Research on Prediction Model of Diesel-Gasoline Contamination Quality Indez Based on the BP Neural Network
The prediction model of diesel-gasoline contamination quality index is set up respectively based on analysis of the basic principle of forward back propagation (BP) neural network. The structure of model is 2-4-1 three-layer BP network. The 31 study samples are trained using the improved BP algorithm to adjust the weight and threshold. The non-linear relationship between flash point of diesel-gasoline contamination and concentration of gasoline was gotten continuously. The distribution in the network connect right was stored, and the concentration of gasoline and the flash point of contamination was recorded finally in the form of the objective function of the complex non-linear relationship knowledge to connect the matrix. The mapping from input to output patterns of arbitrary non-linear model was established. And the flash point of contamination viscosity was predicted under the nonlinear influence of various factors of the concentration to the point.
pipeline BP neural network contamination quality indez prediction model
Zhao Hui-jun Zheng Yu Wang Shu-li Zhou Shi-dong Zhao Shu-hua
Jiangsu Key Laboratory of Oil & Gas Storage and Transportation Technology, Jiangsu Polytechnic University, Changzhou, China,213016
国际会议
第八届国际测试技术研讨会(8th International Symposium on Test and Measurement)
重庆
英文
1516-1519
2009-08-01（万方平台首次上网日期，不代表论文的发表时间）