus, the main causes of radiation exceeding the standard are diagnosed. (3) three groups of confirmatory experiments are carried out. The noise signals of the two common mode models, the double differential modes model and the common-difference mode mixed model are separated blindly, and the energy spectrum of the separated independent signals is analyzed. The reason of radiation exceeding the standard is found, and the feasibility of this method is verified. Finally, through the case analysis of two real complex electronic products, the validity of the method is verified, the reason of radiation exceeding the standard is found out successfully, and the corresponding rectification is taken according to the reason of exceeding the standard, so that the electronic equipment can pass the national standard test. Based on the theoretical analysis and experimental results, the localization method proposed in this paper has important reference value for the study of EMC multi-emitter signal separation in complex electronic equipment. Moreover, it has positive guiding significance to the rectification and reform of EMC.
【学位授予单位】：南京师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN03
【参考文献】
相关期刊论文 前10条
1 陈庆彬;陈为;;开关电源中变压器共模传导噪声抑制能力的评估方法[J];中国电机工程学报;2012年18期
2 王世山;朱叶;崔永生;;基于“负参数”理论的EMI滤波器寄生参数的双重消除[J];电工技术学报;2011年10期
3 赵阳;颜伟;邱晓晖;;电力载波通讯中辐射噪声的分析与抑制[J];中国矿业大学学报;2011年01期
4 赵阳;颜伟;赵波;罗永超;李世锦;;电路辐射干扰机理诊断与特性估计[J];电工技术学报;2010年10期
5 赵阳;董颖华;陆ZK泉;姜宁秋;颜伟;;EMI噪声分离网络在电力线噪声分析中的应用[J];中国电机工程学报;2010年21期
6 赵阳;罗永超;颜伟;赵波;李世锦;陆ZK泉;邱晓晖;;高频电路辐射干扰快速分析与预估方法[J];电波科学学报;2010年03期
7 任小永;姚凯;旷建军;阮新波;;损耗最小化输出滤波电感的设计[J];中国电机工程学报;2008年27期
8 孟进;马伟明;张磊;潘启军;赵治华;;带整流桥输入级的开关电源差模干扰特性[J];电工技术学报;2006年08期
9 陈文洁;杨旭;王兆安;;集成模块用有源电磁干扰滤波器的设计[J];中国电机工程学报;2005年24期
10 和军平,陈为,姜建国;开关电源共模传导干扰模型的研究[J];中国电机工程学报;2005年08期
相关硕士学位论文 前2条
1 夏欢;基于ICA和STFFT算法的辐射EMI噪声分析与诊断方法[D];南京师范大学;2015年
2 平云娟;基于盲源分离的辐射EMI噪声分析与诊断方法研究[D];南京邮电大学;2011年
，本文编号：2237768
本文链接：https://www.wllwen.com/kejilunwen/dianzigongchenglunwen/2237768.html
