Roller wear scoring method and system based on data quality fusion correction

Abstract

The application discloses a kind of based on data quality fusion correction roll wear scoring method and system, it is related to rolling mill wear evaluation technical field, including: based on the mechanism model of the rolling mill to be evaluated is constructed based on tribology principle, and the mechanism predicted wear distribution of roll is calculated based on current process parameters;Based on historical roll wear data, the historical periodic wear trend of roll is calculated, and the historical predicted wear distribution of roll is calculated;The stability index and noise level index of current process parameters are calculated, and the comprehensive quality coefficient is constructed;Based on the comprehensive quality coefficient, the fusion weight coefficient is constructed, and the mechanism predicted wear distribution, historical predicted wear distribution and the roll measured wear distribution of the rolling mill to be evaluated are weighted and fused, to obtain the correction wear distribution, the roll wear score of the rolling mill to be evaluated is calculated.The application alleviates the technical problems that the existing technology introduces uncertainty and misjudgment results to the evaluation results due to the quality difference of input data.

Description

Technical Field

[0001] This invention relates to the field of rolling mill wear evaluation technology, and in particular to a rolling mill wear scoring method and system based on data quality fusion correction. Background Technology

[0002] In high-end electronics, energy storage, and precision manufacturing, copper and its alloy strips, due to their excellent conductivity, ductility, and surface quality requirements, are increasingly being developed towards ultra-thinness, high precision, and high consistency. This trend poses unprecedented challenges to the stability and control precision of rolling mill equipment. As product thickness continues to decrease to the micrometer level, even minor fluctuations in equipment status can cause strip thickness deviations, uneven shape, or surface defects, thereby affecting the reliability and consistency of end products. Against this backdrop, achieving refined status perception and health assessment of rolling mill equipment has become a crucial issue for the copper processing industry to achieve high-quality manufacturing and intelligent management.

[0003] As a key piece of equipment for achieving high-precision rolling of ultra-thin copper alloy strip, the 20-roll mill, with its complex multi-level roll system and multi-degree-of-freedom transmission characteristics, plays a crucial role in ensuring high flatness, high thickness accuracy, and excellent surface quality. However, long-term operation under high load, high friction, high temperature gradient, and high-speed alternating stress environments makes the mill system highly susceptible to degradation phenomena such as work roll wear, surface roughness deterioration, and stress fatigue damage. Work roll wear not only alters the contact stiffness distribution of the roll system, leading to a decrease in strip shape control capability and thickness unevenness, but may also cause abnormal rolling force transmission, resulting in rolling instability or even fatigue detachment of the roll surface. As the surface morphology gradually deviates from the ideal state, the rolls cannot maintain the designed geometry and roughness characteristics, ultimately leading to increased surface defects in the copper strip, decreased shape control accuracy, and a significant reduction in production yield. Especially under the current rolling process conditions for the thinnest copper foil (≤6 μm), any minute wear non-uniformity can be amplified into quality problems such as thickness fluctuations or warping, placing almost stringent requirements on the dynamic stability of the mill system.

[0004] Existing methods for evaluating roll wear mainly fall into two categories: mechanistic modeling methods and data-driven methods. Mechanistic models can reveal the physical mechanisms underlying wear formation, but their parameters are heavily dependent on field measurements and process data. When these data contain noise, are missing, or drift, the model's predictions will be significantly biased. While data-driven methods can learn complex nonlinear relationships, they are extremely sensitive to the quality of input data and are easily affected by outliers or measurement fluctuations, leading to unstable predictions. In actual production environments, data acquisition is affected by factors such as sensor accuracy, installation location, signal interference, and system latency, making it difficult to guarantee high-quality data input. Therefore, directly performing state modeling and prediction without considering data quality differences often introduces uncertainty or even misjudgments. Summary of the Invention

[0005] To address the aforementioned technical problems in existing technologies, embodiments of the present invention provide a method and system for scoring roll wear based on data quality fusion correction. The technical solution is as follows: On the one hand, a roll wear scoring method based on data quality fusion correction is provided. The method includes: constructing a mechanistic model of the roll mill to be evaluated based on tribological principles, and calculating the mechanistic predicted wear distribution of the roll based on the current process parameters of the roll mill to be evaluated and the mechanistic model; calculating the historical periodic wear trend of the roll based on the historical roll wear data of the roll mill to be evaluated, and calculating the historical predicted wear distribution of the roll based on the historical periodic wear trend; calculating the stability index and noise level index of the current process parameters, and constructing a comprehensive quality coefficient based on the stability index and the noise level index; constructing a fusion weighting coefficient based on the comprehensive quality coefficient, and weighting and fusing the mechanistic predicted wear distribution, the historical predicted wear distribution, and the measured wear distribution of the roll mill to be evaluated based on the fusion weighting coefficient to obtain a corrected wear distribution; and calculating the roll wear score of the roll mill to be evaluated based on the corrected wear distribution.

[0006] Optionally, the mechanism model includes the total wear of the rolls; the rolls include work rolls and support rolls; the mechanism model of the roll mill to be evaluated is constructed based on the tribological principle, including: calculating the wear between the work roll and the strip and the wear between the work roll and the support roll based on the tribological principle; dividing the rolls into multiple roll segments along the roll body direction, and calculating the wear value of each roll segment based on the wear amount; and superimposing the wear values ​​of each roll segment to obtain the total wear of the rolls.

[0007] Optionally, the total wear includes:

[0008] In the formula, x is the coordinate of the roll along the roll body direction, M(x) is the total wear amount, M1(x) is the wear amount between the work roll and the strip, M2(x) is the wear amount between the work roll and the support roll, F(x) is the rolling force distribution, µ1 is the wear coefficient between the work roll and the strip, and n is the number of cycles of contact between a point on the roll surface and the strip surface. It is the relative sliding distance on the contact arc, α x denoted as μ2, where μ is the metal lateral flow influence coefficient, L is the rolling mileage, f is the forward slip coefficient, D1 is the work roll diameter, H is the work roll body hardness, µ2 is the relative rolling wear coefficient between the work roll and the strip, and d is the relative rolling distance between the work roll and the strip. denoted by μ3, represents the inter-roller contact pressure distribution; µ3 is the relative sliding wear coefficient between the work roll and the first intermediate roll; µ4 is the relative rolling wear coefficient between the work roll and the support roll; and b is the relative rolling wear coefficient between the contact surfaces of the work roll and the first intermediate roll. It is the radius of the working roller.

[0009] Optionally, the historical predicted wear distribution includes:

[0010] In the formula, M pred (x) represents the historical predicted wear distribution. is the measured average wear value of the historical roll wear data, s is the slope of the historical cyclic wear trend, and Δt is the current wear cycle time point of the roll mill to be evaluated.

[0011] Optionally, the formula for calculating the stability index includes:

[0012] The calculation formula for the noise level index includes:

[0013] The overall quality coefficient includes:

[0014] In the formula, Q stab Q is the stability index. noise Here, q represents the noise level index, and q represents the overall quality coefficient. It is the standard deviation of the current process parameters within the sliding time window. It is the average value of the current process parameters within the sliding time window. It is the noise power in the current process parameters. It is the signal power of the current process parameters. Let be the weighting coefficient, satisfying .

[0015] Optionally, the correction of wear distribution includes:

[0016] In the formula, M adj (x) represents the corrected wear distribution, q represents the comprehensive quality coefficient, and M meas M(x) represents the measured wear distribution of the roll, and M(x) represents the predicted wear distribution based on the mechanism. pred (x) represents the historical predicted wear distribution, and c represents the mechanism confidence level.

[0017] Optionally, the roll wear rating includes:

[0018] In the formula, S is the wear score of the roll. M is the average value of the corrected wear distribution. linmit The maximum permissible average wear.

[0019] On the other hand, a roll wear scoring system based on data quality fusion correction is also provided to implement the roll wear scoring method based on data quality fusion correction provided in the embodiments of the present invention. The system includes: a first prediction module, a second prediction module, a calculation module, a fusion module, and a scoring module. The first prediction module is used to construct a mechanistic model of the roll mill to be evaluated based on tribological principles, and to calculate the mechanistic predicted wear distribution of the roll based on the current process parameters of the roll mill to be evaluated and the mechanistic model. The second prediction module is used to calculate the historical periodic wear trend of the roll based on the historical roll wear data of the roll mill to be evaluated. The system calculates the historical predicted wear distribution of the rolls based on the historical cyclic wear trend; the calculation module calculates the stability index and noise level index of the current process parameters, and constructs a comprehensive quality coefficient based on the stability index and the noise level index; the fusion module constructs a weighting coefficient based on the comprehensive quality coefficient, and performs weighted fusion of the mechanism-predicted wear distribution, the historical predicted wear distribution, and the measured wear distribution of the rolls of the mill to be evaluated based on the weighting coefficient to obtain a corrected wear distribution; the scoring module calculates the roll wear score of the mill to be evaluated based on the corrected wear distribution.

[0020] On the other hand, an electronic device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method provided in the embodiments of the present invention.

[0021] On the other hand, a computer-readable storage medium is also provided, wherein program code is stored in the computer-readable storage medium, and the program code can be called by a processor to execute the method provided in the embodiments of the present invention.

[0022] This invention provides a method and system for scoring roll wear based on data quality fusion correction, directly embedding data quality evaluation into the core processes of wear prediction and condition scoring. The method first performs segmented modeling of the non-uniform wear characteristics of the roll along the axial direction to improve prediction resolution. Then, it quantifies the quality of measured data from two key dimensions: stability and noise level, forming a comprehensive data quality coefficient. Next, a weighted fusion strategy is used to adjust the ratio of the predicted value to the corrected measured value, ensuring that high-quality data dominates the correction and low-quality data automatically has a reduced weight, thereby reducing the interference of noise and abnormal data on prediction accuracy. Finally, the roll wear score is calculated based on the corrected wear distribution, achieving a reliable quantitative evaluation of the roll's health status. This invention, by introducing a data quality factor, makes the wear evaluation results more robust and reliable, providing a scientific basis for precision mill maintenance and roll replacement decisions, and alleviating the technical problems of uncertainty and misjudgment introduced into the evaluation results due to differences in input data quality in existing technologies. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a roll wear scoring method based on data quality fusion correction provided by an embodiment of the present invention; Figure 2 This is a schematic diagram of a segmented work roll body provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the contact between a work roll and a blank provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of a roll wear scoring system based on data quality fusion correction provided in an embodiment of the present invention. Detailed Implementation

[0025] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0026] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0027] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0028] Figure 1 This is a flowchart of a roll wear scoring method based on data quality fusion correction according to an embodiment of the present invention. Figure 1 As shown, the method specifically includes the following steps: Step S102: Construct a mechanism model of the rolling mill to be evaluated based on the tribological principle, and calculate the mechanism-predicted wear distribution of the rolling mill based on the current process parameters of the rolling mill to be evaluated and the mechanism model.

[0029] Step S104: Calculate the historical periodic wear trend of the rolls based on the historical roll wear data of the mill to be evaluated, and calculate the historical predicted wear distribution of the rolls based on the historical periodic wear trend.

[0030] Step S106: Calculate the stability index and noise level index of the current process parameters, and construct a comprehensive quality coefficient based on the stability index and noise level index.

[0031] Step S108: Construct a fusion weighting coefficient based on the comprehensive quality coefficient, and then perform weighted fusion of the mechanism-predicted wear distribution, the historical predicted wear distribution, and the measured wear distribution of the rolls of the mill to be evaluated based on the fusion weighting coefficient to obtain the corrected wear distribution.

[0032] Step S110: Calculate the roll wear score of the mill to be evaluated based on the corrected wear distribution.

[0033] Specifically, the mechanism model includes the total wear of the rolls; the rolls include work rolls and support rolls. Step S102 also includes the following steps: Step S1021: Based on the principle of tribology, calculate the wear between the work roll and the strip, and the wear between the work roll and the support roll. Step S1022: Divide the roll into multiple roll segments along the roll body direction, and assign wear values ​​to each roll segment based on the amount of wear. Step S1023: The wear values ​​of each roll segment are summed to obtain the total wear of the roll.

[0034] Specifically, according to the tribological principle, the amount of wear on mutually contacting frictional surfaces is directly proportional to the contact load and the relative sliding and rolling distances, that is: (1) Where M is the amount of wear. F is the wear coefficient, and F is the normal contact force (N). It is the relative distance of motion (mm).

[0035] Based on equation (1), the wear amount M(x) of the work roll can be expressed by equation (2): (2) Where M1(x) represents the wear amount (μm) between the work roll and the strip, and M(x) represents the wear amount (μm) between the work roll and the first intermediate roll.

[0036] Because when calculating the wear of the work rolls during the rolling process, the rolls are evenly divided into n segments along the roll body direction, such as... Figure 2 As shown, the wear amount of each section of the roll is calculated separately and then superimposed to obtain the wear distribution of the work roll along the entire roll body direction.

[0037] The contact interface between the work roll and the strip is as follows Figure 3 As shown, the section corresponding to where the strip enters the roll gap is E. in F in ,thickness l 0, at the point where the plate exits the roll gap, the corresponding cross section is B. out C out ,thickness l 1. The work roll radius is R1, the angular velocity is β1, and the circumferential velocity of the roll at the exit is v. in The bite angle is θ. At a certain moment t when the strip contacts the work roll, the cross section E... in F in Move to the corresponding plate and strip cross section B x C x At this time, B x The line OB connecting the point to the center of the roll x The angle between the strip and the exit cross section is θ. x .

[0038] Therefore, the relative sliding wear between the work roll and the belt is: (3) Where F(x) is the rolling force distribution, µ1 is the wear coefficient between the work roll and the strip, and n is the number of cycles in which a point on the roll surface contacts the strip surface. It is the relative sliding distance on the contact arc, α x It is the influence coefficient of lateral flow of metal.

[0039] The relative rolling wear between the work roll and the belt is: (4) Therefore, the wear between the work roller and the belt is: (5) Where L is the rolling mileage, f is the forward slip coefficient, D1 is the work roll diameter, H is the work roll body hardness, µ2 is the relative rolling wear coefficient between the work roll and the strip, and d is the relative rolling distance between the work roll and the strip.

[0040] The sliding wear amount m3(x) between the work roll and the first intermediate roll is shown in equation (6): (6) in, This represents the inter-roller contact pressure distribution, and µ3 is the relative sliding wear coefficient between the work roll and the first intermediate roll. This is the relative sliding distance between the work roll and the contact surface of the belt.

[0041] Based on the principles of tribology, the rolling wear m4(x) between the work roll and the first intermediate roll is shown in equation (7): (7) Where µ4 is the relative rolling wear coefficient between the work roll and the support roll, and b is the relative rolling wear coefficient between the contact surfaces of the work roll and the first intermediate roll.

[0042] The wear between the work roll and the support roll is: (8) in, It is the radius of the working roller.

[0043] Therefore, the total wear of the work roll is: (9) Specifically, the parameters of the mechanistic model rely on field-collected data, which often contains measurement errors, missing data, delays, or noise. Direct correction may lead to unstable predictions; therefore, a data fusion evaluation of wear conditions is necessary. In reality, the wear value of the work roll is measured periodically. This invention predicts the current wear value based on historical roll wear data. Specifically, it extrapolates from historical cyclical wear trends to obtain the wear for the next cycle, i.e., using the average measured wear over the past *a* cycles to predict the wear for the next cycle. Specifically, the historical predicted wear distribution includes: (10) In the formula, M pred (x) represents the historical predicted wear distribution. is the measured average wear value of historical roll wear data, s is the slope of the historical cyclic wear trend, and Δt is the current wear cycle time point of the roll mill to be evaluated.

[0044] In equipment condition assessment, the reliability of the data directly determines the accuracy and usability of the assessment results. If the collected measured data has biases, missing data, excessive fluctuations, severe delays, or significant noise interference, even precise modeling and prediction methods may lead to distorted conclusions. Therefore, before conducting condition analysis and prediction correction, it is necessary to quantitatively evaluate the data quality and incorporate the results into subsequent calculations to improve the credibility and robustness of the final assessment results. Data quality typically includes multiple dimensions such as accuracy, completeness, stability, timeliness, and noise level. However, in roll assessment, accuracy and completeness are usually guaranteed by the measurement system and acquisition process, while timeliness mainly affects real-time monitoring rather than performance evaluation. In contrast, stability reflects the fluctuation characteristics of data over different time periods and is closely related to the balance of the roll's working state, while noise level reflects the strength of random disturbances in the signal and directly affects the ability to sensitively capture changes in roll performance. Therefore, focusing on stability and noise level indicators in roll assessment can more accurately reflect its operating characteristics and health status.

[0045] Specifically, stability metrics are used to measure the degree of data fluctuation over a short period. A high stability metric indicates smooth data changes and low noise interference, which helps in extracting reliable trend features; a poor stability metric indicates large short-term fluctuations, which may be signals of measurement noise, environmental disturbances, or equipment malfunctions. In condition assessment, stability metrics are directly related to the model's ability to capture the true trend.

[0046] Specifically, the formula for calculating the stability index includes: (11) In the formula, Q stab As a stability indicator, It is the standard deviation of the current process parameters within the sliding time window, used to measure the magnitude of data fluctuation; It is the average value of the current process parameters within the sliding time window, used to normalize the fluctuation range.

[0047] Noise level metrics reflect the ratio of noise power to signal power in data, describing the relative magnitude of useful information and unwanted interference in measurement results. Low noise level metrics indicate high data fidelity, which helps in the accurate extraction of key features; high noise level metrics, on the other hand, can mask the true signal, reducing the accuracy of parameter estimation and prediction in the model, thus requiring noise reduction processing to improve this.

[0048] Specifically, the formula for calculating the noise level index includes: (12) In the formula, Q noise As an indicator of noise level, It is the noise power in the current process parameters, estimated through filtered residuals or a noise model. It represents the signal power of the current process parameters, and the variance or power of the signal after noise reduction.

[0049] In subsequent status evaluation processes, a single quantitative indicator that can comprehensively reflect the overall data quality is often needed for direct use in calculations. To this end, weights can be assigned to each quality indicator according to its importance in the evaluation task, and a comprehensive quality coefficient can be obtained through weighted summation. This coefficient can numerically and intuitively reflect the overall reliability of the data and provide a unified quality correction factor for subsequent model calculations.

[0050] Specifically, the overall quality coefficient includes: (13) In the formula, q is the overall quality coefficient. Let be the weighting coefficient, satisfying .

[0051] Specifically, correcting the wear distribution includes: (14) In the formula, M adj (x) represents the corrected wear distribution, M meas (x) represents the measured wear distribution of the roll, and M(x) represents the mechanistic predicted wear distribution. pred (x) represents the historical predicted wear distribution, and c represents the reliability of the mechanism. When the overall quality coefficient q of the data is high, the results are mainly corrected by actual measurements (close to actual measurements). When the overall quality coefficient q is low, it reverts to a combination of mechanism and history. c determines the relative share of mechanism and history.

[0052] Specifically, the roll wear rating includes: (15) In the formula, S represents the roll wear score. To correct for the average wear distribution, M linmit The maximum permissible average wear.

[0053] The above methods can be summarized and implemented using the algorithms listed in Table 1 below:

[0054] To verify the effectiveness and practicality of the roll wear scoring method based on data quality fusion correction proposed in this invention, an embodiment of this invention selects a 20-roll mill at a copper alloy strip production site as the experimental object, and evaluates the wear status of its work rolls over a complete service cycle.

[0055] Table 2. Basic data and parameters used in the experiment

[0056] Step 1: Mechanism modeling and segmented calculation of theoretical wear.

[0057] First, based on the tribological model, the work roll is uniformly divided into 21 sections along the roll body direction ( For the rolling process parameters of Period 1, the wear amount M1(x) generated by contact with the strip and the wear amount M2(x) generated by contact with the first intermediate roll in each section are calculated using formulas (2) to (7), and then the theoretical total wear amount M(x) of this cycle is obtained by superimposing them.

[0058] Table 3 Calculation results of theoretical wear distribution for Period 1

[0059] Step 2: Trend prediction based on historical data.

[0060] Collect measured wear data of the work roll over three consecutive production cycles prior to Period 1. Based on formula (7), the wear data is calculated using the rolling mileage. Using the variable as the independent variable, calculate the slope of the historical cyclical wear trend of wear amount for each roller segment. And extrapolate to predict the historical predicted wear distribution M in Period 1 pred (x).

[0061] Step 3: Quantifying the quality of measured data.

[0062] The average stability index of the Period 1 measured data was calculated. Average noise level index .

[0063] Step 4: Data quality integration and correction.

[0064] Overall quality coefficient Because this coefficient is relatively high, the correction result is biased towards the measured value.

[0065] Step 5: Wear condition rating.

[0066] Calculate the average wear after fusion:

[0067] Set maximum allowable wear Calculate the roll wear score:

[0068] The data quality metrics for Period 2 are: Stability Indicators

[0069] Noise level index

[0070] Overall quality coefficient

[0071] Average wear after fusion ,score:

[0072] Table 4 Overall Experimental Results

[0073] The experimental results are shown in Table 4. These results demonstrate that the data quality fusion mechanism based on stability and noise level can adaptively adjust the weights of predicted and measured values: when the data quality in Period 1 is high (q=0.912), the score (74.7 points) mainly reflects the measured wear condition; while when the data quality in Period 2 decreases (q=0.822), the method automatically strengthens its reliance on the mechanistic model and historical trends, still providing a reasonable score (65.3 points). This adaptive data quality fusion strategy effectively avoids the interference of low-quality measurement data on the evaluation results, ensuring the robustness of the scoring system. Furthermore, the score results are consistent with the actual wear degree of the equipment, and the score change between the two periods is smooth and reasonable, proving the applicability of this method in engineering practice and providing a reliable basis for roll condition assessment and predictive maintenance.

[0074] As described above, this invention proposes a roll wear scoring method based on data quality fusion correction. By constructing a unified heat transfer index and roll wear scoring system, it achieves a quantitative assessment of the wear state of continuous annealing furnaces and rolls. This method innovatively incorporates data quality indicators such as stability and noise level into the evaluation process. Through weighted fusion of predicted values ​​and measured data, it significantly improves the reliability and adaptability of the evaluation results. Experimental results show that the proposed method can effectively capture the evolution trend of equipment performance and provide robust state scores under different data quality conditions, providing a scientific basis for predictive maintenance and energy efficiency optimization.

[0075] Figure 4This is a schematic diagram of a roll wear scoring system based on data quality fusion correction according to an embodiment of the present invention. Figure 4 As shown, the system includes: a first prediction module 10, a second prediction module 20, a calculation module 30, a fusion module 40, and a scoring module 50.

[0076] Specifically, the first prediction module 10 is used to construct a mechanism model of the mill to be evaluated based on the tribological principle, and to calculate the mechanism prediction wear distribution of the mill based on the current process parameters and mechanism model of the mill to be evaluated. The second prediction module 20 is used to calculate the historical periodic wear trend of the rolls based on the historical roll wear data of the roll mill to be evaluated, and to calculate the historical predicted wear distribution of the rolls based on the historical periodic wear trend. The calculation module 30 is used to calculate the stability index and noise level index of the current process parameters, and to construct a comprehensive quality coefficient based on the stability index and noise level index. The fusion module 40 is used to construct weighted coefficients based on the comprehensive quality coefficients, and to perform weighted fusion of the mechanism-predicted wear distribution, the historical predicted wear distribution, and the measured wear distribution of the rolls of the mill to be evaluated based on the weighted coefficients to obtain the corrected wear distribution; The scoring module 50 is used to calculate the roll wear score of the mill to be evaluated based on the corrected wear distribution.

[0077] The present invention also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method provided in the embodiments of the present invention.

[0078] The present invention also provides a computer-readable storage medium storing program code, which can be called by a processor to execute the method provided in the embodiments of the present invention.

[0079] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0080] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0081] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0082] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0083] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0084] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0085] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0086] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for scoring roll wear based on data quality fusion correction, characterized in that, The method includes: A mechanistic model of the rolling mill to be evaluated is constructed based on the principles of tribology, and the mechanistic prediction of wear distribution of the rolling mill is calculated based on the current process parameters of the rolling mill to be evaluated and the mechanistic model. The historical periodic wear trend of the rolls is calculated based on the historical roll wear data of the mill to be evaluated, and the historical predicted wear distribution of the rolls is calculated based on the historical periodic wear trend. Calculate the stability index and noise level index of the current process parameters, and construct a comprehensive quality coefficient based on the stability index and the noise level index; Based on the comprehensive quality coefficient, a fusion weight coefficient is constructed, and based on the fusion weight coefficient, the mechanism-predicted wear distribution, the historical predicted wear distribution, and the measured wear distribution of the mill to be evaluated are weighted and fused to obtain the corrected wear distribution; Based on the corrected wear distribution, the roll wear score of the mill to be evaluated is calculated.

2. The method according to claim 1, characterized in that, The mechanism model includes the total wear of the rolls; the rolls include work rolls and support rolls; A mechanistic model of the rolling mill to be evaluated is constructed based on tribological principles, including: Based on the principles of tribology, the wear between the work roll and the belt, and the wear between the work roll and the support roll are calculated separately. The roll is divided into multiple roll segments along the roll body direction, and the wear value of each roll segment is determined based on the wear amount. The wear values ​​of each roll segment are summed to obtain the total wear of the roll.

3. The method according to claim 2, characterized in that, The total wear includes: In the formula, x is the coordinate of the roll along the roll body direction, M(x) is the total wear amount, M1(x) is the wear amount between the work roll and the strip, M2(x) is the wear amount between the work roll and the support roll, F(x) is the rolling force distribution, µ1 is the wear coefficient between the work roll and the strip, and n is the number of cycles of contact between a point on the roll surface and the strip surface. It is the relative sliding distance on the contact arc, α x denoted as μ2, where μ is the metal lateral flow influence coefficient, L is the rolling mileage, f is the forward slip coefficient, D1 is the work roll diameter, H is the work roll body hardness, µ2 is the relative rolling wear coefficient between the work roll and the strip, and d is the relative rolling distance between the work roll and the strip. denoted by μ3, represents the inter-roller contact pressure distribution; µ3 is the relative sliding wear coefficient between the work roll and the first intermediate roll; µ4 is the relative rolling wear coefficient between the work roll and the support roll; and b is the relative rolling wear coefficient between the contact surfaces of the work roll and the first intermediate roll. It is the radius of the working roller.

4. The method according to claim 1, characterized in that, The historical predicted wear distribution includes: In the formula, M pred (x) represents the historical predicted wear distribution. is the measured average wear value of the historical roll wear data, s is the slope of the historical cyclic wear trend, and Δt is the current wear cycle time point of the roll mill to be evaluated.

5. The method according to claim 1, characterized in that, The formula for calculating the stability index includes: The calculation formula for the noise level index includes: The overall quality coefficient includes: In the formula, Q stab Q is the stability index. noise Here, q represents the noise level index, and q represents the overall quality coefficient. It is the standard deviation of the current process parameters within the sliding time window. It is the average value of the current process parameters within the sliding time window. It is the noise power in the current process parameters. It is the signal power of the current process parameters. Let be the weighting coefficient, satisfying .

6. The method according to claim 1, characterized in that, The corrected wear distribution includes: In the formula, M adj (x) represents the corrected wear distribution, q represents the comprehensive quality coefficient, and M meas M(x) represents the measured wear distribution of the roll, and M(x) represents the predicted wear distribution based on the mechanism. pred (x) represents the historical predicted wear distribution, and c represents the mechanism confidence level.

7. The method according to claim 1, characterized in that, The roll wear rating includes: In the formula, S is the wear score of the roll. M is the average value of the corrected wear distribution. linmit The maximum permissible average wear.

8. A roll wear scoring system based on data quality fusion correction, characterized in that, This system is used to implement the roll wear scoring method based on data quality fusion correction as described in any one of claims 1-7; the system comprises: a first prediction module, a second prediction module, a calculation module, a fusion module, and a scoring module; wherein... The first prediction module is used to construct a mechanism model of the mill to be evaluated based on the tribological principle, and to calculate the mechanism prediction wear distribution of the mill based on the current process parameters of the mill to be evaluated and the mechanism model. The second prediction module is used to calculate the historical periodic wear trend of the rolls based on the historical roll wear data of the mill to be evaluated, and to calculate the historical predicted wear distribution of the rolls based on the historical periodic wear trend; The calculation module is used to calculate the stability index and noise level index of the current process parameters, and to construct a comprehensive quality coefficient based on the stability index and the noise level index. The fusion module is used to construct weighting coefficients based on the comprehensive quality coefficients, and to perform weighted fusion of the mechanism-predicted wear distribution, the historical predicted wear distribution, and the measured wear distribution of the mill to be evaluated based on the weighting coefficients to obtain a corrected wear distribution; The scoring module is used to calculate the roll wear score of the mill to be evaluated based on the corrected wear distribution.

9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method as claimed in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that can be invoked by a processor to execute the method as described in any one of claims 1 to 7.

Images