Determination of the conditions of natural gas hydrate formation using Neural Network

knowledge of the conditions of natural gas hydrate formation has various applications in enginnering calculations including natural gas transmission pipelines, hydrocarbon reservoir gas transmission by hydrate method and gas transmission pipeline safety , etc. there are lots of methods including katz method , campbell method parrish - prausnitz method , etc. there are lots of methods including katz method, campbell method parrish -prausnitz method, etc. for calculations and prediction of these conditions. but through investigation and examination results of these methods it is explored that although some of these methods has high speed in calculations they are not enough precise. neural network method has been used to eliminate such king of problems in then it will be provided both enough speed calculation and high precision for the calculation . in this study regarding to the availability of various relations in papers for prediction of hydrate formation conditions first of all a proper relation has been proposed for prediction of these conditions , and then neural network modeling has been done for the determination of correlations of hydrate formaiton equation.

Keywords:

modeling , neural networks , back propagation training , activated function , natural gas , hydrate.