Method for optimizing roller diameter of side supporting roller of eighteen-roller mill based on abrasion of backing bearing

By monitoring and optimizing the side support roll diameter in real time in an 18-roll mill, the problem of uneven wear of the backing bearings was solved, resulting in more efficient production and lower maintenance costs, thus improving production efficiency and product quality.

CN118917005APending Publication Date: 2024-11-08UNIV OF SCI & TECH BEIJING
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410570609.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-05-09
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

When facing diverse and rapidly changing market demands, the existing 18-roll mill's side support roll diameter optimization method lacks flexibility, leading to uneven wear of the backing bearings, affecting product quality and increasing maintenance costs.

Method used

A signal acquisition device is used to monitor the acoustic emission and vibration signals of the backing bearing. Noise is denoised by continuous wavelet transform and EEMD algorithm. Feature extraction and classification are performed by PSO-LSSVM algorithm. The roller diameter of the side support roller is adjusted by fuzzy control method to achieve real-time monitoring and optimization of backing bearing wear.

Benefits of technology

It improves the accuracy of wear detection, reduces wear on side support rollers, extends equipment life, improves production efficiency and product quality, and reduces production costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917005A_ABST
    Figure CN118917005A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of roller diameter design, and discloses a method for obtaining basic parameters of rolling mill equipment and basic parameters of strip steel. Arranging a signal acquisition device at a rolling mill to acquire an acoustic emission signal and a vibration signal of the backing bearing in the operation process; de-noising the acoustic emission signal and the vibration signal by using a continuous wavelet transform method, and then performing feature extraction on the signals; the extracted features are classified by adopting a PSO-LSSVM algorithm, and the wear degree of the backing bearing is divided; and based on the basic parameters of rolling mill equipment, the basic parameters of strip steel and the abrasion degree of a backing bearing of the side supporting roller, the roller diameter of the side supporting roller is adjusted through a fuzzy control method. The roller diameter of the side supporting roller is adjusted based on the fuzzy control method, effective control over abrasion of the backing bearing is achieved, meanwhile, the abrasion degree of the side supporting roller is reduced, and the service life of the side supporting roller and the service life of the backing bearing are prolonged.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of multi-roller mill roll diameter design, and in particular to a roll diameter optimization method for a side support roll of an eighteen-roller mill based on backing bearing wear. Background Art

[0002] Eighteen-roll mills are widely used in the metal rolling industry, especially playing a key role in the high-precision and high-quality rolling requirements. In the eighteen-roll cold rolling mill, due to the influence of strip thickness and production process, the roll diameter of the working roll is small, which leads to a large ratio of roll length to roll diameter (L / d). When the mill is working, the rigidity of the working roll is relatively poor, and radial distortion is prone to occur. In addition, due to the uneven force on the side support rolls and the backing bearings in contact with the side support rolls, the side support rolls and the backing bearings are worn to a greater extent.

[0003] At present, the side support roller diameter optimization technology of rolling mills mainly relies on mechanical design and new material development. Although the mechanical design method reduces wear in design, it does not take into account the impact of on-site working conditions on working roller wear. Especially in the wear problem of backing bearings, uneven wear will lead to imbalance of the rollers, affecting product quality and increasing maintenance costs. Although the new material development method has improved the service life of the rollers to a certain extent, it has not fundamentally solved the wear problem caused by uneven force during long-term or high-load rolling. In addition, although the existing mill configuration methods can cope with some conventional production needs, they are not flexible and responsive enough in the face of diversified and rapidly changing market demands. For the rolling of special materials or drastically changing production specifications, especially in complex mill operating environments, the existing methods have problems with untimely and inaccurate adjustments, and there are still limitations in data analysis and side support roller diameter optimization.

[0004] Therefore, there is a need for improving the roller diameter optimization method in the prior art. Summary of the invention

[0005] In view of this, the purpose of an embodiment of the present invention is to propose a method for optimizing the roll diameter of the side support roll of an eighteen-roll rolling mill based on backing bearing wear. The method of the present invention can monitor and optimize the roll diameter of the side support roll in real time and accurately, which helps to improve production efficiency, reduce production costs, increase production capacity and overall production efficiency, and create greater economic benefits.

[0006] Based on the above purpose, an embodiment of the present invention provides a method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing, comprising the following steps:

[0007] Obtain basic parameters of rolling mill equipment and strip steel;

[0008] A signal acquisition device is provided at the rolling mill to collect acoustic emission signals and vibration signals of the backing bearing during operation;

[0009] Continuous wavelet transform method is used to denoise the acoustic emission signal and vibration signal, and then feature extraction is performed on the signal;

[0010] The PSO-LSSVM algorithm is used to classify the extracted features and divide the wear degree of the backing bearing;

[0011] Based on the basic parameters of rolling mill equipment, basic parameters of strip steel and the wear degree of the side support roller backing bearing, the roller diameter of the side support roller is adjusted using the fuzzy control method.

[0012] In some embodiments, the basic parameters of the rolling mill equipment and the basic parameters of the strip steel include: the roll diameter of the side support rolls of the eighteen-high rolling mill, the maximum rolling force and the strip steel speed.

[0013] In some embodiments, a signal acquisition device is provided at a rolling mill to collect acoustic emission signals and vibration signals of the wear of the backing bearing during operation, including:

[0014] An acoustic emission sensor and a vibration sensor are arranged on the backing bearing saddle of the rolling mill, and the acoustic emission sensor and the vibration sensor are used to collect the acoustic emission signal and the vibration signal of the backing bearing in operation.

[0015] In some embodiments, denoising the acoustic emission signal and the vibration signal using a continuous wavelet transform method includes:

[0016] Using the continuous wavelet transform method, according to the signal characteristics and noise frequency range of the acoustic emission signal and vibration signal, the scale parameters are selected, the high-frequency noise components are removed, the key feature information is retained, the continuous wavelet transform is performed, the noise components are identified and removed, and the denoised signal is reconstructed.

[0017] In some embodiments, extracting features from a signal includes:

[0018] The EEMD algorithm is used to extract features from the signal, white noise is added to the signal, and the noisy data is decomposed using the traditional empirical mode decomposition (EMD) algorithm to obtain a series of intrinsic mode functions (IMFs). The Hilbert transform of each IMF is calculated to obtain the instantaneous frequency and amplitude. Based on the instantaneous frequency and amplitude of the IMF, the Hilbert spectrum of the signal is constructed, and characteristic parameters are extracted from the Hilbert spectrum.

[0019] In some embodiments, the characteristic parameters include: average frequency, standard deviation frequency, and energy concentration frequency.

[0020] In some implementations, classifying the extracted features and categorizing the wear degree of the backing bearing includes:

[0021] Acquire the extracted feature data, which includes key feature indicators of the acoustic emission signal and the vibration signal, and construct a feature vector;

[0022] Format the feature data, use the PSO-LSSVM algorithm to classify the extracted features, and set the initial parameters of the PSO algorithm and the LSSVM parameter range;

[0023] Execute the PSO optimization process and train the LSSVM model based on the performance evaluation results;

[0024] The trained LSSVM model is used to classify the wear degree of the backing bearing.

[0025] In some embodiments, the degree of wear includes: 0-5% for no wear, 5%-15% for light wear, 15%-30% for moderate wear, and greater than 30% for heavy wear.

[0026] In some embodiments, using the fuzzy control method to adjust the roller diameter of the side support roller includes:

[0027] Define the input variables and output variables of the fuzzy control system, set membership functions for the input variables and output variables, determine the corresponding output variables according to different combinations of input variables, and establish fuzzy rules based on the membership functions of the input variables and output variables.

[0028] In some embodiments, the input variables include wear level, strip speed, and rolling temperature, the output variable is the roller diameter adjustment amount, and the wear level includes light wear, medium wear, and heavy wear.

[0029] The present invention has at least the following beneficial technical effects:

[0030] 1. The PSO-LSSVM algorithm is used to classify the collected acoustic emission signals and vibration signals, which improves the wear detection accuracy of the side support roller backing bearing of the 18-high rolling mill, realizes the effective identification and classification of the wear degree of the side support roller backing bearing, and provides accurate and reliable data support for the subsequent roller diameter adjustment.

[0031] 2. The roller diameter of the side support roller is adjusted based on the fuzzy control method, which realizes the effective control of the wear of the backing bearing, reduces the wear degree of the side support roller, and prolongs the service life of the side support roller and the backing bearing.

[0032] 3. By analyzing the wear of the backing bearing and adjusting the roller diameter, the production preparation time caused by roller replacement and maintenance is reduced compared to traditional methods, and the overall production efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other embodiments can be obtained based on these drawings without paying creative work.

[0034] Figure 1 A schematic diagram of an embodiment of a method for optimizing the roll diameter of a side support roll of an 18-high rolling mill based on backing bearing wear provided by the present invention;

[0035] Figure 2 A schematic diagram of an embodiment of the signal acquisition device for the backing bearing of an eighteen-high rolling mill provided by the present invention.

[0036] Description of reference numerals:

[0037] 1- backing bearing; 2- side support roller; 3- saddle; 4- signal acquisition device. DETAILED DESCRIPTION

[0038] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the embodiments of the present invention are further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0039] The terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions; the terms "first", "second", etc. in the specification and claims of the present invention or the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. "Multiple" means two or more, unless otherwise clearly and specifically defined.

[0040] In the specification and claims of the present invention and the above-mentioned drawings, when an element is referred to as being "fixed to" or "mounted on" or "disposed on" or "connected to" another element, it may be directly or indirectly located on the other element. For example, when an element is referred to as being "connected to" another element, it may be directly or indirectly connected to the other element.

[0041] In addition, reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0042] like Figure 1 The figure shows a schematic diagram of an embodiment of a method for optimizing the roll diameter of a side support roll of an 18-high rolling mill based on backing bearing wear provided by the present invention, the method comprising the following steps:

[0043] S1 obtains basic parameters of rolling mill equipment and basic parameters of strip steel;

[0044] S2 A signal acquisition device is arranged at the rolling mill to collect acoustic emission signals and vibration signals of the backing bearing during operation;

[0045] S3 uses continuous wavelet transform method to denoise the acoustic emission signal and vibration signal, and then extracts the features of the signal;

[0046] S4 uses the PSO-LSSVM algorithm to classify the extracted features and divide the wear degree of the backing bearing;

[0047] S5 uses fuzzy control method to adjust the roller diameter of the side support roller based on the basic parameters of the rolling mill equipment, the basic parameters of the strip steel and the wear degree of the side support roller backing bearing.

[0048] Furthermore, in S1, the basic parameters of the rolling mill equipment include: the roll diameter φ of the side support roll of the 18-high rolling mill, the rolling force F and the strip speed v.

[0049] Furthermore, in S2, an acoustic emission signal and vibration signal collection device is arranged, such as Figure 2 Shown is a schematic diagram of an embodiment of the signal acquisition device for the backing bearing of an eighteen-high rolling mill provided by the present invention, wherein a side support roller 2 is arranged on the saddle 3, and a plurality of backing bearings 1 are sleeved on the side support roller 2, and a signal acquisition device 4 is arranged on the saddle 3 to collect acoustic emission signals and vibration signals of defects and damages of the worn backing bearing 1 during operation.

[0050] Among them, in the acoustic signal and vibration signal collection device, the acoustic emission signal device is specifically: the acoustic emission signal detection of the backing bearing wear adopts a resonant acoustic emission sensor, and the surface of the sensor detection head is closely attached to the saddle surface of the 18-roll backing bearing, and the attenuation of the backing bearing acoustic emission signal is minimized while collecting the signal. In some embodiments, a resonant acoustic emission sensor is used, the model is PXR20, the resonant frequency is 150kHz, the 10dB bandwidth is 150-600, the size is φ17×16, and the method is direct magnetic suction. In order to minimize the attenuation of the acoustic emission signal, the sensor surface is closely attached to the saddle surface of the backing bearing to collect the acoustic emission signal of the backing bearing.

[0051] Among them, in the acoustic signal and vibration signal acquisition device, the vibration signal device is specifically: the vibration signal detection of the backing bearing wear adopts a MEMS vibration sensor, and the surface of the sensor detection head is placed on the surface of the backing bearing saddle, and the attenuation of the backing bearing vibration signal is minimized while collecting the signal. In some embodiments, a MEMS vibration sensor is used, the model is VBL12RS485, the vibration frequency range is less than 1000Hz, and the three-axis acceleration time domain is less than 6600Hz. The vibration sensor is placed on the surface of the backing bearing saddle to obtain the vibration signal of the rolling bearing.

[0052] During the operation of the rolling mill, the acoustic emission signals of the defects and damage of the worn backing bearings during operation were collected by resonant acoustic emission sensors and recorded for analysis. The vibration signals of the rolling bearings were collected by MEMS vibration sensors and recorded for analysis. It was found that as the degree of wear increased, the amplitude and frequency domain peak of the time domain signal gradually increased. The frequency domain peak values ​​were particularly different at different degrees of wear, so the frequency domain peak values ​​of the acoustic emission and vibration signals M were selected. AE and M v The feature vector of the classification model is used for verification, and the acoustic emission and vibration signals of different wear degrees are extracted to form the feature vector.

[0053] Furthermore, in S3, denoising using the continuous wavelet transform method includes:

[0054] The Morlet wavelet with good time-frequency localization characteristics is selected as the wavelet basis function.

[0055] Determine the scale parameter: In order to effectively remove noise, it is necessary to conduct a preliminary analysis of the signal to determine the frequency range where the noise is mainly concentrated, and then select a scale set that covers this frequency range. According to the characteristics of the signal and the frequency range of the noise, high-frequency noise components are removed and key feature information is retained.

[0056] Perform continuous wavelet transform. The acoustic emission and vibration signals are processed by continuous wavelet transform respectively, using the selected Morlet wavelet and scale parameters to obtain the time-frequency representation of the signal.

[0057] Identify and remove noise components. By analyzing the time-frequency representation, identify the frequency range where the main noise components are located and remove these components in the wavelet coefficients.

[0058] Reconstruct the denoised signal. Use the wavelet coefficients after noise removal to reconstruct the signal and obtain the processed acoustic emission and vibration signals.

[0059] Extract features from complex acoustic emission and vibration signals using the ensemble empirical mode decomposition method. Decompose the denoised signal into a series of intrinsic mode functions (IMFs), each of which captures different frequency components of the signal.

[0060] Each IMF is analyzed to extract key features, such as average frequency, energy distribution, amplitude, etc. The extracted features are constructed into feature vectors for identification and classification of backing bearing wear.

[0061] The empirical mode decomposition (EEMD) algorithm is used to denoise and decompose the acoustic emission signal and vibration signal. The EEMD algorithm is an adaptive data analysis method that can be used to process nonlinear and non-stationary data. First, white noise is added to the signal, and then the noisy data is decomposed using the traditional empirical mode decomposition (EMD) algorithm to obtain a series of intrinsic mode functions (IMFs). The Hilbert transform of each IMF is calculated to obtain the instantaneous frequency and amplitude. The Hilbert transform formula is as follows:

[0062]

[0063] Where H(t) is the Hilbert transform of the signal x(t) and P represents the principal value integral.

[0064] Based on the instantaneous frequency and amplitude of the IMF, the Hilbert spectrum of the signal is constructed. The Hilbert spectrum is a time-frequency distribution graph that shows how the frequency components of a signal change over time.

[0065] Extract characteristic parameters from the Hilbert spectrum, such as average frequency, standard deviation frequency, energy concentration frequency, etc. These parameters can be used to describe the characteristics of the signal. The specific calculation formula is as follows:

[0066]

[0067] The calculation formula of average frequency is:

[0068]

[0069] The calculation formula of standard deviation frequency is:

[0070]

[0071] The calculation formula of energy concentration frequency is:

[0072]

[0073] Among them, f i is the instantaneous frequency of the ith IMF, A i is the corresponding amplitude and n is the total number of IMFs.

[0074] Furthermore, the denoised signal is subjected to feature extraction using the EEMD algorithm; wherein white noise is added to the signal, and then the noisy data is decomposed using the traditional empirical mode decomposition (EMD) algorithm to obtain a series of intrinsic mode functions (IMFs). The Hilbert transform of each IMF is calculated to obtain the instantaneous frequency and amplitude; wherein, based on the instantaneous frequency and amplitude of the IMF, the Hilbert spectrum of the signal is constructed. The Hilbert spectrum is a time-frequency distribution diagram; wherein characteristic parameters such as the average frequency, standard deviation frequency, energy concentration frequency, etc. are extracted from the Hilbert spectrum to describe the characteristics of the signal.

[0075] The EEMD algorithm is used to extract features from the collected signals, including:

[0076] The collected original acoustic emission signal and vibration signal are denoised using continuous wavelet transform method; assuming ψ(t)∈L 2 (R), whose Fourier transform is Then its wavelet coefficient is:

[0077]

[0078] Where ψ(t) is a basic wavelet. After scaling and translation, ψ(t) can be obtained:

[0079]

[0080] Where ψ(t) is a basic wavelet, a and b represent the scaling factor and translation factor respectively. For the analyzed signal, ψ a,b (t) is equivalent to an observation window, so:

[0081]

[0082] Therefore, for any f(t)∈L 2 (R), the continuous wavelet transform of the signal f(t) is:

[0083]

[0084] Function f,g∈L 2 The convolution of (R) can be expressed as:

[0085]

[0086] make So:

[0087]

[0088] The EEMD algorithm is used to extract effective and significantly different waveform signal features from bearing wear acoustic emission and vibration signals for identification of backing bearing wear.

[0089] Among them, the extraction of effective and significantly different waveform signal features is specifically described as follows: with the increase of wear degree, the amplitude and frequency domain peak of the time domain signal are increasing, and the frequency domain peak is particularly different at different wear degrees. Therefore, the frequency domain peak of the acoustic emission and vibration signal is selected as the classification model feature vector for verification; the acoustic emission and vibration signals of different wear degrees are extracted to form a feature vector, and a scatter plot of the composed feature vector is drawn.

[0090] Further, in S4, the PSO-LSSVM algorithm is used to classify the extracted features and divide the wear degree of the backing bearing into:

[0091] Prepare classified data and organize data. Collect and organize the feature data extracted by the EEMD algorithm, including the key feature indicators of acoustic emission signals and vibration signals, and construct feature vectors, which represent individual features;

[0092] Data formatting: Format the feature data into a form acceptable to the PSO-LSSVM algorithm to ensure that each data point contains the necessary feature information and corresponding labels.

[0093] Set the initial parameters of the PSO algorithm. This includes:

[0094] The number of particles: determines the coverage of the search space; the initial position and speed: affect the starting point and dynamic change process of the algorithm in finding the optimal solution.

[0095] Set the LSSVM parameter range, kernel parameter and regularization parameter, and set the search range of kernel parameter and regularization parameter for the LSSVM model to ensure that the PSO algorithm can find the optimal solution in a suitable space.

[0096] Execute the PSO optimization process.

[0097] Specifically, in this embodiment, executing the PSO optimization process includes the following steps:

[0098] Initialize the particle swarm: Each particle represents a set of LSSVM parameters. Initially, the particles are randomly distributed in the parameter space;

[0099] Fitness evaluation: The fitness of each particle (i.e., the performance of the LSSVM model) is evaluated by cross-validation. A high accuracy means a better fitness;

[0100] Update particle position and velocity: adjust the position and velocity of each particle in the parameter space according to the individual best solution of each particle and the best solution of the entire group;

[0101] Iterative process: It is carried out iteratively. In each iteration, the particle updates its position in the parameter space based on its own experience and the experience of the group. This process aims to find the optimal combination of parameters so that the LSSVM model can achieve the best performance;

[0102] Train the LSSVM model and evaluate the model performance. Train the model: Use the optimal parameters determined by the PSO algorithm, the kernel parameter γ and the regularization parameter C, to train the LSSVM model. The model receives the input feature vector and learns how to map it to the corresponding wear degree category. Evaluate performance: Evaluate the performance of the model on the wear degree classification task through indicators such as accuracy, recall, and F1 score. These indicators help understand the model's classification ability on different categories and its overall performance.

[0103] Based on the performance evaluation results, the LSSVM model is adjusted.

[0104] Specifically, in this embodiment, adjusting the LSSVM model includes the following steps:

[0105] Parameter re-optimization: If the model performance is not ideal, it is necessary to return to the PSO optimization step and adjust the PSO parameters, such as particle speed, learning factor, etc., to explore better LSSVM parameters.

[0106] Model structure adjustment: Depending on the performance of the model, consider adjusting the LSSVM model itself, such as choosing a different kernel function or modifying the loss function.

[0107] Feature selection: Analyzing the performance of a model may reveal that some features are more informative than others. In this case, the feature set can be adjusted to retain only the most useful features.

[0108] Determine the wear classification, record and interpret the results. Wear classification: Use the trained LSSVM model to classify the wear of the backing bearing. Classify the wear into four levels: 0-5% is no wear, 5%-15% is light wear, 15%-30% is moderate wear, and above 30% is heavy wear.

[0109] Result recording and interpretation: Record the classification results in detail, and explain and analyze the degree of wear at each level. This information will serve as an important basis for subsequent roller diameter optimization.

[0110] Further, in S5, based on the basic parameters of the 18-high rolling mill, the strip parameters, the rolling process parameters and the wear degree of the side support roller backing bearing, the fuzzy control method is used to adjust the roller diameter of the side support roller to reduce the wear, specifically:

[0111] According to the characteristics of the 18-high reversible rolling mill and the process requirements, the fuzzy control rules are set to define the input and output variables of the fuzzy control system. The input variables are: wear level, strip speed, rolling temperature; the output variable is: roll diameter adjustment amount.

[0112] The wear levels are divided into three levels: light wear, moderate wear, and heavy wear. Specifically: 5%-15% is light wear, 15%-30% is moderate wear, and more than 30% is heavy wear; no wear does not participate in the input variables.

[0113] In addition to the degree of wear, membership functions are defined for each input and output variable, specifically:

[0114] Rolling speed: low speed (trapezoidal function, covering range 0-0.4), medium speed (trapezoidal function, covering range 0.3-0.7), high speed (trapezoidal function, covering range 0.6-1); Rolling force: small (Gaussian function, center at the low end), medium (Gaussian function, center in the middle area), large (Gaussian function, center at the high end); Roller diameter adjustment: no adjustment required (triangular function, covering range 0-0.2), tiny (triangular function, covering range 0.1-0.4), small (triangular function, covering range 0.3-0.6), medium (triangular function, covering range 0.5-0.8), large (triangular function, covering range 0.7-1).

[0115] Fuzzy rules are established based on the membership functions of input and output variables, as shown in Table 1:

[0116] Table 1 Fuzzy control rules for roller diameter adjustment of 18-high rolling mill

[0117]

[0118]

[0119] In some embodiments, the roller diameter adjustment amount is 2% of the roller diameter, the small roller diameter adjustment amount is 5% of the roller diameter, the medium roller diameter adjustment amount is 8% of the roller diameter, and the large roller diameter adjustment amount is 12% of the roller diameter; wherein the slow speed is 0%-30% of the maximum safe speed of the strip steel, the medium speed is 30%-60% of the maximum safe speed of the strip steel, and the fast speed is 60%-100% of the maximum safe speed of the strip steel; wherein the small rolling force is 0-35% of the maximum rolling force, the medium rolling force is 35-75% of the maximum rolling force, and the large rolling force is 75%-100% of the maximum rolling force;

[0120] This embodiment provides a method for optimizing the roller diameter of the side support roller of an eighteen-roll mill based on the wear of the backing bearing. First, the basic parameters of the rolling mill equipment and the strip parameters are collected, and then the operating signals of the backing bearing are collected using acoustic emission and vibration sensors. Next, the continuous wavelet transform and EEMD algorithm are used to denoise and extract features of these signals, thereby constructing a feature vector for classifying the degree of bearing wear. Subsequently, the feature vectors are classified by the PSO-LSSVM algorithm, and the algorithm parameters are optimized based on the model performance evaluation results. Finally, based on the degree of wear, strip speed and rolling temperature, the roller diameter is adjusted by designing a fuzzy controller to reduce wear and improve product quality and output. The implementation of this method not only improves the operating efficiency of the rolling mill equipment, reduces the work preparation time, but also extends the service life of the equipment, providing an effective solution for the efficient and stable operation of the rolling mill and the management of backing bearing wear in strip production.

[0121] The present invention will be further explained below in conjunction with specific embodiments.

[0122] Based on a reversible single-stand 18-high rolling mill production line in a steel plant, the collected data can be obtained: the roll diameter of the side support roll of the 18-high rolling mill Rolling force F = 9600KN and strip speed v = 500m / min. From the technical parameters of the 18-high rolling mill, it can be seen that the maximum rolling force of the rolling mill is 12000KN and the maximum speed v = 900m / min. After collecting signals through sensors and processing data, the wear degree of the backing bearing of a certain 18-high rolling mill is 11%, which is a light wear.

[0123] First determine the membership of each input variable:

[0124] Among them, the strip speed is medium speed: the strip speed membership function is a trapezoidal function, the speed range is 0 to 1, and the medium speed covers a range of 0.3-0.7. In the medium speed area, the membership is 1. Therefore, the medium speed membership of the strip speed is 1, and the strip speed is medium speed;

[0125] Among them, the rolling force is between the medium rolling force and the large rolling force: the rolling force range is 0-1, and the membership is represented by Gaussian function. Assume that the standard deviation of Gaussian function σ = 0.1, and the center of Gaussian function c of medium rolling force 中 = 0.5, the center of the Gaussian function of the large rolling force c 大 =0.8.

[0126] Then the membership of the rolling force in the medium force and the large force are:

[0127]

[0128]

[0129] Since the rolling force F = 9300KN is between the medium rolling force and the large rolling force, we need to standardize it to the range of 0-1 for calculation, as follows:

[0130]

[0131] Here F norm is the standardized value between 0 and 1, and F can be calculated norm =0.775According to F norm Calculate the membership, where:

[0132] Membership degree μ of medium rolling force 中 :

[0133]

[0134] Membership degree μ of large rolling force 大 :

[0135]

[0136] This shows that at a rolling force of 9300 kN, the force is almost completely classified as a large rolling force, while the correlation with the medium rolling force is very low. These membership values ​​will then be used to apply fuzzy rules and perform rule synthesis and defuzzification to calculate the roll diameter adjustment

[0137] Use the centroid method to defuzzify and convert the fuzzy output into a clear action, where x i is the value of the output variable, μ(x i ) is in x i The output membership at . The defuzzified result x * :

[0138]

[0139] According to the corresponding fuzzy rules: if the strip speed is medium, the rolling force is medium, and the wear degree is light, the roller diameter adjustment amount is small; if the strip speed is medium, the rolling force is large, and the wear degree is light, the roller diameter adjustment amount is small; combined with the coverage range of the triangular function, the peak value of the small adjustment is set to 0.25 and the peak value of the small adjustment is set to 0.45, then the defuzzification formula is:

[0140]

[0141] The calculated roll diameter optimization amount after defuzzification is 0.254597, that is, under the given strip speed, rolling force and wear degree, the fuzzy controller makes a small roll diameter adjustment. According to the site conditions, the hydraulic cylinder is adjusted accordingly to achieve the requirement of 2% roll diameter optimization.

[0142] The above are exemplary embodiments disclosed in the present invention, but it should be noted that various changes and modifications may be made without departing from the scope disclosed in the embodiments of the present invention as defined in the claims. The functions, steps and / or actions of the method claims according to the disclosed embodiments described herein do not need to be performed in any particular order. In addition, although the elements disclosed in the embodiments of the present invention may be described or required in individual form, they may also be understood as multiple unless explicitly limited to the singular.

[0143] It should be understood that, as used herein, the singular forms "a", "an" are intended to include the plural forms as well, unless the context clearly supports an exception. It should also be understood that, as used herein, "and / or" refers to any and all possible combinations including one or more of the associated listed items.

[0144] The serial numbers of the embodiments disclosed in the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0145] A person skilled in the art should understand that the discussion of any of the above embodiments is only exemplary and is not intended to imply that the scope of the disclosure of the embodiments of the present invention (including the claims) is limited to these examples; under the concept of the embodiments of the present invention, the technical features in the above embodiments or different embodiments can also be combined, and there are many other changes in different aspects of the embodiments of the present invention as above, which are not provided in detail for the sake of simplicity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention should be included in the protection scope of the embodiments of the present invention.

Claims

1. A method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing, characterized in that: include: Obtain basic parameters of rolling mill equipment and strip steel; A signal acquisition device is arranged on the rolling mill to collect acoustic emission signals and vibration signals of the backing bearing during operation; Using continuous wavelet transform method to perform denoising on the acoustic emission signal and the vibration signal, and then extracting features from the signal; The PSO-LSSVM algorithm is used to classify the extracted features and divide the wear degree of the backing bearing; Based on the basic parameters of the rolling mill equipment, the basic parameters of the strip steel and the wear degree of the backing bearing of the side support roller, the roller diameter of the side support roller is adjusted by using a fuzzy control method.

2. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 1, characterized in that: The basic parameters of the rolling mill equipment and the basic parameters of the strip steel include: the roll diameter of the side support rolls of the 18-high rolling mill, the maximum rolling force and the strip steel speed.

3. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 1, characterized in that: The signal acquisition device is arranged at the rolling mill to collect the acoustic emission signal and vibration signal of the wear of the backing bearing during operation, including: An acoustic emission sensor and a vibration sensor are arranged on the backing bearing saddle of the rolling mill, and the acoustic emission sensor and the vibration sensor are used to collect the acoustic emission signal and the vibration signal of the backing bearing in operation.

4. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 1, characterized in that: Using continuous wavelet transform method to denoise the acoustic emission signal and the vibration signal includes: Using the continuous wavelet transform method, according to the signal characteristics and noise frequency range of the acoustic emission signal and the vibration signal, the scale parameter is selected, the high-frequency noise component is removed, the key feature information is retained, the continuous wavelet transform is performed, the noise component is identified and removed, and the denoised signal is reconstructed.

5. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 1, characterized in that: Feature extraction of signals includes: The EEMD algorithm is used to extract features from the signal, white noise is added to the signal, and the noisy data is decomposed by the traditional empirical mode decomposition (EMD) algorithm to obtain a series of intrinsic mode functions (IMFs). The Hilbert transform of each IMF is calculated to obtain the instantaneous frequency and amplitude. Based on the instantaneous frequency and amplitude of the IMF, the Hilbert spectrum of the signal is constructed, and characteristic parameters are extracted from the Hilbert spectrum.

6. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 5, characterized in that: The characteristic parameters include: average frequency, standard deviation frequency, and energy concentration frequency.

7. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 1, characterized in that: The PSO-LSSVM algorithm is used to classify the extracted features and divide the wear degree of the backing bearing into: Acquire extracted feature data, the feature data including key feature indicators of the acoustic emission signal and the vibration signal, and construct a feature vector; Format the feature data, use the PSO-LSSVM algorithm to classify the extracted features, and set the initial parameters of the PSO algorithm and the LSSVM parameter range; Execute the PSO optimization process and train the LSSVM model based on the performance evaluation results; The trained LSSVM model is used to classify the wear degree of the backing bearing.

8. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 7, characterized in that: The degree of wear includes: 0-5% for no wear, 5%-15% for light wear, 15%-30% for moderate wear, and greater than 30% for heavy wear.

9. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 1, characterized in that: The method of using fuzzy control to adjust the roller diameter of the side support roller includes: The input variables and output variables of the fuzzy control system are defined, membership functions are set for the input variables and the output variables, corresponding output variables are determined according to different combinations of input variables, and fuzzy rules are established based on the membership functions of the input variables and the output variables.

10. The method for optimizing the roll diameter of the side support roll of an 18-high rolling mill based on the wear of the backing bearing according to claim 9, characterized in that: The input variables include wear level, strip speed, and rolling temperature; the output variable is the roller diameter adjustment amount; and the wear level includes light wear, moderate wear, and heavy wear.