预测钻控影响产量的多元线性回归方法

MULTI-ELEMENT LINEAR REGRESSION METHOD TO PREDICT EFFECT OF SHUT-IN WELL ON PRODUCTION RATE

为了在原油年度规划编制中更加准确地预测钻关对产量的影响，采用多元线性回归理论，对钻关影响产量的因素，如钻关前井网密度、含水、流动系数、采油强度、油水井数比、注水强度以及地层压力等进行了回归分析，采取后退法进行了变量筛选，建立了相对最优的预测方程。通过多元线性回归显著性分析，钻关影响产量的主要因素是钻关前的井网密度，其次为含水和流动系数，影响最小的为采油强度和油水井数比。经验证，建立的回归方程预测精度高，可以满足实际要求，在今后的油田生产中，可以利用该回归方程预测钻关对产量的影响。

To accurately predict effect of shut-in well on production rate in compiling annual plan, by using multi-element linear regression theory, a regression analysis is performed for impacting factors of shut-in well on production rate including well pattern density before shut-in well, water cut, flow coefficient, production intensity, injector-producer ratio, injection intensity and formation pressure. A variable screening is proceeded with a receding method, and a relatively optimal prediction equation is established. Through significance analysis on multi-element linear regression, the main impacting factors of shut-in well on production rate are well pattern density before shut-in, secondly are water cut and flow coefficient, and the least influencing factors are production intensity and injector-producer ratio. Verification shows high predictive accuracy of the regression equation, which can satisfy the actual requirement. The regression equation can be used to predict effect of shut-in well on production rate in future oilfield production.