用遗传算法求解广义S形增长模型

Making a solution for generalized S-shaped increasing model by genetic algorithms

广义S形增长模型在用于预测油气田累积产量和含水率方面功能显著,但该模型是具有多个未知参数的非线性模型,常规求解方法存在的不足可能会造成计算精度变差,从而影响预测效果。以累积产油量广义S形增长模型的求解为例,说明了采用遗传算法求解的方法和效果。遗传算法可提高求解的精度,是求解广义S形增长模型的较好方法。

Generalized S-shaped increasing model is useful in prediction of cumulative production and water-cut in fields, but the conventional approach to make its solution could result in poor calculation accuracy and prediction efficiency because it is a nonlinear model with several unknown parameters. The method and efficiency to make a solution for the model by genetic algorithms are illustrated during prediction of cumulative production for the generalized S-shaped increasing model. It is considered that the genetic algorithms can improve solution accuracy, and this is a better method to make a solution for the increasing model.