T-beam production management system based on artificial intelligence

Abstract

The invention belongs to the technical field of structural member production health detection, and relates to a T-beam production management system based on artificial intelligence, which comprises the following steps: mapping collected prestress data and surface influence data into a three-dimensional mechanical simulation model to truly reflect the deformation and fission states of a member, and applying primary and secondary dynamic loads based on the deformation and fission states, the dynamic load applying precision is improved; carrying out multi-criterion decision analysis by adopting a TOPSIS method, comprehensively evaluating the stress change of each stress distribution node, and positioning a potential failure area under a dynamic load so as to determine an internal sounding position; based on the position features of the potential failure area, establishing a mapping table of the acoustic signal features in theoretical and actual measurement levels, and generating a damage assessment report including damage types and degrees; the damage development trend is predicted through current and historical damage assessment reports, and maintenance priority ranking is performed on potential failure areas, so that the maintenance efficiency is optimized.

Description

Technical Field

[0001] The present invention belongs to the technical field of structural component production health detection, and relates to a T-beam production management system based on artificial intelligence. Background Art

[0002] Prestressed T-beams are a key structural component widely used in engineering fields such as construction and bridges. Their production performance is directly related to the safety and durability of the overall structure during use. In actual use, T-beams are often affected by factors such as complex loads, environmental corrosion, and material aging, resulting in potential internal damage such as microcracks, debonding, and stress concentration. If these damages are not discovered and addressed in a timely manner during the production inspection cycle, they may cause sudden failure of the component and even cause serious safety accidents. To this end, it is possible to detect potential failure areas of T-beams during the production inspection cycle in real time and conduct accurate and scientific assessments of the damage type and extent, thereby providing a basis for maintenance decisions during the actual use of T-beams.

[0003] There are also some solutions related to structural component health monitoring in the existing technology. For example, the Chinese patent with authorization publication number CN111855817B discloses a method for cloud-edge collaborative fatigue crack detection of complex structural components. This method mainly uses acoustic emission sensors and vibration sensors to collect sound velocity data and vibration data to detect cracks and vibrations. The detection capability of prestressed components is limited. Since prestress distribution is a key factor affecting the overall performance of the structure, it is difficult to fully reflect the stress state and changing trend of the component by relying solely on sound velocity and vibration data. In addition, there is a lack of deeper data correlation analysis between the prestress distribution of the component and the internal damage type, which leads to the inability to fully reflect the overall state of the structural component. At the same time, the existing technology uses blind source separation and feature signal extraction to process sound velocity data, and complex tracking to process vibration data. Although it can extract features and vibration modes, it can only detect the damaged parts of the structure. It lacks analysis of stress distribution under dynamic loads, cannot simulate the dynamic stress changes of structural components under actual working conditions, and it is difficult to accurately predict the potential failure areas of structural components. Summary of the Invention

[0004] In view of this, in order to solve the problems raised in the above background technology, an artificial intelligence-based T-beam production management system is proposed.

[0005] The purpose of the present invention can be achieved through the following technical solutions: The present invention provides a T-beam production management system based on artificial intelligence, which includes: a data acquisition module, which collects prestressed data and surface defect data of the T-beam through a detection terminal.

[0006] The load application module constructs a three-dimensional mechanical simulation model of the T-beam based on the collected data, applies dynamic loads to it, calculates the stress distribution in different areas of the T-beam, and screens out potential failure areas where stress exceeds the limit.

[0007] The damage analysis module uses acoustic emission sensors to capture the micro-damage acoustic wave signal characteristics of the potential failure area inside the T-beam and establishes a damage assessment report for the potential failure area that includes the damage type and damage degree.

[0008] The failure assessment module obtains current and historical damage assessment reports of potential failure areas and obtains maintenance priority rankings of potential failure areas.

[0009] The maintenance determination module generates a maintenance plan including operation instructions and risk warnings to the terminal based on the maintenance priority.

[0010] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention maps the collected prestress data and surface defect data into a three-dimensional mechanical simulation model to ensure that the model can truly reflect the actual deformation and fission state of the component, and applies the primary and secondary dynamic loads of the simulation model based on the actual deformation and fission state, thereby increasing the accuracy of applying dynamic loads.

[0011] (2) The present invention adopts the TOPSIS method to perform multi-criteria decision analysis, comprehensively considering multiple evaluation indicators such as the primary and secondary stress progression rates, and comprehensively evaluating the stress changes of each original stress distribution node, thereby locating the potential failure area of ​​the T-beam under dynamic load and determining the internal acoustic detection position, which helps to improve the efficiency of internal damage detection of the T-beam. Moreover, this multi-index comprehensive evaluation capability makes the stress analysis more comprehensive and accurate.

[0012] (3) Based on the positional characteristics of different potential failure areas of the T-beam, the present invention establishes a mapping table between the acoustic wave signal characteristics and the potential failure areas at the theoretical and measured levels, so as to establish a damage assessment report for the potential failure area including the damage type and damage degree, thereby realizing the intelligent location and classification of damage and providing a clear target area for subsequent maintenance.

[0013] (4) By obtaining current and historical damage assessment reports, the present invention can predict damage development trends and prioritize the maintenance of potential failure areas, thereby optimizing the allocation of maintenance resources, improving maintenance efficiency, and reducing maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0015] Figure 1 This is a schematic diagram of the connection of the modules of the present invention.

[0016] Figure 2 The figure is a schematic flow chart of the steps for implementing the method of the present invention.

[0017] Figure 3 This is a schematic diagram of the typical regional location of the present invention.

[0018] Figure 4 Schematic diagram of corresponding position numbers of web areas A1, A2, and A3 of the present invention.

[0019] Figure 5 Schematic diagram of corresponding position numbers of flange area B2 of the present invention.

[0020] Reference numerals: B1, upper flange, 2, web region, B3, lower flange, A1, upper region of the web, A2, middle region of the web, A3, lower region of the web. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0022] See Figure 1 As shown, the present invention provides a T-beam production management system based on artificial intelligence, which includes: a data acquisition module and a load application module, a damage analysis module, a failure assessment module and a maintenance determination module; the load application module is respectively connected to the data acquisition module and the damage analysis module, and the failure assessment module is respectively connected to the maintenance determination module and the damage analysis module.

[0023] See Figures 2 to 5 As shown, the data acquisition module collects the prestress data and surface defect data of the T-beam through the detection terminal.

[0024] In a preferred embodiment, the detection terminal includes a distributed optical fiber sensor, a strain gauge and an industrial camera.

[0025] The distributed optical fiber sensor is arranged along the prestressed tendons of the T-beam and is used to monitor the prestress distribution data in real time. The prestress distribution data includes the magnitude of the prestress, such as the tensioning force value, the distribution of the prestress, such as the stress change along the length of the prestressed tendons, and the uniformity of the prestress, such as whether there is local stress concentration or relaxation.

[0026] The stress gauge is pasted or installed at key positions of the prestressed tendons, such as the flange and web of the T-beam, to collect the prestress values ​​at key positions of the prestressed tendons of the T-beam.

[0027] The industrial camera is used to collect surface defect data of the T-beam, and the surface defect data includes surface deformation data and crack data.

[0028] The surface deformation data includes the surface displacement distance or bulge volume, bending curvature and other related deformation data of the T-beam.

[0029] The crack data includes distribution data such as the width, length and position of the cracks on the T-beam surface.

[0030] The load application module constructs a three-dimensional mechanical simulation model of the T-beam based on the collected data, applies dynamic loads to it, calculates the stress distribution in different areas of the T-beam, and screens out potential failure areas where stress exceeds the limit.

[0031] In a preferred embodiment, the three-dimensional mechanical simulation model of the T-beam includes: based on the actual geometric dimensions, prestress data and surface defect data of the T-beam, a three-dimensional coordinate system is established in the three-dimensional modeling software to obtain the corresponding coordinates of each deformation position and each crack position.

[0032] The three-dimensional coordinate system specifically includes: an X-axis constructed along the length direction of the T-beam, a Y-axis constructed along the width direction of the T-beam, and a Z-axis constructed along the thickness direction of the T-beam. The origin of the three-dimensional coordinate system is usually set at one end or the geometric center of the component to facilitate subsequent loading and analysis.

[0033] The actual geometric dimensions include the geometric shape of the component (including the dimensions of the flange and web), material properties (such as the elastic modulus and Poisson's ratio of concrete, the mechanical properties of prestressed tendons, etc.), the initial prestress state (mapped to the model based on the collected prestress data), and boundary conditions (such as support constraints).

[0034] The corresponding coordinates of each deformation position and each crack position are integrated into the coordinates of each original stress distribution node of the T-beam, and the fission rate of each original stress distribution node is obtained. The fission rate of each original stress distribution node is the deformation degree or crack degree of each deformation position and each crack position.

[0035] The degree of deformation is the product of the concave-convex volume change at each deformation position of the T-beam per unit time and the corresponding bending curvature. For the concave-convex volume change, the shape and size of the concave-convex volume can be obtained through three-dimensional scanning technology, and then the current volume is calculated. The current volume is subtracted from the previous volume, and the absolute value of the difference is used as the concave-convex volume change; for the bending curvature, the curvature value can be calculated by measuring the curve shape of the component surface and using the curvature mathematical formula.

[0036] The crack degree is the product of the increased length and depth of the crack at each crack position of the T-beam per unit time. Based on the industrial camera to collect high-definition images of the T-beam surface, the width and length of the crack are accurately recorded. The crack boundary is identified through the image edge detection algorithm, and the pixel-level width and length are calculated and converted into actual physical dimensions. After obtaining the corresponding increased length and width, the crack degree is calculated.

[0037] Specifically, the method for identifying the degree of deformation and cracking can be obtained by analyzing the surface deformation data and cracking data based on image processing and analysis, morphological processing and other technologies in the prior art, which will not be repeated here.

[0038] The secondary dynamic load of the T-beam is determined according to the fission rates of the original stress distribution nodes. Specifically, the highest fission rate among the fission rates of the original stress distribution nodes is screened, and it is matched with the adaptive dynamic load corresponding to the preset fission rates, and then the obtained adaptive dynamic load is used as the secondary dynamic load of the T-beam.

[0039] In a further preferred embodiment, the screening of potential failure areas with excessive stress includes: defining the stress progression rate as the unit time change amplitude of the stress vector of the stress distribution node under the pressure of dynamic load in space, wherein the stress vector is obtained by integrating the stress direction and stress value. For example, for a certain stress distribution node under the pressure of dynamic load in space, the stress value in the positive direction of the X axis is generated. , stress value in the positive direction of Y axis , Z-axis positive direction stress value , then the stress values ​​in the three directions of XYZ axis are integrated , and obtain the stress vector of the stress distribution node ,in Represents the unit vectors of the X-axis, Y-axis, and Z-axis respectively, reflecting the stress proportion of different coordinate axes in the stress vector integration process; obtains the stress vector of the stress distribution node before being compressed by the dynamic load , and the time step of compression is obtained as , define the stress progression rate of the stress distribution node as , reflecting the stress distribution node at The magnitude of the stress vector change caused by the dynamic load during this time step, Represents the magnitude (Euclidean norm) of the difference between two stress vectors.

[0040] Specifically, the stress progression rate is used to evaluate the rate of stress change at a node under dynamic loads. A higher stress progression rate indicates a faster stress change at the node, meaning the node is more susceptible to failure. A lower stress progression rate indicates a slower stress change at the node, meaning the node is safer.

[0041] In the three-dimensional mechanical simulation model, a first-level dynamic load is applied according to the actual working conditions to generate the first-level stress progression rate of each original stress distribution node. The dynamic load includes the load type, load size, action position, and time step. For example, for the wind load type, a dynamic load of the wind load type is applied, the load size is 200Pa, the action position is at the stress distribution node position in a certain flank area of ​​the T-beam, and the time step is 5 minutes, and then the first-level stress progression rate of the position point is derived from the three-dimensional mechanical simulation model.

[0042] By applying the secondary dynamic load of the T-beam, the secondary stress progression rate of each original stress distribution node is generated.

[0043] The TOPSIS method (topology of closest ideal solutions) is used to calculate the relative closeness of the primary stress progression rate and the secondary stress progression rate of each original stress distribution node to the ideal solution, and to screen out potential failure areas with excessive stress.

[0044] The TOPSIS method is a multi-criteria decision analysis method capable of simultaneously processing multiple evaluation indicators. In the stress progression analysis of T-beams, the primary and secondary stress progression rates involve multiple indicators. The TOPSIS method comprehensively considers these indicators to generate a single composite value, thereby comprehensively evaluating the stress changes at each node of the original stress distribution. This multi-indicator comprehensive evaluation capability gives the TOPSIS method a significant advantage in complex stress analysis.

[0045] The present invention maps the collected prestress data and surface defect data into a three-dimensional mechanical simulation model to ensure that the model can truly reflect the actual deformation and fission state of the component, and applies the primary and secondary dynamic loads of the simulation model based on the actual deformation and fission state, thereby increasing the accuracy of dynamic load application.

[0046] In a further preferred embodiment, the specific implementation steps of the TOPSIS method are as follows: obtain a reference sequence of each original stress distribution node, such as an ideal stress progression rate, calculate the grey correlation between the primary stress progression rate and the secondary stress progression rate of each original stress distribution node and the reference sequence, and record them as , build a decision matrix , Indicates the row element number of each original stress distribution node, .

[0047] The reference sequence of each original stress distribution node is the original setting value, such as 5, 4, 5..., and the grey correlation degree refers to the difference comparison between the primary stress progression rate and the secondary stress progression rate of different original stress distribution nodes and the reference sequence of the corresponding node position. For example, if the reference sequence of a certain original stress distribution node is set to 5 and the primary stress progression rate of the original stress distribution node is 2, then the difference comparison ratio formula is used. , calculate the grey correlation degree between the first-order stress progression rate of the original stress distribution node and the reference sequence as .

[0048] Standardize the decision matrix to obtain the standardized decision matrix ,in Respectively represent The first-order stress progression rate and the second-order stress progression rate of the original stress distribution node are the corresponding standardized values.

[0049] Assign weights to primary and secondary stress progression rates ,satisfy Since the positions of the original stress distribution nodes are different, the load-bearing importance in the T-beam structure is different. Therefore, different weights are given to the positions of the original stress distribution nodes. This can be determined through empirical fitting. The weighted standardized decision matrix R is multiplied by the weight to obtain the weighted standardized decision matrix ,in Respectively represent The first-order stress progression rate and the second-order stress progression rate of the original stress distribution node are the weighted and standardized values.

[0050] Determine the ideal solution corresponding to the primary stress progression rate and the secondary stress progression rate and negative ideal solutions , calculate the distance between the first-order stress progression rate and the second-order stress progression rate of each original stress distribution node and the ideal solution and the negative ideal solution respectively, which is 、 , the ideal solution represents the maximum solution of the stress progression rate, and the negative ideal solution represents the minimum solution of the stress progression rate.

[0051] Calculate the relative closeness between each original stress distribution node and the ideal value , The value range is [0, 1], and the larger the value, the closer it is to the ideal solution.

[0052] Based on relative proximity Sort the original stress distribution nodes. The larger the value, the more significant the stress change. The original stress distribution node positions exceeding the preset proximity threshold are used as the corresponding distribution position points of the potential failure area.

[0053] Specifically, the dimension effect is eliminated and the data range is unified through the standardized decision matrix; the importance of indicators is reflected through the weighted standardized decision matrix to adjust the data impact; the optimal and worst states are defined through the ideal solution and the negative ideal solution to provide a comparison benchmark; finally, the relative proximity quantifies the degree of closeness between the node and the ideal solution to comprehensively screen potential failure areas.

[0054] The present invention adopts the TOPSIS method to perform multi-criteria decision analysis, comprehensively considering multiple evaluation indicators such as the primary and secondary stress progression rates, and comprehensively evaluating the stress changes of each original stress distribution node, thereby locating the potential failure area of ​​the T-beam under dynamic load and determining the internal acoustic test position, which helps to improve the efficiency of internal damage detection of the T-beam. This multi-indicator comprehensive evaluation capability makes stress analysis more comprehensive and accurate.

[0055] The damage analysis module uses acoustic emission sensors to capture the micro-damage acoustic wave signal characteristics of the potential failure area inside the T-beam and establishes a damage assessment report for the potential failure area that includes the damage type and damage degree.

[0056] An acoustic emission sensor is a device used to detect elastic waves released by materials or structures during stress. Its working principle is based on the process in which elastic waves generated by micro-damage inside the material, such as crack expansion and plastic deformation, propagate to the surface of the material, are received by the sensor, and are converted into electrical signals.

[0057] In a preferred embodiment, the use of an acoustic emission sensor to capture the micro-damage acoustic wave signal characteristics of the potential failure area inside the T-beam includes: obtaining the position characteristics of all potential failure areas in the overall structure of the T-beam, and establishing a subtraction relationship mapping table between the position characteristics of the potential failure area and the acoustic wave signal characteristics, wherein the position characteristics include the position number, the area ratio of the area, and the position depth, and the acoustic wave signal characteristics include the acoustic wave propagation time, the acoustic wave amplitude attenuation, and the acoustic wave frequency change.

[0058] According to the micro-damage acoustic wave signal characteristics of the potential failure area inside the T-beam actually captured by the acoustic emission sensor, an acoustic wave measurement relationship mapping table of the potential failure area is established.

[0059] In a further preferred embodiment, the attenuation relationship mapping table between the potential failure area position characteristics and the acoustic signal characteristics includes: the attenuation mapping relationship between the position number and the acoustic signal characteristics, the attenuation mapping relationship between the area ratio and the acoustic signal characteristics, and the attenuation mapping relationship between the position depth and the acoustic signal characteristics.

[0060] See also Figure 2 、 Figure 3 、 Figure 4 As shown in the figure, the position numbers are used to describe the locations of typical regions, such as the web and flange. The locations of typical regions are then identified by dividing the grid or using region numbers, such as "web regions A1, A2, A3" or "flange regions B1, B2, B3". Specifically, the web region refers to the vertical portion of the T-beam connecting the upper and lower flanges, which bears shear forces and bending moments. Web region A1 is located in the upper region of the web and close to the upper flange; web region A2 is located in the middle region of the web and in the middle part of the web; and web region A3 is located in the lower region of the web and close to the lower flange. The flange region refers to the upper and lower horizontal portions of the T-beam, which mainly bear tensile and compressive stresses. Flange region B1 is located in a local area of ​​the upper flange and close to web region A1; flange region B2 is located in a local area of ​​the upper or lower flange and in the middle of the flange; and flange region B3 is located in a local area of ​​the lower flange and close to web region A3.

[0061] The area ratio represents the proportion of the potential failure area to the overall T-beam structure, reflecting the severity of the damage. A higher ratio indicates a larger damaged area and a more significant impact on structural safety. This ratio is expressed by calculating the ratio of the potential failure area to the total area of ​​the T-beam. For example, if the failure area is 0.5 square meters and the total area of ​​the T-beam is 10 square meters, the area ratio is 5%.

[0062] The location depth indicates the depth of the potential failure area within the T-beam, defined as the vertical distance from the component surface to the center of the failure area. This depth information helps determine the hidden nature of the damage and the difficulty of repair. Depth can be expressed using the Z-axis coordinate in millimeters or centimeters. For example, if a failure area is 50 mm below the surface, the location depth is 50 mm.

[0063] The attenuation patterns of sound waves propagating through a T-beam can be determined through theoretical calculations, numerical simulations (such as finite element analysis), or experimental data. For example, the propagation time, amplitude attenuation, and frequency changes of sound waves at different locations, depths, and surface areas can be measured experimentally.

[0064] The attenuation relationship mapping table between the potential failure area position characteristics and the acoustic signal characteristics is established through theoretical or experimental data, reflecting the theoretical attenuation law of the acoustic signal when propagating at different positions in the component, and providing a theoretical basis for the subsequent analysis of measured data.

[0065] For example: Table 1, reduction relationship mapping table

[0066]

[0067] The acoustic wave measured relationship mapping table of the potential failure area includes the measured mapping relationship between the position number and the micro-damage acoustic wave signal characteristics, the measured mapping relationship between the area ratio and the micro-damage acoustic wave signal characteristics, and the measured mapping relationship between the position depth and the micro-damage acoustic wave signal characteristics. The acoustic wave measured relationship mapping table of the potential failure area is established based on the acoustic wave signal actually captured by the acoustic emission sensor, reflecting the actual damage status of the potential failure area.

[0068] For example: Table 2, acoustic wave measurement relationship mapping table

[0069]

[0070] In a further preferred embodiment, the establishment of a potential failure area damage assessment report including damage type and damage degree includes: comparing the differences between a subtraction relationship mapping table of potential failure area position characteristics and acoustic wave signal characteristics and a mapping table of acoustic wave measured relationships of potential failure areas to obtain measured damage characteristics of different potential failure areas, wherein the measured damage characteristics include acoustic wave propagation time deviation, acoustic wave amplitude attenuation deviation, and acoustic wave frequency change deviation. By comparing the theoretical subtraction relationship with the measured relationship, abnormal features in the measured data can be identified, thereby more accurately evaluating the damage status, which can effectively improve the accuracy and reliability of damage detection.

[0071] An empirical fitting method is used to regularly integrate the measured damage characteristics of different potential failure areas to determine the damage types of different potential failure areas. The damage types include crack propagation, interface detachment, material fracture, voids / pores, and corrosion / aging. For example, if the sound wave propagation time in the web area increases significantly and the amplitude attenuates significantly, it may be crack propagation; if the sound wave propagation time in the flange area increases significantly and the amplitude attenuates less, it may be interface debonding.

[0072] The damage types and damage characteristics of different potential failure areas are mapped and associated to generate a potential failure area damage assessment report containing the damage type and damage degree, where the damage degree includes slight damage, moderate damage, and severe damage.

[0073] Table 3. Potential failure area damage assessment report including damage type and damage degree

[0074]

[0075] Specifically, due to the different sensitivities of different areas of the T-beam to damage, different potential failure areas have different deviation responses to the measured damage characteristics, and thus the identification results of their damage degrees may be different. For example, the web area A1 is highly sensitive to damage. Even if the frequency change deviation is small, crack expansion may cause a significant decrease in structural stiffness, so its damage degree is judged to be moderate damage; the flange area B2 mainly bears tensile stress and compressive stress and is less sensitive to damage. When it has interface debonding damage type, it has little impact on the overall structural performance, so its damage degree is judged to be slight damage; when the web area A3 has material fracture damage, it may lead to serious stress concentration and structural failure risk, so its damage degree is judged to be high damage.

[0076] The failure assessment module obtains current and historical damage assessment reports of potential failure areas and obtains maintenance priority rankings of potential failure areas.

[0077] In a preferred embodiment, the current and historical potential failure area damage assessment reports are obtained to obtain the maintenance priority ranking of the potential failure areas, which includes: deriving the various damage types and damage degrees existing in different potential failure areas from the historical potential failure area damage assessment reports, and counting the number of records of various damage degrees for different potential failure areas under the current damage type.

[0078] The damage degree expansion trend judgment formula under the current damage type is as follows: , according to the quantitative calculation result value and the preset damage extension trend of rapid extension, moderate extension, and slow extension range value, the damage extension trend of different potential failure areas under the current damage type is determined. Specifically, Respectively represent the number of records of slight damage, moderate damage, and severe damage in the potential failure area under the current damage type. They represent the preset damage compensation weights of the potential failure area under the current damage type, namely slight damage, moderate damage, and severe damage, which are used to quantify the impact of different damage levels on the performance of the T-beam structure and avoid errors caused by subjective judgment. .

[0079] The damage expansion trend is quantitatively calculated through a weighted summation method. On the one hand, the weight distribution can reflect the actual threat of different damage degrees to structural safety. On the other hand, it can directly integrate the comprehensive impact of the corresponding number of records of slight damage, moderate damage, and severe damage in the failure area under the current damage type at multiple levels.

[0080] The preset damage compensation weights can be set based on industry experience or obtained through a limited number of test data. For example, historical data on T-beam damage types and frequencies can be collected first, and then the correlation coefficients between the impact of different damage degrees and the structural properties of the T-beam can be calculated. Regression analysis or logistic regression analysis can be used to determine the contribution of each damage degree. Finally, after normalization, the contribution is converted into preset damage compensation weights, and their sum is 1.

[0081] All potential failure areas are integrated into a set of potential failure area sequences with rapid expansion, moderate expansion, and slow expansion.

[0082] Based on the positional characteristics of different potential failure areas of T-beams, the present invention establishes a mapping table between acoustic wave signal characteristics and potential failure areas at the theoretical and measured levels, so as to create a damage assessment report for potential failure areas that includes damage type and damage degree, realizes intelligent damage location and classification, and provides a clear target area for subsequent maintenance.

[0083] The maintenance determination module generates a maintenance plan including operation instructions and risk warnings to the terminal based on the maintenance priority.

[0084] In a preferred embodiment, the maintenance plan including operation instructions and risk warnings is generated and sent to the terminal based on the maintenance priority sorting, and the content includes: for the rapidly expanding potential failure area sequence set, the priority is the highest, and immediate repair operation instructions are set based on the corresponding damage types of different potential failure areas in the sequence set, and based on the corresponding positions of different potential failure areas in the sequence set, full inspection instructions are set for the same positions of all production components in this batch, such as material fracture, reinforcement or replacement of materials, crack expansion, grouting or carbon fiber reinforcement, corrosion, surface treatment or coating repair, interface debonding, adhesive repair, and rapidly expanding damage poses a serious risk of structural failure and requires immediate processing.

[0085] For a moderately expanded potential failure area sequence set, the priority is medium. Based on the corresponding positions of different potential failure areas in the sequence set, sampling re-inspection instructions are set for the same positions of the components produced in this batch.

[0086] For the slowly expanding potential failure area sequence set, the priority is the lowest, and position marks are set based on the corresponding positions of different potential failure areas in the sequence set.

[0087] Specifically, all the production components of this batch refer to components produced in the same cycle under the same working condition scenario.

[0088] By acquiring current and historical damage assessment reports, the present invention can predict damage development trends and prioritize maintenance of potential failure areas, thereby optimizing the allocation of maintenance resources, improving maintenance efficiency and reducing maintenance costs.

[0089] In addition, the present invention also provides a T-beam production management method based on artificial intelligence, including: S1, collecting prestressed data and surface defect data of the T-beam through a detection terminal.

[0090] S2. Build a three-dimensional mechanical simulation model of the T-beam based on the collected data, apply dynamic loads to it, calculate the stress distribution in different areas of the T-beam, and screen out potential failure areas with excessive stress.

[0091] S3. Use acoustic emission sensors to capture the micro-damage acoustic wave signal characteristics of the potential failure area inside the T-beam, and establish a damage assessment report for the potential failure area that includes the damage type and damage degree.

[0092] S4. Obtain current and historical potential failure area damage assessment reports and obtain maintenance priority rankings for potential failure areas.

[0093] S5. Generate a maintenance plan including operation instructions and risk warnings to the terminal based on the maintenance priority.

[0094] The above formulas are all dimensionless and numerically calculated, and the preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0095] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all fall within the scope of protection of the present invention.

Claims

1. A T-beam production management system based on artificial intelligence, characterized in that: include: Data acquisition module, which collects the prestress data and surface defect data of the T-beam through the detection terminal; The load application module constructs a three-dimensional mechanical simulation model of the T-beam based on the collected data, applies dynamic loads to it, calculates the stress distribution in different areas of the T-beam, and screens out potential failure areas with excessive stress; The damage analysis module uses acoustic emission sensors to capture the micro-damage acoustic wave signal characteristics of the potential failure area inside the T-beam and creates a damage assessment report for the potential failure area that includes damage type and damage degree; Failure assessment module, which obtains current and historical damage assessment reports of potential failure areas and determines maintenance priority rankings of potential failure areas; The maintenance determination module generates a maintenance plan including operation instructions and risk warnings to the terminal based on the maintenance priority.

2. The artificial intelligence-based T-beam production management system according to claim 1, characterized in that: The detection terminal includes a distributed optical fiber sensor, a strain gauge and an industrial camera; The distributed optical fiber sensor is arranged along the prestressed tendons of the T-beam to monitor the prestress distribution data in real time; The stress gauge is attached or installed at a key position of the prestressed tendons to collect the prestress value at the key position of the prestressed tendons of the T-beam; The industrial camera is used to collect surface defect data of the T-beam, and the surface defect data includes surface deformation data and crack data.

3. The artificial intelligence-based T-beam production management system according to claim 1, characterized in that: The three-dimensional mechanical simulation model of the T-beam includes: establishing a three-dimensional coordinate system in three-dimensional modeling software based on the actual geometric dimensions, prestress data and surface defect data of the T-beam, and obtaining the corresponding coordinates of each deformation position and each crack position; The corresponding coordinates of each deformation position and each crack position are integrated into the coordinates of each original stress distribution node of the T-beam, and the fission rate of each original stress distribution node is obtained. The fission rate of each original stress distribution node is the deformation degree or crack degree of each deformation position and each crack position; wherein the deformation degree is the product of the concave-convex volume change of each deformation position per unit time and the corresponding bending curvature; the crack degree is the product of the length and depth of the crack at each crack position per unit time; The secondary dynamic load of the T-beam is determined according to the fission rate of each original stress distribution node.

4. The artificial intelligence-based T-beam production management system according to claim 3, characterized in that: The method of screening out potential failure areas with excessive stress includes: defining a stress progression rate as a unit time change amplitude of a stress vector of a stress distribution node in space under pressure from a dynamic load, wherein the stress vector is obtained by integrating the stress direction and the stress value; In the three-dimensional mechanical simulation model, a first-level dynamic load is applied according to the actual working conditions to generate the first-level stress progression rate of each original stress distribution node. The dynamic load includes the load type, load magnitude, action location, and time step. By applying the secondary dynamic load of the T-beam, the secondary stress progression rate of each original stress distribution node is generated; The TOPSIS method is used to calculate the relative closeness of the primary stress progression rate and the secondary stress progression rate of each original stress distribution node to the ideal solution, and to screen out potential failure areas with excessive stress.

5. The artificial intelligence-based T-beam production management system according to claim 4, characterized in that: The specific implementation steps of the TOPSIS method are as follows: Obtain the reference sequence of each original stress distribution node, calculate the grey correlation between the primary stress progression rate and the secondary stress progression rate of each original stress distribution node and the reference sequence, and construct a decision matrix; The decision matrix is ​​standardized to obtain a standardized decision matrix; Assign weights to the primary stress progression rate and the secondary stress progression rate, and multiply the standardized decision matrix by the weights to obtain a weighted standardized decision matrix; Determine the ideal solution and negative ideal solution corresponding to the primary stress progression rate and the secondary stress progression rate of each original stress distribution node, and calculate the distance between the primary stress progression rate and the secondary stress progression rate of each original stress distribution node and the ideal solution and the negative ideal solution, respectively, wherein the ideal solution represents the maximum value solution of the stress progression rate and the negative ideal solution represents the minimum value solution of the stress progression rate; Calculate the relative proximity between each original stress distribution node and the ideal value, and the relative proximity value range is [0, 1]; The original stress distribution nodes are sorted according to their relative proximity, and the original stress distribution node positions whose relative proximity exceeds a preset proximity threshold are selected as the corresponding distribution position points of the potential failure area.

6. The artificial intelligence-based T-beam production management system according to claim 1, characterized in that: The method of using an acoustic emission sensor to capture micro-damage acoustic wave signal characteristics of potential failure areas inside the T-beam includes: obtaining position characteristics of all potential failure areas in the overall structure of the T-beam, and establishing a subtraction relationship mapping table between the position characteristics of the potential failure areas and the acoustic wave signal characteristics, wherein the position characteristics include position number, area ratio, and position depth, and the acoustic wave signal characteristics include acoustic wave propagation time, acoustic wave amplitude attenuation, and acoustic wave frequency change; According to the micro-damage acoustic wave signal characteristics of the potential failure area inside the T-beam actually captured by the acoustic emission sensor, an acoustic wave measurement relationship mapping table of the potential failure area is established.

7. The artificial intelligence-based T-beam production management system according to claim 6, characterized in that: The reduction relationship mapping table of the potential failure area position characteristics and acoustic signal characteristics includes: the reduction mapping relationship between the position number and the acoustic signal characteristics, the reduction mapping relationship between the area ratio and the acoustic signal characteristics, and the reduction mapping relationship between the position depth and the acoustic signal characteristics; The position number is used to describe the typical regional position, and then the typical regional position is identified by dividing the grid or the regional number; The area ratio of the region represents the area ratio of the potential failure region in the overall structure of the T-beam; The position depth indicates the buried depth of the potential failure area in the T-beam, that is, the vertical distance from the surface of the component to the center of the failure area; The attenuation relationship mapping table between the potential failure area position characteristics and the acoustic wave signal characteristics is established through theoretical or experimental data, reflecting the theoretical attenuation law of the acoustic wave signal when propagating in the T-beam; The acoustic wave measured relationship mapping table of the potential failure area includes the measured mapping relationship between the position number and the micro-damage acoustic wave signal characteristics, the measured mapping relationship between the area ratio and the micro-damage acoustic wave signal characteristics, and the measured mapping relationship between the position depth and the micro-damage acoustic wave signal characteristics. The acoustic wave measured relationship mapping table of the potential failure area is established based on the acoustic wave signal actually captured by the acoustic emission sensor, reflecting the actual damage status of the potential failure area.

8. The artificial intelligence-based T-beam production management system according to claim 7, characterized in that: The establishment of a potential failure area damage assessment report including damage type and damage degree includes: comparing a subtraction relationship mapping table of potential failure area position characteristics and acoustic wave signal characteristics with a mapping table of acoustic wave measured relationships of potential failure areas to obtain measured damage characteristics of different potential failure areas, wherein the measured damage characteristics include acoustic wave propagation time deviation, acoustic wave amplitude attenuation deviation, and acoustic wave frequency variation deviation; An empirical fitting method is used to integrate the measured damage characteristics of different potential failure areas to determine the damage types of different potential failure areas. The damage types include crack propagation, interface detachment, material fracture, voids / pores, and corrosion / aging. The damage types and damage characteristics of different potential failure areas are mapped and associated to generate a potential failure area damage assessment report containing the damage type and damage degree, where the damage degree includes slight damage, moderate damage, and severe damage.

9. The artificial intelligence-based T-beam production management system according to claim 1, characterized in that: The method of obtaining current and historical potential failure area damage assessment reports and obtaining maintenance priority rankings for the potential failure areas includes: deriving various damage types and damage degrees existing in different potential failure areas from historical potential failure area damage assessment reports, and counting the number of records of various damage degrees for different potential failure areas under the current damage type; The number of records of various damage levels is imported into the damage expansion trend discrimination formula of the potential failure area under the current damage type. The quantitative calculation result value is matched with the pre-set range value of the damage expansion trend of rapid expansion, moderate expansion, and slow expansion to determine the damage expansion trend of different potential failure areas under the current damage type. All potential failure areas are integrated into a set of potential failure area sequences with rapid expansion, moderate expansion, and slow expansion.

10. The artificial intelligence-based T-beam production management system according to claim 9, characterized in that: The maintenance plan including operation instructions and risk warnings is generated and sent to the terminal based on the maintenance priority, including: For the rapidly expanding potential failure area sequence set, the highest priority is given. Based on the corresponding damage types of different potential failure areas in the sequence set, immediate repair operation instructions are set. Based on the corresponding positions of different potential failure areas in the sequence set, full inspection instructions are set for the same positions of all production components in this batch. For a moderately expanded potential failure area sequence set, the priority is medium. Based on the corresponding positions of different potential failure areas in the sequence set, a sampling re-inspection instruction is set for the same position of the components produced in this batch; For the slowly expanding potential failure area sequence set, the priority is the lowest, and position marks are set based on the corresponding positions of different potential failure areas in the sequence set.

Images