用人工神经网络预测天然气水合物的生成

PREDICTING GAS HYDRATE FORMATION BY ARTIFICIAL NEURAL NETWORKS

天然气水合物生成是复杂的结晶过程 ,是一个多组分、多相系统 ,对该系统的生成动力学和热力学规律进行研究非常困难 ,因此很难对天然气水合物的生成进行精确的预测。文中运用人工神经网络理论 ,建立了水—天然气生成天然气水合物的人工神经网络模型 ,并选取大量的实验数据对模型进行了训练 ,通过实验所获得的关键点对模型的测试 ,证明该模型准确。

The formation of natural gas hydrate is a comp li cated crystal process in a multi components and multi phases system.Study of the dynamics and thermodynamics laws for the system forming is very difficult.S o,it is hard to predict gas hydrate formation.Applying the theory of artificial neural networks,the article develops a model of artificial neural networks for gas hydrate forming from water and gas,and trains the model with extensive exp erimental data.The model testing with the key points obtained by experiments pro ves the model is accurate and reliable.