CAJ | 학술논문

经验模态分解（empiricalmodedecomposition，简称EMD）的端点效应使得EMD分解结果产生严重失真，为了减小分解过程中产生的端点效应，将支持向量机（SVM）这一智能算法引入EMD，提出采用SVM模型解决分解中产生的端点效应问题．通过支持向量机对其原始数据两端进行延拓，以获得一个或者多个极大值和极小值．为了使端点处的延拓变得更加合理，引入粒子群（PSO）智能算法对支持向量机算法参数进行优化，使其两个端点处的数据延拓得更加准确，从而使得三次样条曲线在端点处不会发生大的摆动，实现EMD分解的固有模态函数（IMF）更加准确可靠．通过对仿真信号的研究表明，基于PSO—SVM方法的延拓方法能够很好地抑制了分解的端点效应．
경험모태분해（empiricalmodedecomposition，간칭EMD）적단점효응사득EMD분해결과산생엄중실진，위료감소분해과정중산생적단점효응，장지지향량궤（SVM）저일지능산법인입EMD，제출채용SVM모형해결분해중산생적단점효응문제．통과지지향량궤대기원시수거량단진행연탁，이획득일개혹자다개겁대치화겁소치．위료사단점처적연탁변득경가합리，인입입자군（PSO）지능산법대지지향량궤산법삼수진행우화，사기량개단점처적수거연탁득경가준학，종이사득삼차양조곡선재단점처불회발생대적파동，실현EMD분해적고유모태함수（IMF）경가준학가고．통과대방진신호적연구표명，기우PSO—SVM방법적연탁방법능구흔호지억제료분해적단점효응．
End effects of EMD （empirical mode decomposition） make a serious distortion of the decom- position result. In order to reduce the end effects in the process of decomposition, support vector machine （SVM） which is a kind of intelligent algorithm is combined with EMD, then a solution to the end effects problem during the course of decomposition using SVM model is proposed. Firstly, one or more extreme values are obtained by extending two endpoints of the original data with SVM. Moreover, in order to get more reasonable extension at endpoint, SVM algorithm is combined with particle swarm algorithm （PSO） to optimize the parameters, and the extension of two endpoints will be more accurate, then the end-points of cubic spline curve will not have large swing so as to achieve that intrinsic mode functions （IMF） of EMD are more accurate and reliable. Simulation results indicate that the extension method for data based on PSO-SVM method can restrain the end effects effectively.