基于关联维的联合站月输量时间序列混沌特征识别

联合站输量受单井生产动态、前级运行效率、自身处理能力等多因素的影响,且输量变化一般具有高度的非线性.混沌对确定性系统的伪随机性,探索联合站输量时间序列是否具有混沌特征具有积极的意义.在相空间重构基础上,采用G-P算法提取中原油田某联合站60个月的月输量时间序列的关联维.研究发现,联合站输量时间序列饱和嵌入维和吸引子分数维分别为9和8.44,表明该联合站月输量变化具有混沌特征且该联合站系统模型化研究所需的最小变量数目为9,为系统数理模型建立,特别是人工神经网络预测模型的输入层节点数设置提供了参考.

Recognition of chaotic characteristics of the time series of the monthly delivery rate of central gathering station based on correlation dimension

The delivery rate of a central gathering station,controlled by the production performance of individual wells,the operation efficiency of former units and itself handling capacity,is usually highly nonlinear.Chaos is the pseudo-randomness of a deterministic system.To study whether the time series of the monthly delivery rate of the central gathering station has chaotic characteristics is very significant to the prediction of the delivery rate.The correlation dimension of the time series of the 60-month delivery rate of a certain central gathering station in Zhongyuan Oilfield is extracted by G-P algorithm based on phase space reconstruction.It is found that the correlation dimension of the time series is 9,and the attracters dimension is 8.44.This shows that the variation of the monthly delivery rate of this station has chaotic characteristics and the minimum number of the essential variables to the modeling study of the central gathering station system is 9.The results are very helpful to the establishment of the mathematical model of the system,especially the determination of the number of the input layer nodes in ANN (artificial neural network) models.