Stress spectrum analysis and residual life evaluation method and device for shipbuilding portal crane

Abstract

The invention provides a stress spectrum analysis and residual life evaluation method and device for a shipbuilding portal crane. The method provided by the invention comprises the following steps: acquiring dynamic load parameters of the shipbuilding portal crane under various working conditions; determining a lifting dynamic load coefficient based on the lifting state level corresponding to the dynamic load parameter; a stress spectrum prediction model based on the improved relevance vector machine is updated according to the lifting dynamic load coefficient; determining a first value of a crack growth rate parameter through experimental data fitting and finite element analysis; predicting a real-time stress spectrum under a real-time working condition based on the stress spectrum prediction model; determining the degradation degree of the material under the real-time working condition based on the real-time stress spectrum and a shipbuilding portal crane material aging prediction model; the first value is optimized based on the real-time stress spectrum and the degradation degree, and the optimal value of the crack growth rate parameter under the real-time working condition is obtained; and calculating the residual life of the shipbuilding portal crane under the real-time working condition based on the optimal value.

Description

Technical Field

[0001] The present application relates to the field of crane safety technology, and in particular to a method and device for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane. Background Art

[0002] Gantry cranes are important equipment in the shipbuilding industry, and their structural safety is directly related to operating efficiency and personnel safety. Stress spectrum analysis can fully capture the complex loads and dynamic responses of cranes under actual working conditions, identify potential fatigue hazards in the structure, and provide a basis for design and maintenance. For shipbuilding gantry cranes, which have more complex structures among gantry cranes, the working environment of shipbuilding gantry cranes is complex, and they often face severe weather, load fluctuations, etc., and they will be subjected to complex random loads during long-term operation. Through stress spectrum analysis and remaining life assessment, more reliable modeling and prediction of variable working conditions can be carried out, thereby improving the adaptability and durability of shipbuilding gantry cranes in complex environments.

[0003] However, the commonly used analysis methods currently have significant limitations in practical applications. The traditional load combination method is based on a simplified ideal working condition assumption, which makes it difficult to cope with the frequent random load fluctuations and coupling effects in the working environment of shipbuilding gantry cranes, resulting in prediction results that deviate from the actual situation. Although models based on BP neural networks or support vector machines have improved the prediction accuracy to a certain extent, these methods are highly dependent on training data and are prone to insufficient generalization capabilities. In addition, the single kernel function model has limited performance in dealing with multidimensional nonlinear features and cannot fully explore the multi-scale characteristics of stress spectra under complex working conditions; the traditional grid search optimization of kernel parameters is inefficient, which also limits the improvement of model performance.

[0004] Therefore, there is an urgent need for a method to solve the shortcomings of traditional methods in terms of insufficient prediction accuracy under complex working conditions, with stronger generalization ability, to provide reliable guarantees for the safety assessment and life prediction of shipbuilding gantry cranes. Summary of the invention

[0005] In view of this, the present application provides a method and device for stress spectrum analysis and remaining life assessment of shipbuilding gantry cranes to address the shortcomings of traditional methods in insufficient prediction accuracy under complex working conditions. It has stronger generalization ability and provides reliable guarantee for safety assessment and life prediction of shipbuilding gantry cranes.

[0006] Specifically, the present application is implemented through the following technical solutions:

[0007] The first aspect of the present application provides a method for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane, the method comprising:

[0008] Obtain the dynamic load parameters of shipbuilding gantry cranes under various working conditions;

[0009] Based on the lifting state level corresponding to the dynamic load parameter, a lifting dynamic load coefficient is determined; the lifting dynamic load coefficient represents the amplification or correction of the dynamic load of the shipbuilding gantry crane to the static load under each working condition;

[0010] According to the lifting dynamic load coefficient, a stress spectrum prediction model based on an improved correlation vector machine is updated, wherein the stress spectrum prediction model is used to predict the stress conditions at various positions of the shipbuilding gantry crane under preset working conditions to obtain a stress spectrum of the shipbuilding gantry crane; the stress spectrum includes the stress amplitude and the corresponding number of cycles of the shipbuilding gantry crane during operation, and the stress amplitude is linearly related to the lifting dynamic load coefficient;

[0011] Determine a first value of a crack growth rate parameter by experimental data fitting and finite element analysis, wherein the first value is applicable to each working condition;

[0012] Predicting a real-time stress spectrum under real-time working conditions based on the stress spectrum prediction model;

[0013] Determining the degree of material degradation under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model;

[0014] Optimizing the first value based on the real-time stress spectrum and the degradation degree to obtain an optimal value of the crack growth rate parameter under the real-time working condition;

[0015] The remaining life of the shipbuilding gantry crane under the real-time working condition is calculated based on the optimal value.

[0016] The second aspect of the present application provides a device for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane, the device comprising an acquisition module, a determination module, a prediction module, an optimization module and a calculation module; wherein:

[0017] The acquisition module is used to obtain the dynamic load parameters of the shipbuilding gantry crane under various working conditions;

[0018] The determination module is used to determine the lifting dynamic load coefficient based on the lifting state level corresponding to the dynamic load parameter; the lifting dynamic load coefficient represents the amplification or correction of the dynamic load of the shipbuilding gantry crane to the static load under each working condition;

[0019] The prediction module is used to update the stress spectrum prediction model based on the improved correlation vector machine according to the lifting dynamic load coefficient, and the stress spectrum prediction model is used to predict the stress conditions of each position of the shipbuilding gantry crane under preset working conditions to obtain the stress spectrum of the shipbuilding gantry crane; the stress spectrum includes the stress amplitude and the corresponding number of cycles of the shipbuilding gantry crane during operation, and the stress amplitude is linearly related to the lifting dynamic load coefficient;

[0020] The determination module is further used to determine a first value of a crack growth rate parameter by experimental data fitting and finite element analysis, wherein the first value is applicable to various working conditions;

[0021] The prediction module is further used to predict the real-time stress spectrum under real-time working conditions based on the stress spectrum prediction model;

[0022] The determination module is further used to determine the degradation degree of the material under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model;

[0023] The optimization module is used to optimize the first value based on the real-time stress spectrum and the degradation degree to obtain the optimal value of the crack growth rate parameter under the real-time working condition;

[0024] The calculation module is used to calculate the remaining life of the shipbuilding gantry crane under the real-time working condition based on the optimal value.

[0025] The present application provides a method and device for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane. In the first aspect, the dynamic load parameters directly affect the dynamic response of the crane. The dynamic load coefficient of the lifting is dynamically determined according to the dynamic load parameters, which can more accurately reflect the amplification effect of the dynamic load on the static load of the shipbuilding gantry crane under actual working conditions. Since the traditional method usually adopts a fixed lifting dynamic load coefficient, this method is prone to overly conservative design or ignores the influence of special working conditions, and can avoid unnecessary redundancy or potential risks caused by a single parameter. In addition, the lifting dynamic load coefficient is calculated by the actual dynamic load parameters and can be flexibly adjusted under different working conditions. For example, when the lifting speed is fast, the lifting dynamic load coefficient will increase, reflecting a greater dynamic load effect; while under low-speed and light-load conditions, the lifting dynamic load coefficient will be relatively reduced, thereby more realistically describing the load characteristics of different working conditions. In a complex shipbuilding environment, the impact of dynamic loads on static loads may fluctuate over time. By dynamically adjusting the dynamic load coefficient through dynamic load parameters, real-time changes can be better captured, and the timeliness and accuracy of load analysis can be improved. Moreover, since the lifting dynamic load coefficient is linearly related to the stress amplitude in the stress spectrum, the dynamic calculation of the lifting dynamic load coefficient based on the dynamic load parameters can provide accurate input for the stress spectrum prediction model, improve the reliability of the stress spectrum prediction, and ensure the accuracy of the stress spectrum prediction. Secondly, by optimizing the crack growth rate parameters, not only the accuracy of the assessment of the remaining life of the shipbuilding gantry crane is significantly improved, but also the adaptability of the assessment model to complex working conditions is enhanced. Since the crack growth rate parameters are the key constants for the remaining life assessment, in practical applications, these parameters are not only affected by material properties, but also by the combined effects of dynamic loads, stress distribution and environmental conditions. By optimizing the crack growth rate parameters, the remaining life assessment is first made closer to the actual operating conditions of the crane. Moreover, by combining experimental data and finite element analysis, the crack growth rate parameters can effectively reflect the influence of dynamic loads on crack growth behavior in actual work. The experimental test provides real data of crack growth under different working conditions, and the finite element simulation reveals the distribution of stress concentration areas and the evolution of crack paths. The combination of the two can more accurately capture the law of crack growth rate changes with time and working conditions. This method eliminates the evaluation errors caused by overly idealized assumptions in traditional models, thereby improving the scientific nature of life assessment. In addition, the optimization process also fully considers the dynamic changes of stress amplitude and number of cycles, especially the changes in stress intensity factors under complex working conditions. By fitting the crack growth rate model to the actual fatigue crack growth data and introducing the stress distribution and crack growth path obtained by finite element simulation, the optimized crack growth rate parameters can accurately characterize the fatigue behavior of crane materials under complex load conditions. This method not only improves the prediction accuracy of the crack growth rate model, but also enhances the adaptability of the model to a variety of load conditions, providing possibilities for a wider range of engineering applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A flow chart of a method for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane provided in Example 1 of the present application;

[0027] Figure 2 This is a schematic structural diagram of the device for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane provided in Example 2 of the present application. DETAILED DESCRIPTION

[0028] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0029] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

[0030] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0031] Specific embodiments are given below to introduce the technical solution of the present application in detail.

[0032] Figure 1 This is a flow chart of the method for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane provided in Example 1 of the present application. Figure 1 , the method provided in this embodiment may include:

[0033] S101. Obtain dynamic load parameters of a shipbuilding gantry crane under various working conditions.

[0034] Specifically, the dynamic load parameters include lifting speed, load size, etc. The lifting speed represents the operating speed of the shipbuilding gantry crane when lifting or lowering the load, and the load size represents the weight of the lifted object. The lifting speed and load size jointly determine the lifting dynamic load coefficient.

[0035] In specific implementation, a load sensor is installed on the spreader of the shipbuilding gantry crane, and an encoder is installed on the drum of the shipbuilding gantry crane. Under various working conditions of the shipbuilding gantry crane, the encoder calculates the linear speed of the hook by recording the rotation speed of the drum to obtain the lifting speed; the load sensor senses the force change caused by the load, outputs the corresponding electrical signal, and obtains the load size. By reading the data recorded by the encoder and the load sensor, the dynamic load parameters (lifting speed, load size) of the shipbuilding gantry crane under various working conditions are obtained.

[0036] S102. Determine a lifting dynamic load coefficient based on the lifting state level corresponding to the dynamic load parameter; the lifting dynamic load coefficient represents the amplification or correction of the dynamic load of the shipbuilding gantry crane to the static load under each working condition.

[0037] Specifically, the lifting dynamic load coefficient represents the amplification effect of the dynamic load relative to the static load, and reflects the load changes caused by dynamic factors such as acceleration, vibration, and impact under various working conditions of the shipbuilding gantry crane.

[0038] In a specific implementation, the determining of the lifting dynamic load coefficient based on the lifting state level corresponding to the dynamic load parameter includes:

[0039] (1) Based on the lifting speed and load size in the dynamic load parameters, the corresponding lifting state level is determined; the lifting speed is positively correlated with the lifting state level, and the load size is positively correlated with the lifting state level.

[0040] Specifically, the lifting state level is a classification of the lifting state of the shipbuilding gantry crane under different working conditions. The division of the lifting state level is usually related to the lifting speed and load size. Each lifting state level corresponds to a lifting speed and load size within a range. Among them, the lifting speed is positively correlated with the lifting state level, and the load size is positively correlated with the lifting state level, that is, the higher the lifting state level, the greater the corresponding lifting speed and the corresponding load size. For example, the lifting status levels include no load, light load, medium load, full load and overload. The lifting speed range corresponding to no load is 0-20% of the rated speed, and the load size range is 0-20% of the rated load. The lifting speed range corresponding to light load is 20%-40% of the rated speed, and the load size range is 20%-50% of the rated load. The lifting speed range corresponding to medium load is 40%-60% of the rated speed, and the load size range is 50%-80% of the rated load. The lifting speed range corresponding to full load is 60%-80% of the rated speed, and the load size range is 80%-100% of the rated load. The lifting speed range corresponding to overload exceeds 80% of the rated speed, and the load size range exceeds 100% of the rated load.

[0041] In specific implementation, the above corresponding relationship is fitted to determine the weight coefficient corresponding to the lifting speed and the weight coefficient corresponding to the load size, and a mathematical model of the lifting state level is constructed:

[0042] S=αV+βW, wherein S is the lifting state level; V is the lifting speed; W is the load size; α is the weight coefficient corresponding to the lifting speed; and β is the weight coefficient corresponding to the load size.

[0043] Based on the lifting speed and load size in the acquired dynamic load parameters, the range interval of the lifting speed and the range interval of the load size are determined, and substituted into the mathematical model of the above lifting state level to calculate the corresponding lifting state level.

[0044] (2) Based on the lifting state level, determine the corresponding first coefficient and second coefficient; wherein each lifting state level corresponds to a first coefficient and a second coefficient, and the higher the lifting state level, the larger the values ​​of the first coefficient and the second coefficient.

[0045] Specifically, the first coefficient and the second coefficient are set according to the lifting state level, the first coefficient represents the minimum value of the lifting dynamic load coefficient, each lifting state level corresponds to a first coefficient and a second coefficient, and the values ​​of the first coefficient and the second coefficient are positively correlated with the lifting state level. Table 1 is a corresponding table of the lifting state level and the first coefficient and the second coefficient shown in this application:

[0046] Table 1 Correspondence between lifting status level and first coefficient and second coefficient

[0047] Lifting status level The first coefficient The second coefficient HC1 1.05 0.17 HC2 1.10 0.34 HC3 1.15 0.51 HC4 1.20 0.68

[0048] Please refer to Table 1, it can be seen that each lifting state level uniquely corresponds to a first coefficient and a second coefficient. For example, the first coefficient corresponding to the lifting state level HC1 is 1.05, and the corresponding second coefficient is 0.17. The higher the lifting state level, the larger the corresponding first coefficient and second coefficient. For example, the first coefficient corresponding to the lifting state level HC2 is 1.10, and the corresponding second coefficient is 0.34, which is greater than the first coefficient and the second coefficient corresponding to the lifting state level HC1.

[0049] In specific implementation, in this step, based on the determined lifting state level, the first coefficient and the second coefficient corresponding to the lifting state level are obtained by searching a correspondence table between the lifting state level and the first coefficient and the second coefficient.

[0050] (3) Determine a third coefficient based on the product of the rated lifting speed and the second coefficient.

[0051] Specifically, the rated lifting speed represents the standard speed at which the shipbuilding gantry crane can safely and stably lift the maximum rated load under normal working conditions. The rated lifting speed is related to the driving control type and lifting operation method of the lifting mechanism of the shipbuilding gantry crane. The maximum value of the rated lifting speed occurs when the motor or generator is started without load (equivalent to the lifting device, objects and completely relaxed wire ropes are placed on the ground at this time), and the lifting speed has reached the maximum value of the stable lifting speed when the lifting device and objects are lifted off the ground.

[0052] In specific implementation, the rated lifting speed is determined according to the driving control type of the lifting mechanism of the shipbuilding gantry crane and the lifting operation method, and the product of the rated lifting speed and the second coefficient is determined as the third coefficient.

[0053] (4) The sum of the first coefficient and the third coefficient is determined as the lifting dynamic load coefficient.

[0054] Specifically, the sum of the first coefficient and the third coefficient is determined as the lifting dynamic load coefficient. It should be noted that different types of cranes have different corresponding lifting dynamic load coefficients, and the maximum value of the lifting dynamic load coefficient cannot exceed 2.2 for cranes with high lifting speeds such as building tower cranes and port cranes, and cannot exceed 2.0 for other types of cranes.

[0055] In specific implementation, the lifting dynamic load coefficient can be determined based on the following formula:

[0056]

[0057] Among them, the is the lifting dynamic load coefficient; is the first coefficient; 2 is the second coefficient; the v q is the rated lifting speed; 2 v q The third speed.

[0058] For example, combined with the above description, when the lifting state level of the shipbuilding gantry crane is HC4 and the rated lifting speed is 5.1m / min, the lifting dynamic load coefficient is determined as It is 1.20+0.68×5.1 / 60=1.2578.

[0059] S103, updating a stress spectrum prediction model based on an improved correlation vector machine according to the lifting dynamic load coefficient.

[0060] The stress spectrum prediction model is used to predict the stress conditions at various positions of the shipbuilding gantry crane under preset working conditions to obtain the stress spectrum of the shipbuilding gantry crane; the stress spectrum includes the stress amplitude and the corresponding number of cycles of the shipbuilding gantry crane during operation, and the stress amplitude is linearly related to the lifting dynamic load coefficient.

[0061] Specifically, the stress spectrum represents the stress amplitude distribution and the corresponding cycle number distribution under different lifting states of the shipbuilding gantry crane during operation. The stress amplitude represents the maximum stress intensity that the structure can withstand, and the cycle number represents the frequency of occurrence of the stress amplitude.

[0062] Furthermore, the stress amplitude is linearly related to the lifting dynamic load coefficient. When the lifting dynamic load coefficient increases, the stress amplitude borne by the structure also increases. That is, the change of the lifting dynamic load coefficient directly affects the size of the stress amplitude, and then affects the generation of the stress spectrum.

[0063] In the specific implementation, after obtaining the lifting dynamic load coefficient, the lifting dynamic load coefficient is used to train the stress spectrum prediction model of the improved correlation vector machine. The stress spectrum prediction model of the improved correlation vector machine is a machine learning method based on the Bayesian framework, which can learn and predict the output (such as stress spectrum) according to the input training data (lifting dynamic load coefficient). In the training stage, the known load data (such as the lifting dynamic load coefficient) and the corresponding stress data (such as stress amplitude and number of cycles) are input into the stress spectrum prediction model for training. In the prediction stage, the trained stress spectrum prediction model of the improved correlation vector machine is used to input the lifting dynamic load coefficient under the working condition to be predicted, and the corresponding stress amplitude and number of cycles are predicted. By calculating the stress amplitude under different lifting dynamic load coefficients and combining the corresponding number of cycles, a complete stress spectrum can be obtained.

[0064] For example, in combination with the above description, in the obtained stress spectrum, when the lifting dynamic load coefficient is 1.2, the corresponding stress amplitude is 100MPa, and the number of cycles is 2000. When the lifting dynamic load coefficient is 1.5, the corresponding stress amplitude is 120MPa, and the number of cycles is 1500.

[0065] S104. Determine a first value of a crack growth rate parameter through experimental data fitting and finite element analysis, wherein the first value is applicable to various working conditions.

[0066] Specifically, the crack growth rate characterizes the rate at which a crack grows under a unit load cycle, which is usually related to the stress intensity factor of the material and the fatigue characteristics of the material. By calculating the crack growth rate, accumulating the crack growth of each load cycle, and gradually calculating the crack growth, the remaining life of the shipbuilding gantry crane can be predicted. The crack growth rate parameters are material parameters, which represent the proportionality coefficient of the crack growth rate and the exponent of the crack growth, respectively. The crack growth rate parameters are usually obtained through experimental data.

[0067] Optionally, determining the first value of the crack growth rate parameter by experimental data fitting and finite element analysis includes:

[0068] (1) The fatigue crack growth data of the shipbuilding gantry crane under different working conditions are obtained through experimental testing.

[0069] In a specific implementation, the method of obtaining the fatigue crack growth data of the shipbuilding gantry crane under different working conditions through experimental testing includes: analyzing the actual working conditions of the shipbuilding gantry crane to determine the abnormal working conditions of the shipbuilding gantry crane; determining the experimental parameters based on the stress distribution, load characteristics and fatigue damage mechanism of the shipbuilding gantry crane under the abnormal working conditions; simulating the fatigue experiment under the abnormal working conditions based on the experimental parameters to collect the fatigue crack growth data of the shipbuilding gantry crane.

[0070] Specifically, the abnormal working condition is set according to actual needs, and is not limited in this embodiment. For example, the abnormal working condition may be high-frequency lifting, large tonnage load, long-term continuous operation, etc. The fatigue crack growth data includes crack growth rate, crack growth path, fatigue failure mode, etc.

[0071] In the specific implementation, the working conditions of the shipbuilding gantry crane are analyzed, including the working environment (analyzing the working environment and operation scenarios of the shipbuilding gantry crane, such as the specific working conditions of the shipbuilding gantry crane when working in ports, shipyards, docks and other environments, taking into account environmental factors such as temperature, humidity, wind speed, etc. that may affect the fatigue properties of the material), load characteristics (analyzing the loads borne by the shipbuilding gantry crane during work, including maximum load, minimum load, load fluctuation, etc. For example, the load fluctuation characteristics of the shipbuilding gantry crane in different actions such as lifting, lowering, and rotation), working mode (considering the loads and stresses generated by the shipbuilding gantry crane on the structure in different working modes), stress distribution (through finite element analysis or simulation calculation, obtain the stress distribution on the shipbuilding gantry crane structure under different working conditions, especially under high load and dynamic load, etc.

[0072] Furthermore, according to the analyzed working conditions, the most representative abnormal working conditions are selected. The abnormal working conditions should generally cover the common workloads and operation modes that the shipbuilding gantry crane may experience. For example, working conditions that can represent the actual workload, such as full load, overload, long-term operation, high-frequency load fluctuations, etc. are selected. Working conditions that may generate large dynamic loads, such as dynamic actions such as lifting, lowering, traversing, and rotation, are selected. While selecting abnormal working conditions, some extreme working conditions, such as long-term overload work or operation under extreme weather conditions, need to be considered. After determining the specific working conditions, the experimental parameters are set based on the stress distribution, load characteristics, and fatigue damage mechanism of the shipbuilding gantry crane under abnormal working conditions, including stress amplitude and frequency, load characteristics, experimental conditions, and loading methods. After determining the experimental parameters, experimental samples similar to the key parts of the shipbuilding gantry crane are prepared according to the load and stress characteristics of the selected abnormal working conditions. These samples may include steel, welded joints, etc., and the fatigue characteristics of the actual structure are reflected through test data. The specimens are placed in a fatigue testing machine and loaded according to the experimental parameters (such as stress amplitude, loading frequency, etc.) determined in advance. Simulate the load conditions under specific working conditions and record the stress, load and other information of each loading cycle. Use non-destructive testing technology (such as X-ray, ultrasonic testing, surface microscope observation, etc.) to monitor the crack growth process. After each loading cycle, check the length, shape and growth rate of the crack to record the fatigue crack growth data.

[0073] (2) The shipbuilding gantry crane is simulated using finite element analysis software to obtain predicted fatigue crack growth data under different working conditions.

[0074] (3) Determining the stress distribution and crack propagation path of the stress concentration area based on the fatigue crack growth data and the predicted fatigue crack growth data.

[0075] In the specific implementation, FEA finite element analysis software (such as ANSYS, ABAQUS) is used to simulate the stress distribution of shipbuilding gantry cranes under typical working conditions. The stress concentration area is determined, and the initial crack is applied to the location where the crack may occur. The fracture mechanics method is used to analyze the stress intensity factor at the crack tip and determine the crack propagation path. The change in the range of the stress intensity factor during the crack propagation process is simulated to obtain the value of the crack propagation rate. The simulation results are generated into a data table, including the crack length, number of cycles, propagation path, etc.

[0076] (4) Fitting the crack growth rate parameter based on the stress distribution and the crack growth path to obtain a first value.

[0077] In a specific implementation, the crack growth rate parameter is fitted in combination with the stress distribution and the crack growth path to obtain a first value, including: determining the stress intensity factor under different working conditions based on the stress distribution and the crack growth path; matching the fatigue crack growth data under each working condition with the stress intensity factor for different working conditions to establish a crack growth rate model; the crack growth rate model characterizes the relationship between the crack growth rate and the stress intensity factor; fitting the constant term in the crack growth rate model under each working condition to obtain a crack growth rate parameter uniformly used under each working condition.

[0078] Specifically, based on fracture mechanics theory, the stress intensity factor is calculated in the crack tip area. Under different working conditions, the corresponding stress intensity factor range is calculated. The fatigue crack growth data under each working condition is matched with the corresponding stress intensity factor to generate data pairs, and a crack growth rate model is established according to the Paris formula. The crack growth rate model is linearly or nonlinearly fitted using the least squares method. If the crack growth rate model is complex or there are many experimental data, a genetic algorithm or a particle swarm optimization algorithm can be used for global optimization to obtain the optimized crack growth rate parameters.

[0079] Optionally, the crack growth rate parameter is fitted in combination with the stress distribution and the crack growth path to obtain a first value, including: calculating the stress intensity factor under different positions and working conditions based on the stress distribution and the crack growth path; obtaining the crack size and the corresponding number of cycles at different stages based on the fatigue crack growth data; grouping the stress intensity factor, the crack size, and the number of cycles according to different working conditions; for each group of data, performing a preliminary fitting of the crack growth rate parameter using the least squares method; adjusting the first crack growth rate parameter in the crack growth rate parameter based on the preliminary fitting result and the fitting error, refitting the second crack growth rate parameter, and repeatedly iterating and calculating the fitting error until the number of iterations reaches a preset number, and determining the first crack growth rate parameter and the second crack growth rate parameter with the smallest fitting error as the first value.

[0080] In specific implementation, according to the Paris formula in fracture mechanics theory Where ΔK is the range of stress intensity factors, which is closely related to stress distribution and crack propagation path. In this step, the stress intensity factor K under different positions and working conditions is first calculated using finite element analysis software in combination with the obtained stress distribution data and crack propagation path, and then ΔK is obtained. At the same time, by reading the fatigue crack propagation data, the crack size a and the corresponding number of cycles N at different stages are obtained. Furthermore, the stress intensity factor range, crack size and corresponding number of cycles are grouped according to different working conditions (such as lifting load size, lifting speed range, running track conditions, etc.). For example, the working condition data with a lifting load between 500-600t, a lifting speed of 0.3-0.5m / s and a rail joint height difference in the range of 0.5-1mm are grouped together. For each set of data, the crack propagation rate parameters in the Paris formula are preliminarily fitted using fitting methods such as the least squares method. First, the first crack growth rate parameter is fixed to an empirical value (such as 3), and then the second crack growth rate parameter is fitted to obtain the preliminary fitting parameters and corresponding fitting errors of different groups. According to the preliminary fitting results and the fitting error, the value of the first crack growth rate parameter is adjusted, and the second growth rate parameter is refitted. This process is repeated until the number of iterations reaches a large value, and the combination of the first growth rate parameter and the second growth rate parameter that minimizes the overall fitting error of all grouped data is determined. For example, when the first growth rate parameter is adjusted from 3 to 3.2, it is found that the overall fitting error is significantly reduced. At this time, continue to fine-tune in this vicinity and optimize at the same time. After multiple iterative optimizations, the relatively optimal sum value suitable for each working condition is obtained, which is used as the first value of the crack growth rate parameter.

[0081] Optionally, the method of determining the first value of the crack growth rate parameter through experimental data fitting and finite element analysis includes: obtaining fatigue crack growth data of a shipbuilding gantry crane under different working conditions; fitting the fatigue crack growth data to construct a crack growth rate model; performing finite element simulation on the shipbuilding gantry crane to determine the stress distribution and crack growth path in a stress concentration area; extracting a stress intensity factor from the stress distribution and the crack growth path, matching the crack growth rate model based on the stress intensity factor, optimizing the crack growth rate model according to the distribution of the stress intensity factor and the crack growth path, and determining the first value of the crack growth rate parameter.

[0082] In the specific implementation, the fatigue crack growth data of the shipbuilding gantry crane under different working conditions are obtained through actual experiments, including the starting position of the crack, crack growth rate, load level, stress distribution, etc. The fatigue crack growth data are statistically analyzed to determine the mathematical relationship between the crack growth rate and the stress intensity factor (or other influencing factors), and a crack growth rate model is constructed. Usually, the relationship between the crack growth rate and the stress intensity factor is described by the Paris law or a similar model. By fitting the fatigue crack growth data, a crack growth rate model can be obtained, which contains constant terms and coefficients to characterize the basic laws of crack growth. After fitting, the constant term of the crack growth rate under each working condition can be obtained, and the quantitative relationship between the crack growth rate and the stress intensity factor can be further established.

[0083] Furthermore, a three-dimensional finite element model of the shipbuilding gantry crane is established, including the geometry, material properties, loads and boundary conditions of the shipbuilding gantry crane. Dynamic loading and stress distribution analysis is performed using finite element analysis software (such as ANSYS, ABAQUS, etc.). For each working condition, the finite element analysis software calculates the stress field and crack propagation area of ​​the structure according to the actual operating conditions (such as load, lifting speed, etc.). In the results of finite element analysis, special attention will be paid to the stress distribution in the crack initiation and extension areas. Specifically, the stress distribution of each position under the current working condition over time is calculated based on the finite element model, and the change path of the crack under each working condition is determined according to the stress change value under each working condition. For each working condition, the starting position of the crack is taken as the starting point, the horizontal direction and the vertical direction are the X and Y directions, a coordinate system is established, and each stress change value in the coordinate system is obtained to obtain the change path. If there are multiple vertical position points with stress changes at the same horizontal position point, the target position point is selected from the multiple vertical position points according to the vertical position point of the subsequent horizontal position point, and the stress change of the target position point is used as the stress change of the horizontal position point; the crack initiation area and the extension area are located according to the change path. The stress intensity factor at the crack tip is analyzed in detail to obtain the stress distribution and crack extension path of the stress concentration area under different working conditions. After obtaining the stress distribution and crack extension path, the stress intensity factor is calculated using the stress intensity factor formula, and the stress intensity factor under different working conditions is associated with the crack extension path. The stress intensity factor is matched with the crack growth rate model after fitting the experimental data, and the coefficients in the crack growth rate model are adjusted based on the deviation between the experimental data and the simulation data. According to the stress intensity factor obtained by finite element analysis and the crack growth rate model after fitting the experimental data, combined with the specific working conditions and crack growth path, the initial value (first value) of the crack growth rate parameter is accurately determined. If the crack growth rate model obtained by fitting the experimental data is not completely consistent with the finite element simulation results, the crack growth rate model is optimized by adjusting the constants or coefficients in the crack growth rate model, or by adding more working condition data to optimize the crack growth rate model, and the first value of the crack growth rate parameter is determined based on the optimized crack growth rate model.

[0084] The method provided in this embodiment, in the first aspect, can make a detailed analysis of the stress state of the structure under different working conditions by calculating the stress distribution of each position changing with time based on finite element simulation, which provides highly accurate data support for the prediction of crack extension, significantly improves the accuracy of crack extension prediction, and provides a more reliable basis for structural safety assessment. Finite element simulation can reflect the actual stress distribution of the structure under various working conditions, thereby providing an accurate physical basis for the prediction of the crack extension path. Secondly, by taking the starting position of the crack as the starting point under each working condition, establishing a coordinate system, and obtaining the stress change value of each position point, the specific path and process of crack extension can be accurately described. This method can capture the slight differences in stress changes in the structure and provide a comprehensive basis for the simulation of the crack extension path. In particular, in the case of multiple vertical position points, the direction of the crack extension path is determined by screening the target position point, which can avoid inaccurate predictions caused by the locality of stress changes and ensure the accuracy and rationality of the crack extension path. This method of locating the crack initiation area and the extension area based on the changing path helps to gain a deeper understanding of the process from the initiation to the extension of the crack. Through this high-precision analysis and modeling, the trend of crack propagation can be better predicted and the safety risks caused by crack propagation can be reduced. Secondly, by matching the crack propagation rate model fitted with experimental data based on finite element analysis and combining the stress intensity factors under different working conditions, the deviation between the crack propagation rate model and the actual data can be identified, and the correlation coefficient can be adjusted to optimize the crack propagation rate model, continuously improve the accuracy of the crack propagation rate model, and make the values ​​of the crack propagation rate parameters more accurate. This systematic optimization process makes the crack propagation rate prediction not only rely on a single working condition or a single method, but also combine and adjust in many aspects, thereby improving the wide applicability and accuracy of the crack propagation rate model. Thirdly, through actual experiments, crack propagation data under different working conditions, including crack starting position, propagation rate, load level and stress distribution, provide real basic data for the crack propagation rate model. This modeling method based on experimental data makes the crack propagation rate model closer to the actual situation and avoids errors caused by assumptions and simplifications in theoretical derivation. Through statistical analysis of experimental data, the mathematical relationship between crack growth rate and stress intensity factor can be accurately determined, and a crack growth rate model can be established, which provides a scientific basis and data support for subsequent crack growth prediction.

[0085] As an optional embodiment, the determining the first value of the crack growth rate parameter by experimental data fitting and finite element analysis further includes:

[0086] The crack growth rate of each working condition under the first time length is calculated by fitting the experimental data; the predicted crack growth rate of each working condition under the first time length is predicted according to the finite element model; and the finite element model is corrected according to the difference between the crack growth rate and the predicted crack growth rate.

[0087] S105 . Predicting a real-time stress spectrum under a real-time working condition based on the stress spectrum prediction model.

[0088] Specifically, combined with the above description, the trained stress spectrum prediction model of the improved correlation vector machine can predict the input lifting dynamic load coefficient to generate a stress spectrum. By inputting the lifting dynamic load coefficient under real-time working conditions into the trained stress spectrum prediction model, the corresponding stress amplitude and number of cycles are predicted. By calculating the stress amplitude under the lifting dynamic load coefficient and combining it with the corresponding number of cycles, a complete stress spectrum can be obtained.

[0089] Optionally, the stress spectrum prediction model of the improved correlation vector machine includes a particle swarm optimization module, and the particle swarm optimization module optimizes the kernel parameters of the correlation vector machine based on a particle swarm optimization algorithm.

[0090] Specifically, the stress spectrum prediction model of the improved correlation vector machine is improved on the basis of the correlation vector machine, and a particle swarm optimization module is introduced. The particle swarm optimization module optimizes the kernel function of the correlation vector machine based on the particle swarm optimization algorithm. Among them, the particle swarm optimization algorithm is a kind of swarm intelligence algorithm. The particle swarm optimization algorithm finds the optimal solution through the collaboration and information sharing between individuals in the group. By randomly generating a group of particles, the particle swarm iteratively searches in the n-dimensional space. The position Xi of each particle represents a solution to the problem, and the quality of the solution is evaluated by the fitness function. The particles search for solutions by continuously updating their positions. The position of the i-th particle is represented by X i =(x 1 , x 2 , …, x n ) and the speed is expressed by V i =(v 1 , v 2 , …, v n ). The speed of each particle is affected by its own optimal position and the optimal position of the group. The position and speed of each particle are updated by the following formula:

[0091]

[0092] X k+1 =X k +V k+1

[0093] Among them, the V k+1 is the velocity of the particle in the next iteration;k The speed of the current iteration; the X k+1 is the position of the particle in the next iteration; k is the position of the particle in the current iteration; ω is the inertia weight; c 1 、c 2 is the learning factor; 1 、r 2 is a random number; Individual optimal particle position; is the global optimal particle position; k is the current iteration number.

[0094] Optionally, the stress spectrum prediction model of the improved correlation vector machine also includes a combined kernel function, and the combined kernel function is obtained by linearly combining a polynomial kernel function and a radial basis kernel function.

[0095] Specifically, the stress spectrum prediction model of the improved correlation vector machine is further improved on the basis of the correlation vector machine, and a combined kernel function is introduced. The combined kernel function combines multiple types of kernel functions (polynomial kernel function and radial basis kernel function) to better adapt to the nonlinear characteristics of complex data, thereby improving the fitting ability of the stress spectrum prediction model.

[0096] Furthermore, the polynomial kernel function is good at capturing global nonlinear relationships, and is particularly suitable for global, large-scale changes in the data. For example, there may be high-order relationships between data, and the polynomial kernel function can better express this relationship through high-order terms. The radial basis kernel function is good at capturing local nonlinear relationships, and is particularly suitable for situations where there are complex local structures or "local changes" in the data. It can sensitively capture local feature changes in the data by calculating the distance between samples and exponentially weighting the distance. By combining the polynomial kernel function and the radial basis kernel function, the stress spectrum prediction model can focus on both the local and global features of the data, and enhance the ability to express nonlinear patterns.

[0097] In the specific implementation, the polynomial kernel function and the radial basis kernel function are linearly combined based on different weights to obtain the combined kernel function:

[0098] K mix (x,x′)=λK Rbf (x,x′)+(1-λ)K Polly (x,x′)

[0099] =λexp(-γ·||xx′|| 2 )+(1-λ)(x·x′+1) d (0<λ<1)

[0100] Among them, the Kmix (x, x′) is a combined kernel function; Rbf (x, x′) is the radial basis kernel function; Poly (x, x′) is a polynomial kernel function; λ is a weight parameter; γ is a width parameter; d is a feature adjustment parameter; x′ is the center of the kernel function; and x is each sampling point of the kernel function.

[0101] It should be noted that the kernel parameters γ and d are related to the number and degree of correlation of the related vectors, which affect the sparsity and generalization of the stress spectrum prediction model. The weight parameter λ determines the proportion of each kernel function in the stress spectrum prediction model. Therefore, in order to improve the prediction accuracy of the stress spectrum prediction model, combined with the above description, it is necessary to optimize the kernel parameters and weight parameters in combination with the particle swarm optimization algorithm.

[0102] Optionally, the combined kernel function is obtained by linearly combining a polynomial kernel function and a radial basis kernel function, including: respectively calculating a first optimal value of kernel parameters under the polynomial kernel function and a second optimal value of kernel parameters under the radial basis kernel function based on the particle swarm optimization module; comprehensively considering the first optimal value and the second optimal value, and fitting to obtain a third optimal value of the kernel parameter when both the polynomial kernel function and the radial basis kernel function are included; determining a weight parameter based on the third optimal value; and linearly combining the polynomial kernel function and the radial basis kernel function based on the weight parameter to obtain a combined kernel function.

[0103] In specific implementation, combined with the above description, for each kernel function in the combined kernel function, the weight parameter λ is set to 0 or 1, and the optimal values ​​of the kernel parameters γ and d under the single kernel function are calculated. Among them, when the weight parameter λ is 0, the optimal values ​​of the kernel parameters γ and d in the radial basis kernel function are calculated, and when the weight parameter λ is 1, the optimal values ​​of the kernel parameters γ and d in the polynomial kernel function are calculated. For the polynomial kernel function or the radial basis kernel function, the optimal values ​​of the kernel parameters are calculated by the particle swarm optimization module to obtain the first optimal value and the second optimal value. After obtaining the optimal kernel parameters (the first optimal value and the second optimal value) of the individual kernel function, the two optimal kernel parameters are comprehensively considered. The first optimal value and the second optimal value are fitted by weighted average or other optimization methods to obtain a new kernel parameter that comprehensively considers the characteristics of both, that is, the third optimal value. Based on the third optimal value, the weight parameter can be further determined. The determination of the weight parameter needs to consider the relative influence of the polynomial kernel function and the radial basis kernel function in the combined kernel function, and a suitable weight parameter is obtained through the optimization process so that the advantages of the two kernel functions can complement each other without causing overfitting or underfitting. Finally, based on the determined weight parameters, the polynomial kernel function and the radial basis kernel function are combined in a linear combination manner to obtain a new combined kernel function.

[0104] S106. Determine the degree of material degradation under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model.

[0105] In specific implementation, the method of determining the degree of material degradation under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model includes: determining the material performance parameters of the shipbuilding gantry crane; establishing a fatigue cumulative damage model based on the fatigue cumulative damage principle; associating the fatigue cumulative damage model with the material performance parameters according to time to obtain a shipbuilding gantry crane material aging prediction model in which fatigue cumulative damage changes with time; inputting the real-time stress spectrum into the shipbuilding gantry crane material aging prediction model to obtain the current fatigue cumulative damage; and determining the optimized crack growth rate parameters based on the current fatigue cumulative damage and the changing trend of the crack growth rate over time.

[0106] Specifically, in the actual operation of the shipbuilding gantry crane, strain gauges or wireless stress sensors are used to monitor the stress changes of key structural parts, record the distribution of dynamic stress over time, and obtain stress monitoring data. The stress monitoring data is denoised and normalized, and the fatigue performance of the key component materials of the shipbuilding gantry crane is tested under laboratory conditions to obtain material performance parameters, including fatigue strength, crack growth rate parameters, and fatigue crack growth threshold. Based on the fatigue cumulative damage theory, the Miner cumulative damage model is used to integrate the stress monitoring data and the material SN curve to establish a fatigue cumulative damage model. The fatigue cumulative damage model is associated with the actual operating time, and the functional relationship between the fatigue damage and time changes in different operating stages is calculated. The real-time stress spectrum is input into the time change model to calculate the current fatigue cumulative damage at the current moment. Based on the current fatigue cumulative damage and the trend of the crack growth rate over time, the crack growth rate parameters in the crack growth rate model are adjusted, and the crack growth rate parameters are updated by fitting using the least squares method or optimization algorithm.

[0107] In specific implementation, the method determines the optimized crack growth rate parameter based on the current fatigue cumulative damage and the changing trend of the crack growth rate over time, including: determining the growth law of the current fatigue cumulative damage over time based on the data distribution of the current fatigue cumulative damage; constructing a corresponding form of association model based on the growth law and the changing trend of the crack growth rate over time; the association model characterizes the correlation between the crack growth rate parameter and the current fatigue cumulative damage and time; uses historical data, current fatigue cumulative damage, and crack growth rate to perform data fitting on the association model to determine model parameters; verifies the fitted association model based on verification data that does not participate in the fitting, and corrects and optimizes the association model based on the difference between the predicted crack growth rate parameter and the actual crack growth rate to obtain the optimized association model; inputs the current fatigue cumulative damage and time value into the optimized association model to obtain the optimized crack growth rate parameter.

[0108] Specifically, analyze the data distribution of the current fatigue cumulative damage to determine its growth pattern over time, such as linear growth, exponential growth or other complex change patterns. At the same time, analyze the change trend of the crack growth rate over time to determine whether it is gradually stable, accelerating growth or showing periodic fluctuations. Furthermore, according to the growth pattern of the current fatigue cumulative damage over time, the change trend of the crack growth rate over time and the data characteristics, select a suitable mathematical function to construct a correlation model between the crack growth rate parameter and the current fatigue cumulative damage and time. If the current fatigue cumulative damage is approximately linearly related to the crack growth rate, a linear regression model can be tried; if the relationship is more complex, it may be necessary to use nonlinear functions such as polynomial functions, exponential functions or logarithmic functions for fitting.

[0109] Furthermore, the constructed correlation model is fitted with data using the existing historical data and the current fatigue cumulative damage and crack growth rate. The parameter values ​​in the correlation model are calculated using the least squares method, the maximum likelihood estimation method or other appropriate parameter estimation methods. In this process, the initial parameter values ​​and the parameters of the fitting algorithm are continuously adjusted to improve the fitting accuracy of the correlation model. For example, when using the least squares method, the sum of square errors between the model prediction value and the actual observation value is minimized through iterative calculation, thereby obtaining a more accurate parameter estimation value. After determining the parameter values ​​of the correlation model, the obtained correlation model is verified using a part of the verification data that is not involved in the fitting. The verification data is substituted into the correlation model to calculate the predicted crack growth rate parameters and compare them with the actual crack growth rate. If the prediction error is large, it means that there may be problems with the correlation model, and the correlation model needs to be corrected and optimized according to the verification results until the prediction accuracy of the model meets the requirements. For example, if the correlation model structure is unreasonable, the function form needs to be reselected, or there may be outliers or noise in the training data, and the training data needs to be cleaned and preprocessed. After sufficient verification and correction, when the correlation model can accurately reflect the relationship between the crack growth rate parameter and the current fatigue cumulative damage and time, the current fatigue cumulative damage and time values ​​are substituted into the correlation model to calculate the optimized crack growth rate parameter value.

[0110] S107. Optimize the first value based on the real-time stress spectrum and the degree of degradation to obtain an optimal value of the crack growth rate parameter under the real-time working condition.

[0111] Specifically, the real-time stress spectrum provides dynamic stress amplitude, which directly affects the calculation of stress intensity factor. The larger the stress amplitude, the faster the crack growth rate, and the adaptability of the crack growth rate parameter in the crack growth rate model needs to be readjusted. Under high stress amplitude, the material enters the plastic zone, the crack growth rate accelerates, and the crack growth rate parameter increases.

[0112] Furthermore, the degree of degradation mainly affects the material's resistance to crack growth. In the crack growth rate model, the crack growth rate parameter will respond more violently to load changes as the material performance degrades, and the crack growth rate parameter increases.

[0113] In the specific implementation, the real-time stress spectrum and the degradation degree output by the shipbuilding gantry crane material aging prediction model are combined to achieve the optimization of the crack growth rate parameters using the dynamic coupling mechanism. Specifically, the stress amplitude provided by the real-time stress spectrum is input into the crack growth rate model, and the degradation degree is introduced to correct the stress intensity factor and the constant parameters in the crack growth rate model. This process is completed through the fitting optimization method to ensure that the crack growth rate model can simultaneously reflect the dynamic changes of the external load and the evolution of the internal material state.

[0114] S108. Calculate the remaining life of the shipbuilding gantry crane under the real-time working condition based on the optimal value.

[0115] Specifically, after obtaining the optimal value of the crack growth rate parameter, the crack growth rate is first calculated. In specific implementation, the crack growth rate is calculated based on the Paris formula:

[0116]

[0117] Among them, the is the crack growth rate; C and m are crack growth rate parameters; ΔK is the stress intensity range; Y is the shape parameter, which generally refers to the function of crack size a and plate width W. s (a, W). Since the width of the flange plate and web plate of the metal structure of the shipbuilding gantry crane is much larger than the crack length a, the flange plate and web plate with cracks can be regarded as a plate with an infinite central crack or an infinite single-side crack. For a plate with an infinite central crack, Y is 1; for a plate with an infinite single-side crack, Y is 1.12; Δσ is the stress range; a is the crack size.

[0118] It should be noted that in order to eliminate the influence of average stress, according to the equal life principle, the Goodman formula is used to convert all amplitude stresses into stress range Δσ under cyclic characteristics R = 0:

[0119]

[0120] Among them, the σ Ra is the stress amplitude when the cyclic characteristic is R; Rm The stress mean value when the cyclic characteristic is R; -1 When R=1, the yield limit of the material; b is the tensile limit of the material; Δσ is the stress range under the cyclic characteristic R=0, Δσ=σ max -σ min =σ max .

[0121] Furthermore, when the load is a variable amplitude load, the Miner stress amplitude equivalent method should be used according to the equal life principle to convert the variable amplitude load into a constant amplitude load. The equivalent formula is:

[0122]

[0123] Wherein, σ is the stress at each level; a i is the ratio of the number of cycles of each level of stress amplitude to the remaining life; the σ ri is the stress amplitude at each level; and m is the crack growth rate parameter.

[0124] Furthermore, after the crack growth rate is obtained, the remaining life of the shipbuilding gantry crane is calculated based on the crack growth rate, the initial crack size and the critical crack size. Among them, the initial crack size represents the initial crack size of the shipbuilding gantry crane, that is, the tiny cracks caused by fatigue, manufacturing defects, corrosion or environmental factors in the stress concentration area, which are usually obtained through non-destructive testing technology. The critical crack size represents that when the crack size expands to a certain critical value, the material or structure no longer has the bearing capacity due to fracture failure. The critical crack size is determined according to the linear elastic fracture judgment criterion:

[0125]

[0126] Among them, the a l is the critical crack size; the K c is the fracture toughness of the material; Y is the shape parameter; σ max is the maximum cyclic stress.

[0127] Furthermore, after determining the initial crack size, critical crack size and crack growth rate, the remaining life of the shipbuilding gantry crane can be calculated based on the following formula:

[0128]

[0129] Among them, the N f is the remaining life of the shipbuilding gantry crane; is the crack growth rate, C and m are crack growth rate parameters, Y is the shape parameter, Δσ is the stress range; a 0 is the initial crack size; l is the critical crack size.

[0130] The stress spectrum analysis and remaining life assessment method of the shipbuilding gantry crane provided in this embodiment, firstly, the dynamic load parameters directly affect the dynamic response of the crane. The dynamic load coefficient of the lifting is dynamically determined according to the dynamic load parameters, which can more accurately reflect the amplification effect of the dynamic load on the static load of the shipbuilding gantry crane under actual working conditions. Since the traditional method usually adopts a fixed lifting dynamic load coefficient, this method is easy to lead to overly conservative design or ignore the influence of special working conditions, which can avoid unnecessary redundancy or potential risks caused by a single parameter. In addition, the lifting dynamic load coefficient is calculated by the actual dynamic load parameters and can be flexibly adjusted under different working conditions. For example, when the lifting speed is fast, the lifting dynamic load coefficient will increase, reflecting a greater dynamic load effect; while under low speed and light load conditions, the lifting dynamic load coefficient will be relatively reduced, thereby more realistically describing the load characteristics of different working conditions. In a complex shipbuilding environment, the impact of dynamic loads on static loads may fluctuate over time. By dynamically adjusting the dynamic load coefficient through dynamic load parameters, real-time changes can be better captured, and the timeliness and accuracy of load analysis can be improved. Moreover, since the lifting dynamic load coefficient is linearly related to the stress amplitude in the stress spectrum, the dynamic calculation of the lifting dynamic load coefficient based on the dynamic load parameters can provide accurate input for the stress spectrum prediction model, improve the reliability of the stress spectrum prediction, and ensure the accuracy of the stress spectrum prediction. Secondly, by optimizing the crack growth rate parameters, not only the accuracy of the assessment of the remaining life of the shipbuilding gantry crane is significantly improved, but also the adaptability of the assessment model to complex working conditions is enhanced. Since the crack growth rate parameters are the key constants for the remaining life assessment, in practical applications, these parameters are not only affected by material properties, but also by the combined effects of dynamic loads, stress distribution and environmental conditions. By optimizing the crack growth rate parameters, the remaining life assessment is first made closer to the actual operating conditions of the crane. Moreover, by combining experimental data and finite element analysis, the crack growth rate parameters can effectively reflect the influence of dynamic loads on crack growth behavior in actual work. The experimental test provides real data of crack growth under different working conditions, and the finite element simulation reveals the distribution of stress concentration areas and the evolution of crack paths. The combination of the two can more accurately capture the law of crack growth rate changes with time and working conditions. This method eliminates the evaluation errors caused by overly idealized assumptions in traditional models, thereby improving the scientific nature of life assessment. In addition, the optimization process also fully considers the dynamic changes of stress amplitude and number of cycles, especially the changes in stress intensity factors under complex working conditions. By fitting the crack growth rate model to the actual fatigue crack growth data and introducing the stress distribution and crack growth path obtained by finite element simulation, the optimized crack growth rate parameters can accurately characterize the fatigue behavior of crane materials under complex load conditions. This method not only improves the prediction accuracy of the crack growth rate model, but also enhances the adaptability of the model to a variety of load conditions, providing possibilities for a wider range of engineering applications.Thirdly, when the stress spectrum prediction model based on the improved correlation vector machine is used for stress spectrum prediction, the core of the stress spectrum prediction model lies in accurately fitting the relationship between the stress amplitude and the number of cycles in the operation of the shipbuilding gantry crane. By adding a combined kernel function (including a polynomial kernel function and a radial basis kernel function) to the stress spectrum prediction model, the stress spectrum prediction model is enhanced in its ability to express nonlinear and multidimensional data by integrating the advantages of different kernel functions. Among them, the polynomial kernel function is suitable for processing global characteristics and can better capture the long-range correlation between data. The radial basis kernel function is good at modeling local characteristics and can effectively cope with complex nonlinear changes. Through linear combination, the two kernel functions complement each other, significantly improving the stress spectrum prediction model's ability to fit stress data, especially in the case of complex coupling between stress amplitude and load changes. In addition, by adding a particle swarm optimization algorithm to the stress spectrum prediction model, the parameters of the combined kernel function (such as polynomial order, radial basis width, etc.) are optimized through global search capabilities, avoiding the local optimal problem that the traditional method may fall into, thereby further improving the generalization ability and prediction accuracy of the stress spectrum prediction model.

[0131] Corresponding to the aforementioned embodiment of a method for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane, the present application also provides an embodiment of a device for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane.

[0132] Figure 2 This is a schematic diagram of the structure of the device for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane provided in Example 2 of the present application. Figure 2 The device provided in this embodiment includes an acquisition module 210, a determination module 220, a prediction module 230, an optimization module 240 and a calculation module 250; wherein,

[0133] The acquisition module 210 is used to acquire the dynamic load parameters of the shipbuilding gantry crane under various working conditions;

[0134] The determination module 220 is used to determine the lifting dynamic load coefficient based on the lifting state level corresponding to the dynamic load parameter; the lifting dynamic load coefficient represents the amplification or correction of the dynamic load of the shipbuilding gantry crane to the static load under each working condition;

[0135] The prediction module 230 is used to update the stress spectrum prediction model based on the improved correlation vector machine according to the lifting dynamic load coefficient, and the stress spectrum prediction model is used to predict the stress conditions of each position of the shipbuilding gantry crane under the preset working conditions to obtain the stress spectrum of the shipbuilding gantry crane; the stress spectrum includes the stress amplitude and the corresponding number of cycles of the shipbuilding gantry crane during operation, and the stress amplitude is linearly related to the lifting dynamic load coefficient;

[0136] The determination module 220 is further used to determine a first value of a crack growth rate parameter by experimental data fitting and finite element analysis, wherein the first value is applicable to various working conditions;

[0137] The prediction module 230 is further used to predict the real-time stress spectrum under the real-time working condition based on the stress spectrum prediction model;

[0138] The determination module 220 is further used to determine the degradation degree of the material under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model;

[0139] The optimization module 240 is used to optimize the first value based on the real-time stress spectrum and the degradation degree to obtain the optimal value of the crack growth rate parameter under the real-time working condition;

[0140] The calculation module 250 is used to calculate the remaining life of the shipbuilding gantry crane under the real-time working condition based on the optimal value.

[0141] The device of this embodiment can be used to perform Figure 1 The steps, specific implementation principles and implementation processes of the method embodiment shown are similar and will not be repeated here.

[0142] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0143] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0144] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for stress spectrum analysis and remaining life assessment of a shipbuilding gantry crane, characterized in that: The method comprises: Obtain the dynamic load parameters of shipbuilding gantry cranes under various working conditions; Based on the lifting state level corresponding to the dynamic load parameter, a lifting dynamic load coefficient is determined; the lifting dynamic load coefficient represents the amplification or correction of the dynamic load of the shipbuilding gantry crane to the static load under each working condition; According to the lifting dynamic load coefficient, a stress spectrum prediction model based on an improved correlation vector machine is updated, wherein the stress spectrum prediction model is used to predict the stress conditions at various positions of the shipbuilding gantry crane under preset working conditions to obtain a stress spectrum of the shipbuilding gantry crane; the stress spectrum includes the stress amplitude and the corresponding number of cycles of the shipbuilding gantry crane during operation, and the stress amplitude is linearly related to the lifting dynamic load coefficient; Determine a first value of a crack growth rate parameter by experimental data fitting and finite element analysis, wherein the first value is applicable to each working condition; Predicting a real-time stress spectrum under real-time working conditions based on the stress spectrum prediction model; Determining the degree of material degradation under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model; Optimizing the first value based on the real-time stress spectrum and the degradation degree to obtain an optimal value of the crack growth rate parameter under the real-time working condition; The remaining life of the shipbuilding gantry crane under the real-time working condition is calculated based on the optimal value.

2. The method according to claim 1, characterized in that The determining of the lifting dynamic load coefficient based on the lifting state level corresponding to the dynamic load parameter includes: Based on the lifting speed and load size in the dynamic load parameters, the corresponding lifting state level is determined; the lifting speed is positively correlated with the lifting state level, and the load size is positively correlated with the lifting state level; Based on the lifting state level, determining the corresponding first coefficient and second coefficient; wherein each lifting state level corresponds to a first coefficient and a second coefficient, and the higher the lifting state level, the greater the values ​​of the first coefficient and the second coefficient; determining a third coefficient based on the product of the rated lifting speed and the second coefficient; The sum of the first coefficient and the third coefficient is determined as the lifting dynamic load coefficient.

3. The method according to claim 1, characterized in that The first value of the crack growth rate parameter is determined by experimental data fitting and finite element analysis, including: Obtain fatigue crack growth data of the shipbuilding gantry crane under different working conditions through experimental testing; The shipbuilding gantry crane is simulated by using finite element analysis software to obtain predicted fatigue crack growth data under different working conditions; Determining stress distribution and crack propagation path of a stress concentration area based on the fatigue crack propagation data and the predicted fatigue crack propagation data; The crack growth rate parameter is fitted in combination with the stress distribution and the crack growth path to obtain a first value.

4. The method according to claim 3, characterized in that The method of obtaining fatigue crack growth data of the shipbuilding gantry crane under different working conditions through experimental testing includes: Analyzing actual working conditions of the shipbuilding gantry crane to determine abnormal working conditions of the shipbuilding gantry crane; Determining experimental parameters based on the stress distribution, load characteristics and fatigue damage mechanism of the shipbuilding gantry crane under the abnormal working condition; A fatigue experiment under the abnormal working condition is simulated based on the experimental parameters, and fatigue crack growth data of the shipbuilding gantry crane is collected.

5. The method according to claim 3, characterized in that: The combining the stress distribution and the crack propagation path to fit the crack propagation rate parameter to obtain a first value includes: Determining stress intensity factors under different working conditions based on the stress distribution and the crack propagation path; According to different working conditions, the fatigue crack growth data under each working condition is matched with the stress intensity factor to establish a crack growth rate model; the crack growth rate model characterizes the relationship between the crack growth rate and the stress intensity factor; The constant term in the crack growth rate model under various working conditions is fitted to obtain the crack growth rate parameters uniformly used under various working conditions.

6. The method according to claim 1, characterized in that The determining the degradation degree of the material under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model comprises: Determining material performance parameters of the shipbuilding gantry crane; A fatigue cumulative damage model is established based on the fatigue cumulative damage principle; the fatigue cumulative damage model is associated with the material performance parameters according to time to obtain a shipbuilding gantry crane material aging prediction model in which fatigue cumulative damage changes with time; Inputting the real-time stress spectrum into the shipbuilding gantry crane material aging prediction model to obtain the current fatigue accumulated damage; Based on the current fatigue accumulated damage and the change trend of the crack growth rate over time, an optimized crack growth rate parameter is determined.

7. The method according to claim 3, characterized in that Combining the stress distribution and the crack propagation path, fitting the crack propagation rate parameter to obtain a first value includes: Calculating stress intensity factors at different positions and working conditions based on the stress distribution and the crack propagation path; Based on the fatigue crack growth data, obtaining crack sizes at different stages and corresponding cycle numbers; Grouping the stress intensity factor, the crack size, and the number of cycles according to different working conditions; For each set of data, the least square method is used to perform a preliminary fitting on the crack growth rate parameters; Based on the preliminary fitting result and the fitting error, the first crack growth rate parameter in the crack growth rate parameters is adjusted, the second crack growth rate parameter is refitted, and the fitting error is calculated repeatedly and iteratively until the number of iterations reaches a preset number, and the first crack growth rate parameter and the second crack growth rate parameter with the smallest fitting error are determined as the first value.

8. The method according to claim 6, characterized in that The step of determining an optimized crack growth rate parameter based on the current fatigue accumulated damage and the change trend of the crack growth rate over time includes: Based on the data distribution of the current fatigue cumulative damage, determine the growth law of the current fatigue cumulative damage over time; Based on the growth law and the change trend of the crack growth rate over time, a corresponding correlation model is constructed; the correlation model represents the correlation between the crack growth rate parameter and the current fatigue cumulative damage and time; Using historical data, current fatigue cumulative damage, and crack growth rate to perform data fitting on the correlation model to determine model parameters; The fitted correlation model is verified based on the verification data that does not participate in the fitting, and the correlation model is corrected and optimized based on the difference between the predicted crack growth rate parameter and the actual crack growth rate to obtain an optimized correlation model; The current fatigue accumulated damage and time value are input into the optimized correlation model to obtain the optimized crack growth rate parameter.

9. The method according to claim 1, characterized in that: The first value of the crack growth rate parameter is determined by experimental data fitting and finite element analysis, including: Obtain fatigue crack growth data of shipbuilding gantry cranes under different operating conditions; Fitting the fatigue crack growth data to construct a crack growth rate model; Performing finite element simulation on the shipbuilding gantry crane to determine stress distribution and crack propagation path in stress concentration areas; A stress intensity factor is extracted from the stress distribution and the crack propagation path, the stress intensity factor is matched with the crack propagation rate model, the crack propagation rate model is optimized according to the distribution of the stress intensity factor and the crack propagation path, and a first value of the crack propagation rate parameter is determined.

10. A device for stress spectrum analysis and remaining life assessment of shipbuilding gantry crane, characterized in that: The device includes an acquisition module, a determination module, a prediction module, an optimization module and a calculation module; wherein, The acquisition module is used to obtain the dynamic load parameters of the shipbuilding gantry crane under various working conditions; The determination module is used to determine the lifting dynamic load coefficient based on the lifting state level corresponding to the dynamic load parameter; the lifting dynamic load coefficient represents the amplification or correction of the dynamic load of the shipbuilding gantry crane to the static load under each working condition; The prediction module is used to update the stress spectrum prediction model based on the improved correlation vector machine according to the lifting dynamic load coefficient, and the stress spectrum prediction model is used to predict the stress conditions of each position of the shipbuilding gantry crane under preset working conditions to obtain the stress spectrum of the shipbuilding gantry crane; the stress spectrum includes the stress amplitude and the corresponding number of cycles of the shipbuilding gantry crane during operation, and the stress amplitude is linearly related to the lifting dynamic load coefficient; The determination module is further used to determine a first value of a crack growth rate parameter by experimental data fitting and finite element analysis, wherein the first value is applicable to various working conditions; The prediction module is further used to predict the real-time stress spectrum under real-time working conditions based on the stress spectrum prediction model; The determination module is further used to determine the degradation degree of the material under the real-time working condition based on the real-time stress spectrum and the shipbuilding gantry crane material aging prediction model; The optimization module is used to optimize the first value based on the real-time stress spectrum and the degradation degree to obtain the optimal value of the crack growth rate parameter under the real-time working condition; The calculation module is used to calculate the remaining life of the shipbuilding gantry crane under the real-time working condition based on the optimal value.

Images