Data-driven Ozone Process Control And Fault Diagnosis

With the advancement of science and technology,the problem of environmental pollution brought about by the development of industry has become more and more serious,and the party's report on the Nineteenth Congress has also kept a close eye on the key areas of environmental protection.This shows the importance of environmental issues.Ozone is one of the most promising clean and environmentally-friendly materials,and its strong oxidizing property is a prominent feature.Second,ozone does not produce secondary pollution.Industrially,ozone is mainly produced by the dielectric barrier discharge(DBD)method.This subject mainly takes the dielectric barrier discharge ozone generator as the research background.With the development of computer technology and sensor technology,a large amount of data will be generated during the operation of the equipment,and these data will be saved.Effective utilization and mining of these data will solve many problems.Based on the premise of data,this topic focuses on the modeling and control of dielectric barrier discharge ozone generators and fault diagnosis.The specific research content is as follows:(1)First of all,for high power DBD type ozone generators,the inverter power supply and its equivalent circuit are analyzed.Based on this,the electrical parameters(current,voltage,frequency)and process parameters(gas pressure,flow rate,temperature,and concentration)are studied.The dynamic relationship between them uses recursive neural networks to build their data-driven models.(2)Secondly,based on the above model,the data-driven optimal control method based on adaptive dynamic programming is studied,the utility function is selected,the tracking controller is designed;then the control constraints of the system are taken into account,and the utility function is reselected to design the tracking control.The device implements an iterative algorithm under two conditions to achieve tracking control and comparison of ozone concentration,yield,and yield.(3)Finally,the fault features of the high power DBD ozone generator are studied,the failure characteristic index is summarized,and the failure of the high power DBD ozone generator is identified by the LM neural network algorithm.The accuracy is verified by the confusion matrix and the ROC curve,and the identification is analyzed.effect.Then perform online fault detection.