XGB algorithm-based model for predicting sulfate erosion resistance of recycled concrete
Through the prediction model of sulfate resistance resistance performance of recycled concrete based on XGB algorithm, the problem of high uncertainty in the prediction results in the prior art is solved, and the rapid and accurate prediction of the sulfate resistance resistance performance of recycled concrete is achieved, and the use of An effective tool for engineering design.
Patent Information
- Application Number
- CN202410995034.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prediction of the sulfate corrosion resistance of recycled concrete, the prior art has problems such as long test cycle, large workload, low research efficiency and high uncertainty in the prediction results, which leads to poor applicability of empirical models and is difficult to apply in actual engineering.
The corrosion coefficient B is predicted by the XGB algorithm based on the sulfate corrosion resistance performance prediction model of recycled concrete by inputting factors that affect the sulfate corrosion resistance performance of recycled concrete, and the time-varying model of compressive strength corrosion resistance coefficient is calculated to achieve the Fast and accurate prediction of sulfate corrosion resistance of recycled concrete.
This model can effectively handle multiple sulfate corrosion resistance performance indicators, improve the accuracy and reliability of prediction results, and provides a fast and effective tool to support the application of recycled concrete in engineering.
Smart Images

Figure BDA0004959910380000011 
Figure BDA0004959910380000041 
Figure BDA0004959910380000051
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of building materials, and in particular to a prediction model for the sulfate corrosion resistance of recycled concrete based on an XGB algorithm. Background Art
[0002] Recycled concrete is made by partially or completely replacing natural aggregate with recycled aggregate processed by crushing and screening waste concrete. It has received widespread attention because its application can effectively solve problems such as environmental pollution and resource shortages. Because the surface of recycled aggregate is rough and attached with old mortar and has defects such as microcracks, the durability of recycled concrete is generally worse than that of ordinary concrete. One of the important indicators for evaluating durability is resistance to sulfate erosion. Sulfate erosion is the process in which sulfates in the external environment enter the concrete through pores and cracks, react with cement hydration products, and generate substances such as calcium sulfoxide and gypsum with high expansion. The internal expansion stress increases, and microcracks continue to expand, eventually leading to the expansion and cracking of concrete. In the saline-alkali areas and coastal areas of northwest my country, the sulfate content is high, and the concrete structure is severely eroded by sulfate. In order to ensure that recycled concrete is widely used in engineering practice, the research on the prediction of the sulfate erosion resistance of recycled concrete is extremely critical.
[0003] At present, many experts at home and abroad have conducted relevant theoretical and experimental research on the prediction of the sulfate erosion resistance of recycled concrete. However, most of the studies use traditional experimental methods to obtain data such as compressive strength loss and mass loss of recycled concrete under sulfate erosion conditions, statistically analyze the experimental data, and obtain the sulfate erosion empirical model through nonlinear surface fitting. This method has problems such as long test cycle, large workload, and low research efficiency. In addition, there are many uncertain factors that affect the sulfate erosion resistance of recycled concrete, which leads to large discreteness of experimental observation data, poor applicability of the obtained empirical model, and it is difficult to apply it to actual projects. Summary of the invention
[0004] In view of the above problems, the present invention provides a prediction model for the sulfate corrosion resistance of recycled concrete based on the XGB algorithm. The prediction model can be used to quickly predict the sulfate corrosion resistance of recycled concrete. The prediction result is accurate and reliable. It can be used as an effective tool for predicting the sulfate corrosion resistance of recycled concrete, and provide relevant engineering design support and reference for the application of recycled concrete in engineering practice.
[0005] The present invention provides a prediction model for the sulfate erosion resistance of recycled concrete based on the XGB algorithm. The prediction model for the sulfate erosion resistance of recycled concrete is constructed based on the initial prediction model. The expression of the objective function of the prediction model for the sulfate erosion resistance of recycled concrete is as follows:
[0006]
[0007] Among them, F obj is the objective function, L is the loss function, y i is the true value corresponding to the data sample, is the predicted value of the i-th vector in the training set predicted in the first t-1 iterations, f t (x i ) is the predicted value of the i-th vector predicted by the prediction model of the sulfate corrosion resistance of recycled concrete, Ω is the regularization term to avoid overfitting, and f k Represents the mapping relationship of the kth regression tree.
[0008] During the research process, the inventors found that because the time-varying model of compressive strength corrosion resistance coefficient in the field of technology is obtained through theoretical derivation, the key parameter corrosion coefficient B has no clear calculation formula and is difficult to determine, so the prediction and evaluation effect of the sulfate corrosion resistance of recycled concrete is poor. Therefore, the inventors trained the XGBoost algorithm (XGB algorithm for short) in the above manner, so that the trained model (i.e., the prediction model of the sulfate corrosion resistance of recycled concrete) can obtain the predicted value of the corrosion coefficient B by inputting the influencing factors of the sulfate corrosion resistance of recycled concrete, and the prediction effect is accurate and reliable.
[0009] In one embodiment, the initial prediction model is an XGBoost algorithm.
[0010] In one of the embodiments, the construction includes the following steps: taking the factors affecting the sulfate corrosion resistance of the recycled concrete of the sample as input variables, taking the corrosion coefficient of the sample as output variables, training the initial prediction model, and optimizing the parameters to obtain a prediction model for the sulfate corrosion resistance of the recycled concrete.
[0011] In one of the embodiments, the method for obtaining the corrosion coefficient of the sample includes the following steps: collecting test data on the sulfate corrosion resistance of recycled concrete, and obtaining the corrosion coefficient of the sample by fitting a time-varying model of the compressive strength corrosion resistance coefficient.
[0012] In one embodiment, the fitting method is the least squares method.
[0013] In one embodiment, the sulfate corrosion resistance test is a sulfate dry-wet cycle test, and the time-varying model of the compressive strength corrosion resistance coefficient is as follows:
[0014] K f =1-4B(αe -βm +1)(m 0.5 +Bm)
[0015] Among them, K fis the compressive strength corrosion resistance coefficient, which indicates the sulfate corrosion resistance of recycled concrete; B is the corrosion coefficient, which indicates the corrosion degree and speed of recycled concrete by sulfate solution; e is the natural constant; α and β are fitting parameters, and their physical meanings are the influence coefficients of the compressive strength loss rate of recycled concrete when it is in the deterioration stage; m indicates the time when recycled concrete is corroded by sulfate.
[0016] In one embodiment, the training is implemented in Python.
[0017] The prediction model of sulfate erosion resistance of recycled concrete obtained after the above Python training can predict the corrosion coefficient B after inputting the influencing factors of sulfate erosion resistance of recycled concrete (including: water-cement ratio, recycled aggregate replacement rate, fly ash content, and mineral powder content). The compressive strength corrosion resistance coefficient K can be further calculated through the accurate and reliable prediction value of the corrosion coefficient B and the time-varying model of the compressive strength corrosion resistance coefficient. f , and then realize the evaluation of the sulfate corrosion resistance of recycled concrete.
[0018] In one of the embodiments, the factors affecting the sulfate corrosion resistance of the recycled concrete include: at least one of the water-cement ratio, the recycled aggregate replacement rate, the fly ash content, and the mineral powder content.
[0019] The present invention also provides a method for predicting the sulfate corrosion resistance of recycled concrete, comprising the following steps: inputting the influencing factors of the sulfate corrosion resistance of the recycled concrete of the sample to be tested into the recycled concrete sulfate corrosion resistance prediction model to obtain the corrosion coefficient of the sample to be tested.
[0020] In the art, the index for judging the sulfate corrosion resistance of recycled concrete is the compressive strength corrosion resistance coefficient K f , when K f When it is less than 75%, it is judged that the recycled concrete has been damaged by sulfate erosion. The above-mentioned corrosion coefficient B is an important parameter in the time-varying model of compressive strength corrosion resistance coefficient. Although it is impossible to directly judge whether the recycled concrete has been damaged by sulfate erosion by the corrosion coefficient B, technicians in this field can calculate the compressive strength corrosion resistance coefficient K by the corrosion coefficient B. f , and then evaluate the sulfate corrosion resistance of recycled concrete.
[0021] In one of the embodiments, the factors affecting the sulfate corrosion resistance of the recycled concrete include at least one of the water-cement ratio, the recycled aggregate replacement rate, the fly ash content, and the mineral powder content.
[0022] The present invention also provides a system for predicting the sulfate corrosion resistance of recycled concrete, comprising:
[0023] A data storage module, used for storing data information;
[0024] A data analysis module, used to perform analysis according to the prediction method to obtain the corrosion coefficient of the recycled concrete to be evaluated; and
[0025] The data display module is used for outputting and displaying the corrosion coefficient of the recycled concrete to be evaluated.
[0026] In one of the embodiments, the data storage module is used to store the prediction model of the recycled concrete's anti-sulfate corrosion performance and the factors affecting the recycled concrete's anti-sulfate corrosion performance to be evaluated.
[0027] In one of the embodiments, the data analysis module is used to call the factors affecting the anti-sulfate corrosion performance of the recycled concrete to be evaluated and substitute them into the anti-sulfate corrosion performance prediction model of the recycled concrete for analysis.
[0028] Compared with the prior art, the present invention has the following beneficial effects:
[0029] The XGB algorithm-based prediction model for the sulfate erosion resistance of recycled concrete of the present invention quantifies the influence of various uncertain factors on the sulfate erosion resistance of recycled concrete in the sulfate erosion process, can effectively process multiple sulfate erosion resistance performance indicators, and make the prediction results of the sulfate erosion resistance of recycled concrete more accurate and reliable. The prediction method based on the prediction model makes full use of the advantages of intelligent algorithms, avoids the problems of traditional empirical models, solves the problems such as the instability of the prediction results of the sulfate erosion resistance of recycled concrete in actual projects, and provides a quick and effective tool for the prediction of the sulfate erosion resistance of recycled concrete. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 is a flow chart of Example 1;
[0031] Figure 2 It is a flowchart of the XGBoost algorithm of Example 1;
[0032] Figure 3 This is a schematic diagram of the prediction results of the XGBoost prediction model test set in Example 1;
[0033] Figure 4 This is a schematic diagram of a system for predicting the sulfate attack resistance of recycled concrete according to Example 2. DETAILED DESCRIPTION
[0034] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. The preferred embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present invention more thorough and comprehensive.
[0035] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the present invention specification herein are only for describing specific embodiments, rather than limiting the present invention.
[0036] source:
[0037] Unless otherwise specified, the reagents, materials, and equipment used in this example are all commercially available; the experimental methods are all conventional experimental methods in the art unless otherwise specified.
[0038] Example 1
[0039] The prediction model of the sulfate corrosion resistance of recycled concrete based on the XGB algorithm and its construction method are as follows: Figure 1 shown.
[0040] In this embodiment, the XGBoost algorithm is referred to as the XGB algorithm.
[0041] 1. Corrosion coefficient.
[0042] The sulfate dry-wet cycle test data of recycled concrete under different working conditions were collected, and the corrosion coefficient of recycled concrete under different working conditions was obtained by fitting the time-varying model of compressive strength corrosion resistance coefficient.
[0043] In this embodiment, 53 recycled concrete samples were selected, data were obtained through sulfate dry-wet cycle test, and then the corrosion coefficient of recycled concrete under different working conditions was obtained by fitting the time-varying model of compressive strength corrosion resistance coefficient. The time-varying model of compressive strength corrosion resistance coefficient is as follows:
[0044] K f =1-4B(αe -βm +1)(m 0.5 +Bm)
[0045] Among them, K f is the compressive strength corrosion resistance coefficient, which indicates the sulfate corrosion resistance of recycled concrete; B is the corrosion coefficient, which indicates the corrosion degree and speed of recycled concrete by sulfate solution; e is the natural constant; α and β are fitting parameters, and their physical meanings are the influence coefficients of the compressive strength loss rate of recycled concrete when it is in the deterioration stage; m indicates the time when recycled concrete is corroded by sulfate.
[0046] 2. Analyze and determine the influencing factors of the sulfate corrosion resistance of recycled concrete, among which the influencing factors of the sulfate corrosion resistance of recycled concrete include: water-cement ratio, recycled aggregate replacement rate, fly ash content, and mineral powder content; the evaluation index of the sulfate corrosion resistance of recycled concrete is the corrosion coefficient.
[0047] The factors affecting the sulfate corrosion resistance of recycled concrete and the evaluation indicators of the sulfate corrosion resistance of recycled concrete for some recycled concrete samples are shown in the table below.
[0048] Table 1 Factors affecting the sulfate corrosion resistance of recycled concrete and evaluation indicators of the sulfate corrosion resistance of recycled concrete
[0049]
[0050]
[0051] 3. Collect and establish a sample data set of influencing factors for the above 53 recycled concrete samples, and divide them into training set and test set.
[0052] Fourth, a prediction model for the sulfate corrosion resistance of recycled concrete was constructed based on the XGBoost algorithm. The above-mentioned influencing factors were used as input variables, and the corrosion coefficient was used as the output variable. The model was trained and parameters were optimized to finally obtain the prediction results for the sulfate corrosion resistance. The details are as follows:
[0053] 1. Training is based on the initial prediction model based on the XGBoost algorithm. The initial prediction model is the XGBoost algorithm, and its objective function expression is:
[0054]
[0055] Among them, F obj is the objective function, L is the loss function, y i is the true value corresponding to the data sample, is the predicted value of the i-th vector in the training set predicted in the first t-1 iterations, f t (x i ) is the predicted value of the i-th vector predicted by the prediction model of the sulfate corrosion resistance of recycled concrete, Ω is the regularization term to avoid overfitting, and f k Represents the mapping relationship of the kth regression tree.
[0056] 2. The factors affecting the sulfate erosion resistance of recycled concrete are used as the input variables of the XGBoost prediction model, and the corresponding recycled concrete corrosion coefficient is used as the output variable. The model is trained and parameterized by Python to obtain the sulfate erosion resistance prediction model of recycled concrete based on the XGB algorithm. The specific algorithm flow chart is as follows: Figure 2shown.
[0057] 3. Based on the influencing factors of the sulfate erosion resistance of recycled concrete, the test set is predicted using the sulfate erosion resistance prediction model of recycled concrete, and the prediction results are output to verify the prediction effect of the trained sulfate erosion resistance prediction model of recycled concrete.
[0058] The prediction results for the test samples are as follows: Figure 3 As shown, from Figure 3 It can be seen that the predicted value of the recycled concrete anti-sulfate erosion performance prediction model on the test set is closer to the true value.
[0059] 5. Analyze the prediction results to verify the prediction effect of the prediction model for the anti-sulfate corrosion performance of recycled concrete.
[0060] The root mean square error (RMSE) and goodness of fit (R 2 ) is used to further evaluate the prediction accuracy of the model, and the expression is as follows:
[0061]
[0062]
[0063] where y obs is the sample data observation value, y pred is the model prediction value, is the average value of all sample observations, and n is the number of samples. The results are shown in the following table.
[0064] Table 2 RMSE and R of prediction results 2
[0065]
[0066] It can be seen from the above results that the root mean square error of the prediction results of the prediction model for the anti-sulfate erosion performance of recycled concrete is close to 0 and the coefficient of certainty is close to 1, which shows that the model has a good performance in predicting the anti-sulfate erosion performance of recycled concrete and has a certain guiding role in actual engineering.
[0067] Example 2
[0068] A system for predicting the sulfate attack resistance of recycled concrete.
[0069] The system includes a data storage module, a data analysis module, and a data display module. Figure 4 shown.
[0070] The data storage module is used to store the prediction model of the sulfate corrosion resistance of recycled concrete in Example 1 and the factors affecting the sulfate corrosion resistance of the recycled concrete to be evaluated. The factors affecting the sulfate corrosion resistance of the recycled concrete include at least one of the water-binder ratio, the replacement rate of recycled aggregate, the amount of fly ash, and the amount of mineral powder.
[0071] The data analysis module is used to call the influencing factors of the sulfate corrosion resistance of the recycled concrete to be evaluated and substitute them into the prediction model of the sulfate corrosion resistance of the recycled concrete for analysis to obtain the corrosion coefficient of the recycled concrete to be evaluated. The specific method is as follows: the influencing factors of the sulfate corrosion resistance of the sample to be evaluated are input into the prediction model of the sulfate corrosion resistance of the recycled concrete in Example 1 to calculate the corrosion coefficient of the recycled concrete to be evaluated.
[0072] The data display module is used for output display of the corrosion coefficient of the recycled concrete to be evaluated.
[0073] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0074] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.
Claims
1. A prediction model for the sulfate corrosion resistance of recycled concrete based on the XGB algorithm, characterized in that: The prediction model of the sulfate erosion resistance of recycled concrete is constructed based on the initial prediction model. The expression of the objective function of the prediction model of the sulfate erosion resistance of recycled concrete is as follows: Among them, F obj is the objective function, L is the loss function, y i is the true value corresponding to the data sample, is the predicted value of the i-th vector in the training set predicted in the first t-1 iterations, f t (x i ) is the predicted value of the i-th vector predicted by the prediction model of the sulfate corrosion resistance of recycled concrete, Ω is the regularization term to avoid overfitting, and f k Represents the mapping relationship of the kth regression tree.
2. The prediction model for the sulfate corrosion resistance of recycled concrete according to claim 1, characterized in that: The initial prediction model is the XGBoost algorithm.
3. The prediction model for the sulfate corrosion resistance of recycled concrete according to claim 1, characterized in that: The construction includes the following steps: taking the factors affecting the anti-sulfate corrosion performance of the recycled concrete of the sample as input variables, taking the corrosion coefficient of the sample as output variables, training the initial prediction model, and optimizing the parameters to obtain the prediction model for the anti-sulfate corrosion performance of the recycled concrete.
4. The prediction model for the sulfate corrosion resistance of recycled concrete according to claim 3 is characterized in that: The training is implemented in Python.
5. The prediction model for the sulfate corrosion resistance of recycled concrete according to claim 3, characterized in that: The factors affecting the sulfate corrosion resistance of the recycled concrete include: at least one of the water-cement ratio, the replacement rate of recycled aggregate, the amount of fly ash, and the amount of mineral powder.
6. A method for predicting the sulfate corrosion resistance of recycled concrete, characterized in that: The following steps are involved: The factors influencing the sulfate corrosion resistance of the recycled concrete of the sample to be tested are input into the prediction model for the sulfate corrosion resistance of the recycled concrete according to any one of claims 1 to 5 to obtain the corrosion coefficient of the sample to be tested.
7. The prediction method according to claim 6, characterized in that: The factors affecting the sulfate corrosion resistance of the recycled concrete include at least one of the water-cement ratio, the replacement rate of recycled aggregate, the amount of fly ash, and the amount of mineral powder.
8. A system for predicting the sulfate corrosion resistance of recycled concrete, characterized in that: include: A data storage module, used for storing data information; A data analysis module, used to perform analysis according to the prediction method of any one of claims 6-7 to obtain the corrosion coefficient of the recycled concrete to be evaluated; as well as The data display module is used for outputting and displaying the corrosion coefficient of the recycled concrete to be evaluated.
9. The system according to claim 8, characterized in that The data storage module is used to store the prediction model for the sulfate corrosion resistance of recycled concrete according to any one of claims 1 to 5 and the factors affecting the sulfate corrosion resistance of the recycled concrete to be evaluated.
10. The system according to claim 8, characterized in that The data analysis module is used to call the factors affecting the sulfate corrosion resistance of the recycled concrete to be evaluated and substitute them into the prediction model for the sulfate corrosion resistance of the recycled concrete for analysis.
Citation Information
Patent Citations
Recycled aggregate concrete mix proportion optimization method, storage medium and equipment
CN115964957A
Model construction method for sulfate corrosion performance of alkali-activated cementing material
CN117789867A
