A crane metal structure life evaluation method based on digital twinning and non-probability reliability fusion

Abstract

The application discloses a crane metal structure life evaluation method based on digital twinning and non-probability reliability fusion, and belongs to the field of crane safety evaluation; the scheme collects design parameters, monitoring data and environmental data, constructs and calibrates the corresponding digital twinning model of the crane metal structure; then, the material, load, geometry, corrosion and crack uncertainty are characterized based on interval variables, credibility distribution and ellipsoid convex model; the non-probability reliability index is calculated by establishing strength, stiffness, stability and fatigue limit state functions; then, the coupling damage evolution model is constructed in combination with stress spectrum, corrosion thinning and crack propagation, and the residual life is corrected and calculated, and the safety level, early warning information and maintenance suggestions are output; the method can improve the real-time performance, precision and engineering applicability of life evaluation.

Description

Technical Field

[0001] This invention relates to the field of crane safety assessment technology, and in particular to a method for assessing the lifespan of crane metal structures based on the fusion of digital twins and nonprobabilistic reliability. Background Technology

[0002] Crane metal structures, especially the main beams, end beams, support connection areas, and welded joints of bridge cranes, are subjected to alternating loads, impact loads, and environmental corrosion over long periods of service. Their service performance directly affects the operational safety of the crane. As the service life of in-service equipment increases, problems such as degradation of structural material properties, reduction of plate thickness, expansion of weld defects, and decrease in local stiffness gradually accumulate. Therefore, conducting life assessments of crane metal structures is of great significance.

[0003] In existing technologies, most methods for assessing the lifespan of crane metal structures still rely heavily on fatigue theory, often employing stress spectrum statistics, rainflow counting, and fatigue cumulative damage methods to calculate the remaining life of the structure. While these methods have some engineering basis, they are typically based on idealized assumptions, often treating material, load, and structural parameters as deterministic values, or requiring sufficient samples to establish a stable probability distribution. In crane field inspections and in-service assessments, relevant sample data are usually limited, and the service environment is complex. This makes it difficult for existing methods to fully reflect multi-source uncertainties such as material fluctuations, load fluctuations, corrosion effects, crack evolution, and geometric weakening, leading to assessment results that are easily deviated from actual service conditions.

[0004] Furthermore, while some existing solutions incorporate digital twin technology for modeling and visualizing crane structures, they largely remain at the level of model display, state mapping, or localized monitoring. They haven't deeply integrated digital twin models with life assessment and reliability analysis, making it difficult to achieve online updates based on measured data, identification of hazardous areas, and dynamic correction of remaining life. Other existing solutions, while considering reliability analysis, typically focus on static safety checks, lacking joint modeling with service damage evolution, corrosion thinning, and crack propagation processes. This makes it difficult to form a continuous analytical chain for life assessment. Particularly in crane field inspection scenarios, relying solely on periodic inspection results or expert judgment not only makes it difficult to quantify the current safety margin of the structure in a timely manner but also makes it difficult to establish a clear link between the current safety status and future remaining life. This results in problems such as assessment lag, large result dispersion, and insufficient targeted early warning.

[0005] Therefore, there is an urgent need to provide a life assessment method for crane metal structures based on the fusion of digital twins and nonprobabilistic reliability, in order to solve the problems in existing technologies that are difficult to handle small sample uncertainties, difficult to reflect the real-time service status of the structure, difficult to achieve joint assessment of multiple failure modes, or difficult to form a closed loop for life assessment. Summary of the Invention

[0006] In view of this, the purpose of this invention is to propose a life assessment method for crane metal structures based on the fusion of digital twin and nonprobabilistic reliability, which aims to solve the problems in the prior art that are difficult to handle small sample uncertainties, difficult to reflect the real-time service status of the structure, difficult to achieve joint assessment of multiple failure modes, or difficult to form a closed loop for life assessment.

[0007] To achieve the above-mentioned technical objectives, the technical solution adopted by this invention is as follows: A method for assessing the lifespan of crane metal structures based on the fusion of digital twin and nonprobabilistic reliability includes: S1. Obtain relevant data on the crane's metal structure and construct a digital twin model of the crane's metal structure; S2. Based on the relevant data, update the digital twin model to obtain a state-updated digital twin model that matches the actual state of the crane's metal structure. S3. Determine the uncertainty parameter model of the crane's metal structure based on the state-updated digital twin model; S4. Based on the state update digital twin model and the uncertainty parameter model, calculate the non-probabilistic reliability index of the crane metal structure; S5. Based on the state-updated digital twin model, determine the damage evolution model of the crane's metal structure and calculate the cumulative loss-related parameters of the crane's metal structure. S6. Based on the non-probabilistic reliability index, modify the damage evolution model and calculate the remaining life of the crane's metal structure; S7. Determine the safety level based on the remaining lifespan, the non-probabilistic reliability index, and the loss accumulation related parameters, and output the lifespan assessment results for the crane's metal structure.

[0008] As one possible implementation, further, in step S1 of this solution, the crane metal structure includes at least a main beam, end beams, support connection area, and dangerous weld seams.

[0009] As one possible implementation, further, in step S1 of this solution, the relevant data includes structural record data, real-time monitoring data, and environmental data; The structural record data includes one or more of the following: design parameters, material parameters, operating condition parameters, weld layout information, and historical maintenance information. The real-time monitoring data includes at least one of the following: stress data, strain data, displacement data, and vibration data. The environmental data includes at least one of the following: temperature data, humidity data, and corrosive medium concentration data; The digital twin model includes a geometric model, a finite element response model, and / or a monitoring mapping model.

[0010] As a preferred implementation method, step S2 of this scheme preferably includes: preprocessing and multi-source fusion of real-time monitoring data to obtain multi-source monitoring data, and inverting and updating the state parameters in the initial digital twin model according to the deviation between the measured response and the simulation response to obtain a state-updated digital twin model.

[0011] The preprocessing includes one or more of the following: time synchronization, working condition segmentation, and anomaly removal. The working condition segmentation includes cyclically dividing the real-time monitoring data into lifting, translation, braking, and no-load return working conditions, and synchronizing and aligning the multi-source monitoring data based on a unified timestamp.

[0012] As a preferred implementation method, preferably, in step S2 of this scheme, the state parameters include one or more of the following: elastic modulus degradation coefficient, cross section weakening coefficient, weld defect equivalent parameter, boundary stiffness correction coefficient, and residual stress correction coefficient.

[0013] As a preferred implementation method, in step S2 of this scheme, the inversion update adopts a parameter identification method with the goal of minimizing the comprehensive deviation between the measured stress, measured strain, and measured displacement and the simulated stress, simulated strain, and simulated displacement, and updates the elastic modulus degradation coefficient, section weakening coefficient, weld defect equivalent parameter, boundary stiffness correction coefficient, and residual stress correction coefficient.

[0014] As a preferred implementation method, step S3 of this solution preferably includes: extracting multiple feature parameters from the state update digital twin model, representing each feature parameter as an interval variable, and combining multi-source monitoring data to assign credibility to each interval variable, thereby constructing an uncertainty parameter ellipsoidal convex model; The characteristic parameters include one or more of the following: material strength parameters, load parameters, geometric parameters, corrosion parameters, and crack parameters. The interval variables include at least one of the following: material strength interval, load amplitude interval, dynamic load impact coefficient interval, plate thickness weakening interval, corrosion rate interval, initial crack size interval, and residual stress correction interval. The confidence level allocation is determined based on design data, historical inspection data, online monitoring data, and / or non-destructive testing data.

[0015] As a preferred implementation method, step S4 of this solution preferably includes: establishing a limit function based on the state update digital twin model and the uncertainty parameter ellipsoidal convex model, and then calculating the non-probabilistic reliability index of the crane metal structure; The limit function includes one or more of the following: strength limit state function, stiffness limit state function, stability limit state function, and fatigue limit state function; Among them, the comprehensive weights of the strength limit state function, stiffness limit state function, stability limit state function and fatigue limit state function are determined by the analytic hierarchy process based on the monitoring sensitivity correction, and the non-probabilistic reliability index is calculated based on the minimum distance from the failure boundary to the origin in the normalized parameter space.

[0016] As a preferred implementation method, S5 of this scheme includes: generating an equivalent stress spectrum based on the stress time history output by the state-updated digital twin model, and constructing a coupled damage evolution model by combining the corrosion thinning amount and crack propagation amount, and calculating the current damage accumulation value and damage growth rate. The coupled damage evolution model includes fatigue damage components, corrosion damage components, and crack propagation damage components. Among them, the fatigue damage component is calculated based on the equivalent stress spectrum, the corrosion damage component is calculated based on the corrosion rate and service time, and the crack propagation damage component is calculated based on the crack size evolution.

[0017] As a preferred implementation method, in step S6 of this scheme, a lifetime correction function is constructed based on a non-probabilistic reliability index to correct the damage growth rate; wherein, the smaller the non-probabilistic reliability index, the more conservative the remaining lifetime assessment result is.

[0018] As a preferred implementation method, in step S7 of this solution, the safety level is divided according to the combination relationship of the remaining life threshold, the non-probabilistic reliability index threshold and the damage accumulation value threshold, and the corresponding local reinforcement suggestions, weld repair suggestions, load reduction operation suggestions, re-inspection cycle suggestions or shutdown replacement suggestions are output based on the safety level. In addition to outputting life assessment results, the system also outputs early warning information and maintenance recommendations.

[0019] As a preferred implementation method, step S7 of this solution further includes: inputting the newly acquired non-destructive testing results, re-inspection results and maintenance results back to the state update digital twin model, and updating the interval variables, uncertainty parameter ellipsoidal convex model parameters and lifetime correction function to form a lifetime assessment closed loop.

[0020] Based on the above, this solution also proposes a life assessment system for crane metal structures based on the fusion of digital twin and non-probabilistic reliability, which includes: The digital model building unit is used to acquire relevant data about the crane's metal structure and build a digital twin model of the crane's metal structure. The model update unit is used to update the digital twin model based on the relevant data to obtain a state-updated digital twin model that matches the actual state of the crane's metal structure. A parameter reliability allocation unit is used to determine the uncertainty parameter model of the crane's metal structure based on the state-updated digital twin model. The reliability calculation unit is used to calculate the non-probabilistic reliability index of the crane metal structure based on the state-updated digital twin model and the uncertainty parameter model. The damage evolution calculation unit determines the damage evolution model of the crane metal structure based on the state update digital twin model and calculates the cumulative loss-related parameters of the crane metal structure. The life assessment unit is used to modify the damage evolution model based on the non-probabilistic reliability index and calculate the remaining life of the crane's metal structure. The result output unit is used to determine the safety level based on the remaining life, the non-probabilistic reliability index, and the loss accumulation correlation parameters, and output the life assessment results for the crane's metal structure.

[0021] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art: This scheme uses interval variables, confidence allocation, and ellipsoidal convex models to uniformly characterize uncertainties such as materials, loads, geometry, corrosion, and cracks. It does not rely on a large number of samples to fit the probability distribution, so it can adapt to the actual working conditions of limited test data and difficult-to-determine statistical distribution in crane field inspections. It reduces the applicability defects of traditional probability methods in scenarios with insufficient samples and improves the reliability of life assessment under small sample conditions.

[0022] This solution calibrates and updates the digital twin model by monitoring data in real time, gradually correcting the model parameters from design values ​​to service values. Reliability analysis and life calculation are then carried out based on the calibrated model, which can simultaneously reflect the structural stiffness degradation, cross-sectional weakening, weld defect evolution and environmental corrosion effects, thereby significantly improving the consistency between the assessment results and the actual state of the structure.

[0023] This scheme also establishes strength, stiffness, stability, and fatigue limit state functions, and incorporates fatigue damage, corrosion damage, and crack propagation damage into a coupled damage evolution model. Then, it uses non-probabilistic reliability indicators to correct the life results, thereby avoiding the one-sidedness caused by simply relying on fatigue models or static safety factors for evaluation. This makes the remaining life results more consistent with the actual failure laws of in-service crane metal structures, and realizes the joint evaluation of multiple failure modes and multiple damage mechanisms.

[0024] This solution integrates monitoring, modeling, reliability analysis, life assessment, risk classification, and maintenance recommendations into a unified system. It also allows for continuous feedback of subsequent non-destructive testing, re-inspection, and maintenance results, enabling iterative model updates and forming a closed-loop early warning mechanism. This provides quantitative data for crane re-inspection scheduling, local reinforcement, weld repair, reduced-load operation, and shutdown replacement, reducing reliance on purely empirical judgment and demonstrating strong practical engineering value. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a simplified implementation flowchart of the evaluation method for this scheme; Figure 2 This is a simplified connection diagram of the unit modules of the evaluation system for this solution. Detailed Implementation

[0027] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be particularly noted that the following embodiments are for illustrative purposes only and do not limit the scope of the invention. Similarly, the following embodiments are only some, not all, embodiments of the present invention, and all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] like Figure 1 As shown in the figure, this embodiment presents a method for assessing the lifespan of crane metal structures based on the fusion of digital twins and non-probabilistic reliability, which includes: S1. Obtain relevant data on the crane's metal structure and construct a digital twin model of the crane's metal structure; S2. Based on the relevant data, update the digital twin model to obtain a state-updated digital twin model that matches the actual state of the crane's metal structure. S3. Determine the uncertainty parameter model of the crane's metal structure based on the state-updated digital twin model; S4. Based on the state update digital twin model and the uncertainty parameter model, calculate the non-probabilistic reliability index of the crane metal structure; S5. Based on the state-updated digital twin model, determine the damage evolution model of the crane's metal structure and calculate the cumulative loss-related parameters of the crane's metal structure. S6. Based on the non-probabilistic reliability index, modify the damage evolution model and calculate the remaining life of the crane's metal structure; S7. Determine the safety level based on the remaining lifespan, the non-probabilistic reliability index, and the loss accumulation related parameters, and output the lifespan assessment results for the crane's metal structure.

[0029] As one possible implementation method for the metal structure of the crane, in step S1 of this solution, the metal structure of the crane includes at least a main beam, end beams, support connection area, and dangerous weld seams.

[0030] Regarding the collection of relevant data, in step S1 of this solution, the relevant data includes structural record data, real-time monitoring data, and environmental data; The structural record data includes one or more of the following: design parameters, material parameters, operating condition parameters, weld layout information, and historical maintenance information. The real-time monitoring data includes at least one of the following: stress data, strain data, displacement data, and vibration data. The environmental data includes at least one of the following: temperature data, humidity data, and corrosive medium concentration data; The digital twin model includes a geometric model, a finite element response model, and / or a monitoring mapping model.

[0031] Existing crane life assessments are typically based on design drawings or static theoretical models, which fail to reflect the actual service condition of in-service metal structures, especially the simultaneous impact of load fluctuations, boundary stiffness changes, corrosion thinning, and weld defect propagation on structural response. Therefore, this scheme S1 constructs an initial digital twin model that simultaneously maps geometry, mechanical response, monitoring data, and damage status, providing a unified platform for subsequent online calibration, reliability calculations, and life assessments.

[0032] As an example of implementation, in this solution, the crane's metal structure is constantly... The measured state vector (i.e., the vector representation of real-time monitoring data) can be defined as follows: in, Indicates the measured stress; Indicates the measured strain; Indicates the measured displacement or deflection; Represents the characteristic quantity of vibration response; Indicates ambient temperature; Indicates ambient humidity; This indicates the concentration of the corrosive medium.

[0033] When constructing a digital twin model, the parameter vector to be calibrated in the digital twin model can be defined as follows: in, Represents the equivalent elastic modulus; This represents the boundary stiffness correction factor; Indicates the cross-sectional attenuation factor; This indicates the amount of residual stress correction; Indicates the amount of thinning due to corrosion; This indicates the characteristic size of the crack.

[0034] As an example, in step S1 of this scheme, the design parameters, material parameters and working condition parameters of the bridge crane's metal structure are first collected.

[0035] The design parameters include span, main beam cross-sectional dimensions, web thickness, flange thickness, stiffening rib arrangement, support type, and weld distribution.

[0036] Material parameters include yield strength, tensile strength, elastic modulus, Poisson's ratio, and material fatigue parameters.

[0037] Operating parameters include rated lifting capacity, lifting speed, trolley travel speed, trolley travel speed, impact coefficient, and historical load spectrum.

[0038] In addition, step S1 of this solution may include the following sub-steps: Based on the data mentioned above, a three-dimensional geometric model is established, and a finite element response model is further established. Let the displacement vector of the structural nodes be... The external load vector is Then the structural equilibrium equation can be written as: in, Indicates the initial parameters The overall structural stiffness matrix constituted; Indicates time The nodal displacement vector; Indicates time The external load vector.

[0039] The element strain and stress can be further obtained from the structural displacement, i.e.: in, This is the strain-displacement transformation matrix; For the initial elastic modulus Compared to Poisson The constitutive matrix formed; The element strain vector; This represents the element stress vector.

[0040] Subsequently, strain sensors, stress sensors, displacement sensors, vibration sensors, and environmental sensors are installed at the mid-span, end transition zone, support connection zone, weld toe, and stress concentration zone of the main beam to form a data acquisition network. For the strain sensor values, the equivalent stress at the monitoring point can be obtained first based on elastic relationships, and its function definition is as follows: in, Indicates monitoring point In time The measured equivalent stress; This represents the measured strain.

[0041] Considering that crane structural evaluation depends not only on static response but also on service environment and damage state, this scheme simultaneously defines the initial damage vector in step S1, and its function is defined as follows: in, Indicates the initial corrosion thinning amount; This indicates the initial crack size. For areas where no cracks were detected, the initial crack size can be set to... Or it can be set as the smallest detectable size for non-destructive testing.

[0042] Therefore, the structural form of the initial digital twin model output by S1 can be uniformly expressed as follows: in, Represents a geometric model; This represents the physical simulation model, i.e., the finite element response model; This represents the sensor mapping and data acquisition model, i.e., the monitoring mapping model; This indicates the initial damage state.

[0043] In step S1, the initial modeling of this scheme is completed to obtain an initial digital twin model. However, this model still mainly reflects the design state or initial service state and cannot automatically reflect the stiffness degradation, residual stress changes, boundary condition changes and section weakening caused by long-term service.

[0044] Therefore, Solution S2 utilizes real-time on-site monitoring data to continuously calibrate the digital twin model, transforming it from a "design state" to a "service state," thereby ensuring that subsequent reliability analysis and life assessment use the current actual structural condition. The appendix also explicitly states the need to integrate the digital twin model with on-site test data and continuously update it to ensure that model errors are under control.

[0045] As a preferred implementation method, step S2 of this scheme preferably includes: preprocessing and multi-source fusion of real-time monitoring data to obtain multi-source monitoring data, and inverting and updating the state parameters in the initial digital twin model according to the deviation between the measured response and the simulation response to obtain a state-updated digital twin model.

[0046] The preprocessing includes one or more of the following: time synchronization, working condition segmentation, and anomaly removal. The working condition segmentation includes cyclically dividing the real-time monitoring data into lifting, translation, braking, and no-load return working conditions, and synchronizing and aligning the multi-source monitoring data based on a unified timestamp.

[0047] As a preferred implementation method, preferably, in step S2 of this scheme, the state parameters include one or more of the following: elastic modulus degradation coefficient, cross section weakening coefficient, weld defect equivalent parameter, boundary stiffness correction coefficient, and residual stress correction coefficient.

[0048] As a preferred implementation method, in step S2 of this scheme, the inversion update adopts a parameter identification method with the goal of minimizing the comprehensive deviation between the measured stress, measured strain, and measured displacement and the simulated stress, simulated strain, and simulated displacement, and updates the elastic modulus degradation coefficient, section weakening coefficient, weld defect equivalent parameter, boundary stiffness correction coefficient, and residual stress correction coefficient.

[0049] As an example of implementation, step S2 of this solution may include the following sub-steps: First, the multi-source monitoring data collected by S1 is preprocessed, including time synchronization, anomaly removal, noise filtering, operating condition identification, and missing data completion. Let the processed monitoring vector still be denoted as... .

[0050] In the digital twin model, the current parameter vector The predicted response vector is denoted as for: in, This represents the mapping function used to calculate the monitoring response from the digital twin model; The digital twin model representing the previous moment, This is the data vector for the simulation response.

[0051] Define the residual vector between the measured response and the simulated response. for: To obtain the model parameters that best match the current monitoring data, a weighted least squares objective function is constructed: in, The observation weight matrix is ​​used to characterize the importance of data in different dimensions of stress, strain, displacement, and vibration. is the regularization coefficient, used to limit the parameter update magnitude between adjacent time steps and prevent parameter jumps caused by local abnormal data; For the previous moment ( ) The calibrated parameter vector.

[0052] The following solution is performed: The optimal calibration parameters for the current moment can then be obtained. .

[0053] Under the above objective function, the reasoning logic is as follows: First term To make the simulated response approximate the measured response as closely as possible, the second term... This ensures that parameter updates have physical continuity, thus balancing fitting accuracy and engineering stability.

[0054] By updating the parameters, a digital twin model of the service status is obtained: in, This represents the model update operator.

[0055] After the above sub-steps, the output of step S2 of this scheme includes: the calibrated digital twin model. Update parameter vector And corrosion thinning amount obtained based on multi-source monitoring data Crack size Section weakening coefficient and equivalent elastic modulus .

[0056] Crane inspections and life assessments often face challenges such as small sample sizes, discrete data, complex random load disturbances, and difficulties in accurately fitting probability distributions. Directly using probabilistic reliability models often leads to biased results due to insufficient distribution assumptions.

[0057] Therefore, in this scheme, step S3 provides a unified and calculable expression for uncertain factors such as materials, loads, geometry, corrosion, and cracks, under the conditions of insufficient experimental data and unclear statistical distribution.

[0058] As a preferred implementation method, step S3 of this solution preferably includes: extracting multiple feature parameters from the state update digital twin model, representing each feature parameter as an interval variable, and combining multi-source monitoring data to assign credibility to each interval variable, thereby constructing an uncertainty parameter ellipsoidal convex model; The characteristic parameters include one or more of the following: material strength parameters, load parameters, geometric parameters, corrosion parameters, and crack parameters. The interval variables include at least one of the following: material strength interval, load amplitude interval, dynamic load impact coefficient interval, plate thickness weakening interval, corrosion rate interval, initial crack size interval, and residual stress correction interval. The confidence level allocation is determined based on design data, historical inspection data, online monitoring data, and / or non-destructive testing data.

[0059] As an example of implementation, step S3 of this solution may include the following sub-steps: Based on the current service status parameters obtained in step S2, the parameters with the greatest impact on life assessment are selected to form an uncertainty variable vector. Its definition is as follows: Among them, the vector of uncertain variables It may include at least: material yield strength, equivalent elastic modulus, lifting load amplitude, impact coefficient, plate thickness reduction, corrosion rate, initial crack size, and residual stress correction.

[0060] Each parameter is represented in interval form: Thus, the set of interval parameters is obtained. : in, Indicates the first The parameters at time... The lower bound; It indicates its upper bound.

[0061] To integrate design data, historical inspection data, online monitoring data, and non-destructive testing data, a confidence level allocation is performed using evidence theory. Let the... One source of evidence for the interval event The basic credibility allocation is as follows The multiple evidence conflict coefficient Defined as: in, The number of sources of evidence; This represents the empty set.

[0062] After the conflict coefficient is determined, the synthesis confidence level is: Under the above function definition, the reasoning logic is as follows: first identify the contradictory parts among multiple sources of evidence, i.e., conflict terms. The non-conflicting parts are then normalized and merged to obtain a more reliable comprehensive interval evaluation result.

[0063] To further describe the correlation between the uncertain variables, an ellipsoidal convex model is introduced. First, the interval center and interval radius are defined: Based on this, standardized variables are constructed: make Then the convex region of the ellipsoid can be represented as: in, This is a correlation matrix, used to describe the coupling relationships between different uncertain variables.

[0064] After the above steps, we can obtain: the set of intervals. , Integration of credibility allocation Interval center vector Interval radius vector and correlation matrix .

[0065] Even with a service-state digital twin model and uncertainty intervals, if strength failure, stiffness failure, stability failure, and fatigue failure cannot be uniformly mapped to quantifiable safety margin indicators, it is impossible to determine the current safety boundary of the structure.

[0066] Therefore, step S4 of this scheme establishes a unified reliability evaluation index for multiple failure modes under interval uncertainty conditions, and outputs it as the safety margin required for lifetime correction.

[0067] As a preferred implementation method, step S4 of this solution preferably includes: establishing a limit function based on the state update digital twin model and the uncertainty parameter ellipsoidal convex model, and then calculating the non-probabilistic reliability index of the crane metal structure; The limit function includes one or more of the following: strength limit state function, stiffness limit state function, stability limit state function, and fatigue limit state function; Among them, the comprehensive weights of the strength limit state function, stiffness limit state function, stability limit state function and fatigue limit state function are determined by the analytic hierarchy process based on the monitoring sensitivity correction, and the non-probabilistic reliability index is calculated based on the minimum distance from the failure boundary to the origin in the normalized parameter space.

[0068] As an example of implementation, step S4 of this solution may include the following sub-steps: Based on the updated digital twin model in step S2 and the relevant data corresponding to the crane's metal structure, multiple limit state functions are established.

[0069] 1. Strength limit state function Strength limit state function Defined as follows: in, This represents the allowable stress obtained after taking into account corrosion, residual stress, and material degradation. Indicates the structure under uncertainty parameters The equivalent stress under the following conditions.

[0070] 2. Stiffness Limit State Function Stiffness Limit State Function Defined as follows: in, Indicates the allowable deflection; This represents the calculated deflection under the current operating conditions.

[0071] 3. Stability Limit State Function Stability Limit State Function Defined as follows: in, This represents the critical instability load after considering section weakening and boundary correction; This indicates the current equivalent load.

[0072] 4. Fatigue Limit State Function Fatigue limit state function Defined as follows: in, This indicates the allowable number of fatigue cycles in the current state; This indicates the number of loops that have consumed or provided an equivalent demand.

[0073] Since the functions mentioned above have different dimensions, normalization is performed first: in, Indicates the first Limit state function The normalized scaling factor.

[0074] Then, an improved AHP method is used to determine the weights. Let the initial weights obtained from the traditional AHP be... The monitoring sensitivity is The corrected weights are: in, It can be determined based on the sensitivity of the corresponding failure mode to changes in the real-time monitoring response.

[0075] The logic of the above correction function not only considers the subjective importance of expert experience, but also the actual sensitivity reflected by the field data.

[0076] Then, construct the comprehensive limit state function. : in, For the first The weights of each limit state function are summed to 1.

[0077] Next, within the convex ellipsoidal domain constructed by S3, the path from the center point to the failure surface is searched. The minimum standardized distance is defined as the nonprobabilistic reliability index: The reasoning above is as follows: the farther the failure surface is from the center of the interval, the greater the safety margin the structure has within the current uncertainty range; if the failure surface is closer to the center of the interval, the structure is more sensitive to parameter fluctuations and has lower reliability. Therefore, It can characterize the safety margin of a structure under non-probabilistic uncertainty conditions.

[0078] After processing in step S4, the core output quantity is: non-probabilistic reliability index. This indicator will be directly passed to S5 and S6 for hazardous site screening and lifespan correction.

[0079] Judging structural safety solely based on current reliability metrics fails to address the crucial question of how long the crane's metal structure can remain in service. Furthermore, estimating lifespan solely using traditional fatigue models overlooks real damage mechanisms such as corrosion thinning and crack propagation.

[0080] Therefore, this scheme S5 establishes a coupled damage model of fatigue, corrosion and crack propagation based on the real response output by the service-state digital twin model, forming the damage state quantities required for the remaining life calculation.

[0081] As a preferred implementation method, S5 of this scheme includes: generating an equivalent stress spectrum based on the stress time history output by the state-updated digital twin model, and constructing a coupled damage evolution model by combining the corrosion thinning amount and crack propagation amount, and calculating the current damage accumulation value and damage growth rate. The coupled damage evolution model includes fatigue damage components, corrosion damage components, and crack propagation damage components. Among them, the fatigue damage component is calculated based on the equivalent stress spectrum, the corrosion damage component is calculated based on the corrosion rate and service time, and the crack propagation damage component is calculated based on the crack size evolution.

[0082] As an example of implementation, step S5 of this solution may include the following sub-steps: First, the stress time history of key hazardous points is output using the calibrated digital twin model obtained in step S2. The working cycle is decomposed and rainflow is counted to obtain the first... Amplitude of stress cycle Mean stress and number of loops .

[0083] To comprehensively consider the influence of mean stress, the equivalent stress amplitude is constructed as follows: in, Indicates the tensile strength of the material; Indicates the first The equivalent force amplitude of a cyclic type.

[0084] Further data was obtained as follows: 1. Fatigue damage component The fatigue damage component is calculated based on the SN relationship, which is defined as follows: in, Indicates the first The number of fatigue life cycles corresponding to a stress amplitude; and These are material fatigue parameters.

[0085] Based on the Miner linear accumulation criterion, fatigue damage components for: in, Indicates the number of stress cycle categories. The number of loops. To calculate the index.

[0086] Under the above formula reasoning logic, each type of stress cycle will consume a certain share of fatigue life, and the sum of all life shares will result in cumulative fatigue damage.

[0087] 2. Corrosion damage component Let the corrosion rate be... During the inspection period Internal corrosion thinning increment for: The cumulative corrosion thinning amount is: in, , They are time , The cumulative thinning amount over time.

[0088] If the critical thinning amount As the corrosion failure limit, the corrosion damage component Defined as: in, This indicates the critical corrosion thinning amount at which the lower safety limit of the cross-section is reached.

[0089] 3. Crack propagation damage component For welds and areas of high stress concentration, the Paris crack propagation formula is used: in, Indicates the crack size; Indicates the number of loop iterations; , For crack propagation material parameters; This represents the magnitude of the stress intensity factor.

[0090] And: in, Indicates the geometric correction factor; This indicates the stress amplitude at the location of the crack.

[0091] Integrating the above crack propagation relationship yields the current crack size. Further define the crack damage components as: in, Indicates the initial crack size; Indicates the critical crack size.

[0092] 4. Total damage component Unify and couple the three types of damage: in, These represent the weights for fatigue, corrosion, and crack damage, respectively, satisfying the following: .

[0093] To facilitate the calculation of remaining lifetime, the current damage growth rate is redefined: After processing in step S5, the output is: Total Damage and damage growth rate .

[0094] Given the current damage amount and damage growth rate, if the current safety margin of the structure is not considered and the life is simply extrapolated linearly based on the damage, it may still give an overly optimistic conclusion when the reliability is insufficient, or it may give an overly conservative conclusion when the safety margin is sufficient.

[0095] Therefore, in this scheme S6, the non-probabilistic reliability index obtained in S4 is introduced into the lifetime model to make a safety margin correction on the remaining lifetime, so that the lifetime result takes into account both damage accumulation and current structural reliability.

[0096] As a preferred implementation method, in step S6 of this scheme, a lifetime correction function is constructed based on a non-probabilistic reliability index to correct the damage growth rate; wherein, the smaller the non-probabilistic reliability index, the more conservative the remaining lifetime assessment result is.

[0097] As an example of implementation, step S6 of this solution may include the following sub-steps: Define the total damage threshold when the structure reaches its life limit as: Generally acceptable The current damage is obtained from S5. The current damage growth rate is If we temporarily approximate it using a local linear approach, then the future... The damage after a certain period of time is as follows: To bring it to its limit of damage, that is: The remaining lifespan was not corrected. for: Under the reasoning logic of the above formula, when the damage growth rate per unit time is approximately stable, the time required to go from the current damage to the failure threshold is the basic remaining lifetime.

[0098] To reflect the impact of non-probabilistic reliability indicators, a reliability correction function is constructed. Its definition is as follows: in, To correct the sensitivity coefficient; To prevent the use of tiny positive numbers with a denominator of zero; This is the non-probabilistic reliability index output by S4.

[0099] The corrected remaining lifetime is: Under the computational logic of this formula: when A smaller value indicates that the failure surface is closer to the current uncertainty center, and the structural safety margin is insufficient. Enlarging, leading to a longer remaining lifespan Smaller, thus improving the conservatism of early warnings; when A larger value indicates higher reliability, thus the correction range decreases.

[0100] Based on this, lifetime utilization rate can also be defined: in, Indicates the design life; It represents the ratio of remaining service life to design service life.

[0101] After processing in step S6, the core output is: corrected remaining lifetime. and lifespan utilization .

[0102] Even after obtaining reliability metrics and remaining lifespan data, if these cannot be transformed into actionable early warning conclusions and maintenance recommendations, they will still be difficult to directly serve inspection decisions, operation and maintenance management, and regulatory scenarios.

[0103] Therefore, S7 of this scheme transforms the calculation results of S4 and S6 into explicit safety levels, early warning states, and maintenance and handling recommendations, making the assessment results valuable for engineering applications. The appendix also explicitly states that early warnings, assessment reports, and maintenance recommendations will be output when the remaining service life is below the threshold.

[0104] As a preferred implementation method, in step S7 of this solution, the safety level is divided according to the combination relationship of the remaining life threshold, the non-probabilistic reliability index threshold and the damage accumulation value threshold, and the corresponding local reinforcement suggestions, weld repair suggestions, load reduction operation suggestions, re-inspection cycle suggestions or shutdown replacement suggestions are output based on the safety level. In addition to outputting life assessment results, the system also outputs early warning information and maintenance recommendations.

[0105] As an example of implementation, step S7 of this solution may include the following sub-steps: Let the threshold for nonprobabilistic reliability index be... The damage threshold is The remaining lifetime threshold is To comprehensively consider the three core indicators, a comprehensive risk indicator is constructed. : in, , , For risk fusion weighting, to meet ; This serves as a reliability reference constant; comprehensive risk index The larger the value, the higher the risk.

[0106] The reasoning logic of this formula is as follows: Describe the extent to which the damage has been consumed; Describe the degree of reliability deficiency. The smaller the value, the larger the value. Describes how close the remaining lifetime is to the threshold. The smaller the value, the larger the value.

[0107] Then based on comprehensive risk indicators The structural state is classified into levels. For example, it can be classified as follows: The safety level is Level 1, and the system is operating normally. The safety level is Level 2; enhanced monitoring is required. The safety level is level three; load limiting or partial repair is recommended. The safety level is level four. It is recommended to shut down the machine for inspection or replacement.

[0108] Meanwhile, to reflect the early warning output required by claim 1, triggering rules can be set: in, This indicates that an alert has been triggered; This indicates that no warning will be triggered.

[0109] The final assessment report should include at least the following: current ranking of hazardous areas, and non-probabilistic reliability indicators. Current total damage Remaining lifespan Risk indicators Safety level and corresponding maintenance recommendations.

[0110] Maintenance recommendations may include weld repair, local reinforcement, reduced load operation, shortening the re-inspection cycle, or shutdown for replacement.

[0111] To improve the optimization capability of the model, as a preferred implementation method, step S7 of this scheme further includes: feeding back the newly acquired non-destructive testing results, re-inspection results and maintenance results to the state update digital twin model, and updating the interval variables, uncertainty parameters, ellipsoidal convex model parameters and lifetime correction function to form a lifetime assessment closed loop.

[0112] Combination Figure 2 As shown above, this solution also proposes a life assessment system for crane metal structures based on the fusion of digital twin and non-probabilistic reliability, which includes: The digital model building unit is used to acquire relevant data about the crane's metal structure and build a digital twin model of the crane's metal structure. The model update unit is used to update the digital twin model based on the relevant data to obtain a state-updated digital twin model that matches the actual state of the crane's metal structure. A parameter reliability allocation unit is used to determine the uncertainty parameter model of the crane's metal structure based on the state-updated digital twin model. The reliability calculation unit is used to calculate the non-probabilistic reliability index of the crane metal structure based on the state-updated digital twin model and the uncertainty parameter model. The damage evolution calculation unit determines the damage evolution model of the crane metal structure based on the state update digital twin model and calculates the cumulative loss-related parameters of the crane metal structure. The life assessment unit is used to modify the damage evolution model based on the non-probabilistic reliability index and calculate the remaining life of the crane's metal structure. The result output unit is used to determine the safety level based on the remaining life, the non-probabilistic reliability index, and the loss accumulation correlation parameters, and output the life assessment results for the crane's metal structure.

[0113] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0114] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0115] The above description is only a part of the embodiments of the present invention and does not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made based on the content of the present invention specification and drawings, or direct or indirect application in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for assessing the lifespan of crane metal structures based on the fusion of digital twin and non-probabilistic reliability, characterized in that, It includes: S1. Obtain relevant data on the crane's metal structure and construct a digital twin model of the crane's metal structure; S2. Based on the relevant data, update the digital twin model to obtain a state-updated digital twin model that matches the actual state of the crane's metal structure. S3. Determine the uncertainty parameter model of the crane's metal structure based on the state-updated digital twin model; S4. Based on the state update digital twin model and the uncertainty parameter model, calculate the non-probabilistic reliability index of the crane metal structure; S5. Based on the state-updated digital twin model, determine the damage evolution model of the crane's metal structure and calculate the cumulative loss-related parameters of the crane's metal structure. S6. Based on the non-probabilistic reliability index, modify the damage evolution model and calculate the remaining life of the crane's metal structure; S7. Determine the safety level based on the remaining lifespan, the non-probabilistic reliability index, and the loss accumulation related parameters, and output the lifespan assessment results for the crane's metal structure.

2. The crane metal structure life assessment method based on the fusion of digital twin and non-probabilistic reliability as described in claim 1, characterized in that, In step S1, the crane's metal structure includes at least a main beam, end beams, support connection areas, and dangerous weld seams. In step S1, the relevant data includes structural record data, real-time monitoring data, and environmental data; The structural record data includes one or more of the following: design parameters, material parameters, operating condition parameters, weld layout information, and historical maintenance information. The real-time monitoring data includes at least one of the following: stress data, strain data, displacement data, and vibration data. The environmental data includes at least one of the following: temperature data, humidity data, and corrosive medium concentration data; The digital twin model includes a geometric model, a finite element response model, and / or a monitoring mapping model.

3. The crane metal structure life assessment method based on the fusion of digital twin and non-probabilistic reliability as described in claim 2, characterized in that, Step S2 includes: preprocessing and multi-source fusion of real-time monitoring data to obtain multi-source monitoring data, and inverting and updating the state parameters in the initial digital twin model according to the deviation between the measured response and the simulation response to obtain a state-updated digital twin model. The preprocessing includes one or more of the following: time synchronization, working condition segmentation, and anomaly removal. The working condition segmentation includes cyclically dividing the real-time monitoring data according to the lifting working condition, translation working condition, braking working condition and no-load return working condition, and synchronizing and aligning the multi-source monitoring data based on a unified timestamp. In step S2, the state parameters include one or more of the following: elastic modulus degradation coefficient, section weakening coefficient, weld defect equivalent parameter, boundary stiffness correction coefficient, and residual stress correction coefficient.

4. The crane metal structure life assessment method based on the fusion of digital twin and non-probabilistic reliability as described in claim 3, characterized in that, In step S2, the inversion update adopts a parameter identification method with the goal of minimizing the comprehensive deviation between measured stress, measured strain, and measured displacement and simulated stress, simulated strain, and simulated displacement, and updates the elastic modulus degradation coefficient, section weakening coefficient, weld defect equivalent parameter, boundary stiffness correction coefficient, and residual stress correction coefficient.

5. The crane metal structure life assessment method based on the fusion of digital twin and non-probabilistic reliability as described in claim 3 or 4, characterized in that, Step S3 includes: extracting multiple feature parameters from the state update digital twin model, representing each feature parameter as an interval variable, and assigning confidence to each interval variable in combination with multi-source monitoring data to construct an uncertain parameter ellipsoidal convex model; The characteristic parameters include one or more of the following: material strength parameters, load parameters, geometric parameters, corrosion parameters, and crack parameters. The interval variables include at least one of the following: material strength interval, load amplitude interval, dynamic load impact coefficient interval, plate thickness weakening interval, corrosion rate interval, initial crack size interval, and residual stress correction interval. The confidence level allocation is determined based on design data, historical inspection data, online monitoring data, and / or non-destructive testing data.

6. The crane metal structure life assessment method based on the fusion of digital twin and non-probabilistic reliability as described in claim 5, characterized in that, Step S4 includes: establishing a limit function based on the state-updated digital twin model and the uncertainty parameter ellipsoidal convex model, and then calculating the non-probabilistic reliability index of the crane metal structure; The limit function includes one or more of the following: strength limit state function, stiffness limit state function, stability limit state function, and fatigue limit state function; Among them, the comprehensive weights of the strength limit state function, stiffness limit state function, stability limit state function and fatigue limit state function are determined by the analytic hierarchy process based on the monitoring sensitivity correction, and the non-probabilistic reliability index is calculated based on the minimum distance from the failure boundary to the origin in the normalized parameter space.

7. The crane metal structure life assessment method based on the fusion of digital twin and non-probabilistic reliability as described in claim 6, characterized in that, S5 includes: generating an equivalent stress spectrum based on the stress time history output by the updated digital twin model according to the state, and constructing a coupled damage evolution model by combining the corrosion thinning amount and crack propagation amount, and calculating the current damage accumulation value and damage growth rate. The coupled damage evolution model includes fatigue damage components, corrosion damage components, and crack propagation damage components. Among them, the fatigue damage component is calculated based on the equivalent stress spectrum, the corrosion damage component is calculated based on the corrosion rate and service time, and the crack propagation damage component is calculated based on the crack size evolution.

8. The crane metal structure life assessment method based on the fusion of digital twin and non-probabilistic reliability as described in claim 7, characterized in that, In step S6, a lifetime correction function is constructed based on a non-probabilistic reliability index to correct the damage growth rate; wherein, the smaller the non-probabilistic reliability index, the more conservative the remaining lifetime assessment result is.

9. The method for assessing the lifespan of crane metal structures based on the fusion of digital twin and non-probabilistic reliability as described in claim 7, characterized in that, In step S7, the safety level is divided according to the combination relationship of the remaining life threshold, the non-probabilistic reliability index threshold and the damage accumulation value threshold, and the corresponding local reinforcement suggestions, weld repair suggestions, load reduction operation suggestions, re-inspection cycle suggestions or shutdown replacement suggestions are output based on the safety level. In addition to outputting life assessment results, it also outputs early warning information and maintenance recommendations; Step S7 further includes: inputting the newly acquired non-destructive testing results, re-inspection results, and maintenance results back to the state update digital twin model, and updating the interval variables, uncertainty parameters, ellipsoidal convex model parameters, and lifetime correction function to form a lifetime assessment closed loop.

10. A life assessment system for crane metal structures based on the fusion of digital twin and non-probabilistic reliability, characterized in that, It includes: The digital model building unit is used to acquire relevant data about the crane's metal structure and build a digital twin model of the crane's metal structure. The model update unit is used to update the digital twin model based on the relevant data to obtain a state-updated digital twin model that matches the actual state of the crane's metal structure. A parameter reliability allocation unit is used to determine the uncertainty parameter model of the crane's metal structure based on the state-updated digital twin model. The reliability calculation unit is used to calculate the non-probabilistic reliability index of the crane metal structure based on the state-updated digital twin model and the uncertainty parameter model. The damage evolution calculation unit determines the damage evolution model of the crane metal structure based on the state update digital twin model and calculates the cumulative loss-related parameters of the crane metal structure. The life assessment unit is used to modify the damage evolution model based on the non-probabilistic reliability index and calculate the remaining life of the crane's metal structure. The result output unit is used to determine the safety level based on the remaining life, the non-probabilistic reliability index, and the loss accumulation correlation parameters, and output the life assessment results for the crane's metal structure.

Images