The Cargo Oil Heating Operation System Based On The Theory Of Data Mining

Petroleum or called crude oil is one of the most important resources which supports the development of industry and economy in the country strategy level. China has been the2nd biggest oil consumption country and3rd biggest oil importing country according to the development of industry and economy. It means the oil transport has become a significant issue which can affect not only the social development but also national security.The conception of low-carbon ship has been developed due to the idea of environmental protection which is the most important concerning issue in China in recent years, energy-saving and emission-reduction is one of the most important approach, and it has the same idea with cost-reduction for oil transport in the background of high oil price. Thus how to reduce the cost of oil transport by tank becomes an important issue. A cooperation work has been developed to research to heat management of cargo oil in tank with COSCO Dalian, expect to improve the cargo oil heat system in the tank.In this paper, the data mining technology in machine learning has been used to evaluate the influencing factors for cargo oil heating, coupled with the BP neural network algorithm to fit the relationship between the influencing factors of oil heating pressure and heating time. Thus the neural network software package in MATLAB has been used to learn the data from ship records,then the results has been predicted.In this paper, navigation data of LIANPINGHU which is provided by COSCO Dalian has been used to obtain the leaning sample and the detection sample. And MATLAB software has been used to establish the data model to get the prediction. The error of the prediction is supposed to be between±5%. Application methods of the software has been summarized and some further improvements has been discussed for the further works.