模拟退火在非稳态驱替实验计算相对渗透率中的应用

岩石的相对渗透率是储层物理性质的一个重要参数,其实验测量目前主要采用非稳定法.非稳定法通过测量恒流或恒压驱替过程中的时间、流量、压差参数,并根据一定的算法计算出被测样品的相对渗透率曲线.在JBN法计算相对渗透率中,经常采用线性函数或可线性化的非线性函数对实验数据进行滤波处理来提高相对渗透率曲线质量.将模拟退火算法应用于实验数据的拟合滤波处理,使得数据拟合光滑处理可以选择任意符合数据分布规律的函数,尤其是不能转化为线性的非线性函数,拓展了JBN改进方法的适用范围.实际数据处理结果表明,模拟退火算法用于非稳态驱替实验计算相对渗透率曲线是有效的.

Application of Simulation Annealing Algorithm to Calculating Relative Permeability With Unstable-state Displacement Experiment Data

Relative permeability is a important petrophysic parameter of rock. Ususally, the unstablestate flow technique is used to measure it. In the unstable-state method, the time, flowrate and different pressure are measured in the displacement experiment. Based on these parameters, the relative permeability curve of sample is calculated with some algorithms. In the modified JBN method, the linear and those which can be translated into linear function are used for smoothing experiment data before or after calculating the relative permeability to improve the quality of relative permeability curve. In this paper, the simulation annealing algorithm is adopted to regress data for smoothing . It makes selection of function for smoothing data discretional according to the data distribution. The function can be linear and nonlinear. By using the simulation annealing algorithm ,the applicability of modified JBN method becomes wider. The experiment data processing case indicates that the simulation annealing algorithm in calculating relative permeability curve with unstable-state experiment data is valid.