The Design And Achievement Of Algorithm On The Application Of Crude Oil Evaluation Data

To better utilize crude oil, the precious natural resource, it is necessary to be acquainted with its property. Crude evaluation is designed to grasp the crude property based on a series of analysis conducted on the newly-drilled crude under laboratory conditions. The data acquired through crude evaluation is the most fundamental but significant one, because its accuracy has direct effect on production and business and decision-making.However, as the crude variety and import crude surge in recent years, the traditional crude evaluation method cannot keep pace with the development of the situation and fail to satisfy the production needs. Therefore, a powerful, user-friendly and easy-to-use data base is called to come into being, in order to provide the international and domestic crude evaluation data fast, conveniently and systematically and predict detailed data of mixed crude. When improving the data base, the accumulated results of the previous data base will be reprocessed via a particular algorithm, to get a more minute property.This thesis aims to probe into algorithms in crude evaluation data application, in the context of the crude data base and application system developed by a major petrochemical enterprise. Firstly, Lagrange method and Newton interpolation are introduced, and how Lagrange interpolation solves the problem of boiling point curve-fitting is interpreted. Next, the thesis briefs on cluster analysis and cluster analysis algorithm based on mahalanobis distance analysis which realizes the function of crude cluster analysis. In the end, experimental results are employed to verify cluster analysis algorithm based on mahalanobis distance analysis.