Seamless steel pipe dimensional parameter detection and deviation correction system based on digital twin

Abstract

This invention specifically relates to the field of digital twin technology, and discloses a seamless steel pipe dimensional parameter detection and deviation correction system based on digital twins. The system includes a multi-dimensional data acquisition module, a digital twin model construction module, a deviation quantification and evaluation module, a correction scheme generation module, a correction scheme iteration module, and a knowledge model optimization module. By constructing a real-time transmission mechanism of data acquisition, deviation analysis, and correction decision-making, the multi-dimensional data acquisition module collects all dimensions of the seamless steel pipe in real time and synchronously inputs them into the digital model of the steel pipe. A full-process data-driven quantification model is constructed, and the deviation quantification and evaluation model accurately calculates the dimensional deviation degree and surface quality score, classifies the comprehensive deviation level, generates an initial correction scheme based on the deviation type and level, and iteratively optimizes the correction scheme to quantify and adjust parameters. This constructs a closed-loop structure of iterative correction, case accumulation, and model optimization, enabling continuous system evolution and improving production efficiency.

Description

Technical Field

[0001] This invention relates to the field of digital twin technology, and more specifically, to a system for detecting and correcting deviations in the dimensional parameters of seamless steel pipes based on digital twins. Background Technology

[0002] As a key basic component in the industrial field, the dimensional accuracy and surface quality of seamless steel pipes directly determine the safety and reliability of downstream applications. With the continuous improvement of intelligent manufacturing and industrial quality control requirements, traditional testing technologies cannot meet the goals of efficient and accurate quality control, which has driven the development of seamless steel pipe testing technology towards digital twins.

[0003] Existing seamless steel pipe inspection technologies mainly rely on conventional and quality parameters for testing. They focus on the basic geometric dimensions of seamless steel pipes and the surface quality, and develop correction plans based on the operational experience of staff. However, in actual use, these technologies still have some shortcomings. First, they are slow to respond and lack real-time linkage. Existing seamless steel pipe inspection and correction systems have not established a real-time transmission mechanism for inspection data and correction operations. The inspection and correction processes are independent of each other, with a significant time interval. This makes it difficult to respond and correct quickly when inspection deviations or surface defects are found, which can easily lead to the accumulation of unqualified products, affect production efficiency, and increase the difficulty of subsequent corrections.

[0004] Second, the correction scheme lacks quantitative basis. The existing seamless steel pipe inspection and correction system has not established a data-based quantitative model for the formulation of correction schemes. It relies heavily on the work experience of technical personnel. The rationality and effectiveness of the schemes lack unified standards, and the judgment of deviation level lacks clear quantitative indicators, resulting in unstable correction effects. It cannot guarantee the consistency of seamless steel pipe dimensional accuracy and surface quality, and lacks accurate quantitative basis and standardized support.

[0005] Third, the lack of a closed-loop continuous optimization mechanism means that the existing seamless steel pipe inspection and correction system can only complete a single inspection and correction process. It lacks a systematic storage, analysis and feedback mechanism for the entire process data and cannot reuse past successful experiences, resulting in the system being in a static operating state and making it difficult to adapt to the technical requirements of continuous iterative optimization in intelligent manufacturing. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a seamless steel pipe dimensional parameter detection and deviation correction system based on digital twins. By constructing a real-time transmission mechanism for data acquisition-deviation analysis-correction decision, a full-process data-driven quantitative model, and a closed-loop structure of iterative correction-case accumulation-model optimization, the system effectively solves the problems of delayed response and lack of real-time linkage, lack of quantitative basis for correction schemes, and lack of closed-loop continuous optimization mechanism proposed in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a seamless steel pipe dimensional parameter detection and deviation correction system based on digital twins, comprising a multi-dimensional data acquisition module, a digital twin model construction module, a deviation quantification and evaluation module, a correction scheme generation module, a correction scheme iteration module, and a knowledge model optimization module.

[0008] Multi-dimensional data acquisition module: Collects parameters of seamless steel pipes in all dimensions, constructs a comprehensive inspection dataset of seamless steel pipes, and transmits it to the digital twin model construction module;

[0009] Digital twin model construction module: Based on the seamless steel pipe full-dimensional inspection dataset, construct a digital model of the steel pipe, a defect knowledge graph, and a correction process parameter mapping library;

[0010] Deviation Quantification Assessment Module: Calculates the dimensional deviation and surface quality score of the seamless steel pipe based on the full-dimensional inspection dataset, and obtains the comprehensive deviation level based on the dimensional deviation and surface quality score;

[0011] Correction scheme generation module: Generates initial correction schemes based on dimensional deviation, surface quality score, and comprehensive deviation level matching correction process parameter mapping library;

[0012] The iterative correction module verifies the correction effect based on the digital model of the steel pipe and the initial correction scheme, formulates a multi-round correction convergence control strategy, and outputs the final correction result.

[0013] Knowledge model optimization module: Based on the final correction results, case studies and model parameter liberalization are carried out to iteratively update the digital model of steel pipe.

[0014] The technical effects and advantages of this invention are as follows:

[0015] 1. This invention constructs a real-time transmission mechanism for data acquisition, deviation analysis, and correction decision-making through digital twin technology, breaking down the independent barriers between the detection and correction links. It collects all-dimensional parameters of seamless steel pipes in real time through a multi-dimensional data acquisition module, synchronously inputs them into the digital model of the steel pipe, and performs deviation quantification evaluation through a deviation quantification evaluation module. After comprehensively judging the deviation level, an initial correction plan is generated, which effectively avoids the accumulation of unqualified products and improves production efficiency.

[0016] 2. This invention constructs a full-process data-driven quantitative model, accurately calculates dimensional deviation and surface quality scores through a deviation quantitative evaluation model, and classifies comprehensive deviation levels to eliminate the subjective ambiguity in deviation level judgment. It generates an initial correction scheme based on the deviation type and level, and calculates P2 to quantify adjustment parameters, unifying the standards for the rationality and effectiveness of the scheme, and ensuring product quality consistency.

[0017] 3. This invention meets the continuous needs of intelligent manufacturing by constructing a closed-loop structure of iterative correction, case accumulation, and model optimization. By recording the effect value of each correction and iteration data, a case database is built, and the reasons for failure and improvement plans of failed cases are added. The correction algorithm coefficients are optimized regularly, and the digital model of steel pipe and defect knowledge graph are updated to achieve continuous system evolution and adapt to the technical requirements of dynamic optimization. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0019] Figure 2 This is a schematic diagram illustrating the steps of constructing a digital model of a steel pipe according to the present invention.

[0020] Figure 3 This is a schematic diagram illustrating the steps for generating the initial modified scheme of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] As attached Figure 1 The seamless steel pipe dimensional parameter detection and deviation correction system based on digital twin shown includes a multi-dimensional data acquisition module, a digital twin model construction module, a deviation quantification and evaluation module, a correction scheme generation module, a correction scheme iteration module, and a knowledge model optimization module.

[0023] In a more specific application of this invention, the output of the multi-dimensional data acquisition module is connected to the input of the digital twin model construction module, the output of the digital twin model construction module is connected to the input of the deviation quantification evaluation module, the output of the deviation quantification evaluation module is connected to the input of the correction scheme generation module, and the output of the correction scheme generation module is connected to the input of the knowledge model optimization module.

[0024] The specific embodiments of the present invention include the following:

[0025] Multi-dimensional data acquisition module: Collects parameters of seamless steel pipes in all dimensions, constructs a comprehensive inspection dataset of seamless steel pipes, and transmits it to the digital twin model construction module;

[0026] Furthermore, the parameters of seamless steel pipes across all dimensions include conventional dimensional parameters, surface quality parameters, and material performance parameters. Conventional dimensional parameters include the outer diameter D0, inner diameter d0, wall thickness T0, and length L0. Surface quality parameters include surface roughness R. a0 The number, type, and size of defects; material property parameters refer to hardness H. B and material composition.

[0027] In this embodiment, it should be specifically explained that the outer and inner diameters of the steel pipe are collected using a laser diameter gauge, the wall thickness is collected using an ultrasonic thickness gauge, and the length of the steel pipe is measured using a photoelectric encoder. For surface quality parameters, the surface roughness, number of defects, defect type, and defect size of the steel pipe are collected using machine vision. Defect types include dents, protrusions, and scratches, and defect sizes include length, width, and depth. Surface quality parameters are used to quantify the surface quality state. Material performance parameters include hardness collected using an online hardness tester and material composition obtained using a material analyzer, which are used to supplement the physical performance data of the steel pipe.

[0028] It should be further explained that the seamless steel pipe full-dimensional inspection dataset is represented as {D0, d0, T0, L0, R}. a0 Defect type and size, H B Material composition}.

[0029] Digital twin model construction module: Based on the seamless steel pipe full-dimensional inspection dataset, construct a digital model of the steel pipe, a defect knowledge graph, and a correction process parameter mapping library;

[0030] Furthermore, such as Figure 2 As shown, the steps for constructing a digital model of a steel pipe are as follows:

[0031] S1.1: Obtain the seamless steel pipe full-dimensional inspection dataset and filter the core input variables in the seamless steel pipe full-dimensional inspection dataset, including the outer diameter, wall thickness, length and surface roughness of the steel pipe;

[0032] In this embodiment, it should be specifically noted that the conventional dimensional parameters D, T, and L in the seamless steel pipe full-dimensional inspection dataset are the basis for determining the macroscopic geometry of the steel pipe, directly reflecting the diameter, wall thickness uniformity, and overall length of the steel pipe, as well as the surface roughness R. a The microscopic manifestations that affect the geometry of steel pipes are considered. By using the above parameters as core input variables, the model can be ensured to cover both macroscopic structure and microscopic surface features, thus avoiding significant deviations between the model and the actual geometry due to neglecting microscopic characteristics.

[0033] S1.2: Construct a digital model of the steel pipe based on the core input variables and the geometric characteristics of the seamless steel pipe, represented as M=f(D,T,L,R). aThis process transforms the core input variables into digital geometric data.

[0034] In this embodiment, it should be specifically explained that the construction of the digital model of the steel pipe takes the ideal geometric shape as the output target. For example, for a cylindrical seamless steel pipe, the function f calculates the inner diameter d=D-2T based on D and T, generates the three-dimensional coordinate data of the cylindrical surface by combining the length L, and converts the surface roughness Ra into digital parameters of micro-surface undulations. Finally, it is integrated into a digital model of the steel pipe that includes macro-structure and micro-surface features.

[0035] S1.3: Obtain real-time data from the seamless steel pipe full-dimensional inspection dataset, input it into the steel pipe digital model, obtain the initial steel pipe digital model, and verify and calibrate the steel pipe digital model.

[0036] In this embodiment, it is necessary to specifically explain the selection of the actual tested steel pipe parameters D0, T0, L0, and R. a0 Substituting the data into the constructed digital model of the steel pipe, an initial digital model of the steel pipe is generated. Model verification and calibration refers to obtaining the three-dimensional coordinate data of the actual physical shape of the target steel pipe through a high-precision three-dimensional scanning device as the reference data. The geometric parameters of the initial digital model of the steel pipe and the reference data are compared, and the error value is calculated. If the error is >0.1%, the calculation coefficients in the model are adjusted, such as the micro-fluctuation coefficient or the size conversion coefficient. The comparison and adjustment process is repeated until the error is <0.1%, ensuring that the model accurately reproduces the geometric shape of the steel pipe and provides reliable support for subsequent correction decisions.

[0037] Furthermore, the defect knowledge graph includes a defect type classification system, defect morphology quantitative dimension features, and defect identification standards. The correction process parameter mapping library is a structured database that associates seamless steel pipe deviation type, deviation level, and correction parameter.

[0038] In this embodiment, it should be specifically noted that the defect type classification system clearly covers typical surface and near-surface defects in seamless steel pipe production, including but not limited to dents, protrusions, scratches, cracks, inclusions, porosity, scales, folds, pits, abrasions, and craters. Each defect type corresponds to a commonly used defect definition in the industry, ensuring the standardization and universality of defect classification. The defect morphology quantification dimension features each defect type are matched with multi-dimensional quantification features, including basic attributes, size features, morphological features, and distribution features. The basic attributes include the location of the defect and the process in which the defect occurred, and the size features... The defects include length, width, depth, area, and volume; morphological characteristics include defect regularity, edge smoothness, and surface flatness; and distribution characteristics include defect distribution density and distribution direction. Defect identification standards refer to setting clear identification thresholds and risk levels for each defect type and its corresponding quantitative characteristics. For example, for scratch defects, when a defect length of 5-50mm, width of 0.5-5mm, and depth of 0.1-0.5mm is detected, it is judged as a normal scratch; if the depth is greater than 0.5mm or the length is greater than 50mm, it is judged as a severe scratch, corresponding to a high-risk level.

[0039] It should be further explained that the modified process parameter mapping library specifically includes a deviation type and level classification system, modified process parameters, correction intensity, and deviation-parameter association rules. The deviation type and level classification system divides deviations into dimensional deviation categories and surface defect deviation categories. The dimensional deviation category includes diameter deviation, wall thickness deviation, and length deviation. The surface defect deviation category corresponds to the defect type in the defect knowledge graph, which is divided into four levels: minor, general, severe, and fatal.

[0040] Correction process parameters refer to the corresponding correction process parameters matched for different deviation types and levels, including dimensional deviation correction parameters and surface defect deviation correction parameters. Among them, the diameter / wall thickness deviation in the dimensional deviation correction parameters corresponds to the straightening roll pressure, straightening speed, rolling force, and rolling temperature; the length deviation corresponds to the cutter feed speed and cut-off position speed; the scratch / dent / protrusion in the surface defect deviation correction parameters corresponds to the grinding head feed amount, grinding time, polishing speed, and grinding wheel force; and the oxide scale / scab corresponds to the pickling concentration, pickling time, and passivation temperature. The deviation-parameter association rules are used to clarify the correspondence between each deviation type and level and the correction parameters, forming directly searchable association entries, providing direct data support for the formulation of correction schemes.

[0041] Deviation Quantification Assessment Module: Calculates the dimensional deviation and surface quality score of the seamless steel pipe based on the full-dimensional inspection dataset, and obtains the comprehensive deviation level based on the dimensional deviation and surface quality score;

[0042] Furthermore, calculating the dimensional deviation requires obtaining the outer diameter D0, wall thickness T0, and length L0 of the seamless steel pipe from the full-dimensional inspection dataset, based on the outer diameter and the target value D0. t Substitute into the formula:

[0043] ,

[0044] The outer diameter deviation ΔD is calculated, and based on the wall thickness and the target wall thickness T. t Length L t The wall thickness deviation ΔT and length deviation ΔL are calculated based on the target length value, respectively; the surface quality score requires obtaining the surface quality parameter roughness score S. R Defect density score S M and defect depth score S S The surface quality score Q is obtained by weighted calculation, where the weight coefficients of roughness score, defect density score and defect depth score are w1, w2 and w3, respectively.

[0045] In this embodiment, it is necessary to specifically explain D. t T t and L t It is a standard value preset according to the seamless steel pipe production standards or downstream application requirements; roughness score S R The calculation requires the surface roughness R from the surface quality parameters. a0 Based on the application scenarios of seamless steel pipes, industry quality standards, and customer needs, a surface roughness rating table was developed. The grading standards are 0.8μm, 1.6μm, and 3.2μm, assigning scores of 100, 80, 60, and 40 to the surface roughness within these four ranges, respectively. The defect density rating S... M The calculation requires calculating the steel pipe area S based on the number of defects in the surface quality parameters, the outer diameter and length of the steel pipe, and then calculating the defect density. The defect density grading standard is 1 defect / m. 2 3 / m 2 And 5 / m 2 The defect density in each of the four intervals is assigned a score of 100, 80, 60, and 40, respectively; the defect depth score is S. S The calculation requires the defect depth in the surface quality parameters and the maximum defect depth value in the target steel pipe is selected for scoring. The defect depth grading standards are 0.1mm, 0.2mm and 0.3mm, and the defect depth in the four intervals is assigned scores of 100, 80, 60 and 40 respectively. The defect density and defect depth scoring tables are obtained in the same way as the surface roughness scoring tables.

[0046] It should be further explained that the weight coefficients of roughness score, defect density score and defect depth score are all equal to 1. They are set according to the importance of surface quality indicators in different scenarios and are determined by process engineers in combination with production experience and standards.

[0047] Furthermore, the determination of the overall deviation level requires a dual-index matrix grading method based on dimensional deviation and surface quality scores, dividing the overall deviation level into four levels: the first deviation level, the second deviation level, the third deviation level, and the fourth deviation level.

[0048] In this embodiment, the steps for determining the overall deviation level are as follows:

[0049] S2.1: When ΔD≤1% and Q≥90 points, it is the first level of deviation, which is a slight deviation;

[0050] S2.2: When 1% < ΔD ≤ 3% and 80 ≤ Q < 90 points, it is the second deviation level, which belongs to the general deviation.

[0051] S2.3: When 3% < ΔD ≤ 5% and 70 ≤ Q < 80 points, it is the third level of deviation, which is a serious deviation;

[0052] S2.4: When ΔD > 5% and Q < 70 points, it is the fourth level of deviation, which is a fatal deviation.

[0053] It should be further explained that, by combining the wall thickness deviation ΔT and the length deviation ΔL, any deviation exceeding the standard is judged according to the highest deviation level, covering all geometric dimensional risks. The dimensional deviation scoring threshold should be based on the application scenario of the seamless steel pipe and the critical value of dimensional deviation set by industry standards. The smaller the threshold, the higher the dimensional accuracy of the corresponding product. The surface quality scoring threshold is the critical value of surface quality matched with the dimensional deviation, used to avoid quality risks caused by one dimension meeting the standard but another dimension exceeding the standard.

[0054] Correction scheme generation module: Generates initial correction schemes based on dimensional deviation, surface quality score, and comprehensive deviation level matching correction process parameter mapping library;

[0055] Furthermore, such as Figure 3 As shown, the steps for generating the initial correction scheme are as follows:

[0056] S3.1: Based on dimensional deviation and surface quality score, the deviation types are decomposed and classified according to the modified process parameter mapping library;

[0057] In this embodiment, it should be specifically explained that the deviation types include dimensional deviation types and surface defect deviation types. The dimensional deviation degree is judged one by one to determine whether the dimensional deviation degree exceeds the preset qualified threshold. If the deviation degree of a certain dimension exceeds the standard, it is determined to be the corresponding dimensional deviation type. The surface defect type is determined based on the surface quality score. The dimensional deviation and surface defect deviation in the modified process parameter mapping library are matched and classified with the deviation type.

[0058] S3.2: Based on the categorized deviation type, retrieve the associated corrected process parameters and ranges corresponding to the deviation type in the corrected process parameter mapping library;

[0059] In this embodiment, it should be specifically explained that to retrieve the associated correction parameters and ranges for the corresponding deviation type, it is necessary to search for entries that completely match the deviation type in the deviation type and correction process parameter associated entries in the correction process parameter mapping library, extract the core correction process parameters and preset ranges corresponding to the deviation type from the entries, and if there are multiple deviation types, integrate the correction process parameters and ranges corresponding to the deviation type into a parameter range list.

[0060] S3.3: Determine the core parameter values ​​based on the comprehensive deviation level and the associated correction process parameters and ranges, and generate the initial correction scheme P1.

[0061] In this embodiment, it is necessary to specifically explain that the comprehensive deviation level and corresponding correction intensity are matched based on the correction process parameter mapping library. The first deviation level corresponds to the weak correction intensity, the second deviation level corresponds to the medium correction intensity, the third deviation level corresponds to the strong correction intensity, and the fourth deviation level corresponds to the extremely strong correction intensity. Based on the correction intensity, the sub-range of the correction process parameters is divided to obtain the initial correction scheme P1, which is represented as P1={F1,V1,G1,t1}, where F1 represents the straightening roller pressure, V1 represents the straightening speed, G1 represents the grinding head feed amount, and t1 represents the processing time.

[0062] The iterative correction module verifies the correction effect based on the digital model of the steel pipe and the initial correction scheme, formulates a multi-round correction convergence control strategy, and outputs the final correction result.

[0063] Furthermore, the verification of the correction effect refers to executing the initial correction scheme P1, collecting all-dimensional parameters of the seamless steel pipe after the first correction, constructing a digital model M1 of the first-corrected steel pipe, and substituting the initial steel pipe digital model M0 and the first-corrected steel pipe digital model into the formula:

[0064] ,

[0065] The first correction effect value E1 is calculated, where M represents the ideal digital model of the steel pipe.

[0066] In this embodiment, it should be specifically explained that executing the initial correction scheme refers to sending the core parameters in P1 to the control system of the seamless steel pipe production equipment. The equipment automatically performs the correction operation based on the parameters. After the correction operation is completed, the multi-dimensional data acquisition module collects the full-dimensional parameters of the seamless steel pipe after the correction; where M=f(D,T,L,R) a M1=f(D1,T1,L1,R) a1 M0=f(D0,T0,L0,R) a0 The deviation reduction ratio is calculated based on the ratio of the absolute value of the geometric deviation between the corrected model and the ideal model to the absolute value of the deviation between the initial model and the ideal model. The higher the value, the better the correction effect.

[0067] Furthermore, the steps for formulating a multi-round correction convergence control strategy are as follows:

[0068] S4.1: The correction level is determined based on the effect value of a single correction, and is divided into correction level 1, correction level 2, and correction level 3;

[0069] In this embodiment, it should be specifically noted that the determination of the correction level is based on the correction effect value E1. When E1 ≥ 90%, it is in the first correction level; when 70% ≤ E1 < 90%, it is in the second correction level; and when E1 < 70%, it is in the third correction level.

[0070] S4.2: Perform secondary correction verification for the second level of correction, substituting the initial correction scheme P1, the first correction effect value E1, the initial diameter deviation ΔD0, and the first correction diameter deviation ΔD1 into the formula:

[0071] ,

[0072] The optimization correction scheme P2 is obtained, where K1 and K2 are correction algorithm coefficients, representing the correction effect deviation adjustment coefficient and the size deviation rate adjustment coefficient, respectively. Based on the optimization correction scheme, the secondary correction effect value E2 is calculated, and the correction level is determined. If it is at the first correction level, the correction is stopped; otherwise, the correction and verification are iterated until it is at the first correction level or the convergence control is triggered.

[0073] In this embodiment, it should be specifically explained that for the first level of correction, the correction is completed and the final correction result is output, including the compliant correction scheme, the corrected parameters, and the correction effect value; for the third level of correction, it is necessary to start the expert system reconstruction scheme, combine the knowledge defect map and the correction process parameter mapping library, redesign the core parameters, generate the correction scheme for correction verification; K1 and K2 are preset dynamic coefficients, which are periodically iterated and updated by the knowledge model optimization module.

[0074] S4.3: Set the maximum number of iterations and calculate the correction improvement rate I based on the correction effect value. n A dual convergence control strategy is formulated based on the maximum number of iterations and the correction improvement rate.

[0075] In this embodiment, it should be specifically noted that the maximum number of iterations is 5, and the improvement rate is calculated using the formula:

[0076] ,

[0077] The calculations show that, where n > 2, E n This represents the correction effect value of the nth round of correction; the dual convergence control strategy means that when the number of iterations reaches 5, the iteration stops. If E5 ≥ 90%, the final correction result is output; if E5 < 90%, it is marked as a failure case and pushed to the knowledge model optimization module to record the reason for failure. If the correction improvement rate I... n If the result is less than 5%, it is considered a convergence stagnation, requiring recalibration of the testing equipment, supplementation of material performance parameters, and adjustment of local parameters of the digital model of the steel pipe.

[0078] Knowledge model optimization module: Based on the final correction results, case studies and model parameter liberalization are carried out to iteratively update the digital model of steel pipe.

[0079] Furthermore, case study refers to building a case database based on the compliant correction schemes in the final correction results, and supplementing the storage of failed cases; model parameter liberalization refers to optimizing the correction algorithm coefficients K1 and K2 based on the case database, and updating the digital model of steel pipes and the defect knowledge graph.

[0080] In this embodiment, it is necessary to specifically explain that building the case database requires extracting core deviation features from the final correction results, including dimensional deviation degree, surface defect type and quantification parameters, and comprehensive deviation level, forming deviation feature labels, storing the core parameters of the compliant correction scheme, and recording the final correction effect value, parameter comparison before and after correction, and iteration process data; supplementary storage of failed cases refers to analyzing the failure reasons for non-compliant cases, determining the failure reasons in conjunction with equipment operation logs, and formulating targeted improvement suggestions for the failure reasons, storing them as failed cases; for the correction algorithm coefficients K1 and K2, the correction effect value is compared with the currently used correction algorithm coefficient K 1old and K 2old Through the formula:

[0081] ,

[0082] ,

[0083] The optimized correction algorithm coefficient K was calculated. 1new and K 2newη represents the learning rate, preset to 0.01-0.1, controlling the update magnitude of the parameter to avoid system instability due to excessively large single adjustments. Gradient calculations are performed on the K values ​​and correction effect values ​​of all cases over the past week, automatically triggered every Sunday, to optimize the K value. 1new and K 2new Replace the old value for use in subsequent correction schemes.

[0084] It should be further noted that updating the digital model of the steel pipe requires collecting parameters from the past month's compliance correction scheme {D}. n ,T n ,L n ,R an The average deviation from the ideal parameters is calculated. If the average deviation exceeds 0.1%, the coefficients of the digital model of the steel pipe are fine-tuned, and the micro-surface mapping rules of the model are updated simultaneously, such as optimizing the correspondence between roughness and micro-undulations. The defect knowledge graph update refers to integrating new defects detected in the past month, supplementing them to the defect classification system, defining the quantitative feature dimensions of new defects, and updating the association rules between defects and correction parameters.

[0085] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0086] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. Seamless steel pipe size parameter detection and deviation correction system based on digital twinning, characterized by, The method comprises a multi-dimensional data acquisition module, a digital twin model construction module, a deviation quantification evaluation module, a correction scheme generation module, a correction scheme iteration module, and a knowledge model optimization module. The multi-dimensional data acquisition module acquires full-dimensional parameters of the seamless steel pipe, constructs a full-dimensional detection data set of the seamless steel pipe, and transmits the full-dimensional detection data set to the digital twin model construction module. The digital twin model construction module constructs a steel pipe digital model, a defect knowledge graph, and a correction process parameter mapping library based on the full-dimensional detection data set of the seamless steel pipe. The defect knowledge graph comprises a defect type classification system, defect shape quantification dimension characteristics, and a defect identification standard. The correction process parameter mapping library is a structured database associating the deviation type, the deviation level, and the correction parameter of the seamless steel pipe. The deviation quantification evaluation module calculates the size deviation degree and the surface quality score of the steel pipe based on the full-dimensional detection data set of the seamless steel pipe, and obtains a comprehensive deviation level based on the size deviation degree and the surface quality score. The correction scheme generation module generates an initial correction scheme by matching the correction process parameter mapping library based on the size deviation degree, the surface quality score, and the comprehensive deviation level. The generation steps of the initial correction scheme are as follows: S3.1: Based on the size deviation degree and the surface quality score, the deviation type is decomposed, and the deviation type is classified based on the correction process parameter mapping library. S3.2: Based on the classified deviation type, the associated correction process parameters and ranges of the corresponding deviation type are searched in the correction process parameter mapping library. S3.3: Based on the comprehensive deviation level and the associated correction process parameters and ranges, the core parameter value is determined, and the initial correction scheme P1 is generated. The correction scheme iteration module verifies the correction effect based on the steel pipe digital model and the initial correction scheme, formulates a multi-round correction convergence control strategy, and outputs a final correction result. The formulation steps of the multi-round correction convergence control strategy are as follows: S4.1: Based on the first correction effect value, the correction level is judged, and is divided into a first correction level, a second correction level, and a third correction level. ， S4.2: For the second correction level, a second correction verification is performed, and the initial correction scheme P1, the first correction effect value E1, the initial diameter deviation degree ΔD0, and the first correction diameter deviation degree ΔD1 are substituted into the formula: S4.3: Set the maximum number of iterations, and calculate the correction improvement rate I based on the correction effect value n Formulate a double convergence control strategy based on the maximum number of iterations and the correction improvement rate; The optimized correction scheme P2 is obtained, wherein K1 and K2 are correction algorithm coefficients, respectively representing a correction effect deviation adjustment coefficient and a size deviation rate adjustment coefficient.

2. The seamless steel pipe size parameter detection and deviation correction system based on digital twinning according to claim 1, characterized in that: The seamless steel pipe full-dimension parameters include conventional size parameters, surface quality parameters and material performance parameters, wherein the conventional size parameters include steel pipe outer diameter D0, inner diameter d0, wall thickness T0 and length L0, the surface quality parameters include surface roughness R a0 , defect number, defect type and defect size, and the material performance parameters refer to hardness H B and material composition.

3. The seamless steel pipe size parameter detection and deviation correction system based on digital twinning according to claim 1, characterized in that: Based on the optimized correction scheme, the second correction effect value E2 is calculated, the correction level is judged, if it is in the first correction level, the correction is stopped, otherwise, the iterative correction verification is performed until it is in the first correction level or the convergence control is triggered. The knowledge model optimization module performs case learning and model parameter liberalization based on the final correction result, and iteratively updates the steel pipe digital model. The construction steps of the steel pipe digital model are as follows: S1.1: Obtain the full-dimensional detection data set of the seamless steel pipe, and screen the core input variables in the full-dimensional detection data set of the seamless steel pipe, including the pipe diameter, the wall thickness, the length, and the surface roughness. S1.2: Based on the core input variables, combined with the seamless steel pipe geometric characteristics, a steel pipe digital model is constructed, represented as M=f(D,T,L,R a ), which converts the input core variables into digital geometric data; S1.3: The real-time data in the full-dimension detection data set of the seamless steel pipe is substituted into the steel pipe digital model to obtain an initial steel pipe digital model, and the steel pipe digital model is verified and calibrated.

4. The seamless steel pipe size parameter detection and deviation correction system based on digital twinning according to claim 1, characterized in that: The calculation of the size deviation degree requires obtaining the outer diameter D0, wall thickness T0 and length L0 of the seamless steel pipe in the full-dimension detection data set of the seamless steel pipe, and calculating the size deviation degree based on the outer diameter of the steel pipe and the outer diameter target value D t Substitute the formula: ， The outer diameter dimensional deviation degree ΔD is calculated, and the wall thickness, the wall thickness target value T t , the length L t , and the length target value are used to calculate the wall thickness dimensional deviation degree ΔT and the length dimensional deviation degree ΔL, respectively; the surface quality score is calculated by obtaining the roughness score S R , the defect density score S M , and the defect depth score S S , and the surface quality score Q is obtained by weighted calculation, wherein the weight coefficients of the roughness score, the defect density score, and the defect depth score are w1, w2, and w3, respectively.

5. The seamless steel pipe size parameter detection and deviation correction system based on digital twinning according to claim 1, characterized in that: The judgment of the comprehensive deviation level needs to adopt a double-index matrix grading method based on the size deviation degree and the surface quality score to divide the comprehensive deviation level into four levels, i.e., a first deviation level, a second deviation level, a third deviation level and a fourth deviation level.

6. The seamless steel pipe size parameter detection and deviation correction system based on digital twinning according to claim 1, characterized in that: The correction effect verification refers to executing an initial correction scheme P1, collecting full-dimension parameters of the seamless steel pipe after one correction, constructing a one-time correction steel pipe digital model M1, and substituting the initial steel pipe digital model M0 and the one-time correction steel pipe digital model into a formula: ， to calculate a one-time correction effect value E1, wherein M represents an ideal steel pipe digital model.

7. The seamless steel pipe size parameter detection and deviation correction system based on digital twinning according to claim 1, characterized in that: The case learning refers to constructing a case database based on the pass correction scheme in the final correction result and performing supplementary storage of failed cases; and the model parameter liberalization refers to optimizing correction algorithm coefficients K1 and K2 based on the case database, and updating the steel pipe digital model and the defect knowledge graph.

Images