论文部分内容阅读
瓦斯水合物生成是复杂的结晶过程,不同组分和浓度瓦斯生成水合物时压力等热力学参数的获取对水合物技术的应用具有非常重要的意义。鉴于此,利用径向基神经网络方法对瓦斯水合物生成压力进行了预测。针对瓦斯水合物生成边界条件,确定了RBF神经网络的输入、输出向量,建立了RBF神经网络瓦斯水合物生成压力计算及预测模型,并用实验数据进行了验证。结果表明,该模型对瓦斯水合物生成压力的拟合和预测具有计算精度高、速度快等优点。RBF神经网络研究为瓦斯水合物生成压力预测提供了一种新途径。
The formation of gas hydrate is a complex crystallization process. The acquisition of thermodynamic parameters such as pressure at different composition and concentration of gas hydrate is very important for hydrate application. In view of this, the pressure of gas hydrate formation is predicted by using RBF neural network. According to the boundary conditions of gas hydrate formation, the input and output vectors of RBF neural network are determined. The pressure calculation and prediction model of gas hydrate formation in RBF neural network are established and verified by experimental data. The results show that the model has the advantages of high accuracy and high speed in fitting and predicting the formation pressure of gas hydrate. RBF neural network provides a new way to predict the gas hydrate formation pressure.