油井水力压裂后产量变化的混沌特征分析

Chaos analysis of oil production rate variation of hydraulically-fractured well

李小刚 杨兆中 伍晓妮 蒋 海

西南石油大学研究生部，四川  成都  610500 长城钻探工程公司工程技术研究院，辽宁  盘锦  124010

混沌理论的产生与发展打破了基于牛顿科学思想的拉普拉斯决定论，一旦预测对象具有混沌特征，其预测的基本原理和方法可能发生根本改变。目前还未见有关油气井产量变化混沌特征研究的报道。采用相空间重构和关联维数提取技术，对西部某油田一口油井水力压裂后60 d的产油量变化特征进行了研究。研究发现，该井产油量时间序列的饱和嵌入维数和对应的吸引子维数分别为9和3.44，由此发现该井压裂后产油量变化具有混沌特征；其主要影响因素，亦即影响油井产油量变化的基本变量为4~9个。从而为单井产油量预测模型的建立，特别是人工神经网络预测模型输入层节点输量的选择提供了新的参考依据。

Origination and development of chaos theory has changed people’s view of Laplace determinism based on Newton’s scientific thoughts. Once there is chaos in a system, the principles and ways to predict its future must be changed fundamentally. However, it is not known whether variation of oil well production rate is chaos. In this case, techniques of phase space reconstruction and correlation dimension extraction were used to investigate the oil production rate variation during 60 days of an individual hydraulically-fractured well in west China. It is discovered that the correlation dimension of the time series of oil production rate is
9, and the attracters dimension is 3.44. Thus, the oil production rate variation of this well is chaos. Moreover the number of basic factors, in other words, essential variables of the production system ranges from 4 to 9. The results of this study are very helpful for presenting accurate prediction models, especially for defining correct number of input layer nodes in ANN (artificial neural network) models.