论文部分内容阅读
针对凝析气藏露点压力与其影响因子之间的复杂非线性关系,难以建立具体的精确模型。本文应用人工神经网络的建模方法、多层感知器的模型结构、自适应学习速率的BP学习算法,辨识凝析气藏露点压力的功能模型,并把辨识模型的仿真结果与实验室测得的实际露点压力数据相对比,开展了与多元回归、经验公式的比较研究,以检验神经网络模型的可靠性。实验结果表明,这种新的凝析气藏露点压力建模方法具有很高的精度。
In view of the complex nonlinear relationship between the dew point pressure of condensate gas reservoir and its influencing factors, it is difficult to establish a precise model. In this paper, artificial neural network modeling method, multi-layer perceptrons model structure, adaptive learning rate of BP learning algorithm, condensate gas reservoir dew point pressure identification function model, and the identification model simulation results and laboratory measurements Compared with the actual dew point pressure data, a comparative study with multiple regression and empirical formula was conducted to test the reliability of the neural network model. Experimental results show that this new condensate gas reservoir dew point pressure modeling method with high accuracy.