Method and system for establishing a degree of defect evaluation model for a concrete-filled steel tube structure based on ultrasonic detection data

Abstract

The application relates to a method and system for evaluating the degree of defects such as voids and debonding of a steel pipe concrete structure, which comprises the following steps: acquiring test detection data and engineering parameters of a steel pipe concrete arch bridge construction, wherein the detection data and engineering parameters comprise waveform and spectrum diagrams, ultrasonic wave velocity V, concrete 28d compressive strength f c , age t, void rate delta; qualitatively analyzing the defect type according to the waveform and spectrum diagram information, inputting the detection data and engineering parameters into a deterministic model to generate deterministic model parameters, and then inputting the detection data, engineering parameters and model parameters into a pre-built model; the built model is obtained by iteratively optimizing the model using an interactive parallel multi-process algorithm; the interactive model outputs results as dense, void and debonding according to the to-be-detected data, and the degree of voids and debonding is obtained through model inversion, which can further guide the treatment of voids and debonding grouting in the construction process; the method is simple in steps, convenient to operate and good in effect.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of steel pipe concrete structure, and particularly relates to a method and system for establishing a steel pipe concrete structure void and debonding defect degree evaluation model based on ultrasonic detection data. BACKGROUND

[0002] The synergistic performance of steel-concrete composite structure has become a key factor restricting the development of steel pipe concrete arch bridges towards larger spans. However, due to factors such as concrete material quality, pouring process, temperature change, etc., the core concrete is prone to different degrees of debonding and voiding. The existing steel pipe concrete still has the following deficiencies in quality detection.

[0003] The massive "first-hand" detection data has low utilization rate and insufficient interpretability, and the existing evaluation system lacks a set of key technologies with high reliability and strong adaptability.

[0004] Due to the heterogeneous nature of concrete materials and the influence of construction deviations, the influencing factors have significant objective uncertainty. The existing model cannot quantitatively describe the degree of voiding and debonding defects, and cannot accurately establish the relationship between the ultrasonic wave velocity for detection and the compaction condition of the concrete in the pipe.

[0005] In view of this, in order to accurately and comprehensively analyze the compaction state of steel pipe concrete, Metropolis-Hastings algorithm, adaptive Metropolis-Hastings algorithm and other random uncertainty models have been applied in the field of civil engineering, and it is verified that the random uncertainty model may be a suitable tool for solving the field of concrete materials.

[0006] However, the existing random uncertainty model is limited by the defects of the prediction model itself, such as the defect of not setting the adaptive iteration interval, which makes the result fall into a local optimal solution. The second-order acceptance probability criterion improves the rejection rate of model iteration, greatly reduces the calculation efficiency, and single-process iterative analysis cannot guarantee that the iteration process is a global optimal solution, which may result in extremely low accuracy of the result.

[0007] At present, some more advanced random uncertainty models such as DRAM model have gradually achieved successful application in various industries by virtue of their advantages such as fast processing speed, good adaptive iteration effect, and good control of overfitting effect.

[0008] However, the DRAM model has the disadvantages of complex calculation and low calculation efficiency, and the single-process manual adjustment of parameters may not necessarily find a global optimal solution. In view of this, how to fully utilize the number of computer processor cores to improve the calculation efficiency and accuracy has become an urgent problem to be solved to bring the DRAM model to its full potential. SUMMARY

[0009] The present application aims at the problems existing in the prior art, and provides a method and system for establishing a steel pipe concrete structure void and debonding defect degree evaluation model based on ultrasonic detection data, and a method for realizing automatic optimization of DRAM model iterative analysis by means of an interactive parallel multi-process algorithm.

[0010] In order to achieve the above-mentioned purpose, the technical scheme adopted by the present application is as follows:

[0011] In the first aspect, the present application provides a method for establishing a steel pipe concrete structure void and debonding defect degree evaluation model based on ultrasonic detection data, comprising the following steps:

[0012] Obtaining test detection data and engineering parameters of a steel pipe concrete arch bridge construction, the detection data and engineering parameters including waveform and spectrum diagrams, ultrasonic wave velocity V, concrete 28d compressive strength f c , age t, and void rate δ;

[0013] After qualitatively evaluating the defect type according to the waveform and spectrum diagram information, inputting the detection data and engineering parameters into a deterministic model to generate deterministic model parameters, the deterministic quantitative model comprising:

[0014] A void defect model:

[0015] A debonding defect model:

[0016] In the formula, V is a predicted value of ultrasonic wave velocity, f c is a concrete 28d compressive strength, t is an age, δ is a void rate, i.e., a ratio of a void height to a steel pipe diameter, l is a debonding arc length, L is a steel pipe inner circumference, and α=[α1, α2, α3, α4, α5, α6] T is a deterministic model parameter;

[0017] Inputting the engineering parameters and measured ultrasonic wave velocity into a pre-built interactive parallel multi-chain DRAM model; the pre-built interactive parallel multi-chain DRAM model is a model obtained after iterative optimization of the DRAM model by using a parallel multi-process algorithm, and the model comprising:

[0018] A void defect model:

[0019] A debonding defect model:

[0020] In the formula, θ=[θ1, θ2, θ3, θ4, θ5, θ6] TThe measured ultrasonic detection data is updated to the uncertainty model parameter vector, β1ε1 is objective uncertainty caused by geometric, material, environmental parameters and the like, β2ε2 is subjective uncertainty caused by incomplete consideration factors, β1 and β2 are standard normal random distribution values, and ε1 and ε2 are system errors caused by objective and subjective factors respectively;

[0021] According to the interactive parallel multi-chain DRAM model, and in combination with the cases that the confidence level (1-a) is 0.50 and 0.95 respectively, the degrees of the void and debonding defects are quantitatively evaluated.

[0022] As the preferred technical scheme of the present application, the qualitative defect type evaluation on the waveform diagram and the spectrum diagram information comprises the following steps:

[0023] A1, detecting by using the non-metal ultrasonic detector in the opposite measurement method, collecting the waveform and spectrum diagram information after detection;

[0024] A2, when the first wave of the waveform diagram is clear, the first wave amplitude is large, the waveform period is a half envelope diagram, the waveform has no obvious distortion, and the main frequency number of the spectrum diagram is less than or equal to 2, the dense section can be qualitatively judged;

[0025] A3, when the first wave of the waveform diagram is not clear, the first wave amplitude is small, the wave peak and valley interval is small, the waveform has obvious distortion, and the main frequency number of the spectrum diagram is less than or equal to 2, the void can be qualitatively judged;

[0026] A4, when the first wave of the waveform diagram is not clear, the first wave amplitude is small, the wave peak and valley interval is large, the waveform has obvious distortion, and the main frequency number of the spectrum diagram is greater than 2, the debonding can be qualitatively judged.

[0027] As the preferred technical scheme of the present application, the quantitative evaluation on the defect type according to the interactive parallel multi-chain DRAM model comprises the following steps:

[0028] B1, inputting the engineering parameters, the measured ultrasonic wave velocity and the model parameters α=[α1, α2, α3, α4, α5, α6] T to the pre-built interactive parallel multi-chain DRAM model as prior information;

[0029] B2, performing iterative update training on the DRAM model by using the interactive parallel multi-process algorithm, and preferably selecting, in the first iteration, the parallel chain with better Markov chain stability, lower rejection rate and higher calculation efficiency;

[0030] B3, performing second iteration update on the Markov chain selected in the first iteration, so as to exclude the case of "false steady state", when the steady state index of the Markov chain does not change in the interval of 10000 iterations, the second iteration is terminated, the best parameter combination of the DRAM model is determined, and the updated DRAM model is obtained.

[0031] B4. Based on the parallel multi-chain DRAM model, combined with the confidence levels (1-a) of 0.50 and 0.95, the debonding rate and debonding rate of each section are quantitatively calculated to evaluate the compaction performance of the concrete in the tube. The specific settings are:

[0032] When the test data is within the 50% confidence interval, it is an acceptable section for void and debonding defects and does not require grouting reinforcement treatment;

[0033] When the test data is within the 95% confidence interval, it is a critical debonding and delamination defect section, and grouting reinforcement treatment is carried out as needed;

[0034] When the test data is outside the 95% confidence interval and the void and debonding defect sections are unacceptable, grouting reinforcement treatment shall be carried out in a timely manner according to the quantitatively calculated void rate and debonding rate.

[0035] As a preferred technical solution of the present invention, the interactive parallel multi-chain includes:

[0036] C1, given model parameters α = [α1, α2, α3, α4, α5, α6] T The initial value and iterative initial value of both obey the standard random normal distribution, and the iterative standard deviation is taken as large as possible on the basis of satisfying the physical laws. Here, 10000 is used.

[0037] C2, set the number of parallel multi-chain chains n, the number of chains ≥ 8;

[0038] C3, in the parallel calculation process of n Markov chains, when a Markov chain reaches a steady state first, automatically select the current information corresponding to the Markov chain, including: the mean, standard deviation and acceptance probability of the current iteration;

[0039] C4, after taking the current information corresponding to the Markov chain that reaches the steady state first as the current information input of the remaining n-1 Markov chains, continue to iterate;

[0040] C5, until the total number of iterations is met, a group of parallel chains with the highest acceptance probability and computational efficiency are selected from the n Markov chains, which have good Markov chain stability, low rejection rate and high computational efficiency.

[0041] In a second aspect, the present invention further provides a system for establishing a model for assessing the degree of voids and debonding defects in concrete-filled steel tube structures based on ultrasonic testing data, comprising:

[0042] a memory having a computer program stored thereon;

[0043] 8 A processor with 8 or more cores for executing a program in a memory to implement the method of establishing a defect degree evaluation model of a concrete-filled steel tube structure based on ultrasonic detection data according to any one of the preceding aspects.

[0044] In a third aspect, the application further provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method of establishing a defect degree evaluation model of a concrete-filled steel tube structure based on ultrasonic detection data according to any one of the preceding aspects.

[0045] In summary, due to the adoption of the technical solutions described above, the application has the following beneficial effects:

[0046] 1. The method and system of establishing a defect degree evaluation model of a concrete-filled steel tube structure based on ultrasonic detection data according to the application qualitatively judge the defect type based on ultrasonic detection data in combination with a waveform diagram and a spectrum diagram, not only solve the problems of traditional data analysis and processing methods, such as being subject to personnel experience, work efficiency and the like, but also have the outstanding advantages of being intuitive, fast and high-precision;

[0047] 2. The method and system of establishing a defect degree evaluation model of a concrete-filled steel tube structure based on ultrasonic detection data according to the application make use of the method of automatic optimization of a DRAM model Markov chain iteration by means of an interactive parallel multi-process algorithm, update the model parameters by using prior information through the adoption of an adaptive delay rejection algorithm, i.e. a DRAM model, and optimize the method of Markov chain iteration optimization of the DRAM algorithm by means of the multi-core synchronous operation and good global search capability of the interactive parallel multi-process algorithm, thereby making up for the defects of the DRAM model, such as more model parameters, slow convergence, and the fact that the human parameter adjustment is not a global optimal solution to some extent, and improving the efficiency and precision of the application of the DRAM model in the field of arch bridge in-pipe concrete compactness monitoring technology;

[0048] 3. The method and system of establishing a defect degree evaluation model of a concrete-filled steel tube structure based on ultrasonic detection data according to the application combine the powerful nonlinear fitting capability of the DRAM model and the operation mechanism of the interactive multi-process algorithm, establish the method and system of establishing a defect degree evaluation model of a concrete-filled steel tube structure based on ultrasonic detection data, and overcome the fact that the traditional deterministic prediction model cannot reasonably consider the objective uncertainty of factors such as material characteristic parameters of the concrete-filled steel tube, construction and construction pouring process, and cannot consider artificial measurement errors and subjective uncertainty caused by factors such as incomplete consideration of factors or introduction of model assumptions in the model establishment process, improve the discrimination calculation and efficiency by continuously updating the model parameters based on the detection data in view of different design and construction parameters of different concrete-filled steel tube arch bridges, and the method and system are more suitable for practical engineering application and have stronger guidance;

[0049] 4. The method and system for establishing a model for evaluating the degree of voids and debonding defects in steel tube concrete structures based on ultrasonic detection data described in the present invention are compared with the prediction accuracy of other random uncertainty models used in the field of predicting the compaction performance of concrete in arch bridge tubes. Except for the interactive parallel multi-chain DRAM model and DRAM model research of this application, the rest are non-adaptive models, that is, in the iterative process, the covariance matrix, reception probability and other parameter adjustment indicators are not adjusted with the stability of the Markov chain iteration. The MH model (Metropolis-Hastings algorithm) performs iterative analysis by using deterministic fitting parameters as the initial value of the iteration, and the Markov chain has high stability. However, the disadvantage of this model is that the Markov chain often has a "false steady state", that is, it falls into the misunderstanding of the local optimal solution, resulting in the calculation results being only suitable for predicting the ultrasonic wave velocity values ​​of a small number of measuring points that are less affected by subjective and objective uncertainties. The AM model (Adaptive The Metropolis-Hastings algorithm builds on the MH model by introducing an adaptive covariance matrix for iterative updates, which introduces a degree of randomness. However, its overall performance is similar to that of the MH model. The interactive parallel multi-chain DRAM model uses a parallel multi-process algorithm to generalize the parameters of the DRAM model's initial values ​​and iterative process. It also uses automatic optimization to find the Markov chain that first reaches a relatively stable state and then integrates it with the remaining chains. Compared to existing models, this model significantly improves computational accuracy and efficiency, achieving more accurate predictions in the shortest possible time. It is more advanced in terms of both model complexity and performance.

[0050] 5. The method and system for establishing a model for evaluating the degree of voids and debonding in steel tube concrete structures based on ultrasonic detection data described in the present invention use an interactive parallel multi-process algorithm for iterative analysis, establish multiple iteration initial values ​​based on the principle of randomness, and replace the model parameter values ​​of the remaining parallel chains with the model parameter values ​​of the Markov chain that first reaches a better "steady state" during the iteration process of each parallel chain, thereby further reducing the waste of computing resources caused by the error in the iteration direction of the parallel chain; in addition, by setting a threshold condition for terminating the iteration, the iterative analysis is terminated after the iterative steady state reaches the threshold, thereby further improving the computing efficiency while maintaining the prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is the waveform diagram of the steel tube concrete under compaction;

[0052] Figure 2 This is the spectrum diagram of steel tube concrete under compaction;

[0053] Figure 3 This is the waveform diagram of the void defect of concrete-filled steel tube;

[0054] Figure 4 Waveform diagram for steel pipe concrete debonding defect;

[0055] Figure 5 Waveform diagram for steel pipe concrete debonding defect;

[0056] Figure 6 Waveform diagram for steel pipe concrete debonding defect;

[0057] Figure 7 Pipe concrete defect identification diagram based on 95% and 50% confidence interval;

[0058] Figure 8 Optimal Markov chain iteration path for parallel multi-chain DRAM model;

[0059] Figure 9 Second iteration analysis diagram for parallel multi-chain DRAM model;

[0060] Figure 10 Markov chain model parameter information diagram for second iteration;

[0061] Figure 11 Interactive parallel multi-chain DRAM algorithm code diagram;

[0062] Figure 12 Flowchart of the method of the present application. DETAILED DESCRIPTION

[0063] The present application will be described in detail below with reference to the accompanying drawings.

[0064] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0065] Example 1

[0066] As shown in Figure 12 , the method for establishing a steel pipe concrete structure debonding and debonding defect degree evaluation model based on ultrasonic detection data according to the present application comprises the following steps:

[0067] Step one, collect detection data, and qualitatively evaluate the arch rib cross section defect type of the steel pipe concrete arch bridge.

[0068] Obtain the test detection data and engineering parameters of the steel pipe concrete arch bridge construction, and the detection data and engineering parameters include the waveform diagram and frequency spectrum diagram of the detected ultrasonic wave, the ultrasonic wave velocity V, the compressive strength f c of the concrete (28 days), the age t, and the debonding rate δ.

[0069] The detection is performed by a non-metal ultrasonic detector using the opposite measurement method, the waveform and spectrum information after detection are collected, and the qualitative defect type evaluation is performed according to the waveform and spectrum information.

[0070] As shown in Figure 1 and Figure 2 , when the first wave of the waveform diagram is clear, the first wave amplitude is large, the waveform is a half envelope diagram, the waveform has no obvious distortion, and the main frequency number of the spectrum diagram is ≤2, the dense section can be qualitatively judged.

[0071] As shown in Figure 3 and Figure 4 , when the first wave of the waveform diagram is not clear, the first wave amplitude is small, the wave peak and valley interval is small, the waveform has obvious distortion, and the main frequency number of the spectrum diagram is only 1, it can be qualitatively judged as void.

[0072] As shown in Figure 5 and Figure 6 , when the first wave of the waveform diagram is not clear, the first wave amplitude is small, the wave peak and valley interval is large, the waveform has obvious distortion, and the main frequency number of the spectrum diagram is >2, it can be qualitatively judged as debonding.

[0073] Step two, establish a quantitative evaluation model for the void and debonding defects of the concrete in the steel pipe of the concrete-filled steel pipe arch bridge.

[0074] According to the qualitative evaluation of the defect type according to the waveform and spectrum information, the defect degree is further quantitatively analyzed, the detection data and engineering parameters are input into the deterministic model to generate deterministic model parameters, and the deterministic quantitative model includes:

[0075] Void defect model:

[0076] Debonding defect model:

[0077] In the formula, V is the predicted value of the ultrasonic wave speed;

[0078] f c is the 28d compressive strength of the concrete;

[0079] t is the age;

[0080] δ is the void rate, i.e. the ratio of the void height to the diameter of the steel pipe;

[0081] l is the debonding arc length;

[0082] L is the inner circumference of the steel pipe;

[0083] α = [α1, α2, α3, α4, α5, α6] T is the deterministic model parameter.

[0084] The engineering parameters, the measured ultrasonic wave speed and alpha = [alpha1, alpha2, alpha3, alpha4, alpha5, alpha6] T to the pre-built interactive parallel multi-chain DRAM model; the pre-built interactive parallel multi-chain DRAM model is a model obtained by iteratively optimizing the DRAM model using a parallel multi-process algorithm, and the model includes:

[0085] Void defect model:

[0086] Debinding defect model:

[0087] In the formula, theta = [theta1, theta2, theta3, theta4, theta5, theta6] T The uncertainty model parameter vector is updated for the measured ultrasonic wave detection data;

[0088] beta1epsilon1 is the objective uncertainty caused by geometric, material and environmental parameters;

[0089] Beta2epsilon2 is the subjective uncertainty caused by incomplete consideration factors;

[0090] Beta1 and beta2 are standard normal random distribution values;

[0091] Epsilon1 and epsilon2 are system errors caused by objective and subjective factors respectively.

[0092] Step three, adaptive iterative updating of model parameters based on interactive parallel multi-chain DRAM model.

[0093] This embodiment is based on the adaptive iterative updating of model parameters based on the interactive parallel multi-chain DRAM model, and further can establish an adaptive defect evaluation model that matches each arch bridge "unique", and the method includes the following steps:

[0094] A, the engineering parameters, the measured ultrasonic wave speed and model parameters alpha = [alpha1, alpha2, alpha3, alpha4, alpha5, alpha6] T The prior information is input to the pre-built interactive parallel multi-chain DRAM model.

[0095] B, the DRAM model is iteratively updated and trained by the interactive parallel multi-process algorithm, and in the first iteration, the parallel chain with better Markov chain stability, lower rejection rate and higher calculation efficiency is selected, and the specific steps are as follows.

[0096] B1, the iteration parameters (adaptive step size Delta t, burning period T, receiving probability order N, initial error epsilon, iteration number A) of the DRAM model are given, and the value range of the iteration parameters is set.

[0097] B2, Given the initial parameters θ0of parallel multi-process algorithm, here let θ0obeys the standard random normal distribution.

[0098] B3, Given the initial covariance matrix C 1 = C 0 , assuming that the parameters of the probability model are independent of each other, and according to the initial standard deviation θpre-set, the initial covariance matrix is constructed, and the adaptive iteration step is set, and the covariance matrix is updated as follows:

[0099] C t = s d cov(θ t-99 , θ t-98 , …, θ t ) + s d εI d , t ≥ N0

[0100] In the formula, ε is a small variable, s d is a proportional scaling factor, I d is a d-dimensional unit matrix; cov(θ t-99 , θ t-98 , …, θ t ) is the covariance matrix of the historical sample.

[0101] B4, Given the prior value, here let the engineering parameters and the measured ultrasonic detection data of the site test cubic block be the prior value of the model calculation.

[0102] B5, According to the following steps, until the rejection rate is low and tends to be stable Markov chain.

[0103] B51, Set the number of first iteration parallel chains n groups (n ≥ 8), then calculate the initial parameters of n groups of different values according to step B2, then construct the recommended value θ t from the first order proposal distribution q(θ i , ·), that is, according to θ i = θ t + σZ, θ i is obtained, wherein Z is a random number obeying the standard normal distribution, σ is the standard deviation of the first order proposal distribution q(θ t , ·), and · represents the variable of the first order proposal distribution q(θ t , ·).

[0104] B52, Calculate the acceptance probability α1(θ t , θ i ) of the first order:

[0105]

[0106] In the formula, Q(α *) and Q (a t ) is a prior distribution function.

[0107] B53, a number μ is randomly selected from 0 to 1, if μ ≤ a1 (θ t , θ i ), the model parameter suggestion value θ i is accepted, i.e. a t+1 = a i , and then step B57 is entered, otherwise step B54 is entered.

[0108] B54, the second order acceptance probability discrimination is entered, and the covariance matrix C t is adjusted, a new suggestion value θ t is constructed from the second order suggestion distribution q (θ i , θ 2i , ·), i.e. θ 2i is calculated according to θ t = θ 2i + Zσ', wherein σ' is the standard deviation of the second order suggestion distribution q (θ t , θ i , ·), and • represents the variable of the second order suggestion distribution q (θ t , θ i , ·).

[0109]

[0110] B55, similarly, a number u is randomly selected from 0 to 1, if u ≤ a1 (θ t , θ i , θ 2i ), the model parameter suggestion value θ 2i is accepted, i.e. θ t+1 = θ 2i , and then step B57 is entered, otherwise step B56 is entered.

[0111] B56, by analogy, the N (N > 2) order acceptance probability discrimination is entered, and the corresponding covariance matrix C t is adjusted, a new model parameter candidate value θ t is constructed from the N (N > 2) order suggestion distribution q (θ i , θ Ni ,..., ·), i.e. θ Ni is calculated according to θ t = θ Ni + Zσ", wherein σ" is the standard deviation of the N (N > 2) order suggestion distribution q (θ t , θ i ,..., ·).

[0112]

[0113] Similarly, a number u is randomly drawn from 0 to 1, and if u≤α N (θ t ,θ * ,θ 2i ,...,θ Ni ), then accept the recommended value of the model parameter θ Ni , that is, let θ t+1 =α Ni , then go to step B57, otherwise let N=N+1 and repeat step B56.

[0114] B57. If t+1>n, stop iterating, otherwise increase the number of iterations and return to step B51, where θ t+1 is the current variable.

[0115] B58. During the iteration of the parallel multi-process algorithm, the Markov chain that first reaches a steady state can be used as the initial value for recalculating the remaining n-1 groups of chains.

[0116] C. The Markov chain optimized in the first iteration is updated for a second iteration to eliminate the "false steady state" situation. When the steady state indicator of the Markov chain does not change within an interval of 10,000 iterations, the second iteration is terminated, the optimal parameter combination of the DRAM model is determined, and the updated DRAM model is obtained.

[0117] Step 4: Quantitatively evaluate the degree of void and debonding defects.

[0118] According to the parallel multi-chain DRAM model, combined with the confidence levels (1-a) of 0.50 and 0.95 respectively, the debonding rate and debonding rate of each section are quantitatively calculated to evaluate the compaction performance of the concrete in the tube, such as Figure 7 As shown, the specific settings are:

[0119] When the test data is within the 50% confidence interval, it is an acceptable section for void and debonding defects and does not require grouting reinforcement treatment;

[0120] When the test data is within the 95% confidence interval, it is a critical debonding and delamination defect section, and grouting reinforcement treatment is carried out as needed;

[0121] When the test data is outside the 95% confidence interval and the void and debonding defect sections are unacceptable, grouting reinforcement treatment shall be carried out in a timely manner according to the quantitatively calculated void rate and debonding rate.

[0122] Example 2

[0123] The method for establishing a steel pipe concrete structure void and debonding defect degree evaluation model based on ultrasonic detection data, in the embodiment 1, the interactive parallel multi-chain of step B51 is n groups of step B5 synchronous, specifically including:

[0124] B511, the initial value and the iterative initial value of the given model parameter alpha = [alpha1, alpha2, alpha3, alpha4, alpha5, alpha6] are subject to standard random normal distribution, and the iterative standard deviation is as large as possible on the basis of meeting the physical law, here 10000 is adopted. T

[0125] B512, set the number of chains n of parallel multi-chain, the number of chains is greater than or equal to 8.

[0126] B513, in the parallel calculation process of n Markov chains, when a certain Markov chain reaches the steady state first, the current information corresponding to the Markov chain is automatically selected, specifically including: the mean value, standard deviation and receiving probability of the current iteration, as shown in Figure 8 .

[0127] B514, after the current information corresponding to the Markov chain which reaches the steady state first is input as the current information of the remaining n-1 Markov chains, iteration is continued;

[0128] B515, until the total number of iterations is met, the parallel chain with the highest receiving probability and calculation efficiency is selected from the n Markov chains, which has good stability, low rejection rate and high calculation efficiency, and is substituted into step C, as shown in Figure 9 and Figure 10 , and the specific code is shown in Figure 11 .

[0129] Embodiment 3

[0130] A system for establishing a steel pipe concrete structure void and debonding defect degree evaluation model based on ultrasonic detection data, comprising:

[0131] A memory having a computer program stored thereon;

[0132] A processor with 8 or more cores, for executing the program in the memory to realize the method for establishing a steel pipe concrete structure void and debonding defect degree evaluation model based on ultrasonic detection data as described in embodiment 1 or embodiment 2.

[0133] As a preferred scheme of the embodiment, the electronic device can include a processor, a memory, and one or more of a multimedia component, an input / output (I / O) interface, and a communication component.

[0134] ​The processor is configured to control the overall operation of the electronic device to complete all or part of the steps of the method for establishing a degree evaluation model of a concrete-filled steel tube structure void and debonding defect based on ultrasonic detection data.

[0135] The memory is configured to store various types of data to support the operation of the electronic device, which can include, for example, instructions for operating any application or method on the electronic device, and application-related data; the memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0136] The multimedia component can include a screen and an audio component, where the screen can be, for example, a touch screen, and the audio component is configured to output and / or input audio signals; for example, the audio component can include a microphone configured to receive external audio signals, and the received audio signals can be further stored in the memory or transmitted through the communication component; the audio component also includes at least one speaker configured to output audio signals.

[0137] The I / O interface provides an interface between the processor and other interface modules, which can be a keyboard, a mouse, a button, etc.; these buttons can be virtual buttons or physical buttons.

[0138] The communication component is configured to perform wired or wireless communication between the electronic device and other devices; wireless communication, such as Wi-Fi, Bluetooth, near field communication (NFC), 2G, 3G, 4G or 5G, or a combination of one or more of them, so the corresponding communication component can include a Wi-Fi module, a Bluetooth module, an NFC module, and a mobile communication module.

[0139] As a preferred scheme of the embodiment, the electronic device can be implemented by one or more Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), controller, microcontroller, microprocessor or other electronic elements, to execute the method for establishing the defect degree evaluation model of the concrete-filled steel tube structure based on ultrasonic detection data.

[0140] In addition, the computer readable storage medium provided by the embodiment of the present disclosure can be the above-mentioned memory including program instructions, and the program instructions can be executed by the processor of the electronic device to complete the method for establishing the defect degree evaluation model of the concrete-filled steel tube structure based on ultrasonic detection data.

[0141] Embodiment 4

[0142] A computer readable storage medium, which stores a computer program, the program being executed by a processor to implement the method for establishing the defect degree evaluation model of the concrete-filled steel tube structure based on ultrasonic detection data as described in Embodiment 1 or Embodiment 2.

[0143] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement and improvement within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for establishing a model for evaluating the degree of voids and debonding defects in concrete-filled steel tube structures based on ultrasonic testing data, characterized in that: The following steps are involved: Obtain test data and engineering parameters for steel tube concrete arch bridge construction, including waveforms, spectrum, and ultrasonic wave velocity. , 28d compressive strength of concrete , age , hollow rate ; After qualitatively assessing the defect type based on the waveform and spectrum information, the inspection data and engineering parameters are input into the deterministic model to generate deterministic model parameters. The deterministic quantitative model includes: Void defect model: Debonding defect model: Where, Unsonic wave velocity prediction value, is the 28d compressive strength of concrete, For age, is the void ratio, that is, the ratio of void height to steel pipe diameter, is the debonding arc length, is the inner circumference of the steel pipe, are deterministic model parameters; The engineering parameters and the measured ultrasonic wave velocity are input into a pre-built interactive parallel multi-chain DRAM model; the pre-built interactive parallel multi-chain DRAM model is a model obtained by iteratively optimizing the DRAM model using a parallel multi-process algorithm, and the model includes: Void defect model: Debonding defect model: Where, The uncertainty model parameter vector updated for the measured ultrasonic detection data, is the objective uncertainty caused by geometric, material and environmental parameters, In order to consider the subjective uncertainty caused by incomplete factors, and is a standard normal random distribution value. and are the systematic errors caused by objective and subjective factors respectively; According to the interactive parallel multi-chain DRAM model, combined with the confidence level The cases of 0.50 and 0.95 respectively are used to quantitatively evaluate the degree of void and debonding defects; The interactive parallel multi-chain includes: C1, given model parameters The initial value and iterative initial value of both obey the standard random normal distribution, and the iterative standard deviation is taken as large as possible on the basis of satisfying the physical laws. Here, 10000 is used. C2, set the number of parallel multi-chain chains , the number of chains ≥ 8; C3, in The parallel calculation process of Markov chains, when a Markov chain reaches the steady state first, automatically selects the current information corresponding to the Markov chain, including: the mean, standard deviation and acceptance probability of the current iteration; C4, takes the current information corresponding to the Markov chain that reaches steady state first as the rest After the current information of the Markov chain is input, the iteration continues; C5, until the total number of iterations is met, A group of Markov chains with the highest acceptance probability and computational efficiency are selected from the Markov chains with good stability, low rejection rate and high computational efficiency.

2. The method for establishing a model for assessing the degree of voids and debonding defects in concrete-filled steel tube structures based on ultrasonic testing data according to claim 1, characterized in that: Qualitative defect type assessment of the waveform and spectrogram information involves the following steps: A1. Use a non-metallic ultrasonic detector to perform a test using the pair-test method and collect the waveform and spectrum information after the test; A2. When the first wave of the waveform is clear, the amplitude of the first wave is large, the waveform period is a semi-envelope diagram, the waveform has no obvious distortion, and the main frequency number of the spectrum is ≤2, it can be qualitatively judged that the cross section is dense; A3. When the first wave of the waveform is unclear, the amplitude of the first wave is small, the interval between the peak and the trough is small, the waveform is obviously distorted, and the number of main frequencies in the spectrum is ≤2, it can be qualitatively judged as a gap. A4. When the first wave of the waveform is unclear, the amplitude of the first wave is small, the interval between the peak and the trough is large, the waveform is obviously distorted, and the main frequency number of the spectrum is greater than 2, it can be qualitatively judged as debonding.

3. The method for establishing a model for assessing the degree of voids and debonding defects in concrete-filled steel tube structures based on ultrasonic testing data according to claim 1, characterized in that: Quantitatively assessing the degree of voids and disbonds based on the interactive parallel multi-chain DRAM model includes the following steps: B1, engineering parameters, measured ultrasonic wave velocity and model parameters Prior information is input into the pre-built interactive parallel multi-chain DRAM model; B2, iteratively updating and training the DRAM model through an interactive parallel multi-process algorithm, and selecting a parallel chain with better Markov chain stability, lower rejection rate and higher computational efficiency in the first iteration; B3. The Markov chain optimized in the first iteration is updated for a second iteration to eliminate "false steady-state" conditions. When the steady-state indicator of the Markov chain does not change within 10,000 iterations, the second iteration is terminated, and the optimal parameter combination of the DRAM model is determined to obtain an updated DRAM model. B4. Based on the parallel multi-chain DRAM model, combined with the confidence level The debonding rate and de-bonding rate of each section are quantitatively calculated for the cases of 0.50 and 0.95, respectively, to evaluate the compaction performance of the concrete in the pipe. The specific settings are: When the test data is within the 50% confidence interval, it is an acceptable section for void and debonding defects and does not require grouting reinforcement treatment; When the test data is within the 95% confidence interval, it is a critical debonding and delamination defect section, and grouting reinforcement treatment is carried out as needed; When the test data is outside the 95% confidence interval and the void and debonding defect sections are unacceptable, grouting reinforcement treatment should be carried out in a timely manner based on the quantitatively calculated void rate and debonding rate.

4. A system for establishing a model for assessing the degree of voids and debonding defects in concrete-filled steel tube structures based on ultrasonic testing data, characterized in that: include: a memory having a computer program stored thereon; A processor with 8 cores or more, used to execute a program in a memory to implement the method for establishing a model for assessing the degree of voids and debonding defects in a steel tube concrete structure based on ultrasonic detection data as described in any one of claims 1 to 3.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for establishing a model for evaluating the degree of voids and debonding defects in a steel tube concrete structure based on ultrasonic detection data as described in any one of claims 1 to 3 is implemented.

Images