Research And Design Of Process Monitoring And Fault Diagnosis System For Annealing Furnace Of Hot Dip Galvanizing Line

Annealing furnace is the key equipment of hot galvanizing production line, the annealing process is the most important process of hot galvanizing process, the effect of annealing directly affects the quality of the galvanizing products. If a fault occurs, not only directly affects the quality of galvanized sheet, but also will influence the operation of the production line, and may cause huge economic losses, and even cause critical damage to the equipment, endanger the safety of workers. Process monitoring and fault diagnosis of annealing furnace, can effectively find and remedying faults. So, Study on monitoring and fault diagnosis method of annealing furnace have important theoretical and practical significance.Here takes a hot galvanizing production line as the research background. In view of the complexity of the annealing process and the nonlinear relation between the fault causes and symptoms, we use nonlinear methods kernel principal component analysis (KPCA) and support vector machine (SVM) to monitor the condition and diagnosis fault for the annealing process. The main work of the thesis includes:Firstly, according to the production process’s characteristics and requirements, proposed solutions for some practical problems in the application of KPCA in the annealing process of hot dip galvanizing production line such as data preprocessing, realize the kernel principal component analysis of normal sampling data, calculate the monitoring statistic and it’s limits to establish the monitoring model. Then, the results of simulation study for annealing furnace process of galvanizing line monitoring using the process monitoring model verify the effectiveness of the proposed method. Using a method based on data reconstruction to distinguish the fault variables, the results of simulation experiments show that this method is useful. In order to further resolve the problem of fault diagnosis, using support vector machine to classify fault, so as to realize fault diagnosis. Firstly, Training support vector machine using the training set, using cross validation method to improve the generalization ability of the model, then, using the fault diagnosis model to diagnosis fault for hot dip galvanizing annealing process. The results of the simulation verify the validity of the proposed method. Based on the study of theory and methods of KPCA and SVM mentioned above, aim at the annealing furnace of hot dip galvanizing line we designed a system for hot dip galvanizing annealing process’s condition monitoring and fault diagnosis using SQL databasea and technology of hybrid programming of MATLAB and.NET. The system realized the functions such as display of monitoring statistics T2 and SPE, identification of the fault variables, fault diagnosis, inquery of the history fault, laied the software foundation for the practical application of the theory and method.