Development Of Intelligent Monitoring System For Magnesium Alloy Sheet Rolling Mill Based On LabVIEW

With the foundation of the National Science and Technology Support ProgramProject Key technology research of magnesium alloy deep-processing and developmentof large sets of equipment (No.2012BAF09B01), this paper focuses on the development ofmonitoring system of rolling mill for wide high performance magnesium alloy sheet. Themain contents of this paper include monitoring program design of magnesium alloy sheetmill, researches of signal processing and fault diagnosis technology, software and hardwaredesign of an intelligent monitoring system for magnesium alloy sheet rolling mill based onLabVIEW, and simulation tests and field experiments.Based on the related domestic and foreign literatures, the industrial application ofmagnesium alloy is summarized. The significance of monitoring system for magnesiumalloy sheet rolling mill are described. Current problems of rolling mill monitoring system arepointed out. The current domestic and international developments in the field of monitoringsystem is introduced. Application of intelligent monitoring and fault diagnosis method, andvirtual instrument technology based on LabVIEW in monitoring systems is studied, whichare used in monitoring system design for magnesium alloy sheet rolling mill.According to the difference between magnesium alloy sheet rolling and steel oraluminum sheet rolling, monitoring program for magnesium alloy sheet rolling mill isdesigned. The program is based on rolling mills and rolling process of magnesium alloysheet, and typical failure in the rolling process. Physical quantities to be monitored includemain motor current, temperature of working rolls, rotational speed of working rolls, rollinggap, vibration of working rolls, temperature of the sheet, and rolling force.Monitoring signals are analyzed by time domain analysis, amplitude domain analysis,and time-frequency analysis. Feature frequency based on frequency domain analysis, featurevectors based on amplitude domain analysis and energy feature vectors based on waveletpacket analysis are extracted. With the feature vectors based on amplitude domain analysisand energy feature vectors based on wavelet packet analysis as the input, a monitoring and fault diagnosis model based on BP neural network improved by Levenberg-Marquardtmethod is established.On the basis of the monitoring program and signal processing and fault diagnosistechnology, the monitoring system is established. It consists of a hardware system based onsensors, NI data acquisition devices and computer, and a software system based onLabVIEW and MATLAB. Functions like data input, signal processing, real-time monitoring,fault diagnosis and data storage are achieved.With the1725mm four roller reversible magnesium alloy sheet hot rolling mill as thetest object, field experiments of the monitoring system for magnesium alloy sheet rollingmill based on LabVIEW are conducted. The practicality, accuracy and reliability of thesystem is verified.The design of this monitoring system for magnesium alloy sheet rolling mill provides areference to monitoring system development of magnesium alloy sheet rolling mills andother large equipment.