Research On Fault Diagnosis Of Crude Oil Distillation Process Based On SDG And Its Application

The petrochemical industry is one of the important pillar industries of our country,concern a lot of respects,such as national economy energy,material, etc.,it is one of the dangerous and extremely high trades too.Science and technology and economic development make the course of petrochemical industry maximize,complicate day by day.Atmospheric and vacuum distillation process is indispensable for refining plant,this is a complicated procedure of transferring heat and mass with complex mechanism.If some small troubles can't be got rid of in time,may cause the enormous disaster in production.So the security and dependability of the production system seem extremely important.There are some kinds of methods in improving systematic security and dependability,one of important method among them is the fault diagnosis technology.Fault diagnosis means to measure the unusual state of running mechanical equipment,to discern and predict the reason of the unusual state. In recent years,the fault diagnosis of petrochemical industry production process gradually becomes the hot problem of every relevant study,in which certain achievement has been made and some effective methods and technology have been summarized.SDG method based on the deep knowledge model is becoming more remarkable in numerous methods because of its characteristics,such as simple and easy,popularization and completeness,etc.In this paper,the present fault diagnosis technology is summarized at first then the author introduces the principle of SDG and the advantage of using it in the fault diagnosis of petrochemical process.The characters of SDG model, reasoning mechanism and modeling method are deeply analyzed.An example of atmospheric tower is introduced,and modeling method based on SDG is well studied.At last,the backward reasoning and forward verification algorithm is applied to fault diagnosis of atmospheric distillation.