摘要
基于现代时间序列分析原理及系统辨识方法，建立了多变量自适应预测模型，并用该模型对某注水油田的原油产量进行了实际预测。结果表明，预测精度比传统的预测方法大大提高，该方法可为油田编制原油稳产计划提供科学依据。
Multivariate adapting prediction model was established based on the principle ofmodern time squence analysis and method of system identification。The yield of crude oil insome oilfield was predicted with the prediction model.The prediction result shows that themodel is more suitable than the class prediction methods,and can provide the scientific basesfor making the plan of stable production。
出处
《石油大学学报（自然科学版）》
CSCD
1994年第2期102-107,共6页
Journal of the University of Petroleum,China(Edition of Natural Science)
关键词
自适应预测
原油
采油量
Adaptive forecasting
Pararmeters identification
Dynamic system
Time-varing parameters
Crudeoil yield
