Low-carbon pervious concrete and preparation and intelligent screening method thereof

Abstract

This invention belongs to the technical field of permeable concrete, and provides a low-carbon permeable concrete and its preparation and intelligent screening method. The low-carbon permeable concrete comprises, by weight, 75-85 parts red mud, 175-200 parts slag, 800-1400 parts coarse aggregate 1, 600-1200 parts coarse aggregate 2, 38-48 parts water, 3-4 parts NaOH, and 85-95 parts water glass. The preparation method is as follows: coarse aggregate 1 and coarse aggregate 2 are mixed to form a dual-graded coarse aggregate; red mud and slag are mixed to form a red mud-slag geopolymer; an alkali activator solution is prepared using water, NaOH, and water glass; the dual-graded coarse aggregate is added to a mixer, followed by the sequential addition of 50% alkali activator solution, red mud-slag geopolymer, and the remaining 50% alkali activator solution, with stirring performed after each addition; the intelligent screening method utilizes normalized product combined with machine learning algorithms to quickly screen out two concrete experimental groups with superior performance.

Description

Technical Field

[0001] This invention relates to a low-carbon permeable concrete, its preparation, and an intelligent screening method, belonging to the technical field of permeable concrete, specifically a low-carbon permeable concrete, its preparation, and a data-driven intelligent screening method. Background Technology

[0002] Permeable concrete is an eco-friendly building material. Its porous structure allows rainwater to quickly infiltrate, reducing surface runoff and the risk of urban flooding. It also replenishes groundwater and filters pollutants, contributing to the construction of sponge cities. The evaporation of water significantly reduces the urban heat island effect. Furthermore, its permeable and anti-slip properties improve driving and pedestrian safety in rainy weather, its rough surface reduces noise, and it improves the acoustic environment. Combining economy and aesthetics, it is easy to construct, has low maintenance costs, and can be customized with colors or textures, making it suitable for light-load applications such as sidewalks, plazas, and gardens.

[0003] Permeable concrete has significant comprehensive value in ecological restoration, resource recycling, and sustainable urban development, but its load-bearing capacity is limited. In terms of material systems, traditional technologies rely excessively on silicate cement, resulting in high carbon emissions and the consumption of large amounts of non-renewable resources. Its application potential requires further exploration and development. Furthermore, traditional methods for designing and selecting permeable concrete mix proportions mainly rely on empirical methods or single-objective control methods. These designs often overlook crucial aspects, making it difficult to find a precise balance between mechanical strength, permeability coefficient, durability, and pore structure, leading to large fluctuations in material properties. Summary of the Invention

[0004] To address one of the aforementioned technical deficiencies, this invention provides a low-carbon permeable concrete and its preparation and intelligent screening method. This method makes full use of industrial waste, is green, low-carbon, and environmentally friendly, and utilizes an alkali activator to activate polymer activity, enabling the permeable concrete to maintain a certain strength while having high permeability. Combined with machine learning algorithms, this invention achieves a data-driven and scalable method for the preparation and intelligent screening of low-carbon permeable concrete.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: a low-carbon permeable concrete, comprising the following components by weight:

[0006] 75-85 parts red mud;

[0007] 175-200 parts of slag;

[0008] 1,800-1,400 parts coarse aggregate;

[0009] Coarse aggregate 2,600-1,200 parts;

[0010] 38-48 parts water;

[0011] 3-4 parts of NaOH;

[0012] 85-95 parts of water glass.

[0013] The red mud mentioned is Bayer process red mud.

[0014] The slag mentioned is S95 slag.

[0015] The modulus of the water glass is 2.23, and the density of the water glass solution is 1.53 g / cm³. 3 The SiO2 content in the water glass solution is 32%, and the Na2O content in the water glass solution is 14.34%.

[0016] The coarse aggregate 1 is limestone crushed stone with a particle size of 4.75mm-9.5mm, and the coarse aggregate 2 is limestone crushed stone with a particle size of 9.5mm-19mm.

[0017] A method for preparing low-carbon permeable concrete includes the following steps:

[0018] S1. Prepare coarse aggregate 1 and coarse aggregate 2 according to the mass ratio, and mix them thoroughly to obtain coarse aggregate with two particle size distribution;

[0019] S2. Prepare red mud and slag according to the mass ratio, and mix them thoroughly to form a red mud-slag geopolymer;

[0020] S3. Prepare an alkaline activator solution by mixing water, NaOH, and water glass according to the mass ratio;

[0021] S4. Add the double-graded coarse aggregate to the mixer and add 50% alkali activator solution, and stir for 30 seconds;

[0022] S5. Add the red mud-slag geopolymer to the mixer and continue mixing for 90 seconds;

[0023] S6. Add the remaining 50% of the alkaline activator solution to the mixer and continue stirring for 90 seconds;

[0024] S7. Observe the mixing process and discharge the material.

[0025] The specific steps for preparing the alkaline activator solution using water, NaOH, and water glass in step S3 are as follows:

[0026] S31. Slowly pour water into NaOH solid granules and stir slowly until completely dissolved. Heat is released during dissolution. Let stand for 10 minutes to allow the NaOH solution to cool.

[0027] S32. Slowly pour the NaOH solution into the liquid water glass and stir until the two are fully mixed to obtain a mixed solution of alkali activator. Let it stand for 2 hours.

[0028] A smart screening method for low-carbon permeable concrete includes the following steps:

[0029] Step 1: Within the raw material ratio range, based on the porosity, the particle size ratio of coarse aggregate 1 and coarse aggregate 2, and the water-cement ratio, the low-carbon permeable concrete is divided into several groups for sample preparation.

[0030] Step 2: Measure the compressive strength and permeability coefficient of several groups of concrete after 28 days through experiments, and select the maximum compressive strength from these groups. and minimum value Similarly, select the maximum value of the permeability coefficient. and minimum value ;

[0031] Step 3: Based on the normalization principle, normalize the compressive strength of all obtained concrete samples.

[0032] (1);

[0033] in, Let be the measured compressive strength of the i-th group of concrete;

[0034] Based on the normalization principle, the permeability coefficients of all obtained concrete were normalized:

[0035] (2);

[0036] in, Let be the measured permeability coefficient of the i-th group of concrete;

[0037] Further, a normalized comprehensive evaluation index for the compressive strength and permeability coefficient of concrete was obtained:

[0038] (3);

[0039] The comparison showed that the group of concrete with the largest S value had the best overall performance in terms of compressive strength and permeability coefficient.

[0040] Step 4: Define the design porosity, particle size, and water-cement ratio as input feature vectors, and define the compressive strength and permeability coefficient as output variables. Perform performance prediction and model verification based on the XGBoost algorithm.

[0041] The low-carbon permeable concrete and its preparation and intelligent screening method provided by this invention have the following beneficial effects:

[0042] 1. The low-carbon permeable concrete of this invention uses red mud (alumina industrial waste residue) and slag (metallurgical by-product) as the main cementitious materials, completely replacing the traditional cement-based cementitious system, significantly reducing the carbon footprint, solving the problem of red mud disposal in a highly alkaline environment, and realizing the synergistic and complementary advantages of red mud and slag. It improves the shortcomings of traditional single use of red mud, which has high alkalinity but low strength and requires activation of activity, and slag, which has high strength but too fast setting time and strength activation relies on alkali.

[0043] 2. An innovative dual-size composite gradation design strategy is adopted, selecting aggregates in two typical size ranges: 4.75-9.5mm (fine-grained) and 9.5-19mm (coarse-grained). By adjusting the dual-size ratio, a gradient packing system is constructed. While ensuring the interconnected porosity, the mechanical interlocking points between aggregates are maximized, breaking through the physical bottleneck of permeable concrete where strength and permeability are difficult to achieve simultaneously from a physical structure perspective.

[0044] 3. The intelligent screening method for low-carbon permeable concrete involved in this invention utilizes normalization to eliminate the influence of dimensions and numerical ranges, making strength and permeability coefficient comparable within the range of 0-1. This truly reflects the effect of joint optimization of both, enabling rapid screening of two concrete experimental groups with superior performance. The XGBoost model has stronger nonlinear fitting ability compared to traditional regression models; it has lower data requirements and a better anti-overfitting mechanism (regularization) compared to neural network models; and it has higher prediction accuracy (based on residual learning) compared to random forests. Using this model, a high-precision prediction model between mix proportion parameters and performance indicators is established, significantly reducing the number of experiments and lowering design costs, providing an intelligent and digital solution for the customized design of permeable concrete. Attached Figure Description

[0045] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the specific embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 : Flowchart of a method for preparing low-carbon permeable concrete in this invention;

[0047] Figure 2 : Flowchart of an intelligent screening method for low-carbon permeable concrete in this invention;

[0048] Figure 3 Comparison chart of normalized comprehensive performance indicators of concrete in various groups according to the present invention;

[0049] Figure 4: Comparison chart of measured and true values ​​of normalized performance index of concrete in this invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments; based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] This invention discloses a low-carbon permeable concrete, comprising the following components by weight:

[0052] 75-85 parts red mud;

[0053] 175-200 parts of slag;

[0054] 1,800-1,400 parts coarse aggregate;

[0055] Coarse aggregate 2,600-1,200 parts;

[0056] 38-48 parts water;

[0057] 3-4 parts of NaOH;

[0058] 85-95 parts of water glass.

[0059] This invention relates to low-carbon permeable concrete, which uses red mud (alumina industrial waste residue) and slag (metallurgical by-product) as the main cementitious materials to replace the traditional cement-based cementitious system. This reduces the high carbon emissions from cement production and the risk of pollution from red mud stockpiling. At the same time, it achieves the synergistic and complementary advantages of red mud and slag, and improves the shortcomings of traditional single-use red mud, which has high alkalinity but low strength and requires activation of activity, and slag, which has high strength but too fast setting time and strength activation relies on alkali.

[0060] The red mud mentioned is Bayer process red mud, an industrial waste residue from the alumina industry, and is reddish-brown in color.

[0061] The slag mentioned is S95 slag, a metallurgical byproduct.

[0062] The modulus of the water glass is 2.23. The modulus of water glass is defined as the molar ratio of silicon dioxide (SiO2) to alkali metal oxides (such as Na2O). The density of the water glass solution (standard state) is 1.53 g / cm³. 3The SiO2 content (mass percentage of SiO2 in the total weight of the water glass solution) in the water glass solution is 32%, and the Na2O content (mass percentage of Na2O in the total weight of the water glass solution) in the water glass solution is 14.34%.

[0063] The coarse aggregate 1 is limestone crushed stone with a particle size of 4.75mm-9.5mm, and the coarse aggregate 2 is limestone crushed stone with a particle size of 9.5mm-19mm. Addressing the technical bottleneck of difficulty in synergistically optimizing permeability and strength performance in traditional mix design, this invention adopts a dual-particle-size composite gradation design strategy. It selects aggregates in two typical particle size ranges: 4.75-9.5mm (fine-grained) and 9.5-19mm (coarse-grained). By adjusting the proportions of the two particle sizes, a gradient packing system is constructed. The design principle is that the fine-grained aggregate fills the gaps in the coarse-grained skeleton, improving density; while the coarse-grained aggregate maintains the main skeleton structure, ensuring the continuity of permeable channels. While maintaining effective porosity, it significantly enhances the mechanical interlocking effect between aggregates, achieving both high permeability and mechanical load-bearing capacity, thus overcoming the performance contradiction between strength and permeability in permeable concrete.

[0064] Red mud and slag require an alkaline activator to activate their activity. This alkaline activator solution consists of liquid water glass, solid particles of analytical grade NaOH, and distilled water.

[0065] A method for preparing low-carbon permeable concrete, such as Figure 1 As shown, it includes the following steps:

[0066] S1. Wash the required aggregates in advance, air dry them, and prepare coarse aggregate 1 and coarse aggregate 2 according to the mass ratio. Mix the two thoroughly to obtain dual-size coarse aggregate. The dual-size composite gradation design strategy is adopted to realize the transformation of the pore structure from "single through type" to "hierarchical network type", which significantly enhances the mechanical interlocking effect between aggregates while maintaining the effective porosity.

[0067] S2. Prepare red mud and slag according to the mass ratio, and mix them thoroughly to form a red mud-slag geopolymer. The synergistic and complementary advantages of red mud and slag improve the shortcomings of traditional single use of red mud, which has high alkalinity but low strength and needs to be activated, and slag, which has high strength but too fast setting time and strength relies on alkali to be activated.

[0068] S3. Prepare an alkaline activator solution by mixing water, NaOH, and water glass according to the mass ratio;

[0069] The specific steps are as follows:

[0070] S31. Slowly pour water into NaOH solid particles and stir slowly until completely dissolved. Heat is released during dissolution. Let stand for 10 minutes to allow the NaOH solution to cool.

[0071] S32. Slowly pour the NaOH solution into the liquid water glass and stir until the two are fully mixed to obtain a mixed solution of alkali activator. Let it stand for 2 hours.

[0072] NaOH is used to adjust the modulus of water glass, and the two are mixed to prepare an alkaline activator solution to activate the strength and activity of red mud-slag polymer. Moreover, the preparation process does not require high-temperature calcination, and the energy consumption of the alkaline activation reaction at room temperature can be reduced by more than 60%.

[0073] S4. Add the dual-graded coarse aggregate into the mixer and add 50% alkali activator solution. Stir for 30 seconds to make the alkali activator solution evenly coat the surface of the coarse aggregate to achieve a pre-wetting effect, so that the red mud-slag geopolymer can be better combined with the aggregate in the next step.

[0074] S5. Add the red mud-slag geopolymer into the mixer and continue mixing for 90 seconds to ensure that the polymer is evenly coated on the surface of the pre-wetted aggregate.

[0075] S6. Add the remaining 50% alkali activator solution to the mixer and continue mixing for 90 seconds to further combine the aggregate and polymer, firmly wrapping and bonding them so as to achieve higher strength in the later stage.

[0076] S7. Observe the mixing process and discharge the material.

[0077] The intelligent screening method for low-carbon permeable concrete prepared according to the above method is as follows: Figure 2 As shown, it includes the following steps:

[0078] Step 1: Within the raw material ratio range, based on the porosity, the particle size ratio of coarse aggregate 1 and coarse aggregate 2, and the water-cement ratio, the low-carbon permeable concrete is divided into several groups for sample preparation.

[0079] Step 2: Measure the compressive strength and permeability coefficient of several groups of concrete after 28 days through experiments, and select the maximum compressive strength from these groups. and minimum value Similarly, select the maximum value of the permeability coefficient. and minimum value ;

[0080] Step 3: Based on the normalization principle, normalize the compressive strength of all obtained concrete samples.

[0081] (1);

[0082] in, Let be the measured compressive strength of the i-th group of concrete;

[0083] Based on the normalization principle, the permeability coefficients of all obtained concrete were normalized:

[0084] (2);

[0085] in, Let be the measured permeability coefficient of the i-th group of concrete;

[0086] The purpose of the two formulas above is to eliminate dimensions, achieve fair comparison, and convert absolute performance values ​​into relative performance scores. Essentially, these formulas perform a linear transformation, mapping the original performance data of each formulation to a dimensionless value between 0 and 1: a result of 0 indicates that the performance of this formulation is the worst among all experimental groups; a result of 1 indicates that the performance of this formulation is the best among all experimental groups; and a result of 0.5 indicates that the performance of this formulation is at the middle level among all experimental groups.

[0087] The original dimensional "strength" and "permeability" were transformed into a unified "performance score". The higher the score, the better the relative performance of that performance among all the experimental schemes in this study.

[0088] Further, a normalized comprehensive evaluation index for the compressive strength and permeability coefficient of concrete was obtained:

[0089] (3);

[0090] The comparison showed that the group of concrete with the largest S value had the best overall performance in terms of compressive strength and permeability coefficient.

[0091] For permeable concrete, compressive strength and permeability are often contradictory. Increasing strength often requires a denser structure, which reduces porosity and permeability; conversely, increasing porosity to improve permeability weakens its strength.

[0092] The advantage of this method is that it constructs a quantitative indicator through a multiplicative model. This is used to characterize the comprehensive performance level of permeable concrete's contradictory attributes of "strength and permeability," not to pursue the extreme of any single indicator, but to seek a balance point where the overall performance of both is best. The goal of maximizing permeability means, in its physical sense, selecting the most balanced and ideal mix ratio that achieves the best permeability (highest strength) while ensuring sufficient strength.

[0093] Step 4: Define the design porosity, particle size, and water-cement ratio as input feature vectors, and define the compressive strength and permeability coefficient as output variables. Perform performance prediction and model verification based on the XGBoost algorithm.

[0094] The intelligent screening method for low-carbon permeable concrete in this invention will be described in detail below with reference to the embodiments.

[0095] Example

[0096] Step 1: Based on the porosity, the particle size ratio of coarse aggregate 1 and coarse aggregate 2, and the water-cement ratio, the candidate low-carbon permeable concrete is divided into several groups. The particle size ratio refers to the ratio of the amount of coarse aggregate 1 and coarse aggregate 2 used, and the water-cement ratio refers to the ratio of the total water consumption to the total mass of all cementitious materials. As shown in Table 1, a total of 36 groups are divided into groups for sample preparation (the preparation method uses the low-carbon permeable concrete preparation method of this invention).

[0097] Step 2: The compressive strength and permeability coefficient of several groups of concrete after 28 days were measured by experiments, as shown in Table 1.

[0098] Table 1 Concrete parameters and properties for each group

[0099]

[0100] Select the maximum compressive strength of several groups of concrete (25.75) and the minimum value (5.31), similarly select the maximum value of the permeability coefficient. (11.24) and minimum value (2.55);

[0101] Step 3: Based on the normalization principle, normalize the compressive strength of all obtained concrete samples.

[0102] (1);

[0103] in, Let be the measured compressive strength of the i-th group of concrete;

[0104] Based on the normalization principle, the permeability coefficients of all obtained concrete were normalized:

[0105] (2);

[0106] in, Let be the measured permeability coefficient of the i-th group of concrete;

[0107] Further, a normalized comprehensive evaluation index for the compressive strength and permeability coefficient of concrete was obtained:

[0108] (3);

[0109] The comparison shows that the concrete with the largest S value is the group with the best overall performance in terms of compressive strength and permeability coefficient, as shown in Table 2.

[0110] Table 2 Normalized results of concrete properties for each group

[0111]

[0112] like Figure 3 The figure shows a line graph comparing the normalized comprehensive performance indicators of each group of concrete. It can be seen from the figure that the concrete in group 18 has the best comprehensive performance, that is, both the compressive strength and the permeability coefficient are relatively high.

[0113] Step 4: Considering the characteristics of the small sample experimental data in this invention, and the nonlinear coupling relationship between the mix proportion parameters and performance indicators of permeable concrete, the XGBoost model has stronger nonlinear fitting ability compared to traditional regression models; it has lower data requirements and a better anti-overfitting mechanism (regularization) compared to neural network models; and it has higher prediction accuracy (based on residual learning) compared to random forests. Therefore, choosing XGBoost can obtain the most reliable mix proportion screening and verification results with limited experimental costs. The XGBoost prediction performance evaluation indicators are shown in Table 3. It can be seen from the table that the absolute error and root mean square error of the model's prediction of permeability and strength are both at a low level, and the goodness of fit R² has reached a high level, indicating that the model has a good fitting effect.

[0114] Table 3 XGboost Predictive Performance Evaluation Metrics

[0115]

[0116] Table 4 shows the ranking of the normalized results of XGboost's predicted performance and measured performance. It can be concluded that the ranking of XGboost's predicted performance and measured performance is basically consistent, and for the combination ratio, the 18th group is the best, thus verifying the reliability of the results.

[0117] Table 4. Ranking and comparison of normalized results of XGboost predicted and measured performance.

[0118]

[0119] Table 4 shows that the ranking of XGboost's predicted performance is basically consistent with the measured performance. Figure 4 It can also be seen intuitively that the results obtained by the two are very close; for the combination ratio, the 18th group is the best, thus verifying the reliability of the results.

[0120] According to the requirements and proportions of various raw materials for the 18th group of concrete, with a porosity of 15%, coarse aggregate 1 and coarse aggregate 2 with a particle size ratio of 1:1 and a water-cement ratio of 0.35, the raw materials are selected and proportioned. The raw materials include: 926 parts of coarse aggregate 1, 926 parts of coarse aggregate 2, 43 parts of water, 92 parts of water glass, 3.3 parts of NaOH, 79 parts of red mud, and 184.5 parts of slag. The concrete is then produced according to the preparation method of low-carbon permeable concrete of this invention.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A low-carbon permeable concrete, characterized in that, By weight, it includes the following components: 75-85 parts red mud; 175-200 parts of slag; 1,800-1,400 parts coarse aggregate; Coarse aggregate 2,600-1,200 parts; 38-48 parts water; 3-4 parts of NaOH; 85-95 parts of water glass.

2. The low-carbon permeable concrete according to claim 1, characterized in that, The red mud mentioned is Bayer process red mud.

3. The low-carbon permeable concrete according to claim 1, characterized in that, The slag mentioned is S95 slag.

4. The low-carbon permeable concrete according to claim 1, characterized in that, The modulus of the water glass is 2.23, and the density of the water glass solution is 1.53 g / cm³. 3 The SiO2 content in the water glass solution is 32%, and the Na2O content in the water glass solution is 14.34%.

5. The low-carbon permeable concrete according to claim 1, characterized in that, The coarse aggregate 1 is limestone crushed stone with a particle size of 4.75mm-9.5mm, and the coarse aggregate 2 is limestone crushed stone with a particle size of 9.5mm-19mm.

6. A method for preparing low-carbon permeable concrete as described in any one of claims 1-6, characterized in that, Includes the following steps: S1. Prepare coarse aggregate 1 and coarse aggregate 2 according to the mass ratio, and mix them thoroughly to obtain coarse aggregate with two particle size distribution; S2. Prepare red mud and slag according to the mass ratio, and mix them thoroughly to form a red mud-slag geopolymer; S3. Prepare an alkaline activator solution by mixing water, NaOH, and water glass according to the mass ratio; S4. Add the double-graded coarse aggregate to the mixer and add 50% alkali activator solution, and stir for 30 seconds; S5. Add the red mud-slag geopolymer to the mixer and continue mixing for 90 seconds; S6. Add the remaining 50% of the alkaline activator solution to the mixer and continue stirring for 90 seconds; S7. Observe the mixing process and discharge the material.

7. The method for preparing low-carbon permeable concrete according to claim 6, characterized in that, The specific steps for preparing the alkaline activator solution using water, NaOH, and water glass in step S3 are as follows: S31. Slowly pour water into NaOH solid granules and stir slowly until completely dissolved. Heat is released during dissolution. Let stand for 10 minutes to allow the NaOH solution to cool. S32. Slowly pour the NaOH solution into the liquid water glass and stir until the two are fully mixed to obtain a mixed solution of alkali activator. Let it stand for 2 hours.

8. The intelligent screening method for low-carbon permeable concrete as described in any one of claims 1-6, characterized in that, Includes the following steps: Step 1: Within the raw material ratio range, based on the porosity, the particle size ratio of coarse aggregate 1 and coarse aggregate 2, and the water-cement ratio, the low-carbon permeable concrete is divided into several groups for sample preparation. Step 2: Measure the compressive strength and permeability coefficient of several groups of concrete after 28 days through experiments, and select the maximum compressive strength from these groups. and minimum value Similarly, select the maximum value of the permeability coefficient. and minimum value ; Step 3: Based on the normalization principle, normalize the compressive strength of all obtained concrete samples. （1）； in, Let be the measured compressive strength of the i-th group of concrete; Based on the normalization principle, the permeability coefficients of all obtained concrete were normalized: （2）； in, Let be the measured permeability coefficient of the i-th group of concrete; Further, a normalized comprehensive evaluation index for the compressive strength and permeability coefficient of concrete was obtained: （3）； The comparison showed that the group of concrete with the largest S value had the best overall performance in terms of compressive strength and permeability coefficient. Step 4: Define the design porosity, particle size, and water-cement ratio as input feature vectors, and define the compressive strength and permeability coefficient as output variables. Perform performance prediction and model verification based on the XGBoost algorithm.

Images