Research on the application of gasoline endpoint soft-sensing in hydroforming unit based on SVM

The application of Support Vector Machines (SVM) to the soft-sensing modeling technology was studied. To solve the problem that the endpoint of a refinery hydroforming unit can´t be monitored real-time on line, the soft-sensing model based on SVM was established and the gasoline endpoint was predicted. The experimental results show that the model has some characters such that quick calculating rate and high forecast accuracy. The indices are satisfied with the user´s requirements, and the predicting effects are good in the practice.