Method for evaluating residual bearing capacity of component based on entropy fuzzy analysis method

Abstract

The invention discloses a component residual bearing capacity evaluation method based on an entropy fuzzy analysis method, and belongs to the field of structural engineering. Existing evaluation methods such as subjectivity of an expert experience method and high cost of a load test method, a reliability theory depends on mass data, and weight of an existing fuzzy evaluation method is too subjective. The method comprises the following steps: simulating different damage rates of a component through a physical test, and establishing a dynamic critical value adjustment formula set; correcting the normalized matrix of the traditional entropy method to obtain dynamic entropy weight sets at different rates; collecting actual measurement damage data to construct a current and pre-judgment fuzzy judgment matrix, and combining the current and pre-judgment fuzzy judgment matrix into a compensation matrix; and a target layer fuzzy evaluation matrix is calculated in combination with the compensation matrix and the dynamic weight, a final reduction coefficient is obtained through correction, and then the residual bearing capacity is calculated. The problem of dynamic weight distribution under multi-damage coupling is solved, subjective dependence is reduced, damage rate changes are adapted, evaluation precision is improved, and the method can be used for structure maintenance decision making and emergency study and judgment after disasters.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of structural engineering, more particularly to a component residual bearing capacity evaluation method based on entropy value fuzzy analysis method. BACKGROUND

[0002] In the field of civil engineering, the residual bearing capacity evaluation of structural components such as reinforced concrete beams and columns is a technical link to ensure the service safety of infrastructure such as buildings and bridges. With the increase of service life and the deterioration of service environment, these components inevitably face multi-dimensional damage such as cross-section loss, crack propagation, concrete strength degradation, and longitudinal reinforcement corrosion. These damage factors have nonlinear coupling effects, which together determine the actual bearing capacity of the components. Therefore, accurate evaluation of residual bearing capacity has important practical significance for structural maintenance decision-making and post-disaster safety judgment.

[0003] The current mainstream component residual bearing capacity evaluation methods have obvious technical shortcomings: the expert experience method relies on the engineering accumulation of the evaluators to determine the damage level and estimate the bearing capacity, which is affected by subjective cognitive differences and has large consistency error in the evaluation of multi-damage coupling scenarios, making it difficult to meet the engineering precision requirements; the load test method obtains the actual bearing limit of the component through on-site loading, which is intuitive, but requires interrupting the normal use of the structure, has high cost for single component testing, and may exacerbate existing damage during the loading process, making it unsuitable for post-disaster emergency evaluation scenarios; the reliability theory evaluation method builds a damage and bearing capacity correlation model based on probability statistics, but it requires a large amount of historical damage data and component mechanical property parameters, and the model calibration period is long, with significant response lag for sudden damage such as rapid crack propagation after an earthquake.

[0004] To improve the above problems, a bridge bearing capacity evaluation method based on multi-level fuzzy comprehensive evaluation method is disclosed in Chinese patent CN120372907A, which integrates four factors such as material performance degradation and structural damage, uses the analytic hierarchy process to determine the index weight, and establishes a multi-level evaluation system, which has obvious improvement in efficiency compared with traditional static load test. However, this method still has inherent defects. The weight distribution depends on the expert scoring step in the analytic hierarchy process, and the subjective assignment leads to insufficient matching between the weight and the actual damage influence, especially in the multi-damage coupling scenario, the error is obvious; at the same time, this method does not consider the dynamic influence of damage evolution rate on the importance of indicators, and the evaluation model constructed is a static framework, which cannot adapt to the working conditions with rapid changes in damage over time, and still cannot break through the contradiction between evaluation accuracy and calculation complexity and subjectivity.

[0005] In summary, the existing evaluation methods cannot effectively solve the problem of dynamic weight allocation caused by the nonlinear coupling of multiple damage factors, and the technical demand of realizing accurate evaluation without a large amount of prior data and relying on subjective judgment has not been met. In view of the above problems, a component residual bearing capacity evaluation method based on entropy fuzzy analysis method is proposed. SUMMARY

[0006] In view of the problems in the prior art, the purpose of the present application is to provide a component residual bearing capacity evaluation method based on entropy fuzzy analysis method, which can solve the problem of dynamic weight allocation under multiple damage coupling, reduce subjective dependence, adapt to damage rate changes and improve evaluation accuracy.

[0007] To solve the above problems, the technical scheme adopted by the present application is as follows.

[0008] A component residual bearing capacity evaluation method based on entropy fuzzy analysis method, comprising: Step S1, simulate different damage rates of the target component through physical tests, monitor the changes of vertical crack width and cross-section loss rate with time in real time, record the critical bearing capacity values of each rate from level one to level five damage, and establish a mathematical relationship between the rate and the critical values of various damage indicators; Step S2, generate multiple groups of bearing capacity data at each rate using the dynamic critical values of step S1, introduce a rate correction coefficient into the normalization matrix of the traditional entropy method and substitute it into the corresponding data of the dynamic critical values for correction; based on the corrected data, recalculate the entropy values, difference coefficients and weights of each indicator to obtain a dynamic entropy weight set at different rates; Step S3, collect the measured damage data of the current component, combine the dynamic critical values of step S1 to determine the level, and use the dynamic weights of S2 to calculate the current fuzzy evaluation matrix; according to the damage rate prediction of step S1, determine the level and calculate the predicted fuzzy matrix; build a compensation matrix by combining the current matrix and the predicted matrix according to the predetermined weight; Step S4, combine the compensation matrix of step S3 with the dynamic entropy weights of step S2 to calculate the target layer fuzzy evaluation matrix; calculate the initial reduction coefficient according to the matrix, and combine the damage rate correction of step S1 to obtain the final reduction coefficient; calculate the residual bearing capacity according to the design bearing capacity and the final reduction coefficient.

[0009] Further, the mathematical relationship between the rate and the critical values of various damage indicators comprises: Step 11, simulate the micro-crack propagation behavior, crack morphology evolution and dynamic change of cross-section loss rate of the component under different damage evolution rates through high-precision experiments, and introduce a dynamic aging correction factor of crack propagation to analyze the damage evolution path at each rate; Step S12, combined with the provided experimental data, a piecewise curve fitting method is used to describe the dynamic relationship between different damage rates and critical values, and the fitting process is optimized through adaptive weight correction algorithm, considering the time effect of crack propagation and the response difference of component characteristics to damage rate; Step S13, according to the fitting results, a dynamic adjustment formula set is established, considering the influence of different damage rates, and introducing a correction factor based on damage history, to adjust the critical value in real time; Step S14, through real-time monitoring system to collect damage rate and component damage state data, dynamically adjust the critical value of damage index, combined with the adaptive optimization algorithm of damage rate prediction model, adjust the evaluation path, and output real-time damage grade evaluation results.

[0010] Further, using the dynamic critical value of step S1 to generate multiple sets of bearing capacity data under each rate, including: Step S21, by combining experiments and numerical simulation, using local damage reconstruction algorithm, generate local damage mode of component according to different damage rates, and based on the simulation results of crack propagation behavior, calculate the local stress distribution and bearing capacity data under different position crack conditions; Step S22, using time-dependent adaptive incremental correction algorithm, gradually correct the local damage data under each damage rate to update the corresponding bearing capacity data; Step S23, based on the generated bearing capacity data, using dynamic weighted optimization algorithm, according to damage rate, crack propagation speed, material degradation rate and structural geometric characteristic factors, real-time adjust the weight, optimize the bearing capacity data under each rate; Step S24, based on the optimized bearing capacity data, construct real-time iterative updating mechanism; by real-time monitoring the change of damage rate to trigger the dynamic adjustment of correction parameters, and based on the adjusted parameters to re-correct and optimize the iteration of bearing capacity data.

[0011] Further, introduce rate correction coefficient into traditional entropy value normalization matrix and substitute it into dynamic critical value corresponding data for correction, including: Step S31, based on the dynamic correlation between damage rate and each type of damage index, design the mapping mechanism of rate correction coefficient, and use damage rate adaptive prediction model to dynamically adjust the rate correction coefficient according to real-time data; Step S32, based on the generated rate correction coefficient, through the innovative rate sensitivity correction algorithm, dynamically optimize the traditional normalization matrix, and introduce adaptive weighting mechanism, adjust the weight of normalization matrix according to the nonlinear influence of damage rate on bearing capacity; Step S33, map the rate-corrected normalized matrix with the dynamic threshold value, and use a feedback adjustment mechanism to adjust the bearing capacity data in fine granularity according to the real-time damage rate, damage development stage and the change of the threshold value; Step S34, construct a real-time iterative adjustment mechanism to continuously monitor the changes of the damage rate and the threshold value, and automatically update the weights and correction factors in the entropy value calculation process.

[0012] Further, based on the corrected data, the entropy values, difference coefficients and weights of each index are recalculated to obtain a dynamic entropy weight set under different rates, including: Step S41, design an entropy dynamic adjustment mechanism based on the damage rate, and adjust the calculation parameters of the entropy value in real time by introducing a rate correction factor and damage evolution path analysis; Step S42, based on the generated dynamic entropy value, design a rate response type difference coefficient adjustment mechanism, and dynamically correct the difference coefficient of each damage index according to the relationship between the damage rate and the entropy value; Step S43, by obtaining the dynamic entropy value and the difference coefficient, use a weighted optimization mechanism to update the weight of each damage index in real time, and dynamically adjust according to the change of the damage rate; Step S44, by monitoring the changes of the damage rate and the bearing capacity in real time, automatically adjust each weight in the entropy weight set, and optimize the entropy weight set in each evaluation period.

[0013] Further, the current fuzzy evaluation matrix is calculated according to the dynamic weight, including: Step S51, map the dynamic entropy weight with the measured data of each damage index of the component, and establish a nonlinear mapping relationship through the index rate sensitivity to correspond the real-time damage level of each index with its dynamic weight; Step S52, interactively correct the obtained preliminary index membership degree, adjust the membership degree according to the sensitivity relationship between the real-time damage rate of each index and the dynamic weight, to generate the dynamically corrected index membership degree; Step S53, based on the obtained dynamically corrected membership degree, construct the fuzzy evaluation matrix at the current time, wherein the rows of the matrix correspond to the damage indexes, and the columns correspond to the damage levels, and the matrix elements are determined by comprehensively considering the dynamic weight and the interactively corrected membership degree.

[0014] Further, according to the damage rate prediction of the future damage and the judgment of the level in step S1, a prediction fuzzy matrix is calculated, including: Step S61, based on the damage evolution rate and the dynamic threshold value provided in step S1, construct a future rate prediction model of each damage index, and quantitatively represent the damage development of each index in the future time period by analyzing the current rate change trend, historical evolution trajectory and microscopic crack propagation behavior; Step S62, on the basis of the future damage rate output by the prediction model, the future damage level of each index is determined according to the preset grade division standard; Step S63, using the future damage level determination result, a pre-judgment fuzzy matrix is constructed, each row corresponds to a damage index, and each column corresponds to a damage level, and the matrix elements are formed by mapping the prediction level to the membership value; Step S64, on the basis of the generated pre-judgment matrix, a dynamic timeliness correction factor is introduced to nonlinearly adjust the matrix elements.

[0015] Further, a compensation matrix is constructed, and the current matrix and the pre-judgment matrix are combined according to a predetermined weight, including: Step S71, a pre-judgment fuzzy matrix at a future time point is generated through the generated current fuzzy evaluation matrix and the provided damage evolution rate, and a dynamic timeliness correction factor is introduced in the pre-judgment process to nonlinearly adjust the state of each damage index in the prediction period; Step S72, the generated pre-judgment matrix and the generated current fuzzy evaluation matrix are nonlinearly fused according to a predetermined combination weight, an index sensitivity factor and a rate dependence coefficient are introduced in the fusion process to weight adjust the matrix elements; Step S73, on the basis of the fused preliminary combination matrix, iterative optimization is performed, and the matrix elements are gradually adjusted according to the damage rate change, the time decay characteristics of the matrix elements and the weight deviation.

[0016] Further, the step S3 compensation matrix is combined with the step S2 dynamic entropy weight to calculate the target layer fuzzy evaluation matrix, including: Step S81, the compensation matrix generated in step S3 and the dynamic entropy weight in step S2 are nonlinearly mapped to form a preliminary target layer mapping matrix, wherein the mapping relationship is established based on the dynamic weight of each index and the membership value of the matrix element; Step S82, on the basis of the preliminary mapping matrix in step S81, the matrix elements are weighted and adjusted according to the dynamic entropy weight sensitivity of each index to form a correction matrix; Step S83, on the basis of the correction matrix in step S82, the nonlinear mapping values and the sensitivity adjustment results of each index are integrated to construct the target layer fuzzy evaluation matrix, wherein each row of the matrix corresponds to an index, and each column corresponds to a damage level.

[0017] Further, an initial reduction coefficient is calculated according to the matrix, and the final reduction coefficient is obtained in combination with the damage rate correction in step S1, including: S91, based on the target layer fuzzy evaluation matrix, a nonlinear mapping mechanism is established to combine the membership of each index in the matrix with the damage level, to accumulate and weight process the contribution degree of each index, and calculate the initial reduction coefficient; S92, the provided damage evolution rate is coupled with the initial reduction coefficient, a rate sensitive correction mechanism of the reduction coefficient is established, the initial reduction coefficient is nonlinearly adjusted, and a rate correction reduction coefficient is generated; S93, on the basis of the output rate correction reduction coefficient, a final reduction coefficient generation mechanism is constructed, and continuity verification and time sequence consistency adjustment are performed.

[0018] Compared with the prior art, the beneficial effects of the present application are: (1) The scheme effectively solves the problem of dynamic weight allocation in the multi-damage coupling scene of the existing evaluation method, discards the defects of subjective judgment of traditional expert experience method and weight dependence of analytic hierarchy process, establishes a dynamic critical value system through physical test, optimizes the traditional entropy value method normalization matrix by combining the rate correction coefficient, and calculates the dynamic entropy weight under different damage rates in real time, without the support of massive historical data, reduces the interference of subjective factors, accurately matches the influence degree of each damage index with the rate change, and improves the objectivity and precision of the residual bearing capacity evaluation.

[0019] (2) The scheme breaks through the limitation that the existing static evaluation model cannot adapt to the dynamic change of damage rate, dynamically adjusts the critical value, weight and reduction coefficient through real-time monitoring system and iterative updating mechanism, and constructs a compensation matrix by fusing the current measured damage and future damage prediction, taking into account the current state and future damage development trend of the component, which can not only avoid the problem of interrupting the use of structure in load test method and high cost, but also quickly respond to sudden damage, and provide timely and reliable technical support for structure maintenance decision and post-disaster emergency research. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0021] Figure 1 The flowchart of the component residual bearing capacity evaluation method based on the entropy value fuzzy analysis method of the present application. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application; obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0023] Referring to Figure 1 A component residual bearing capacity evaluation method based on entropy value fuzzy analysis method, comprising: Step S1, simulate different damage rates of the target component through physical tests, monitor the vertical crack width and the cross-section loss rate in real time, record the critical bearing capacity values of each rate from the first to the fifth damage, and establish a mathematical relationship between the rate and the critical values of each type of damage index; Step S2, generate multiple groups of bearing capacity data at each rate using the dynamic critical values of step S1, introduce a rate correction coefficient into the normalization matrix of the traditional entropy value method, and modify the corresponding data of the dynamic critical values; based on the modified data, recalculate the entropy values, difference coefficients and weights of each index to obtain a dynamic entropy weight set at different rates; Step S3, collect the measured damage data of the current component, combine the dynamic critical values of step S1 to judge the level, and calculate the current fuzzy evaluation matrix using the dynamic weights of S2; predict the future damage and judge the level according to the damage rate of step S1, calculate the predicted fuzzy matrix; build a compensation matrix, and combine the current matrix and the predicted matrix according to the predetermined weight; Step S4, combine the compensation matrix of step S3 with the dynamic entropy weight of step S2 to calculate the target layer fuzzy evaluation matrix; calculate the initial reduction coefficient according to the matrix, and combine the damage rate correction of step S1 to obtain the final reduction coefficient; calculate the residual bearing capacity according to the design bearing capacity and the final reduction coefficient.

[0024] Step 1 further comprises: step 11, simulating the micro-crack propagation behavior, crack morphology evolution and dynamic change of cross-section loss rate of the component under different damage evolution rates through high-precision experiments, and introducing a dynamic time-effect correction factor of crack propagation to analyze the damage evolution path at each rate; Step S12, combine the provided experimental data, use the piecewise curve fitting method to describe the dynamic relationship between different damage rates and critical values, and optimize the fitting process through an adaptive weight correction algorithm, considering the time effect of crack propagation and the response difference of component characteristics to damage rate; Step S13, according to the fitting result, establish a dynamic adjustment formula set, consider the influence of different damage rates, and introduce a correction factor based on damage history to adjust the critical value in real time; Step S14, collect the data of damage rate and component damage state through a real-time monitoring system, dynamically adjust the critical value of the damage index, combine the adaptive optimization algorithm of the damage rate prediction model to adjust the evaluation path, and output the real-time damage level evaluation result.

[0025] In this embodiment, firstly, the correlation data between the component damage and the critical bearing capacity under different damage evolution rates need to be obtained through experiments. The operation logic of this process is as follows. In the experimental design stage, the micro and macro damage behaviors of the component under various damage evolution rates are simulated, specifically covering the micro crack propagation path, the dynamic evolution law of the macro crack morphology, and the variation characteristics of the cross-section loss rate with time. Considering that there is a significant time effect in crack propagation, such as the difference in the influence mechanism of slow damage and rapid damage on the mechanical properties of the component, a dynamic time aging correction factor for crack propagation needs to be introduced in the experiment. This factor quantifies the effect of the time dimension on damage development, and then more accurately analyzes the damage evolution path corresponding to different damage rates, ensuring that the obtained damage process data can truly reflect the nature of the component damage under different stress and environmental conditions.

[0026] Based on the original data obtained through the above experiments, a quantitative correlation model between the damage rate and the critical bearing capacity is established. At this time, the piecewise curve fitting method is adopted for analysis. The reason for choosing piecewise fitting is that in different damage rate intervals, the mapping relationship between damage indicators such as vertical crack width, cross-section loss rate, and critical bearing capacity often presents non-single characteristics. A single curve cannot accurately describe the dynamic correlation in the whole rate range, while piecewise fitting can establish a dedicated fitting relationship for the characteristics of different rate intervals, improving the accuracy of the correlation description. In the fitting process, an adaptive weight correction algorithm needs to be optimized. This algorithm adjusts the fitting weight dynamically in combination with two key influencing factors. One is the time effect of crack propagation, which corrects the deviation of fitting parameters under long-term slow damage and short-term rapid damage. The other is the response difference of component characteristics, which adjusts the weight proportion of fitting results under different rates according to the inherent characteristics of component material properties, geometric dimensions, and stress forms. Finally, the fitting results can accurately reflect the dynamic relationship between damage rate and critical value under different scenarios.

[0027] Based on the piecewise fitting results, a dynamic adjustment formula set is further constructed. The design goal of this formula set is to realize the real-time adaptation of the critical value. The formula set needs to clearly include the influencing factors of different damage rates to ensure that the critical value output by the formula can be dynamically adjusted with the change of damage rate, avoiding the evaluation deviation caused by the use of fixed critical value. At the same time, in order to consider the cumulative effect of component damage, a correction factor based on damage history needs to be introduced. This factor can quantify the influence of the past damage state of the component on the current critical value. For example, the critical bearing capacity of a component with certain initial damage under the same damage rate is different from that of an initially undamaged component. Through the damage history correction factor, this difference can be included in the formula calculation, and finally the dynamic adjustment formula set has the ability to update the critical value in real time according to the damage history of the component.

[0028] To apply the dynamic formula set to the actual assessment scenario, a closed-loop adjustment mechanism needs to be formed relying on the real-time monitoring system, which continuously collects two types of data, i.e., the current damage rate of the component, such as the growth value of the crack width per unit time and the change rate of the cross-section loss rate, and the real-time damage state of the component, such as the current vertical crack width and cross-section loss rate. Using the collected data, first, the damage index critical value in the dynamic adjustment formula set is calibrated in real time to ensure that the critical value is consistent with the current damage development trend of the component. At the same time, combined with the adaptive optimization algorithm of the damage rate prediction model, the algorithm can predict the trend of the subsequent damage rate according to the historical damage rate data and the real-time monitoring data, and adjust the evaluation path of the damage level based on the trend, for example, when it is predicted that the damage rate will increase, the algorithm can optimize the adjustment range of the critical value and the evaluation parameters in advance, and finally ensure that the output damage level evaluation result can accurately reflect the current and near-term damage state of the component.

[0029] In some embodiments, the dynamic critical value of step S1 is used to generate multiple groups of bearing capacity data under each rate, including: Step S21, by combining experiments and numerical simulations, using a local damage reconstruction algorithm, generating local damage patterns of the component according to different damage rates, and based on the simulation results of crack propagation behavior, calculating the local stress distribution and bearing capacity data under different crack conditions at different positions; Step S22, using a time-dependent adaptive incremental correction algorithm, step-by-step correction of local damage data under each damage rate to update the corresponding bearing capacity data; Step S23, based on the generated bearing capacity data, using a dynamic weighted optimization algorithm, real-time adjustment of the weight according to the damage rate, crack propagation speed, material degradation rate and structural geometric characteristic factors, optimizing the bearing capacity data under each rate; Step S24, based on the optimized bearing capacity data, constructing a real-time iterative updating mechanism; by real-time monitoring of the change of damage rate to trigger dynamic adjustment of correction parameters, and based on the adjusted parameters to re-correct and optimize the iteration of bearing capacity data.

[0030] In this embodiment, firstly, analysis is carried out through the synergy of experiments and numerical simulation, and data refinement is realized by relying on the local damage reconstruction algorithm. The experimental part mainly collects the macro mechanical response and key part damage observation data of the component under different damage rates; numerical simulation is based on the experimental data to build a refined finite element model of the component, which makes up for the limitations of local microscopic damage observation in experiments. The local damage reconstruction algorithm can restore the damage morphology characteristics of different regions of the component, such as tension zone, shear compression zone, corner region, etc., according to the experimental observation data under different damage rates and the local stress and strain distribution of numerical simulation, including the crack initiation position, propagation direction, branch morphology and specific range of cross-section loss, and then simulate the influence of crack propagation behavior at each local damage position on the overall stress transfer path of the component, for example, the local crack propagation in a certain tension zone may change the stress concentration degree in this region, and then change the attenuation trend of the overall bearing capacity, and finally the bearing capacity data refined to different damage positions and damage morphologies are obtained through this local and overall correlation analysis, rather than the overall bearing capacity average.

[0031] After obtaining the refined local damage and bearing capacity correlation data, the local damage data under each type of damage rate needs to be corrected to eliminate the deviation caused by the timeliness factor. This process uses a timeliness adaptive incremental correction algorithm. The timeliness of this algorithm lies in fully considering the time dependence of damage development. Under the same damage rate, the influence of local damage at different damage stages on the bearing capacity is different. For example, the influence of local micro-cracks on the bearing capacity is small at the initial damage stage, while the influence is significant when the local cracks are penetrated at the later damage stage. The adaptability of the algorithm is that it can automatically adjust the direction and amplitude of correction according to the real-time change trend of local damage data, without the need for artificial preset fixed correction parameters. The incremental method means that the correction process is not completed at one time, but is adjusted step by step according to the incremental steps of damage development, such as an incremental step of 0.1 mm increase in local crack width or 0.5% increase in cross-section loss rate. Each step is based on the damage data at the current incremental stage and the correction result of the previous stage, and dynamically updates the bearing capacity data under the damage rate. Through this incremental correction in stages, the real-time influence of damage rate on bearing capacity is accurately adjusted, ensuring that the bearing capacity data matches the time process of damage development.

[0032] Considering that the influence of damage on the bearing capacity is the result of the comprehensive action of multiple factors, further optimization is needed based on the generated bearing capacity data. At this time, a dynamic weighting optimization algorithm is adopted. According to the four key influencing factors of damage rate, crack propagation rate, material degradation rate and structural geometric characteristics, the weight proportion of each factor in the bearing capacity calculation is adjusted in real time. The damage rate directly determines the overall pace of bearing capacity attenuation. The faster the rate, the higher the corresponding weight. The crack propagation rate reflects the speed of local damage development. The area with faster crack propagation rate has more significant short-term impact on the bearing capacity, and the weight needs to be increased accordingly. The material degradation rate, such as the corrosion rate of steel and the carbonation rate of concrete, determines the degree of attenuation of the mechanical properties of the component material. The faster the degradation, the greater the long-term impact on the bearing capacity, and the weight needs to be dynamically adapted. The structural geometric characteristics, such as the cross-sectional size, span and reinforcement ratio of the component, determine the bearing potential of the component itself. For example, the bearing capacity attenuation amplitude of a large cross-section component is different from that of a small cross-section component under the same local damage, and the difference needs to be reflected through the geometric characteristic weight adjustment. The algorithm continuously optimizes the weight coefficients of each factor by real-time collection of the current state data of the above four types of factors, and recalculates the bearing capacity data under each type of damage rate based on the optimized weights, so that the optimized bearing capacity data can comprehensively reflect the synergistic effect of multiple factors, rather than the result dominated by a single factor.

[0033] To ensure that the bearing capacity data can continuously match the real-time damage state of the component, a real-time iterative updating mechanism needs to be established based on the optimized bearing capacity data. This mechanism continuously tracks the current damage rate change of the component by real-time monitoring of the system. If the monitoring finds that the damage rate has changed from 0.2 mm / day to 0.5 mm / day, the mechanism triggers the dynamic adjustment of the correction parameters, including the incremental step in the time-dependent adaptive incremental correction algorithm and the weight coefficients of each factor in the dynamic weighting optimization algorithm. After the parameters are adjusted, the mechanism will call the aforementioned correction and optimization algorithms again, and iteratively calculate the bearing capacity data based on the updated damage rate data, forming a closed-loop process of real-time monitoring of the damage rate, adjusting the correction parameters, iteratively optimizing the bearing capacity data, and monitoring again. Through this continuous iterative updating, even if the damage rate of the component changes due to changes in the external environment, the bearing capacity data can be dynamically adjusted to always match the actual damage state of the component.

[0034] In some embodiments, a rate correction coefficient is introduced to the traditional entropy value normalization matrix and substituted into the dynamic critical value corresponding data for correction, including: Step S31, based on the dynamic correlation of damage rate and various damage indicators, a mapping mechanism of rate correction coefficient is designed, and a damage rate adaptive prediction model is used to dynamically adjust the rate correction coefficient according to real-time data; Step S32, based on the generated rate correction coefficient, the traditional normalization matrix is dynamically optimized by the innovative rate sensitivity correction algorithm, and an adaptive weighting mechanism is introduced to adjust the weight of the normalization matrix according to the nonlinear influence of the damage rate on the bearing capacity; Step S33, map the rate corrected normalization matrix with the dynamic threshold value, and adopt the feedback adjustment mechanism to adjust the bearing capacity data in fine granularity according to the real-time damage rate, damage development stage and the change of the threshold value; Step S34, construct a real-time iterative adjustment mechanism to continuously monitor the changes of the damage rate and the threshold value, and automatically update the weight and correction factor in the entropy value calculation process.

[0035] In this embodiment, the damage rate and the damage index are not in a fixed corresponding relationship, such as vertical crack width and cross-section loss rate, and the damage development stage presents dynamic changes, for example, in the early stage of damage, a lower damage rate may only cause a slight increase in crack width, while in the middle stage of damage, the same rate may cause faster expansion of the crack width. Based on this dynamic correlation, a mapping mechanism of the rate correction coefficient can be designed, which can convert the change range of the damage index under different damage rates into the corresponding correction coefficient range, ensuring that the correction coefficient matches the actual correlation characteristics of the damage development. At the same time, in order to cope with the possible fluctuations of the damage rate, such as temporary increase in rate caused by external load mutation, an adaptive prediction model of damage rate is needed, which can predict the evolution direction of the damage rate in the short term according to the real-time collected damage rate data and historical change trend, and then dynamically adjust the rate correction coefficient, avoiding the lag of the correction coefficient caused by rate mutation, and ensuring that the correction coefficient always matches the current and recent damage rate state.

[0036] The traditional normalization matrix usually only standardizes the data based on the absolute value of the damage index, without distinguishing the influence difference of the same index value on the bearing capacity under different damage rates, for example, the same crack width value usually means faster bearing capacity decay under high damage rate, but the traditional matrix cannot reflect this difference. The rate sensitivity correction algorithm can integrate the rate correction coefficient into the calculation process of the normalization matrix, and standardize the damage index data under different damage rates, so that the standardized data can reflect the rate sensitivity characteristics. At the same time, in order to further adapt to the nonlinear influence of the damage rate on the bearing capacity, such as the increase in the bearing capacity decay amplitude when the damage rate exceeds a certain threshold, an adaptive weighting mechanism is needed, which can dynamically adjust the weight proportion of each damage index in the normalization matrix according to the nonlinear correlation between the real-time analyzed damage rate and the bearing capacity decay, for example, the weight of the damage index more sensitive to the bearing capacity will be increased accordingly in the high damage rate interval, so as to optimize the adaptive ability of the normalization matrix to the nonlinear influence.

[0037] The rate-corrected normalized matrix reflects the standardized state of the damage indicators under different damage rates, and the dynamic threshold represents the indicator threshold corresponding to each damage level. Mapping the two can determine the relative position of the current standardized damage indicator and the critical value of each damage level, providing a basis for adjusting the bearing capacity data. On this basis, a feedback adjustment mechanism is needed, which can collect three data in real time: the change of the current damage rate, the damage development stage of the component, and the update result of the dynamic threshold. By analyzing the synergistic effect of these three data, the bearing capacity data is adjusted in stages and at different rates, for example, when the damage level is in the transition stage and the damage rate is rising, the adjustment range of the bearing capacity data will be increased accordingly according to the proximity of the critical value, avoiding sudden changes or deviations in bearing capacity evaluation, and ensuring that the adjusted bearing capacity data can accurately match the current damage state.

[0038] To ensure the accuracy and timeliness of subsequent entropy value calculation, a real-time iterative adjustment mechanism needs to be built. This mechanism continuously tracks two key parameters by real-time monitoring: the real-time change of damage rate and the update of dynamic threshold. Once any parameter is found to have changed beyond the preset fluctuation range, the mechanism will automatically trigger the update process of the weight and correction factor in the entropy value calculation. The mechanism will recalibrate the rate correction coefficient based on the new damage rate data, and then adjust the weight of the normalized matrix. At the same time, it will reanalyze the entropy value contribution of each damage indicator in combination with the change of the dynamic threshold, to update the difference coefficient and the final indicator weight, ensuring that each parameter in the entropy value calculation process can be iterated in real time as the damage state changes. This continuous iterative update mechanism can avoid the entropy value calculation deviation caused by fixed parameters.

[0039] In some embodiments, the entropy values, difference coefficients, and weights of each indicator are recalculated based on the correction data to obtain a dynamic entropy weight set under different rates, including: Step S41, design an entropy dynamic adjustment mechanism based on damage rate, adjust the calculation parameters of entropy in real time by introducing rate correction factor and damage evolution path analysis; Step S42, based on the generated dynamic entropy value, design a rate-responsive difference coefficient adjustment mechanism, dynamically correct the difference coefficient of each damage indicator according to the relationship between damage rate and entropy value; Step S43, by obtaining the dynamic entropy value and difference coefficient, use a weighted optimization mechanism to update the weight of each damage indicator in real time, and dynamically adjust according to the change of damage rate; Step S44, by monitoring the changes of damage rate and bearing capacity in real time, automatically adjust each weight in the entropy weight set, and optimize the entropy weight set in each evaluation period.

[0040] In this embodiment, since the traditional entropy calculation parameters are mostly fixed values, it is difficult to reflect the difference in the contribution of information of each damage index under different damage rates. Therefore, by introducing a rate correction factor, the influence of damage rate on index information entropy is quantified, and the development law of each damage index under different rates is clarified through damage evolution path analysis. For example, under rapid damage, the information change of the crack width index is more significant, and under slow damage, the information accumulation of the cross-section loss rate index is more critical. Based on these analysis results, the basic parameters in the entropy value calculation process are adjusted in real time, so that the calculated entropy value can dynamically match the index information characteristics under the current damage rate, rather than remaining unchanged.

[0041] The difference coefficient, as an important parameter for measuring the information utility value of each damage index, is negatively correlated with the entropy value. Since the entropy value has been associated with the damage rate through the aforementioned mechanism, the difference coefficient also needs to be dynamically corrected following the damage rate. This mechanism determines the information utility change trend of each damage index under the current rate by analyzing the corresponding relationship between different damage rates and dynamic entropy values. For example, when the damage rate increases, the entropy value of a certain type of index may decrease rapidly, and the corresponding difference coefficient needs to increase accordingly. In this way, the difference coefficient of each damage index is dynamically corrected to ensure that the difference coefficient accurately reflects the information contribution value of the index to the bearing capacity evaluation under the current rate.

[0042] The determination of the weight needs to consider two data, the dynamic entropy value reflects the information disorder degree of the index, and the difference coefficient reflects the information utility of the index. The weighting optimization mechanism calculates the initial weight distribution result by constructing the correlation logic between the two and the weight. At the same time, this mechanism has the ability to dynamically respond to changes in damage rate. When the damage rate changes, the weight proportion of each index will be adjusted in time according to the change of entropy value and difference coefficient under the new rate, so that the weight can accurately match the influence of each index on the remaining bearing capacity evaluation of the member under different rates.

[0043] To ensure the continuous effectiveness of the entropy weight set, the real-time monitoring system needs to track the dynamic changes of damage rate and bearing capacity. Once it is detected that the fluctuations of the two exceed the preset range, the system will automatically trigger the weight adjustment process to make targeted corrections to each index weight in the entropy weight set. At the same time, in each fixed evaluation period, whether the parameter fluctuation exceeds the threshold or not, the optimization program of the weight set will be started, and the weight set will be calibrated comprehensively based on the damage data and bearing capacity change law accumulated in this period, to ensure that the weight set always adapts to the actual damage state and bearing capacity of the member.

[0044] In some embodiments, the current fuzzy judgment matrix is calculated according to the dynamic weight, including: Step S51, mapping the dynamic entropy weight with the measured damage index data of the component, establishing a nonlinear mapping relationship through the index rate sensitivity, and corresponding the real-time damage level of each index with its dynamic weight; Step S52, interactive correction of the obtained preliminary index membership degree, adjusting the membership degree according to the sensitivity relationship between the real-time damage rate of each index and the dynamic weight, to generate the dynamically corrected index membership degree; Step S53, constructing the fuzzy evaluation matrix of the current time based on the obtained dynamically corrected membership degree, wherein the rows of the matrix correspond to the damage indexes, and the columns correspond to the damage levels, and the matrix elements are determined by comprehensively considering the dynamic weight and the interactive corrected membership degree.

[0045] In this embodiment, since the sensitivity of different damage indexes to the damage rate is different, the correlation rule between the real-time damage level of the same damage index at different rates and the dynamic weight is also nonlinear. For example, at a higher damage rate, a slight increase in the vertical crack width may have a more significant impact on the bearing capacity of the component, and the corresponding dynamic weight needs to be adjusted more significantly. Therefore, by establishing a nonlinear mapping relationship through the index rate sensitivity, the real-time damage level of each index, such as the current measured crack width, can be corresponded with its dynamic weight, avoiding the problem of mismatch between the weight and the actual damage influence caused by linear correlation.

[0046] After the above mapping is completed, the preliminary membership degree of each damage index can be obtained, which reflects the preliminary possibility that the real-time damage level of the index belongs to a certain damage level. However, it may not fully consider the deviation caused by the sensitivity correlation between the real-time damage rate and the dynamic weight. Therefore, the preliminary index membership degree needs to be interactive corrected. According to the sensitivity relationship between the current real-time damage rate of each index, such as the change value of the cross-section loss rate per unit time, and the dynamic weight of the index, if the real-time damage rate of a certain index is significantly higher than that in the early stage, and its dynamic weight has been adjusted upward due to the change in rate, then the preliminary membership degree of the index needs to be adjusted appropriately to a higher damage level, otherwise to a lower level. Through this interactive correction based on the sensitivity relationship, a dynamically corrected membership degree that is more consistent with the current damage state is generated.

[0047] Based on the dynamic corrected membership, a fuzzy evaluation matrix of the current time can be constructed. The row dimension of the matrix corresponds to each damage index of the component, such as the vertical crack width, the cross-section loss rate, etc. The column dimension corresponds to the damage levels from level one to level five. Each element in the matrix needs to be determined by comprehensively combining the dynamic weight and the inter-corrected membership. The dynamic weight reflects the importance of the damage index in the current evaluation, and the inter-corrected membership reflects the possibility of the index belonging to the corresponding damage level. Through reasonable comprehensive calculation, such as weighted summation, the specific value of each element is obtained, and finally the fuzzy evaluation matrix is formed. The fuzzy evaluation matrix can comprehensively and accurately reflect the evaluation results of each damage index of the component corresponding to different damage levels at the current time.

[0048] In some embodiments, according to step S1, the future damage is predicted and the level is judged, and a prediction fuzzy matrix is calculated, including: Step S61, based on the damage evolution rate and the dynamic threshold value provided in step S1, a future rate prediction model of each damage index is constructed. By analyzing the current rate change trend, the historical evolution trajectory and the micro crack propagation behavior, the damage development of each index in the future time period is quantitatively represented; Step S62, based on the future damage rate output by the prediction model, the future damage level of each index is determined according to the preset level division standard; Step S63, using the future damage level determination result, a prediction fuzzy matrix is constructed. Each row corresponds to a damage index, and each column corresponds to a damage level. The prediction level is mapped to the membership value to form the matrix elements; Step S64, based on the generated prediction matrix, a dynamic time effectiveness correction factor is introduced to nonlinearly adjust the matrix elements.

[0049] In this embodiment, first, a future rate prediction model of each damage index needs to be constructed. The construction of the model is based on the damage evolution rate and the dynamic threshold value determined in step S1, ensuring that the prediction logic is consistent with the dynamic threshold system established earlier. In the model construction process, three types of information need to be analyzed comprehensively. The first is the change trend of the current damage rate, which is determined by the recent monitoring data to judge whether the rate is rising, falling or stable. The second is the historical evolution trajectory of the damage, which combs the fluctuation rule of the damage rate and the cumulative change characteristics of the damage index in the past period of time. The third is the micro crack propagation behavior, which grasps the influence of the dynamic process of crack initiation, expansion and penetration on the overall damage rate from the micro level. Through the coordinated analysis of the three types of information, the model can quantitatively present the damage development of each damage index in the future time period, for example, the growth amplitude of the vertical crack width or the specific change value of the cross-section loss rate in the next 7 days.

[0050] After the future damage rate prediction model is built and the future damage rate is obtained, the future damage level of each damage index needs to be determined based on the predicted rate. The determination process is based on the preset damage level division standard. The standard has clearly defined the critical range of each index corresponding to different damage levels. For example, the vertical crack width corresponding to the first damage level is 0.1-0.2mm, the cross-section loss rate is 0.5%-1.0%, the second damage level corresponds to wider cracks and higher loss rate, etc. By converting the future damage rate output by the prediction model into the specific quantitative value of each damage index at the end of the future time period, and then comparing these quantitative values with the critical range in the preset level division standard, the damage level of each damage index at the future time can be determined, and the preliminary determination of the future damage level is completed.

[0051] Based on the determination result of the future damage level, a pre-judgment fuzzy matrix can be further constructed. The structure of the matrix is consistent with the current fuzzy evaluation matrix. The row dimension corresponds to each damage index of the component, the column dimension corresponds to the damage level from the first to the fifth level, and the matrix elements are determined by mapping the predicted level to the membership value. If a damage index is determined to belong to the second damage level in the future, the membership value of the index corresponding to the second damage level column will be set to a higher level, such as 0.8, while the membership values corresponding to the first and third damage level columns will be set to a lower level, such as 0.1 and 0.1. Through this membership mapping, the discrete level determination result is converted into continuous fuzzy matrix elements, so that the matrix can more subtly reflect the uncertainty and fuzziness of the future damage level.

[0052] Considering the uncertainty in the time dimension of future damage prediction, and the damage development may be affected by unforeseen factors, a dynamic time-dependent correction factor needs to be introduced based on the generated pre-judgment matrix. The role of the correction factor is to non-linearly adjust the matrix elements according to the proximity of the prediction time. For example, the closer the prediction time is to the current time, the smaller the uncertainty of damage development, and the smaller the adjustment amplitude of the correction factor to the matrix elements. The farther the prediction time, the greater the uncertainty, and the adjustment amplitude increases accordingly. At the same time, this adjustment presents a nonlinear characteristic because the damage development is not a uniform process. In some stages, such as when the damage is close to the critical value, it may appear to accelerate or decelerate. The dynamic time-dependent correction factor can make the pre-judgment matrix elements more consistent with the actual time effect of damage development through nonlinear adjustment, improving the accuracy and reliability of the pre-judgment matrix.

[0053] In some embodiments, a compensation matrix is constructed by combining the current matrix and the pre-judgment matrix with a predetermined weight, including: Step S71, generate a pre-judgment fuzzy matrix at the future time point through the generated current fuzzy evaluation matrix and the provided damage evolution rate, and introduce a dynamic time-dependent correction factor in the pre-judgment process to non-linearly adjust the state of each damage index in the prediction period; Step S72, the generated preliminary judgment matrix and the generated current fuzzy judgment matrix are nonlinearly fused according to a predetermined combination weight, and in the fusion process, an index sensitivity factor and a rate dependence coefficient are introduced to weight and adjust the matrix elements; Step S73, on the basis of the fused preliminary combination matrix, iterative optimization is performed, and the matrix elements are gradually adjusted according to the damage rate change, the time decay characteristics of the matrix elements and the weight deviation.

[0054] In the embodiment, the current fuzzy judgment matrix accurately reflects the membership degree state of each damage index corresponding to different damage levels at the current time of the component, and is a basic reference for the preliminary judgment; the damage evolution rate provides a dynamic trend of damage development, and determines the direction and speed of evolution of each damage index to the future state. In order to eliminate the prediction uncertainty caused by the time dimension, a dynamic time-dependent correction factor needs to be introduced in the preliminary judgment process. The factor will nonlinearly adjust the state of each damage index in the prediction period according to the length of the prediction time span and the nonlinear characteristics of damage development, such as different stages of slow initial damage, accelerated intermediate damage and stable later damage. For example, the farther the prediction time, the higher the uncertainty of damage development, and the greater the adjustment amplitude of the correction factor to the membership degree. The adjustment process is not uniform, but conforms to the nonlinear law of actual damage evolution, and finally generates a preliminary fuzzy matrix that can reasonably reflect the future damage state.

[0055] The predetermined combination weight needs to be set according to specific evaluation requirements. For example, in short-term safety evaluation, the current fuzzy judgment matrix can be given a higher weight, and in long-term durability evaluation, the weight proportion of the preliminary fuzzy matrix can be appropriately increased. In order to further improve the fusion accuracy, an index sensitivity factor and a rate dependence coefficient need to be introduced in the fusion process. The index sensitivity factor is used to distinguish the sensitivity of different damage indexes to the bearing capacity of the component. For example, the influence of vertical crack width on the bearing capacity of a flexural component is usually higher than that of surface micro defects, and the sensitivity factor value is higher, and the weight adjustment amplitude of the corresponding matrix element is larger. The rate dependence coefficient is related to the damage evolution rate. When the damage rate is fast, the coefficient increases the contribution of the preliminary judgment matrix element in the fusion, and vice versa. Through the weighted adjustment of the two types of parameters, the preliminary combination matrix after fusion can not only consider the authenticity of the current state, but also reflect the development trend of future damage.

[0056] The preliminary combined matrix obtained by fusion may not be accurate due to changes in external conditions or deviations in parameter settings, and therefore an iterative optimization process needs to be performed on this basis. The basis for iterative optimization mainly comes from three aspects: first, real-time changes in damage rate. If a deviation is monitored between the damage rate after fusion and the initial rate when the damage is predicted, the matrix elements need to be adjusted accordingly to adapt to the new damage development rhythm. Second, the time decay characteristics of the matrix elements. As time goes by, the utility of historical data, especially the current matrix data, used as a reference in the early fusion period for the current evaluation will gradually decrease, and the element values need to be adjusted according to the decay law. Third, weight deviation. There may be deviations in the parameters such as the predetermined combination weight, index sensitivity factor, etc. during the fusion process. Through iteration, the difference between the matrix output result and the actual damage monitoring data can be compared, and the weight deviation can be gradually corrected. Through multiple rounds of iterative adjustment, the matrix elements are constantly approaching the real damage state correlation characteristics of the component, and finally a more accurate compensation matrix is formed.

[0057] In some embodiments, the step S3 compensation matrix is combined with the step S2 dynamic entropy weight to calculate the target layer fuzzy evaluation matrix, including: Step S81, the compensation matrix generated in step S3 is nonlinearly mapped with the dynamic entropy weight in step S2 to form a preliminary target layer mapping matrix, wherein the mapping relationship is established based on the dynamic weight of each index and the membership value of the matrix element; Step S82, based on the preliminary mapping matrix in step S81, the matrix elements are weighted and adjusted according to the dynamic entropy weight sensitivity of each index to form a correction matrix; Step S83, based on the correction matrix in step S82, the nonlinear mapping values and sensitivity adjustment results of each index are integrated to construct a target layer fuzzy evaluation matrix, wherein each row of the matrix corresponds to an index, and each column corresponds to a damage level.

[0058] In this embodiment, the compensation matrix has integrated the information of the current damage state of the component and the future damage prediction, and its matrix elements represent the membership degrees of each damage index corresponding to different damage levels. The dynamic entropy weight reflects the importance of different damage indices in the current damage rate evaluation of bearing capacity. Since the dynamic weight of each damage index and the membership degree of the corresponding element in the compensation matrix are not in a linear relationship, for example, when the dynamic weight of a certain damage index is high, the influence of a small change in its membership degree on the overall evaluation result will be greater than that of a lower weight index. Therefore, a nonlinear mapping is needed to establish the correlation between the two, to ensure that the dynamic weight can reasonably act on the membership data of the compensation matrix, and the preliminary target layer mapping matrix formed finally can preliminarily reflect the synergistic effect of the index weight and the membership degree.

[0059] The dynamic entropy weight sensitivity refers to the sensitivity of the weight of different damage indicators to the change of damage rate and damage state. The weight of some indicators may fluctuate due to a slight change in damage rate, and such indicators have a more critical influence on the overall evaluation result. Based on this sensitivity characteristic, the matrix element corresponding to the indicator with high weight sensitivity needs to be given a greater adjustment weight when performing weighting adjustment. For example, if the weight sensitivity of a certain indicator is significantly higher than that of other indicators, the element value of the indicator in the preliminary matrix needs to be further adjusted according to the sensitivity degree to amplify the reasonable influence of the weight change of the indicator on the matrix element, so as to avoid the problem of mismatch between the matrix element and the actual importance of the indicator due to the difference in weight sensitivity, and make the corrected matrix more consistent with the influence characteristics of the indicators under the current damage scenario.

[0060] After the weighting adjustment of the matrix elements is completed, the nonlinear mapping values and the sensitivity adjustment results of the indicators need to be integrated on the basis of the corrected matrix. The integration process needs to fully retain the weight and membership basis established by the early nonlinear mapping, and at the same time, integrate the precision optimization brought by the sensitivity adjustment, so as to ensure that each matrix element can reflect not only the dynamic weight influence of the corresponding damage indicator, but also the differentiated adjustment effect brought by the weight sensitivity. The row dimension of the target layer fuzzy evaluation matrix corresponds to each damage indicator of the component, and the column dimension corresponds to the damage levels from the first level to the fifth level. Each element in the matrix is the result of double optimization of nonlinear mapping and sensitivity adjustment, and can comprehensively and accurately reflect the comprehensive evaluation state of each damage indicator corresponding to different damage levels under the current damage rate.

[0061] In some embodiments, the initial reduction coefficient is calculated according to the matrix, and the final reduction coefficient is obtained by combining the damage rate correction of step S1, including: S91, based on the target layer fuzzy evaluation matrix, a nonlinear mapping mechanism is established to combine the membership of each indicator in the matrix with the damage level, accumulate and weight the contribution of each indicator, and calculate the initial reduction coefficient; S92, the provided damage evolution rate is coupled with the initial reduction coefficient for analysis to establish a rate-sensitive correction mechanism of the reduction coefficient, and the initial reduction coefficient is nonlinearly adjusted to generate a rate-corrected reduction coefficient; S93, based on the output rate-corrected reduction coefficient, a final reduction coefficient generation mechanism is constructed to perform continuity verification and time sequence consistency adjustment.

[0062] In this embodiment, the target layer fuzzy evaluation matrix has clearly defined the membership degree of each damage indicator corresponding to different damage levels, and the influence of membership degree and damage level on the reduction of member bearing capacity is not a linear relationship. For example, when the damage level is close to the critical state, the reduction amplitude of bearing capacity corresponding to the same membership degree change will be significantly greater than that in the low damage level stage. Therefore, the association between the two can be accurately described through a nonlinear mapping mechanism. In this process, the contribution of each damage indicator needs to be accumulated and weighted. The weighting basis is the dynamic importance of each indicator in the current damage scenario, and the accumulation is the integrated effect of all indicators on the reduction of bearing capacity. Through this systematic processing, the initial reduction coefficient is finally calculated.

[0063] Under different damage evolution rates, even if the initial reduction coefficient is the same, the actual decay trend of the member bearing capacity also differs. For example, when the damage is rapid, the member bearing capacity may show a more obvious decline in a short period of time. In this case, the initial reduction coefficient needs to be further adjusted to match the actual decay law. Based on this coupling analysis, a rate-sensitive correction mechanism for establishing the reduction coefficient is established. This mechanism will nonlinearly adjust the initial reduction coefficient according to the size and change characteristics of the damage evolution rate. Generally, the higher and more drastic the damage rate, the greater the adjustment amplitude, so as to generate a rate correction reduction coefficient that can reflect the rate influence.

[0064] To ensure the reliability and data consistency of the reduction coefficient, a final reduction coefficient generation mechanism needs to be established based on the rate correction reduction coefficient. The continuity check is used to ensure that the change of the reduction coefficient conforms to the continuous characteristics of the member damage development, avoiding the occurrence of illogical mutations in the reduction coefficient due to fluctuations in the monitoring data, and ensuring that it is consistent with the actual law of gradual development of damage. The time series consistency adjustment is to compare the current reduction coefficient with historical damage data, past rate records and previous evaluation results, eliminate possible data contradictions, and make the final reduction coefficient consistent with the overall damage assessment system in the time dimension.

[0065] After the final reduction coefficient is determined, the residual bearing capacity of the member can be calculated. The design bearing capacity of the member is the upper limit of its bearing capacity in the undamaged state determined according to the relevant design specifications, and the final reduction coefficient has comprehensively covered the influence of the current damage state, future damage prediction and damage rate. Therefore, through the multiplication operation of the design bearing capacity and the final reduction coefficient, the residual bearing capacity of the member in the current and foreseeable future period of time can be obtained. This result can directly provide technical basis for the safety performance evaluation and operation strategy formulation of the member.

[0066] The above merely provides the preferred embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art, according to the technical solution of the present application and the improved concept thereof, makes equivalent replacement or change within the technical range disclosed by the present application, and should be covered within the protection scope of the present application.

Claims

1. A method for evaluating the residual bearing capacity of structural members based on entropy-value fuzzy analysis, characterized in that, include: Step S1: Simulate different damage rates of the target component through physical experiments, monitor the changes in vertical crack width and cross-sectional loss rate over time in real time, record the critical bearing capacity values ​​of damage from level one to level five at each rate, and establish the mathematical relationship between the rate and the critical values ​​of various damage indicators. Step S2: Using the dynamic critical value from step S1, generate multiple sets of bearing capacity data at each rate. Introduce a rate correction coefficient into the normalization matrix of the traditional entropy method and substitute it into the data corresponding to the dynamic critical value for correction. Based on the corrected data, recalculate the entropy value, difference coefficient, and weight of each index to obtain the dynamic entropy value weight set at different rates. Step S3: Collect the measured damage data of the current component, combine it with the dynamic critical value in step S1 to determine the level, and use the dynamic weight in S2 to calculate the current fuzzy evaluation matrix. Based on the damage rate in step S1, predict future damage and determine its level, and calculate the prediction fuzzy matrix; Construct a compensation matrix and combine the current matrix and the prediction matrix according to predetermined weights; Step S4: Combine the compensation matrix from step S3 with the dynamic entropy weights from step S2 to calculate the fuzzy evaluation matrix of the target layer. The initial reduction coefficient is calculated based on the matrix, and the final reduction coefficient is obtained by combining it with the damage rate correction in step S1. The remaining bearing capacity is calculated based on the design bearing capacity and the final reduction factor.

2. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 1, characterized in that, Establish mathematical relationships between rate and critical values ​​of various damage indicators, including: Step 11: Through high-precision experiments, simulate the microcrack propagation behavior, crack morphology evolution, and dynamic changes in cross-sectional loss rate of the component under different damage evolution rates, and introduce a dynamic time-dependent correction factor for crack propagation to analyze the damage evolution path at each rate. Step S12: Based on the provided experimental data, a piecewise curve fitting method is used to describe the dynamic relationship between different damage rates and critical values. The fitting process is optimized by an adaptive weight correction algorithm, taking into account the time effect of crack propagation and the difference in the response of component characteristics to damage rate. Step S13: Based on the fitting results, establish a dynamic adjustment formula set, consider the influence of different damage rates, and introduce a correction factor based on damage history to adjust the critical value in real time. Step S14: Collect data on damage rate and component damage status through a real-time monitoring system, dynamically adjust the critical value of the damage index, adjust the evaluation path by combining the adaptive optimization algorithm of the damage rate prediction model, and output the real-time damage level evaluation result.

3. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 2, characterized in that, Multiple sets of bearing capacity data at various rates are generated using the dynamic critical values ​​from step S1, including: Step S21: By combining experiments and numerical simulations, a local damage reconstruction algorithm is used to generate local damage modes of components according to different damage rates, and based on the simulation results of crack propagation behavior, the local stress distribution and bearing capacity data under crack conditions at different locations are calculated. Step S22: The time-adaptive incremental correction algorithm is used to gradually correct the local damage data at each damage rate in order to update the corresponding bearing capacity data. Step S23: Based on the generated bearing capacity data, a dynamic weighted optimization algorithm is used to adjust the weights in real time according to the damage rate, crack propagation rate, material degradation rate and structural geometric characteristics, and optimize the bearing capacity data under each rate. Step S24: Based on the optimized bearing capacity data, a real-time iterative update mechanism is constructed; the dynamic adjustment of correction parameters is triggered by real-time monitoring of changes in damage rate, and the bearing capacity data is iterated and optimized again based on the adjusted parameters.

4. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 3, characterized in that, The normalization matrix of the traditional entropy method is corrected by introducing a rate correction coefficient and substituting the data corresponding to the dynamic critical value, including: Step S31: Based on the dynamic correlation between damage rate and various damage indicators, design a mapping mechanism for rate correction coefficient, and adopt an adaptive damage rate prediction model to dynamically adjust the rate correction coefficient according to real-time data. Step S32: Based on the generated rate correction coefficient, the traditional normalization matrix is ​​dynamically optimized through an innovative rate sensitivity correction algorithm, and an adaptive weighting mechanism is introduced to adjust the weight of the normalization matrix according to the nonlinear influence of the damage rate on the bearing capacity. Step S33: Map the normalized matrix after rate correction to the dynamic critical value, and use a feedback adjustment mechanism to finely adjust the bearing capacity data according to the real-time damage rate, damage development stage and critical value changes. Step S34: Construct a real-time iterative adjustment mechanism to continuously monitor changes in damage rate and critical value, and automatically update the weights and correction factors in the entropy calculation process.

5. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 4, characterized in that, Based on the corrected data, the entropy values, difference coefficients, and weights of each indicator are recalculated to obtain a dynamic entropy weight set under different rates, including: Step S41: Design an entropy dynamic adjustment mechanism based on damage rate, and adjust the calculation parameters of entropy in real time by introducing a rate correction factor and damage evolution path analysis. Step S42: Based on the generated dynamic entropy value, design a rate-response-type difference coefficient adjustment mechanism to dynamically correct the difference coefficient of each damage index according to the relationship between damage rate and entropy value. Step S43: Using the obtained dynamic entropy value and difference coefficient, a weighted optimization mechanism is adopted to update the weight of each damage index in real time and make dynamic adjustments according to the change of damage rate. Step S44: By monitoring the changes in damage rate and bearing capacity in real time, automatically adjust each weight in the entropy weight set, and optimize the entropy weight set in each evaluation cycle.

6. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 5, characterized in that, The current fuzzy evaluation matrix is ​​calculated based on dynamic weights, including: Step S51: Map the dynamic entropy value weights to the measured damage index data of the component, establish a nonlinear mapping relationship through the index rate sensitivity, and correspond the real-time damage level of each index to its dynamic weight. Step S52: Perform interactive correction on the obtained preliminary index membership degree, and adjust the membership degree according to the sensitivity relationship between the real-time damage rate and dynamic weight of each index to generate dynamically corrected index membership degree. Step S53: Construct the fuzzy evaluation matrix for the current time based on the obtained dynamic correction membership degree, where the rows of the matrix correspond to the damage index and the columns correspond to the damage level. The matrix elements are determined by combining the dynamic weights and the membership degree after interactive correction.

7. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 8, characterized in that, Based on the damage rate in step S1, predict future damage and determine its level, and calculate the prediction fuzzy matrix, including: Step S61: Based on the damage evolution rate and dynamic critical value provided in step S1, construct a future rate prediction model for each damage index. By analyzing the current rate change trend, historical evolution trajectory and microcrack propagation behavior, quantify the damage development of each index in the future time period. Step S62: Based on the future damage rate output by the prediction model, determine the future damage level of each indicator according to the preset level classification standard. Step S63: Using the future damage level determination results, construct a prediction fuzzy matrix, with each row corresponding to the damage index and each column corresponding to the damage level. The matrix elements are formed by mapping the predicted level to the membership value. Step S64: Based on the generated prediction matrix, a dynamic timeliness correction factor is introduced to nonlinearly adjust the matrix elements.

8. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 6, characterized in that, Construct a compensation matrix by combining the current matrix and the prediction matrix according to predetermined weights, including: Step S71: Generate a predictive fuzzy matrix for future time points using the generated current fuzzy evaluation matrix and the provided damage evolution rate, and introduce a dynamic timeliness correction factor during the prediction process to nonlinearly adjust the state of each damage index during the prediction period. Step S72: The generated prediction matrix and the generated current fuzzy evaluation matrix are nonlinearly fused according to a predetermined combination weight. During the fusion process, an index sensitivity factor and a rate dependence coefficient are introduced to adjust the weight of the matrix elements. Step S73: Based on the initial combined matrix after fusion, perform iterative optimization, and gradually adjust the matrix elements according to the damage rate change, the time decay characteristics of the matrix elements and the weight deviation.

9. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 7, characterized in that, The compensation matrix from step S3 is combined with the dynamic entropy weights from step S2 to calculate the fuzzy evaluation matrix for the target layer, including: Step S81: Perform a nonlinear mapping between the compensation matrix generated in step S3 and the dynamic entropy weights in step S2 to form a preliminary target layer mapping matrix, wherein the mapping relationship is established based on the dynamic weights of each index and the membership values ​​of the matrix elements. Step S82: Based on the preliminary mapping matrix in step S81, the matrix elements are weighted and adjusted according to the dynamic entropy values ​​and weight sensitivity of each indicator to form a correction matrix. Step S83: Based on the correction matrix in step S82, integrate the nonlinear mapping values ​​of each indicator with the sensitivity adjustment results to construct a fuzzy evaluation matrix for the target layer, where each row of the matrix corresponds to an indicator and each column corresponds to a damage level.

10. The method for evaluating the residual bearing capacity of a component based on entropy fuzzy analysis according to claim 8, characterized in that, The initial reduction coefficient is calculated based on the matrix, and combined with the damage rate correction in step S1, the final reduction coefficient is obtained, including: S91, based on the target layer fuzzy evaluation matrix, establish a nonlinear mapping mechanism, combine the membership degree of each indicator in the matrix with the damage level, accumulate and weight the contribution of each indicator, and calculate the initial reduction coefficient; S92, the provided damage evolution rate is coupled with the initial reduction coefficient for analysis, a rate-sensitive correction mechanism for the reduction coefficient is established, the initial reduction coefficient is nonlinearly adjusted, and a rate-corrected reduction coefficient is generated; S93, based on the output rate correction reduction coefficient, constructs a final reduction coefficient generation mechanism, and adjusts it through continuity verification and time series consistency.

Images