System and method for predicting parameter of hydrocarbon with spectroscopy and neural networks

A method for predicting parameters of hydrocarbons, comprising the steps of:

• generating an NMR spectrum of a sample of a hydrocarbon having different hydrogen or carbon types;
• dividing said NMR spectrum into regions corresponding to said different hydrogen or carbon types;
• selecting at least one of said regions based upon a desired parameter to be predicted;
• quantifying a signal intensity of said at least one region, so as to provide at least one spectrum extracted quantity;
• and
• applying said at least one spectrum extracted quantity to a trained neural network trained to correlate spectrum extracted quantities with hydrocarbon parameters so as to predict said desired parameter from said spectrum extracted quantities.

A system for predicting a parameter of a hydrocarbon, comprising:

• means for generating an NMR spectrum from a hydrocarbon sample;
• a processor member communicated with said means for generating and adapted to provide at least one spectrum extracted quantity;
• and
• a trained neural network communicated with said processor member so as to receive said at least one spectrum extracted quantity, said trained neural network being programmed to correlate spectrum extracted quantities with hydrocarbon parameters so as to predict said desired parameter from said spectrum extracted quantity.