径向基神经网络的瓦斯水合物相平衡预测  

RBF neural network prediction of gas hydrate phase equilibrium

作　　者：卢斌[1] 吴强[2] 

机构地区：[1]黑龙江科技学院信息网络中心 [2]黑龙江科技学院安全工程学院,哈尔滨150027

出　　处：《黑龙江科技学院学报》2009年第3期169-172,共4页Journal of Heilongjiang Institute of Science and Technology

基　　金：国家自然科学基金资助项目(50874040);黑龙江省自然科学基金资助项目(B2007-10)

摘　　要：瓦斯水合物生成是复杂的结晶过程,获取不同组分和浓度的瓦斯水合物相平衡等热力学参数对水合物技术的应用具有非常重要的意义。针对瓦斯水合物相平衡条件,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相平衡预测模型,并用实验数据进行了验证。结果表明,该模型对瓦斯水合物相平衡的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相平衡预测提供了一种新途径。Gas hydrate formation involves a complex crystallization process. Access to thermodynamic parameters, such as gas hydrate phase equilibrium of different components and concentrations is of greater significance to the application of hydrate technology. In addition to introducing the use of RBF neural network method to predict gas hydrate phase equilibrium, this paper discusses the determination of the input and output vector of the RBF neural network, the establishment of the RBF neural network prediction model of gas hydrate phase equilibrium, and the verification of the experimental data, according to equilibrium conditions for gas hydrates. The results indicates that when used to predict gas hydrate phase equilibrium, the model boasts many advantages, such as high calculation precision and fast speed. The research on RBF neural network provides a new approach to predicting the gas hydrate phase equilibrium.

关 键 词：瓦斯 水合物 相平衡 神经网络 

分 类 号：TD712[矿业工程—矿井通风与安全] TP183[自动化与计算机技术—控制理论与控制工程]