Study On The Extracting Method Characteristics Of Rolling Force Signal Information Of The Rolling Mill

High precision sheet rolling is an important development direction of modernrolling technology, rolling force signal is the essential signal for the gauge control. Theresponse on the surface of the rolling workpiece, such as rolling force, frequency andother information is important parameters for the rolling process controlling.Particularly, the fundamental frequency response is an significant basis for the choicefor the selection and allocation of the components’ time constants during the rollingprocess control; To obtain accurate rolling force response frequency information, toavoid rolling force due to insufficient sample size in response to numerical problems isalso very important. To obtain accurate rolling force response frequency information isalso very important to avoid rolling force due to insufficient sample size for the lack ofresponse.Rolling force sensor is usually installed between the roller bearing seat and thearch rather than on. The response of the rolling force sensor is the difference betweenthe workpiece’s surface rolling force response and rolling force sensor response in rolls.Therefore, as long as the response in rolls is separated, the surface responseinformation will be got.After studying the related literatures of signal analysis, in this paper the measuredsignals from the first frame of a1700mm five-stand continuous rolling mill duringactual production have been frequency extracted, the basic fundamental frequency andthe triple frequency is solved. The main works are as follows:(1) The local mean decomposition is a signal analysis method proposed byJonathan S. Smith in2005, which is now widely used in acoustic signal analysis, thespecific steps include moving average data preprocessing, method for local envelopefunction and local mean function, the method for the PF (Product function) and theamplitude and frequency information that each PF components contain and the methodfor the instantaneous frequency. In order to validate the effectiveness of the local meandecomposition, a combined signal with several non harmonic signals as test signals to test, results show that, this method can accurately extract the frequency test signalvalues.(2)8groups of measured signals are analyzed, these signals are taken from therolling force sensors, including the operating side and driving side roll rolling forcesensor. Through the extracting results by analyzing multiple signals from differentrolling roll on the same rolling conditions, found that the basic fundamental frequencyis8.5Hz, and the triple frequency is25.5Hz.(3) As a control, using empirical mode decomposition (EMD) for the signalprocessing, the results further verify the correctness of the response frequency. Thecomparative study was made on the error that EMD produced during thedecomposition and find out the cause of it. The scope of applying two methods inmechanical signal processing has been described.