Research On The Signal Detection And Information Extraction In EMC Site Test

With the rapid development of electronic technology, the complexity and density of electronic equipment has attached a level which have never seen before and the problems of Electromagnetic Compatibility(EMC) stand out increasing. The research of improving electromagnetic compatibility performance and testing methods is been more and more concerned. The Electromagnetic Compatibility test is always an effective way to have a good command of its EMC performance. The product-forming or the certification testing are usually completed within a standard laboratory, but with the increasing complexity of the device application environment and the increment of the application scale, in some cases, standard test has been unable to reflect the equipment EMC in real work environment, due to the huge system volume and the standard laboratory space is limited, the site test is essential. The site test faced the impact of the external environment and the peripheral equipment, has more difficulties to obtain the real emission information of the equipment, the comprehensively studying the EMC site test signal detection and the information extraction technology is of great significance, which can obtain the EMC of the system with the equipment and provide an important basis for fault diagnosis and information leakage protection.This paper regards the site test signal detection and the information extraction as the research target. First, the EMC site test and standard test are compared in test methods, test content and applicability etc. this paper analyzed the application advantage of the site test in the EMC test field and introduced the classification of the environment electromagnetic interference and the frequency distribution, this paper and then pointed out the site test existing problems by analyzing the limitation of the traditional EMI test.Secondly, according to the needs of the site test, a kind of hardware and software combination method has been adopted, this paper developed a site test program, adopted the improved portable test components, with the optimization of test equipment setup parameters, it solves the anti-saturation test and weak-signal test effectively and ensure maximum access to test information from the hardware side. According to the characteristic of the data acquisition from the site EMC test, with the reference of the time-domain signal processing method, this paper analyzed and compared thede-noising method, analyzed the synchronization cumulative noise elimination method and the Singular Value Decomposition(SVD) noise elimination method. The synchronization cumulative noise elimination method has simple calculation and high fault-tolerance, but multiple measurements are needed, it is suitable for the partial band test; SVD noise elimination method can more truly reflect the overall fluctuation of test data, the data obtained from SVD can be the threshold to extract the characteristic of EMC spectral test, it can improve the accuracy of information extraction. A pretreatment of site test spectral data method based on the synchronization cumulative noise elimination method and SVD noise elimination method has been proposed and the workflow of EMC testing data processing has been presented.In the final, this paper proposed characteristic extraction method based on the de-noising threshold, for the spectral characteristic test, mainly includes the peak information and broadband envelope characteristic, to distinguish different devices emission characteristic. It has effectively solved the peak error extraction or the leakage in the traditional characteristic extraction methods and the broadband envelope characteristic mistakenly truncation, thus provides the basis for the EMC fault analysis or leakage analysis, and then verified the validity of the hardware test system and the information extraction algorithms through examples in the end.