基于改进灰色verhulst模型的稠油产量预测

Prediction of Oil Production for Heavy Oil Filed Based on Optimized Verhulst Model

稠油油田开发是一个复杂的过程,在地下信息较少的情况下可以将其视为灰色系统.以辽河油田某稠油研究区为例,基于Verhulst模型建模理论对该区2005-2011年的年产油量建模发现后期预测误差呈增大趋势.在Verhulst模型的基础上结合G(1,1)灰色建模理论建立了残差修正年产油量Verhulst模型,模型精度由原来的96.0994%提高到97.6763%.2012-2014年的年产油量预测结果显示残差修正模型的预测精度达到了99.2281%,且关联度检验结果显著提高.将模型预测结果与数值模拟预测结果进行了对比,结果表明残差修正的Verhulst模型是一种快速、简洁、精准的预测方法,在一定的条件下使用残差修正Verhulst模型,不仅可以预测稠油油田近期的年产油量,还可对研究区后期开发方案调整等问题提供决策依据.

The development of heavy oil field is a complex process, it can be regarded as gray systems under the circumstances of less information. The oil production model of the research area in Liaohe oilfield has been established based on the Verhulst model theoretical, the result of Verhulst model for oil production from 2005 to 2014 shows an increasing trend of error. The residual modification model has been established by combining the Verhulst model with G（1,1） grey model, and the model accuracy increases from 96.0994% to 97.6763%. Oil production prediction results from 2012 to 2014 shows that prediction accuracy of residual modification model reached 99.2281%, and the correlation increases significantly. Comparing predicted results with the numerical simulation results shows that residual modification Verhulst model is a quick, concise and accurate forecasting method that can be used on heavy oil fields development, under certain conditions, the residual modification Verhulst model not only can be used to predict the oil production, but also can provide basis for decision making.