作　　者： ; ;

机构地区： 西南石油学院计算机科学学院

出　　处： 《西南石油学院学报》 1999年第2期5-8,共4页

摘　　要： 综合集成的试井解释模型和解释模型的自动识别一直是试井解释研究的热点，文中提出了一种基于句法模式识别的试井模型识别新技术，它克服了现有技术在曲线形态识别与模型诊断推理方面的困难，将模型识别的复杂过程分为特征抽取、形态跟踪、模型推断等简单过程；利用试井分析和非线性最优化技术形成了一套集成多种解释模型的高效率解释软件。 omprehensively integrated well testing interpretation model and the automatic recognization of the model have being a hard point for the study on the area. In order to solve the problems of curve pattern recognization and model diagnosing inference, a new method of well testing model recognization is presented in the paper based on syntactic pattern recognization. The complicated process recognizing pattern is simplified into characteristic extraction, shape tracing and model inference. A set of highly effective interpretation software integrating interpretation models is developed by well testing analyses and non-linear optimization technology.

关 键 词： 人工智能 模式识别 试井解释 曲线拟合 井底压力

领　　域： [石油与天然气工程] [石油与天然气工程]