CAJ | 학술논문

经验模态分解在处理非线性，多分量信号方面有其独特的优势，但是受信号采样率以及频程大小的影响，传统经验模态分解之后，各固有模态分解函数并非由单一频率成分组成。采用屏蔽信号和采样率内插相结合的方法来改进传统固有模态分解，得到的固有模态函数较传统EMD分解有很大改善。以改善分解后的固有模态函数为基础，利用合成孔径聚焦成像，成像结果可以很清楚的发现目标体的位置所在，证明了此种改进方法的可行性。
경험모태분해재처리비선성，다분량신호방면유기독특적우세，단시수신호채양솔이급빈정대소적영향，전통경험모태분해지후，각고유모태분해함수병비유단일빈솔성분조성。채용병폐신호화채양솔내삽상결합적방법래개진전통고유모태분해，득도적고유모태함수교전통EMD분해유흔대개선。이개선분해후적고유모태함수위기출，이용합성공경취초성상，성상결과가이흔청초적발현목표체적위치소재，증명료차충개진방법적가행성。
Empirical mode decomposition has its unique advantages in dealing with non-linear,multi-component signal.Subjected to the sampling rate and the fact that frequencies of the individual components usually lie close to each other or within an octave,each intrinsic mode function is not composed of a single frequency component.With the help of the combination effect of masking signal and sampling rate interpolation,the result of intrinsic mode functions is significantly improved relative to the traditional EMD decomposition.The im-proved decomposed intrinsic mode functions are used to image the defects inside a concrete structure with the synthetic aperture focusing technique.The location of the defects can clearly be found,and practice has proved the feasibility of this improved method.