A hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration degree

Abstract

The application provides a kind of hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration degree, it is related to hot continuous rolling mill rolling technical field, including the following steps: S1, frequency concentration degree model, deformation resistance calculation model and the dynamics model of key components are established;S2, based on the deformation resistance calculation model and the historical data after processing, the rolling mill production line database is constructed;S3, vibration evaluation standard library is established based on the classified roll system state;S4, according to the model, the key parameters of measured data are solved;S5, the key parameters of measured data and the key parameters of historical data are matched;S6, the key parameters of measured data and the key parameters of historical data are compared to judge the state of rolling mill system, and the key parameters are determined by comparing the data of theoretical model solution to determine the fault position.The application ensures the rolling stability of hot continuous finishing rolling mill group and process model optimization through the state monitoring and fault diagnosis of rolling mill vibration in rolling process, ensures the product quality and the stability of production.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of hot continuous rolling mill rolling technology, in particular, especially relates to a hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration. BACKGROUND

[0002] The hot continuous rolling mill fault real-time diagnosis technology is the current front technology in the field of hot rolling strip at home and abroad, which can fully utilize the process and vibration data of the production line to evaluate the rolling mill running state in real time, and can verify the reliability of the rolling process model, the assembly precision of the rolling mill and the fault problem diagnosis on the efficient and compact production line. Thus, according to the diagnosis result, the best process model can be selected, which helps to improve the product precision and the stability of the rolling mill equipment of the hot rolling strip production line, thereby reducing energy consumption and environmental pollution.

[0003] However, since the current hot continuous rolling production line mainly adopts the process model setting based on the traditional theoretical model or empirical formula, the real-time rolling state data of the finishing rolling mill group in the rolling process is difficult to quantitatively process, which cannot guarantee the adaptability of the process setting model under the dynamic rolling state of the rolling mill, thereby affecting the stability of the rolling mill equipment in the rolling process. In order to ensure product quality and not to destroy the continuity, patent CN115099284A discloses a hot rolling mill vibration data processing method, system, terminal and storage medium, which is based on the comparison of the peak values of a plurality of data sub-blocks with the average peak values of the data sub-blocks, the determination of the limit and the segmentation of the original vibration data. At the same time, since the process data and the vibration data are closely related, the matching relationship between the two types of data changes when rolling different product specifications, and the prediction calculation based on the theoretical model is complex, has poor real-time performance and insufficient calculation accuracy. In order to meet the requirements of equipment stability and process adaptability when rolling various product specifications, the production line needs to analyze the rolling process and the vibration history database to realize the optimization of the rolling process. Patents CN114091211A, CN108038553A and CN111530943A etc. propose some methods for data processing and diagnosis of rolling mill vibration and process data, which ensure the stability of production.

[0004] The existing rolling mill vibration prediction and equipment state monitoring method is a prediction means based on the processing and comparison of rolling mill vibration measured data and historical data, which considers that the rolling mill vibration data can reflect the stability of the rolling process under the premise of long-term stable rolling mill state, and establishes the threshold value of different parameter indexes of the vibration data according to the historical data. However, these methods all ignore the influence of rolling different product specifications and their process models on the rolling mill vibration, and with the increasing trend of thin and wide specifications and high strength performance of rolling strip products in the steel industry, the further optimization of the process model and the accurate judgment of the vibration state are restricted, which leads to the difficulty of the rolling equipment and process to meet the stringent requirements of future high-quality strip production. SUMMARY

[0005] Therefore, the present application aims to provide a hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration degree to solve the problem that the prior art ignores the influence of rolling different product specifications and their process models on rolling mill vibration.

[0006] The technical means adopted by the present application are as follows:

[0007] A hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration degree comprises the following steps:

[0008] S1, based on the hot continuous rolling mill data, a frequency concentration degree model, a deformation resistance calculation model and a key component dynamics model are established;

[0009] S2, the historical data of the hot continuous rolling mill are acquired, the historical data are processed, a rolling mill production line database is constructed based on the deformation resistance calculation model and the processed historical data, the first key parameter of the historical data is calculated based on the deformation resistance model, and the second key parameter of the vibration data is calculated based on the frequency concentration degree model;

[0010] S3, the roll system state in the historical data is classified, and a vibration evaluation standard library is established based on the classified roll system state and the second key parameter;

[0011] S4, real-time measured data are acquired, the measured data are analyzed, and the key parameter of the measured data is solved according to the model;

[0012] S5, the production line database is divided into sub-databases, the data in the sub-databases are selected to participate in the roll system state evaluation, and the key parameter of the measured data is matched with the first key parameter of the historical data based on the basic indexes of the production line database;

[0013] S6, the rolling mill system state is judged based on the comparison between the key parameter of the measured data and the second key parameter of the historical data, and the fault position is determined by comparing the key parameter with the data solved by the key component dynamics model.

[0014] Further, S2 comprises the following steps:

[0015] S21, the hot rolling grade, strip width and thickness, rolling speed, rolling force, finishing temperature, roll diameter and material composition in the historical data are called to be the basic indexes of the production line database;

[0016] S22, the process model data are taken to match the rolling mill data with the strip data, and the roll gap, rolling speed, rolling force distribution and hot rolling strip grade of each stand are determined;

[0017] S23, the strip temperature change rule and the deformation resistance increment between different stands are calculated according to the deformation resistance calculation model.

[0018] Further, S3 comprises the following steps:

[0019] S31, call the rolling mill vibration data and classify the roll system state: the biting steel stage, the passing steel stage, the waiting steel stage and the throwing steel stage are taken as the rolling state; the pressing process stage, the roll changing impact stage and the parking stage are taken as the adjustment state;

[0020] S32, according to the frequency concentration degree model, analyze the peak-to-peak value, the frequency concentration degree and the main frequency of the different states of the historical vibration data corresponding to the process model data and establish the vibration evaluation standard library;

[0021] S33, count the data in the vibration evaluation standard library obtained in S32, determine the threshold value for distinguishing the stages of different states under the condition interval of 0.95, and obtain the distinguishing threshold value of conditions one to five as the evaluation standard of the measured vibration data and the roll system state.

[0022] Further, in S4, the analysis of the measured data comprises the following steps:

[0023] S41, solve the key parameters using the frequency concentration degree model;

[0024] S42, judge the roll system state: according to the union of condition one and condition two and the adjustment state condition threshold value according to the motor running state, otherwise determine as the rolling state;

[0025] S43, the rolling state stage includes: biting steel stage, passing steel stage, waiting steel stage and throwing steel stage; condition three is met for the biting steel stage; condition four is met for the waiting steel stage; condition five is met for the throwing steel stage; otherwise, it is the passing steel stage;

[0026] S44, the adjustment state stage includes: the pressing process stage, the roll changing impact stage and the parking stage; condition one is met for the pressing process stage; condition two is met for the roll changing impact stage; otherwise, it is the parking stage;

[0027] S45, take the process model data corresponding to the measured vibration data, determine the roll gap, rolling speed, rolling force distribution and hot strip grade of each stand;

[0028] S46, calculate the strip temperature change rule and the deformation resistance increment between different stands according to the deformation resistance calculation model.

[0029] Further, in S31:

[0030] The condition one is:

[0031] ; ;

[0032] wherein, Ppeak is the peak-to-peak value of the measured vibration signal data labg(Δti), Fmain is the main frequency of the measured vibration signal data labg(Δti), Fconcentration is the frequency concentration of the measured vibration signal data labg(Δti); and Ppeak is the peak-to-peak value of the historical data l’abg of the pressing stage, and Fconcentration is the frequency concentration of the historical data l’abg of the pressing stage, Fmain is the main frequency of the historical data l’abg;

[0033] The second condition is:

[0034] ;

[0035] wherein, Ppeak is the peak-to-peak value of the measured vibration signal data labg(Δti), Fconcentration is the frequency concentration of the measured vibration signal data labg(Δti); and Fconcentration is the frequency concentration of the historical data l’abg of the roll changing impact stage, Ppeak is the lower limit of the peak-to-peak value of the roll changing impact stage,

[0036] The third condition is:

[0037] ;

[0038] wherein, Ppeak is the peak-to-peak value of the measured vibration signal data labg(Δti), Fconcentration is the frequency concentration of the measured vibration signal data labg(Δti); , Ppeak is the upper limit of the peak-to-peak value of the historical data l’abg of the biting stage, , Fconcentration is the upper limit of the frequency concentration of the historical data l’abg of the biting stage;

[0039] The fourth condition is:

[0040] ;

[0041] wherein, Ppeak is the peak-to-peak value of the measured vibration signal data labg(Δti), the frequency concentration of the measured vibration signal data labg(Δti) segment; the upper and lower limits of the peak-to-peak value of the historical data l’abg in the stage of the steel to be cast, the upper and lower limits of the frequency concentration of the historical data l’abg in the stage of the steel to be cast;

[0042] the fifth condition is:

[0043]

[0044] wherein, the peak-to-peak value of the measured vibration signal data labg(Δti) segment, the frequency concentration of the measured vibration signal data labg(Δti) segment; the upper and lower limits of the peak-to-peak value of the historical data l’abg in the stage of the steel to be cast, the upper and lower limits of the frequency concentration of the historical data l’abg in the stage of the steel to be cast.

[0045] Further, in S5, the step of establishing the sub-library comprises:

[0046] S51, according to the real-time data basic index as the index data in S45, within the deviation allowable range, the basic index and the first key parameter of the historical database in S22 are screened to complete the first matching;

[0047] S52, according to the index data priority in S51, the historical database L’abg composed of the historical data basic index, the first key parameter and the second key parameter is established;

[0048] S53, the credibility of each level of sub-library and its data l’abg is determined according to the percentage of deviation of the index data in S51 to complete the second matching; the sub-library of l’abg data and its credibility is composed of the historical data sub-library.

[0049] Further, in S5, the matching of the measured data key parameter and the historical data key parameter comprises the following steps:

[0050] S54, according to the first principle of the first key parameter being equal, the sub-library with the highest credibility is selected in the L’abg library, and further matching is performed in the sub-library to obtain the historical data l’abg with the highest credibility;

[0051] ​​​​​S55, extracting the data segment labg(Δtj) containing abnormal vibration state from the measured data labg compared with the historical data l'abg described in S51.

[0052] Further, S6 includes the following steps:

[0053] S61, using the dynamics model of the key components, calculating the dynamics parameters under various system states, and theoretically calculating the dynamics model response labg(Δtk) under different states of the rolling mill system;

[0054] S62, according to the frequency concentration model, calculating the peak-to-peak value and frequency concentration of the dynamics response labg(Δtk) of the theoretical model and the measured data labg(Δtj);

[0055] S63, according to the peak-to-peak value and frequency concentration parameters, matching the measured fault data segment labg(Δtj) and the dynamics model response data labg(Δtk) of the theoretical model, realizing the fault position judgment and credibility evaluation of the rolling mill system.

[0056] Further, the frequency concentration model formula is as follows:

[0057]

[0058]

[0059] wherein, is the frequency concentration of the fundamental frequency and its multiple frequency; is the i-th dominant frequency of the original signal after Fourier transform in descending order of amplitude, is the corresponding amplitude; judges the multiple frequency relationship in ; is the multiple frequency degree, taking values and its reciprocal, is the multiple frequency capacity, taking values ;

[0060] The deformation resistance calculation model formula is as follows:

[0061]

[0062]

[0063] wherein, is the deformation resistance, G FiDjg1 is the finishing mill inlet temperature, G FiDjg2 is the rolling speed, GFiDjg3 is the reduction, G FiDjg4 is the material force parameter, is the Kurochkov formula, representing the first i increment of deformation resistance of the stand relative to the first stand, is the rate of temperature change per unit time caused by the cooling water, L is the first i distance between the strip of the stand and the cooling water; is the Yoshino formula, representing the first i increment of deformation resistance of the stand relative to the previous stand; representing the first i increment of deformation resistance of the stand relative to the previous stand; representing the first i increment of deformation resistance of the stand relative to the previous stand;

[0064] The kinetic model formula of the key components is as follows:

[0065]

[0066] wherein, is the system mass matrix, is the damping matrix, is the first order stiffness matrix, is the third order stiffness matrix; is the acceleration of the motion of each degree of freedom of the system, is the velocity of the motion of each degree of freedom of the system, is the displacement of the motion of each degree of freedom of the system, is the third order displacement of the motion of each degree of freedom of the system; is the dynamic rolling force.

[0067] The application also provides a storage medium, which comprises a stored program, wherein the program performs any one of the above real-time fault diagnosis methods of hot rolling mill considering vibration frequency concentration degree when running.

[0068] Compared with the prior art, the application has the following advantages:

[0069] The application is based on a large amount of historical data and theoretical research, combined with a vibration frequency concentration calculation method and a vibration data and process data matching method, fully considers the dynamic characteristics in the rolling process of each rack, and proposes a hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration. By establishing an online diagnosis historical database and evaluation index and method of measured data and combining the corresponding dynamic model, the rolling mill state diagnosis process is executed according to the parameter matching relationship. In order to ensure the improvement of stable rolling stability and the optimization of rolling process mathematical model, through the invention of data matching means and data evaluation means, the sustainability of the plate and strip rolling mill production line is improved, and the steel production brings higher application value. BRIEF DESCRIPTION OF DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0071] Figure 1 The present application is a general flowchart.

[0072] Figure 2 The present application is a process data and vibration data matching relationship diagram.

[0073] Figure 3 The present application is a historical data sub-library grading flowchart. DETAILED DESCRIPTION

[0074] In order to make the person skilled in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0075] It is to be understood that the terms "first", "second", and the like, used in the description and the claims of the application and the above drawings, are used to distinguish between similar objects, and are not necessarily used to describe a particular sequential or chronological order. It is to be understood that the use of data so designated is not meant to limit, and will not serve to limit, the described embodiments of the application to only such potentially substitutable embodiments, but is meant to cover "normal process, method, system, product, or apparatus" variations of the application that are or can become insubstantial to the application. Furthermore, the terms "comprising" and "including" and any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product, or apparatus that includes a list of steps or units not necessarily limited to those specifically listed, but can include other steps or units not expressly listed or inherent to such process, method, system, product, or apparatus.

[0076] The application firstly establishes a calculation model of frequency concentration degree, and establishes a historical database based on data characteristics and process parameters; secondly, real-time vibration data are screened and further matched with the vibration historical database and process, and abnormal vibration data are selected for subsequent analysis; finally, a dynamic analysis method is used to analyze faults of different vibration data. Through the data matching of vibration and process by the application, the optimization of the process model setting can be realized.

[0077] The application achieves the above-mentioned purpose by the following technical solutions.

[0078] A hot continuous rolling mill finishing mill set fault diagnosis method considering vibration frequency concentration degree. In the rolling process of the hot continuous rolling seven-stand finishing mill set, due to the differences in assembly precision, hydraulic pressure reduction, control frequency and rolling load configuration of each stand, different vibration characteristics of each stand roll system are presented when rolling different grades and different specifications of plate strip products. The application establishes a vibration signal analysis model considering vibration frequency concentration degree based on the rolling plate strip force and energy characteristics, and proposes a hot continuous rolling mill finishing mill set fault diagnosis method based on rolling state recognition, which can diagnose the fault position and assembly precision of each stand through the rolling vibration state, including the following steps.

[0079] Step 1: Establish a vibration evaluation standard library. Collect the information Fa of each finishing stand, the information Db of the plate strip and the corresponding process data Gab, a represents the number of each stand, b represents the number of the plate strip, ab is the process number when the a stand rolls the plate strip b, and the rolling mill production line database L'abg is constructed according to the historical data, g represents the number of the adopted process model.

[0080] 1.1 Retrieve the hot rolling grade, plate strip width and thickness, rolling speed, rolling force, finishing rolling temperature, roll diameter and material composition database basic indicators in the rolling performance;

[0081] 1.2 Collect the rolling mill vibration data and classify the roll system state: the biting steel stage, the passing steel stage, the waiting steel stage and the throwing steel stage are taken as the rolling state; the pressing process stage, the roll changing impact stage and the parking stage are taken as the adjustment state.

[0082] 1.3 Match the rolling mill Fa and the plate strip Db data with the process model data Gab, determine the roll gap, the rolling speed, the rolling force distribution and the hot rolling plate strip grade; calculate the plate strip temperature variation law and the deformation resistance increment between different stands according to the model two.

[0083] 1.4 According to the model one, analyze the peak-to-peak value, the frequency concentration degree and the main frequency of each stage in different states.

[0084] 1.5 Determine the threshold value of distinguishing the state of each stage in different states under the state of the confidence interval being 0.95, and obtain the threshold value (conditions one to five) as the standard of real-time vibration.

[0085] Step 2: Real-time analysis of the process parameter model labg of the measured data. Judge whether labg∩L'abg is empty, if it is empty, directly execute step 2.4, otherwise, refer to the L'abg historical database, execute steps 2.1~2.3.

[0086] 2.1 Classify the measured vibration signal data labg(Δti) collected in the Δt time period according to step 1.2, and analyze the signal data according to the model one.

[0087] 2.2 Judge the roll system state and its stage according to step 4.

[0088] 2.3 Select the highest confidence sub-database in the L'abg database. According to the first principle of equal deformation resistance increment ΔN, further match in the sub-database according to step 5, and obtain the highest confidence historical data l'abg.

[0089] 2.4 Compare the highest confidence sub-database (priority P4>P3>P2>P1) of P1~P4 of the historical data l'abg, and extract the data segment labg(Δtj) containing abnormal vibration state from the measured data labg.

[0090] Step 3: Judge the rolling mill system fault based on the vibration data and the theoretical model.

[0091] Divide the rolling condition into normal state and fault state; the system state is divided into vertical vibration abnormality, horizontal vibration abnormality, torsional vibration abnormality and axial vibration abnormality; the fault position is divided into horizontal gap being too large, transmission centering deviation, rolling mill two side stiffness deviation and hydraulic stiffness deviation.

[0092] ​3.1 Calculate the dynamic parameters of the rolling mill system under various system states using the dynamic model of the key components, and theoretically calculate the dynamic model response labg(Δtk) of the rolling mill system under different states.

[0093] 3.2 Calculate the peak-to-peak value and frequency concentration of the theoretical model's dynamic response labg(Δtk) and the measured data labg(Δtj) according to Model One.

[0094] 3.3 Match the measured fault data segment labg(Δtj) with the theoretical model's dynamic model response data labg(Δtk) according to the peak-to-peak value and frequency concentration parameters, and realize the fault location judgment and reliability evaluation of the rolling mill system.

[0095] Step 4: Classify the rolling state.

[0096] 4.1 Determine the roll state: according to the union of Condition One and Condition Two, and according to the motor operating state as the adjustment state condition threshold, otherwise determine the rolling state.

[0097] 4.2 Rolling state stages include: steel biting stage, steel passing stage, steel waiting stage, and steel throwing stage; Condition Three is met for the steel biting stage; Condition Four is met for the steel waiting stage; Condition Five is met for the steel throwing stage; otherwise, it is the steel passing stage.

[0098] 4.3 Adjustment state stages include: pressing process stage, roll changing impact stage, and parking stage; Condition One is met for the pressing process stage; Condition Two is met for the roll changing impact stage; otherwise, it is the parking stage.

[0099] Step 5: Perform parameter matching between measured data and historical data sub-library, and divide the historical data sub-library into P1~P4 levels according to the following steps, select data in the historical data sub-library to participate in state evaluation, and the reliability of each level sub-library is different, respectively , .

[0100] 5.1 In the historical data sub-library, select the same hot rolling brand series steel to establish P1 sub-library, and set the P1 sub-library reliability according to the similarity of the corresponding hot rolling material force and energy properties of the corresponding brand of the measured data .

[0101] 5.2 In the historical data P1 sub-library, select the same rolling force range steel to establish P2 sub-library, and set , P2 sub-library reliability according to the percentage deviation of historical data rolling force and measured data.

[0102] 5.3 In the historical data P2 sub-library, select steel grades with the same rolling speed range to create the P3 sub-library, and set the parameters according to the percentage deviation between the historical data rolling speed and the measured data. P3 sub-library credibility .

[0103] 5.4 In the historical data P3 sub-library, select steel grades with the same reduction range to create the P4 sub-library, and set the reliability of the P4 sub-library based on the percentage deviation between the historical data reduction and the measured data. .like Step 1 involves creating new historical data.

[0104] Model 1: Frequency Concentration Model

[0105]

[0106]

[0107] Frequency concentration of the fundamental frequency and its harmonics; The i-th dominant frequency after performing a Fourier transform on the original signal, ordered in descending order of amplitude. for The corresponding amplitude; judge middle The harmonic relationship; For the frequency multiple, the value can be 10 ... and its reciprocal, For frequency doubling tolerance, the value can be 1.

[0108] Model 2: Deformation Resistance Calculation Model

[0109]

[0110]

[0111] For deformation resistance, These are parameters from the process database, represented as finishing mill inlet temperature, rolling speed, reduction, and material mechanical energy parameters, respectively. The Kurkov formula represents the increase in deformation resistance of the i-th stand relative to the first stand caused by different rolling temperatures in each stand. Let L be the rate of temperature change per unit time caused by the cooling water, and L be the distance between the i-th frame strip and the cooling water. The formula is Inoue's formula, which represents the increase in the deformation resistance of the i-th stand relative to the previous stand caused by the different rolling speeds of each stand; is the formula, indicating the deformation resistance increment of the i-th stand relative to the previous stand caused by the different reduction of each stand; is the formula, indicating the deformation resistance increment of the i-th stand relative to the previous stand caused by the change of the metal material structure of each stand.

[0112] Model three:

[0113]

[0114] 、 、 、 are the system mass matrix, damping matrix, linear stiffness matrix and cubic stiffness matrix, respectively; 、 、 、 are the acceleration, velocity, displacement and its cubic of each degree of freedom of the system; is the dynamic rolling force.

[0115] Condition one:

[0116] ; ;

[0117] 、 and are the peak-peak value, dominant frequency and frequency concentration of the measured vibration signal data labg(Δti) segment, respectively; 、 and 、 are the upper and lower limits of the peak-peak value and frequency concentration of the historical data l’abg in the pressing stage; is the dominant frequency of the historical data l’abg segment.

[0118] Condition two

[0119] ;

[0120] and are the peak-peak value and frequency concentration of the measured vibration signal data labg(Δti) segment, respectively; 、 are the upper limit of the frequency concentration of the historical data l’abg in the roll changing impact stage, is the lower limit of the peak-peak value of the vibration peak-peak value.

[0121] Condition three

[0122] ;

[0123] and are respectively the peak-peak value and the frequency concentration of the measured vibration signal data labg (Δti) segment; , and , are respectively the upper and lower limits of the peak-peak value and the frequency concentration of the historical data l'abg in the biting stage.

[0124] Condition four

[0125] ;

[0126] and are respectively the peak-peak value and the frequency concentration of the measured vibration signal data labg (Δti) segment; , and , are respectively the upper and lower limits of the peak-peak value and the frequency concentration of the historical data l'abg in the waiting stage.

[0127] Condition five

[0128] ;

[0129] and are respectively the peak-peak value and the frequency concentration of the measured vibration signal data labg (Δti) segment; , and , are respectively the upper and lower limits of the peak-peak value and the frequency concentration of the historical data l'abg in the throwing stage.

[0130] The application also provides a storage medium comprising a stored program, wherein the program, when executed, performs the real-time diagnosis method for hot rolling mill faults considering vibration frequency concentration.

[0131] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for real-time diagnosis of a fault in a hot tandem mill taking into account the concentration of vibration frequencies, characterized in that, It comprises the following steps: S1, based on the hot continuous rolling mill data, a frequency concentration degree model, a deformation resistance calculation model and a key component dynamics model are established; The frequency concentration degree model formula is as follows: wherein, is the frequency concentration of the fundamental frequency and its multiple frequencies; is the i-th dominant frequency of the original signal after Fourier transform in descending order of amplitude, is the i-th dominant frequency of the original signal after Fourier transform in descending order of amplitude, is the corresponding amplitude; determines in the multiple frequency relationship; is the multiple frequency degree, taking values and its reciprocal, is the multiple frequency capacity, taking values ; The deformation resistance calculation model formula is as follows: wherein, is the deformation resistance, G FiDjg1 is the finishing entry temperature, G FiDjg2 is the rolling speed, G FiDjg3 is the reduction, G FiDjg4 is the material force-energy parameter, is the Kolmogorov formula, which represents the first i deformation resistance increment of the stand relative to the first stand, is the temperature change rate per unit time caused by the cooling water, L is the first i distance of the strip in the stand from the cooling water; is the Yoshida formula, which represents the first i deformation resistance increment of the stand relative to the previous stand due to the difference in rolling speed; represents the first i deformation resistance increment of the stand relative to the previous stand due to the difference in reduction; represents the first i deformation resistance increment of the stand relative to the previous stand due to the change in the metal material structure of the stand; The key component dynamics model formula is as follows: wherein, is the system mass matrix, is the damping matrix, is the linear stiffness matrix, is the cubic stiffness matrix; is the acceleration of the motion of the system in each degree of freedom, is the velocity of the motion of the system in each degree of freedom, is the displacement of the motion of the system in each degree of freedom, is the third power of the displacement of the motion of the system in each degree of freedom; is the dynamic rolling force; S2, the historical data of the hot continuous rolling mill are acquired, the historical data are processed, a rolling mill production line database is constructed based on the deformation resistance calculation model and the processed historical data; a first key parameter of the historical data is calculated based on the deformation resistance model, and a second key parameter of the vibration data is calculated based on the frequency concentration degree model; S3, the roll system state in the historical data is classified, and a vibration evaluation standard library is established based on the classified roll system state and the second key parameter; S4, real-time measured data are acquired, the measured data are analyzed, and a key parameter of the measured data is solved according to the model; S41, the key parameter is solved using the frequency concentration degree model; S42, the roll system state is judged: according to the union of condition one and condition two and the motor running state as the adjustment state condition threshold, otherwise, it is judged as the rolling state; S43, the rolling state stage includes: the steel biting stage, the steel passing stage, the steel waiting stage and the steel throwing stage; condition three is met for the steel biting stage; condition four is met for the steel waiting stage; condition five is met for the steel throwing stage; otherwise, it is the steel passing stage; S44, the adjustment state stage includes: the pressing process stage, the roll changing impact stage and the parking stage; condition one is met for the pressing process stage; condition two is met for the roll changing impact stage; otherwise, it is the parking stage; S45, the process model data corresponding to the measured vibration data are taken to determine the roll gap, the rolling speed, the rolling force distribution and the hot rolling strip grade of each stand; S46, the strip temperature change rule and the deformation resistance increment between different stands are calculated according to the deformation resistance calculation model; The condition one is: ； ； wherein, is the peak-to-peak value of the measured vibration signal data labg(Δti) segment, is the dominant frequency of the measured vibration signal data labg(Δti) segment, is the frequency concentration of the measured vibration signal data labg(Δti) segment; and is the upper and lower limit of the peak-to-peak value of the historical data l’abg compression phase, and is the upper and lower limit of the frequency concentration of the historical data l’abg compression phase, is the vibration dominant frequency of the historical data l’abg segment; The condition two is: ； wherein, is the peak-to-peak value of the measured vibration signal data labg(Δti) segment, is the frequency concentration of the measured vibration signal data labg(Δti) segment; and is the upper and lower limit of the frequency concentration of the historical data l’abgroll change impact phase, is the lower limit of the vibration peak-to-peak value of the roll change impact phase; The condition three is: ； wherein, is the peak-to-peak value of the segment of the measured vibration signal data labg(Δti), is the frequency concentration of the segment of the measured vibration signal data labg(Δti); , is the upper and lower limit of the peak-to-peak value of the historical data l’abg biting stage, , is the upper and lower limit of the frequency concentration of the historical data l’abg biting stage; The condition four is: ； wherein, is the upper limit of the peak-to-peak value of the historical data l'abgfor the stage of the steel to be cast, is the upper limit of the frequency concentration of the historical data l'abgfor the stage of the steel to be cast, and is the upper limit of the peak-to-peak value of the historical data l'abgfor the stage of the steel to be cast, , is the upper limit of the frequency concentration of the historical data l'abgfor the stage of the steel to be cast. The condition five is: ； wherein, is the peak-to-peak value of the segment of the measured vibration signal data labg(Δti), is the frequency concentration of the segment of the measured vibration signal data labg(Δti); and are the upper and lower limits of the peak-to-peak value of the historical data l’abg for the strand-off phase, and are the upper and lower limits of the frequency concentration of the historical data l’abg for the strand-off phase. S5, the production line database is divided into sub-databases, the data in the sub-databases are selected to participate in the roll system state evaluation, and the measured data key parameter is matched with the first key parameter of the historical data based on the basic indexes of the production line database; S6, the rolling mill system state is judged based on the comparison between the measured data key parameter and the second key parameter of the historical data, and the fault position is determined by comparing the key parameter with the data solved by the key component dynamics model.

2. The method of claim 1, wherein the method is characterized by, S2 comprises the following steps: S21, the hot rolling grade, the strip width and thickness, the rolling speed, the rolling force, the finishing rolling temperature, the roll diameter and the material composition in the historical data are taken as the basic indexes of the production line database; S22, the process model data are taken to match the rolling mill data and the strip data, and the roll gap, the rolling speed, the rolling force distribution and the hot rolling strip grade of each stand are determined; S23, the strip temperature change rule and the deformation resistance increment between different stands are calculated according to the deformation resistance calculation model.

3. The method of claim 1, wherein the method is characterized by, S3 comprises the following steps: S31, call the rolling mill vibration data and classify the roll system state: the bite steel stage, the over steel stage, the waiting steel stage and the throw steel stage are taken as the rolling state; the pressing process stage, the roll changing impact stage and the parking stage are taken as the adjustment state; S32、According to the frequency concentration model, analyze the peak-to-peak value and the frequency concentration of each stage of different states of the historical vibration data corresponding to the process model data and the main frequency, establish a vibration evaluation standard library; S33, the data in the vibration evaluation standard library obtained in S32 is counted, the threshold value for distinguishing different state stages is determined under the condition that the confidence interval is 0.95, and the threshold value for distinguishing conditions one to five is obtained as the evaluation standard of the measured vibration data and the roll system state.

4. The method of claim 1, wherein the method is characterized by, In S5, the step of establishing the sub-library includes: S51, according to the real-time data basic index in S45 as index data, the basic index and the first key parameter of the historical database in S22 are screened in the deviation allowable range, and the first matching is completed; S52, according to the index data priority in S51, a historical database L'abg composed of historical data basic index, first key parameter and second key parameter is established; S53, the credibility of each level sub-library and its data l'abg is determined according to the percentage of deviation of the index data in S51, the second matching is completed; the sub-library of l'abg data and its credibility is composed of the historical data sub-library.

5. The method for real-time diagnosis of faults in a hot rolling mill taking into account the frequency concentration of vibrations according to claim 4, characterized in that, In S5, the matching of the key parameters of the measured data and the historical data includes the following steps: S54, according to the first principle that the first key parameter is equal, the sub-library with the highest credibility is selected in the L'abg library, and the historical data l'abg with the highest credibility is obtained by further matching in the sub-library; S55, comparing the historical data l'abg in S51, the data segment labg(Δtj) containing abnormal vibration state is extracted from the measured data labg.

6. The method of claim 1, wherein the method is characterized by, In S6, the following steps are included: S61, the dynamics model of the key components is used to calculate the dynamics parameters under multiple system states, and the dynamics model response labg(Δtk) under different states of the rolling mill system is theoretically calculated; S62, according to the frequency concentration model, the peak-peak value and the frequency concentration of the dynamics response labg(Δtk) of the theoretical model and the measured data labg(Δtj) are calculated; S63, according to the peak-peak value and the frequency concentration parameters, the measured fault data segment labg(Δtj) and the dynamics model response data labg(Δtk) of the theoretical model are matched, and the fault position judgment and credibility evaluation of the rolling mill system are realized.

7. A storage medium, characterized by The storage medium includes a stored program, wherein when the program runs, the hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration in any one of claims 1 to 6 is executed. The storage medium includes a stored program, wherein when the program runs, the hot continuous rolling mill fault real-time diagnosis method considering vibration frequency concentration in any one of claims 1 to 6 is executed.

Images