Bridge fatigue life intelligent prediction method, terminal and storage medium

Abstract

The invention relates to the technical field of bridge monitoring, and discloses a bridge fatigue life intelligent prediction method, a terminal and a storage medium. The method comprises the following steps: firstly, calculating a stress spectrum of a bridge in a set period so as to count the stress amplitude and cyclic frequency of the bridge; secondly, optimizing parameters of a logarithmic Gaussian mixture model by adopting a particle swarm improved mixture information criterion, and performing probability density modeling on the stress spectrum by utilizing the optimized logarithmic Gaussian mixture model so as to obtain actual cycle times of a plurality of levels of stress amplitudes; establishing a mapping relation between the stress amplitude and the fatigue life according to an S-N curve of the bridge member; and finally, according to the actual cycle index of each stage of stress amplitude and the corresponding fatigue life, based on a Miner criterion, predicting the accumulated damage amount of the bridge in a set period so as to obtain the predicted fatigue life of the bridge. The prediction precision of the fatigue life of the bridge is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge monitoring, and specifically to an intelligent prediction method for bridge fatigue life, as well as a computer terminal and a computer-readable storage medium applying the method. Background Art

[0002] With the acceleration of the urbanization process, as a key transportation facility, the safety and reliability of bridges have become particularly important. However, during long-term service, bridges gradually develop fatigue damage due to the environment, load, and material aging, seriously threatening their structural safety. In recent years, accidents caused by bridge fatigue damage have occurred frequently, posing a serious threat to life and property safety and social stability.

[0003] Traditional methods for evaluating bridge fatigue life rely on design specifications and laboratory data. These methods are difficult to reflect the changes in the environment and traffic load in actual operation in real time, resulting in insufficient prediction accuracy. With the development of the Internet of Things, sensing technology, and big data, bridge health monitoring systems can collect stress data in real time, more accurately reflect the state of bridges, thus breaking through the limitations of traditional methods and providing a scientific basis for life prediction and decision-making.

[0004] The existing methods for evaluating bridge fatigue life in research mainly rely on model fitting of a single feature, such as normal distribution, Weibull distribution, and gamma distribution, etc. These methods often fail to fully capture the multi-feature problems in the actual stress spectrum, resulting in poor fitting effects. In addition, most of the research relies on the experience and subjective judgment of users in the process of model fitting optimization, lacking automated and intelligent support. This is not only time-consuming but also easily affected by personal judgment, which may lead to the selection result not being scientific or optimal enough. Summary of the Invention

[0005] To solve the technical problems existing in the prior art, the present invention provides an intelligent prediction method for bridge fatigue life, a terminal, and a storage medium. By using a particle swarm improved hybrid information criterion to optimize the number of the best probability density model groups, and using a log Gaussian mixture model to accurately fit the probability density, the prediction accuracy of bridge fatigue life is significantly improved.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention discloses an intelligent prediction method for bridge fatigue life, including:

[0008] Calculating the stress spectrum of the bridge within a set period to statistically obtain the stress amplitude and its cycle frequency of the bridge;

[0009] The parameters of the logarithmic Gaussian mixture model are optimized by using a hybrid information criterion improved by a particle swarm, and the probability density modeling of the stress spectrum is performed by using the optimized logarithmic Gaussian mixture model, so as to obtain the actual number of cycles of several levels of stress amplitude;

[0010] The mapping relationship between stress amplitude and fatigue life is established based on the SN curve of bridge components;

[0011] According to the actual number of cycles of each level of stress amplitude and the corresponding fatigue life, the cumulative damage of the bridge within the set period is predicted based on the Miner criterion, thereby obtaining the predicted fatigue life of the bridge.

[0012] As a further improvement of the above solution, the calculation of the stress spectrum of the bridge within a set period includes:

[0013] Preprocess the bridge strain monitoring data to obtain stress time series data;

[0014] The stress spectrum is generated by a rain flow counting method based on the stress time series data.

[0015] As a further improvement of the above scheme, the preprocessing includes:

[0016] Collect the full strain time series data of the strain monitoring equipment on the bridge structure, detect and fill in the missing data, and thus obtain the original strain data;

[0017] The original strain data is decomposed into several eigenmode functions and a residual term through empirical mode decomposition, the eigenmode functions related to the temperature effect are eliminated, the temperature-corrected strain data are reconstructed, and the corrected strain data are converted into stress time series data according to the elastic modulus of the bridge structure material.

[0018] As a further improvement of the above solution, generating the stress spectrum by rain flow counting method includes:

[0019] Identify local maxima and minima in stress time series data; calculate the amplitude of each stress cycle average value And set the period T = t2-t1; where σ max and σ min are the local maximum and local minimum in the stress time series data respectively; t2 and t1 are the start time and end time of the stress cycle respectively;

[0020] All identified stress cycles are statistically analyzed according to their amplitude and frequency to generate a stress spectrum.

[0021] As a further improvement of the above scheme, the optimization of the parameters of the logarithmic Gaussian mixture model using the particle swarm improved hybrid information criterion includes:

[0022] Construct the Mixed Information Criterion (MIC), whose expression is:

[0023] MIC = α × AIC + (1 - α) × BIC

[0024] Wherein, AIC = 2k - 2ln(L); BIC = ln(q)k - 2ln(L); α is the weight coefficient, α ∈ [0, 1]; k is the number of parameters of the log - Gaussian mixture model, L is the maximum likelihood estimate of the model; q is the sample size;

[0025] Take the Mixed Information Criterion (MIC) as the objective function, and use the particle swarm algorithm to perform minimization optimization on the objective function. The optimization process is as follows:

[0026] Initialize the particle swarm, and update the velocity and position of the particles. The expression formulas are:

[0027] v i (t + 1)= ωv i (t)+ c1r1(P i - x i (t))+ c2r2(g - x i (t))

[0028] x i (t + 1)= x i (t)+ v i (t + 1)

[0029] Wherein, the position x i and velocity v i of each particle i are randomly generated; ω is the inertia weight; c1, c2 are learning factors; r1, r2 are random numbers; P i is the historical best position of particle i; g is the global best position of the population; t is the number of iteration steps;

[0030] Evaluate the objective function value MIC(x i ) corresponding to the position x i of each particle, and update the historical best position P i and the global best position g; Iterate until the termination condition is met to obtain the optimal mixed model parameters that minimize MIC.

[0031] As a further improvement of the above - mentioned scheme, the expression of the log - Gaussian mixture model is as follows:

[0032]

[0033] Wherein, x and y are random variables, x is defined on the real number set, and y > 0; f Xf(x) is the probability density function of the mixed normal distribution of the random variable x; Y f(y) is the probability density function of the mixed lognormal distribution of the random variable y, X f(x) and Y f(y) can be transformed by the exponential mapping; π k is the weight of the k-th Gaussian component, μ X,k and σ X,k are the mean and variance of the k-th Gaussian component respectively, k ∈ [1, K]; K is the number of mixed components.

[0034] As a further improvement of the above solution, the expression of the mapping relationship between the stress amplitude and the fatigue life is as follows:

[0035] lgS = B + DlgN

[0036] In the formula, both B and D are constants, obtained by logarithmic fitting of the S-N curve; S represents the stress, and N represents the number of fatigue cycles.

[0037] As a further improvement of the above solution,

[0038] The expression of the cumulative damage amount Damage of the bridge within the set period of T days is as follows:

[0039]

[0040] In the formula, m is the number of levels of the stress amplitude, n j is the actual number of cycles of the j-th level of the stress amplitude, N j is the fatigue life mapped by the S-N curve under the corresponding stress amplitude; the predicted fatigue life of the bridge is L pre = T / Damage.

[0041] The present invention also discloses a computer terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned intelligent prediction method for the bridge fatigue life are realized.

[0042] The present invention also discloses a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the steps of the above-mentioned intelligent prediction method for the bridge fatigue life are realized.

[0043] Compared with the prior art, the beneficial effects of the present invention are:

[0044] The intelligent prediction method for the fatigue life of bridges disclosed in the present invention can predict the fatigue life of in-service bridges based on strain monitoring data, and thus provide auxiliary decision-making for bridge management and maintenance units. On the one hand, the particle swarm optimization hybrid mutual information criterion algorithm, PSO-MIC, is proposed, which can intelligently optimize the model parameters and avoid the problems of relying on user experience and subjective judgment. On the other hand, compared with the existing single-feature model fitting methods, the present invention proposes a method for multi-feature probability density fitting using the LGM model, which effectively improves the fitting accuracy of the model, can more comprehensively capture the multi-feature problems in the actual stress spectrum, and has better fatigue prediction accuracy.

[0045] The method proposed in the present invention can automatically optimize the model parameters, thereby more accurately fitting the multi-feature stress spectrum of the bridge and predicting the fatigue life performance of the bridge; by using the PSO-MIC algorithm and the LGM model, the fatigue damage distribution of the bridge can be intuitively displayed, which is beneficial to solving the problems of fatigue damage assessment and life prediction in bridge structural health monitoring. Brief Description of the Drawings

[0046] Figure 1 It is a flowchart of the intelligent prediction method for the fatigue life of bridges in Embodiment 1 of the present invention.

[0047] Figure 2 It is the decomposition of strain data based on EMD in Embodiment 1 of the present invention;

[0048] Figure 3 It is the comparison of bridge strain data before and after processing in Embodiment 1 of the present invention;

[0049] Figure 4 It is the bridge stress spectrum diagram in Embodiment 1 of the present invention;

[0050] Figure 5 It is the optimization process of MIC in Embodiment 1 of the present invention;

[0051] Figure 6 It is the modeling result of the LGM model in Embodiment 1 of the present invention;

[0052] Figure 7 It is the comparison of the modeling results of each model in Embodiment 1 of the present invention;

[0053] Figure 8 It is the bridge S-N curve in Embodiment 1 of the present invention.

[0054] Figure 9 It is the structure diagram of the computer terminal in Embodiment 2 of the present invention. Detailed Embodiments

[0055] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0056] Embodiment 1

[0057] Please refer to Figure 1 , this embodiment provides an intelligent prediction method for bridge fatigue life, including:

[0058] S1. Preprocess the bridge strain monitoring data to obtain stress time series data.

[0059] In step S1, the specific processing process of monitoring data preprocessing is as follows:

[0060] S11: Collect the full amount of strain time series data of the front-end strain monitoring equipment on the bridge structure;

[0061] S12: Conduct a missing item test on the full amount of strain time series data. If there is data missing, perform data filling;

[0062] S13: Since the strain monitoring data is affected by the temperature effect and has fluctuating interference, it is necessary to remove the temperature effect from the data. Empirical Mode Decomposition (EMD) can be used to remove the temperature effect. The specific operation is: First, decompose the original strain time series data by EMD, and decompose it into several Intrinsic Mode Functions (IMFs) and a residual term.

[0063] Given an original signal x(t), the goal of EMD is to decompose it into several IMFs and a residual term. The decomposition formula is as follows:

[0064]

[0065] Among them, IMF i (t) is the i-th intrinsic mode function; r n (t) is the residual term; n is the number of decompositions. On this basis, identify the IMFs related to the temperature effect and remove these components from them. Then, recombine the IMFs and the residual term after removing the temperature effect to obtain strain data not affected by temperature interference. Finally, according to the elastic modulus of the bridge structure material, convert the corrected strain data into stress time series data:

[0066] σ(t) = E·ε(t)

[0067] Among them, σ(t) is the stress; E is the elastic modulus; ε(t) is the strain.

[0068] S2. Based on the stress time series data, calculate the stress spectrum of the bridge within a set period to statistically analyze the stress amplitude and its cyclic frequency of the bridge.

[0069] In the step S2, in order to obtain the standard daily stress spectrum meter of the bridge, the rain flow counting method is adopted. All local maxima (peaks) and local minima (valleys) are identified in the strain time series data. The amplitude A of each stress cycle is determined by the local maximum σ max and the local minimum σ min and is calculated as follows:

[0070]

[0071] And the average value of the stress cycle is the average of the maximum value and the minimum value:

[0072]

[0073] Calculate the period according to the time interval in the stress time series data. Assume that the stress cycle starts at time t1 and ends at time t2, then the set period T is:

[0074] T = t2 - t1

[0075] Statistically analyze all the identified stress cycles according to their amplitudes and frequencies, so as to generate a stress spectrum.

[0076] S3. Optimize the parameters of the Log-Gaussian Mixture (LGM) using the Particle Swarm Optimization for Mixed Information Criterion (PSO-MIC).

[0077] In step S3, the optimization of the parameters of the Log-Gaussian Mixture using the Particle Swarm Optimization for Mixed Information Criterion includes:

[0078] S31. Construct the Mixed Information Criterion (MIC), and its expression is:

[0079] MIC = α × AIC + (1 - α) × BIC

[0080] In the formula, AIC = 2k - 2ln(L); BIC = ln(q)k - 2ln(L); α is the weight coefficient, α ∈ [0, 1]; k is the number of parameters of the Log-Gaussian Mixture; L is the maximum likelihood estimate of the model; q is the sample size;

[0081] S32. Take the Mixed Information Criterion (MIC) as the objective function, and use the Particle Swarm Optimization algorithm to minimize and optimize the objective function. The optimization process is as follows:

[0082] Initialize the particle swarm, and update the velocity and position of the particles. The expression formulas are:

[0083] v i (t + 1)= ωv i (t)+ c1r1(P i - x i (t))+ c2r2(g - x i (t))

[0084] x i (t + 1)= x i (t)+ v i (t + 1)

[0085] In the formulas, the position x i and velocity v i of each particle i are both randomly generated; ω is the inertia weight; c1, c2 are learning factors; r1, r2 are random numbers; P i is the historical best position of particle i; g is the global best position of the population; t is the number of iteration steps;

[0086] S33. Evaluate the objective function value MIC(x i ) corresponding to the position x i of each particle, and update the historical best position P i and the global best position g; Iterate until the termination condition is met to obtain the optimal mixed model parameters that minimize MIC.

[0087] S4. Use the optimized log - Gaussian mixture model to perform probability density modeling on the stress spectrum, so as to obtain the actual number of cycles of several stress amplitudes.

[0088] In step S4, the expression of the log - Gaussian mixture model is as follows:

[0089]

[0090] The probability density function of Y can be obtained by taking the logarithmic transformation of the density of X:

[0091]

[0092] In the formulas, x, y are random variables, where x is defined on the real number set and y > 0; f X (x) is the probability density function of the mixed normal distribution of the random variable x; f Y(y) is the probability density function of the mixed lognormal distribution of the random variable y, f X (x) and f Y (y) can be transformed by the exponential mapping; π k is the weight of the k-th Gaussian component, μ X,k and σ X,k are the mean and variance of the k-th Gaussian component respectively, k ∈ [1, K]; K is the number of mixture components.

[0093] S5. Establish the mapping relationship between the stress amplitude and the fatigue life according to the S-N curve of the bridge component.

[0094] In the step S5, the S-N curve of the bridge structure can be obtained by referring to relevant materials to represent the relationship between the stress S and the fatigue cycle times N:

[0095] lgS = B + DlgN

[0096] In the formula, both B and D are constants obtained by logarithmic fitting of the S-N curve; S represents the stress and N represents the fatigue cycle times.

[0097] S6. Based on the actual cycle times of each stress amplitude level and the corresponding fatigue life, predict the cumulative damage amount of the bridge within the set period based on the Miner criterion, so as to obtain the predicted fatigue life of the bridge.

[0098] In step S6, the fatigue life of the bridge can be predicted based on the Miner criterion. Among them, this criterion is a linear cumulative damage mathematical expression based on the principle of the net work absorbed during material damage. First, the stress amplitude levels need to be divided into several levels, each level being:

[0099] Δσ1, Δσ2, Δσ3, …, Δσ m

[0100] The corresponding cycle times are:

[0101] n1, n2, n3, …, n m

[0102] The corresponding fatigue life is:

[0103] N1, N2, N3, …, N m

[0104] Therefore, the damage rate occupied by each level is:

[0105]

[0106] The expression of the cumulative damage amount Damage of the bridge within the set period of T days is as follows:

[0107]

[0108] Wherein, m is the number of levels of stress amplitude, and n j is the actual number of cycles of the j-th level of stress amplitude, and N j is the fatigue life mapped by the S-N curve under the corresponding stress amplitude; the predicted fatigue life of the bridge is L pre = T / Damage.

[0109] In this embodiment, taking an actual bridge as an example, the all-weather real-time strain monitoring data of the bridge is selected, and the steps of the process Figure 1 are used to verify the embodiment.

[0110] I. Pretreatment of monitoring data

[0111] 1. Conduct a comprehensive timestamp check on the strain monitoring data of the bridge to ensure the integrity of the data. For the missing data due to sensor failure or signal loss during the monitoring process, the average value of adjacent data is used for filling to ensure data continuity and the reliability of subsequent analysis;

[0112] 2. Use the EMD method to decompose the filled strain monitoring data into several IMFs to facilitate further analysis of the influence of each frequency component on the strain, as Figure 2 shown; 3. Generally, the last IMF end represents the long-term trend in the strain data, usually reflecting the influence of temperature changes. Therefore, discard the IMF end as the characterization item of the temperature effect, and only retain the short-term fluctuation IMFs from IMF1 to IMF end-1 . Then, recombine these retained IMFs to restore the strain monitoring data after removing the temperature effect, as Figure 3 shown.

[0113] II. Stress spectrum analysis

[0114] After the data pretreatment is completed, apply the rain-flow counting method to the processed strain monitoring data for statistical analysis of the stress range and the number of cycles. The rain-flow counting method can effectively identify the number of stress cycles and the amplitude, which is crucial for the fatigue life analysis of the bridge structure. The stress spectrum is as Figure 4 shown.

[0115] III. Parameter optimization based on PSO-MIC

[0116] According to the formula proposed in the present invention, use MIC as the objective function for modeling, and adopt the PSO optimization algorithm to search for the global optimal solution. In this process, the PSO algorithm automatically adjusts the parameter positions of each particle and gradually approaches the minimum value of MIC. The results are as Figure 5As shown, at the 5th iteration, the value of the objective function reaches the optimum, indicating that the optimal parameter of the LGM model is 5.

[0117] IV. Probability Density Modeling of the LGM Model

[0118] After obtaining the optimal parameters, the probability density of the stress spectrum of the bridge structure was modeled based on the LGM model, as Figure 6 shown. On this basis, the results of the LGM model were compared and analyzed with the normal distribution and gamma distribution of traditional single features, as Figure 7 shown. From the comparison of the K-S statistical results, it can be seen that the stress spectrum and its probability density have significant multi-feature characteristics. Traditional single-feature models such as the normal distribution and gamma distribution cannot effectively capture these multi-peak features, resulting in a poor fitting effect, as shown in Table 1.

[0119] Table 1. Comparison of the fitting effects of three groups of different models

[0120] Model Name K-S Statistic p-Value Evaluation Normal Distribution Model 0.14815 0.90525 Fair Fitting Effect Gamma Distribution Model 0.18519 0.69758 Poor Fitting Effect Logarithmic Gaussian Mixture Model 0.11111 0.99364 Best Fitting Effect

[0121] V. Establishing the S-N Curve of the Bridge

[0122] To accurately evaluate the fatigue performance of the bridge under different stress amplitudes, the fatigue detail classification in the AASHTO standard was incorporated in this invention. By referring to relevant literature and standard specifications, the fatigue detail level corresponding to the bridge structure was determined to be Class E’, where B = 3.70453; D = -1 / 3. Based on this, the corresponding S-N curve was constructed, as Figure 8 shown.

[0123] VI. Fatigue Life Prediction Based on the Miner Criterion

[0124] Based on the S-N curve obtained in step S5 and the probability density function fitted in step S4, the Miner criterion was used to predict the fatigue life of the bridge. According to the Miner criterion, the cumulative damage amount of the bridge under the standard daily stress spectrum was calculated to be D = 1.158×10 -5 . By taking the reciprocal of this value, the fatigue life of the bridge was obtained as 86,325 days, approximately 236 years.

[0125] It should be noted that the prediction results of the bridge fatigue life are affected by the number of cycles and stress range in the standard daily stress spectrum. To improve the prediction accuracy, the strain data for the whole year can be fully utilized to average the stress spectrum during calculation, and the influence of traffic flow growth on the stress spectrum can be considered. This will help to more accurately reflect the fatigue situation under actual use conditions, thus providing a more reliable life prediction.

[0126] Example 2

[0127] This embodiment provides a computer terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the intelligent prediction method for the fatigue life of a bridge as described in Embodiment 1 are implemented.

[0128] As Figure 9 shown, the computer terminal provided in this embodiment includes: at least one processor 101 and a memory 102 connected to at least one processor 101. In this embodiment, the specific connection medium between the processor 101 and the memory 102 is not limited. Figure 9 Here, it is taken as an example that the processor 101 and the memory 102 are connected through a bus 100. The bus 100 is represented by a thick line Figure 9 in. The connection manners between other components are only for illustrative purposes and are not to be taken as limiting. The bus 100 can be divided into an address bus, a data bus, a control bus, etc. For the sake of easy representation, Figure 9 it is only represented by a thick line in, but it does not mean that there is only one bus or one type of bus. Alternatively, the processor 101 can also be called a controller, and there is no limit to the name.

[0129] In this embodiment, the memory 102 stores instructions executable by at least one processor 101. By executing the instructions stored in the memory 102, at least one processor 101 can execute the foregoing method.

[0130] Among them, the processor 101 is the control center of the device. It can connect various parts of the entire control device through various interfaces and lines. By running or executing the instructions stored in the memory 102 and calling the data stored in the memory 102, various functions of the device and process data, so as to monitor the device as a whole.

[0131] In a possible design, the processor 101 may include one or more processing units. The processor 101 may integrate an application processor and a modem processor. Among them, the application processor mainly processes the operating system, user interface, application programs, etc., and the modem processor mainly processes wireless communication. It can be understood that the foregoing modem processor may not be integrated into the processor 101. In some embodiments, the processor 101 and the memory 102 can be implemented on the same chip. In some embodiments, they can also be separately implemented on independent chips.

[0132] The processor 101 can be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments. The general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of the intelligent prediction method for the bridge fatigue life disclosed in conjunction with Embodiment 1 can be directly embodied as being executed by a hardware processor, or executed by a combination of hardware and software modules in the processor 101.

[0133] The memory 102, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The memory 102 can include at least one type of storage medium. For example, it can include flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (RAM), a static random access memory (SRAM), a programmable read-only memory (PROM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic memory, a magnetic disk, an optical disk, and so on. The memory 102 is any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 102 in this embodiment can also be a circuit or any other device capable of implementing a storage function, for storing program instructions and / or data.

[0134] By programming the design of the processor 101, the code corresponding to the security verification method introduced in the foregoing embodiments can be solidified into the chip, so that the chip can execute Figure 1 the steps of the intelligent prediction method for the bridge fatigue life as shown. How to program the design of the processor 101 is a well-known technology to those skilled in the art and will not be elaborated here.

[0135] Embodiment 3

[0136] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the steps of the intelligent prediction method for the bridge fatigue life as described in Embodiment 1.

[0137] The computer-readable storage medium may include flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the storage medium may be an internal storage unit of a computer device, such as the hard disk or memory of the computer device. In other embodiments, the storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the computer device. Of course, the storage medium may also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the memory is generally used to store the operating system and various application software installed in the computer device. In addition, the memory may also be used to temporarily store various data that have been output or are to be output.

[0138] As described above, only the preferred specific embodiments of the present invention are given, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent prediction method for the fatigue life of bridges, characterized in that Including: Calculating the stress spectrum of the bridge within a set period to statistically analyze the stress amplitude of the bridge and its cyclic frequency; Optimizing the parameters of the lognormal mixture model using a particle swarm improved hybrid information criterion, and performing probability density modeling on the stress spectrum using the optimized lognormal mixture model to obtain the actual number of cycles for several levels of stress amplitude; Establishing the mapping relationship between stress amplitude and fatigue life based on the S-N curve of the bridge component; Based on the actual number of cycles for each level of stress amplitude and the corresponding fatigue life, predicting the cumulative damage amount of the bridge within the set period based on the Miner criterion, thereby obtaining the predicted fatigue life of the bridge.

2. The intelligent prediction method for the fatigue life of a bridge according to claim 1, characterized in that The calculation of the stress spectrum of the bridge within the set period includes: Preprocessing the bridge strain monitoring data to obtain stress time series data; Generating the stress spectrum based on the stress time series data through the rainflow counting method.

3. The intelligent prediction method for the fatigue life of a bridge according to claim 2, wherein, The preprocessing includes: Collecting the full amount of strain time series data of the strain monitoring equipment on the bridge structure, detecting and filling in the missing data to obtain the original strain data; Decomposing the original strain data into several intrinsic mode functions and a residual term through empirical mode decomposition, removing the intrinsic mode functions related to the temperature effect, reconstructing the temperature-corrected strain data, and converting the corrected strain data into stress time series data according to the elastic modulus of the bridge structure material.

4. The intelligent prediction method for the fatigue life of a bridge according to claim 2, wherein The generation of the stress spectrum through the rainflow counting method includes: Identify local maxima and local minima in stress time-series data; calculate the amplitude of each stress cycle Average value and set the period T = t2 - t1; where, σ max and σ min are the local maximum and local minimum in the stress time-series data respectively; t2 and t1 are the start time and end time of the stress cycle respectively; Statistically analyzing all identified stress cycles according to their amplitude and frequency to generate the stress spectrum.

5. The intelligent prediction method for the fatigue life of a bridge according to claim 1, characterized in that The optimization of the parameters of the lognormal mixture model using a particle swarm improved hybrid information criterion includes: Constructing the hybrid information criterion MIC, and its expression is: MIC = α × AIC + (1 - α) × BIC In the formula, AIC = 2k - 2ln(L); BIC = ln(q)k - 2ln(L); α is the weight coefficient, α ∈ [0, 1]; k is the number of parameters of the lognormal mixture model, L is the maximum likelihood estimate value of the model; q is the sample size; Taking the hybrid information criterion MIC as the objective function, and using the particle swarm algorithm to perform minimum optimization on the objective function. The optimization process is: Initializing the particle swarm, and updating the velocity and position of the particles. The expression formula is: v i (t + 1)= ωv i (t)+ c1r1(P i -x i (t))+ c2r2(g - x i (t)) x i (t + 1)= x i (t)+ v i (t + 1) where the position x of each particle i i and the velocity v i are both randomly generated; ω is the inertia weight; c1, c2 are learning factors; r1, r2 are random numbers; P i is the historical best position of particle i; g is the global best position of the swarm; t is the number of iteration steps; Evaluate the position x of each particle i The corresponding objective function value MIC(x i ), update the historical best position P of the particle i and the global best position g; Iterate until the termination condition is met to obtain the optimal hybrid model parameters that minimize MIC.

6. The intelligent prediction method for the fatigue life of a bridge according to claim 1, characterized in that, The expression of the lognormal mixture model is as follows: where x and y are random variables, x is defined on the set of real numbers, and y > 0; f X (x) is the probability density function of the mixture normal distribution of the random variable x; f Y (y) is the probability density function of the mixture lognormal distribution of the random variable y, f X (x) and f Y (y) can be transformed by the exponential mapping; π k is the weight of the k-th Gaussian component, μ X,k and σ X,k are the mean and variance of the k-th Gaussian component respectively, k ∈ [1, K]; K is the number of mixture components.

7. The intelligent prediction method for the fatigue life of a bridge according to claim 1, wherein The expression of the mapping relationship between stress amplitude and fatigue life is as follows: lgS = B + DlgN In the formula, both B and D are constants, obtained by logarithmic fitting of the S-N curve; S represents stress, and N represents the number of fatigue cycles.

8. The intelligent prediction method for the fatigue life of a bridge according to claim 1, wherein The expression of the cumulative damage amount Damage of the bridge within the set period T days is as follows: where m is the number of levels of stress amplitude, and n j is the actual number of cycles at the j-th level of stress amplitude, and N j is the fatigue life mapped by the S-N curve under the corresponding stress amplitude; the predicted fatigue life of the bridge is L pre = T / Damage.

9. A computer terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it realizes the steps of the intelligent prediction method for the bridge fatigue life as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it realizes the steps of the intelligent prediction method for the bridge fatigue life as described in any one of claims 1 to 8.

Images