Research On Fault Prediction Algorithm Based On DiPCA

With the increasing complexity of modern industrial production process,the security of the system is facing severe challenges.Fault diagnosis and fault-tolerant control technology emerged as the times require,and has become a research hotspot in the field of control science and engineering.Fault prediction,as a kind of fault diagnosis technology,which can realize the "unknown prophet" of faults and can intervene in advance,has attracted more and more attention in recent years.In this paper,a fault prediction algorithm based on dynamic internal principal component analysis(DiPCA)is proposed,and the Tennessee-Eastman(TE)chemical process is simulated.Different from the traditional principal component analysis(PCA)algorithm,DiPCA considers the dynamic change of data along time dimension.By Constructing Dynamic Implicit variables and their mathematical models,it captures the rmain dynamic characteristics of data,and then deduces the future changes of the system based on historical data,which is more suitable for fault prediction of the system.Firstly,the research background and significance of fault prediction are elaborated,and the research status of fault detection and prediction is analyzed.Secondly,the basic theory of PCA and DiPCA algorithm is introduced.Then,the fault detection of TE industrial process is studied by using PCA and DiPCA algorithm.Finally,a fault prediction algorithm based on DiPCA is proposed,aiming at different types of faults in TE process.The simulation experiments are carried out and the results are analyzed.The results show that the algorithm can predict the occurrence of faults in advance under different types of faults,such as step change,random change and viscous,and verify the effectiveness of the algorithm.