多模型自适应原油产量精确预测方法

AN ACCURATE METHOD TO PREDICT OIL PRODUCTION WITH ADAPTIVE MULTI-MODEL

原油产量的精确预测是原油优化配产的前提。在总结前人大量的产量预测方法的基础上，提出了多模型自适应产量预测方法。该方法是单模型自适应预测技术的自然推广。它不仅在不同时间段具有不同的轨迹方程与预测对象的变化规律进行最佳匹配，而且该最佳匹配方程的参数在预测时也是时变的，是一个完善的自适应预测系统。该方法具有预测精度高、适应性强、计算简单、方便的特点。

It is considered as a prerequisite for the optimized oil production planning to accurately predict oil production.On the basis of reviewing numerous existing methods for prediction of oil production,a multimodel adaptive prediction method is developed.This method derives from the single-model adaptive prediction technique.It can not only best match the variation regularities of the predicted object with the different trajectory equations existing at different time intervals,but also make the parameters involved in the best matching equations timevarying.Hence it represents a perfect adaptive system,and possesses features of high prediction accuracy, good adaptability,simplicity of calculation and convenience.