CAJ | 학술논문

水平井中油水两相流是一个复杂的非线性系统,针对传统的统计特征方法难以应用于油水两相流流型流态的识别,因此,本文基于HHT变换与SVM相结合的方法,对水平模拟管道测量得到的非线性复杂信号进行流型流态的识别。首先对利用HHT具有自适应的分解方式对原始信号进行变换,提取出信号的特征变量,解决信号的非线性和随机性问题,再利用SVM进行流型流态的分类识别。利用该方法在水平模拟井上对油水两相流型进行了研究,仿真结果表明,HHT和SVM在多相流流型流态的识别中可以起到良好的应用,对流型的识别的正确率也较高。
수평정중유수량상류시일개복잡적비선성계통,침대전통적통계특정방법난이응용우유수량상류류형류태적식별,인차,본문기우HHT변환여SVM상결합적방법,대수평모의관도측량득도적비선성복잡신호진행류형류태적식별。수선대이용HHT구유자괄응적분해방식대원시신호진행변환,제취출신호적특정변량,해결신호적비선성화수궤성문제,재이용SVM진행류형류태적분류식별。이용해방법재수평모의정상대유수량상류형진행료연구,방진결과표명,HHT화SVM재다상류류형류태적식별중가이기도량호적응용,대류형적식별적정학솔야교고。
Oil-water two phase flow is a complex nonlinear system. for the sake of the statistical characteristics of the traditional method which is difficult to be used in oil-water two phase flow type identification of flow regime, therefore, in this paper, based on HHT transform combined with SVM, it will identify flow type flow regime of nonlin-ear and complex signal measured in horizontal pipe. Firstly, using HHT with adaptive decomposition to analyze the o-riginal signal to extract the signal characteristic variables, which solves the problems of nonlinear and stochastic sig-nal, and then using the SVM in identification of flow pattern. Using this method in the historical simulation well, studying in oil-water two-phase flow pattern, the simulation results show that the HHT and SVM make a good per-formance in multiphase flow model of flow recognition and have higher accuracy.