Method for detecting shrinkage performance of low-shrinkage concrete and low-shrinkage concrete

Abstract

The invention discloses a low-shrinkage concrete shrinkage performance detection method and low-shrinkage concrete, and relates to the field of concrete performance detection.The method comprises the steps that a drying shrinkage test method and a compression testing machine are adopted for conducting a drying shrinkage test and a compressive strength test on the low-shrinkage concrete, obtaining a drying shrinkage value and a mechanical property value based on a test result; obtaining a mixing ratio set corresponding to the drying shrinkage value and the mechanical property value, and analyzing the relevance between the drying shrinkage performance and the mixing ratio based on a relevance analysis method; establishing a shrinkage performance prediction model of the low-shrinkage concrete based on the mixing proportion set, and outputting the shrinkage performance of the low-shrinkage concrete under different mixing proportions; and combining the shrinkage performance with the optimization target, and constructing a strong classifier by using a particle swarm optimization algorithm to output the optimal preparation ratio of the low-shrinkage concrete. According to the method, the optimal concrete proportion meeting the target performance can be automatically explored and output, and the design efficiency and the actual effect of the low-shrinkage concrete are improved.

Description

Technical Field

[0001] The present invention relates to the field of concrete performance detection, and in particular to a method for detecting shrinkage performance of low-shrinkage concrete and low-shrinkage concrete. Background Art

[0002] The Sichuan-Tibet Railway bridge project uses a large amount of cast-in-place concrete, especially in the piers, towers and columns, which are mostly cast in large volumes. In order to ensure the smooth construction and long-term durability of the Sichuan-Tibet Railway bridge project, it is necessary to ensure that the performance of the concrete material is excellent and long-term stable. However, the natural climate environment along the Sichuan-Tibet Railway is very complex and harsh. The environmental effects along the Sichuan-Tibet Railway include significant large temperature differences, strong winds and drying, strong radiation and other complex effects. Such construction and service environments may have serious effects on the performance of concrete. The early and long-term shrinkage of concrete materials constructed and served in complex environments increases significantly, and the structure has a significant risk of cracking. In addition, the long-term performance degradation rate of concrete materials will also be significantly accelerated, and more complex long-term performance degradation problems of concrete may occur, which need to be paid special attention to.

[0003] In the early shrinkage of concrete, drying shrinkage and autogenous shrinkage occupy the main parts, and the biggest difference between drying shrinkage and autogenous shrinkage lies in whether the concrete material exchanges water with the external environment. For the shrinkage and cracking of concrete in complex environments, at present, in the preparation process, for the early shrinkage of concrete, on the one hand, low-heat cement is often used to inhibit the drying shrinkage of concrete, and on the other hand, shrinkage reducers are usually used to inhibit the autogenous shrinkage of concrete. In the curing process, repeated watering, covering with plastic film or geotextile + sprinkling, automatic spraying and other methods are often used for curing.

[0004] Moreover, due to the low hydration heat release of low-heat cement, the temperature deformation of the concrete structure and the generated temperature stress can be reduced, so that the concrete can better resist the risk of cracking caused by changes in environmental temperature differences. In projects under windy, dry and large temperature difference climate conditions, low-heat cement is beneficial to inhibit early cracking of concrete. Although ordinary shrinkage reducers have a strong shrinkage compensation effect, they are less suitable for complex environments such as large temperature differences, extreme dryness, and strong radiation. Excessive dosage may also lead to a significant reduction in concrete strength. Existing technologies only specifically regulate one type of shrinkage, such as using ordinary cement and special shrinkage reducers, or special cement and ordinary shrinkage reducers to regulate the shrinkage of concrete. Existing technologies cannot maximize the use of existing achievements to control the shrinkage of concrete.

[0005] Furthermore, the climate in plateau areas is complex. In an environment of low pressure and large temperature difference, the moisture in concrete evaporates faster. Maintenance methods such as repeated watering and film-covered sprinkling maintenance are difficult to maintain high humidity on the concrete surface for a long time, and automatic spraying maintenance methods are prone to uneven humidity on the concrete surface. These maintenance methods are not effective and cannot meet the requirements for the crack resistance of concrete in complex environments.

[0006] At the same time, in the prior art, after the concrete is prepared, when the performance is tested, the mix ratio in the preparation process cannot be optimized according to the performance test results, which reduces the design efficiency and actual effect of the concrete.

[0007] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0008] 1. Technical issues to be resolved

[0009] In view of the deficiencies in the prior art, the present invention provides a method for detecting the shrinkage performance of low shrinkage concrete and low shrinkage concrete, which have the advantages of being able to automatically explore and output the optimal concrete mix ratio that meets the target performance, thereby solving the problem of optimizing the mix ratio in the preparation process according to the performance test results.

[0010] (II) Technical solution

[0011] In order to achieve the above-mentioned advantage of being able to automatically explore and output the optimal concrete mix ratio that meets the target performance, the specific technical solution adopted by the present invention is as follows:

[0012] According to one aspect of the present invention, a method for detecting shrinkage performance of low shrinkage concrete is provided, and the performance detection method comprises:

[0013] S1. Use drying shrinkage test method and pressure testing machine to perform drying shrinkage test and compressive strength test on low shrinkage concrete, and obtain drying shrinkage value and mechanical property value based on the test results;

[0014] S2. Obtaining a set of mixing ratios corresponding to drying shrinkage values ​​and mechanical property values, and analyzing the correlation between drying shrinkage performance, mechanical properties and mixing ratio based on a correlation analysis method;

[0015] S3. establishing a shrinkage performance prediction model for low shrinkage concrete based on the correlation results and the mix ratio set, and using the shrinkage performance prediction model to output the shrinkage performance of low shrinkage concrete under different mix ratios;

[0016] S4. Combining shrinkage performance with optimization objectives, a strong classifier is constructed using the particle swarm optimization algorithm, and the optimal preparation ratio of low shrinkage concrete is output based on the strong classifier.

[0017] Preferably, obtaining a set of mixing ratios corresponding to drying shrinkage values ​​and mechanical property values, and analyzing the correlation between drying shrinkage performance, mechanical properties and mixing ratios based on a correlation analysis method includes:

[0018] S21, obtaining the preparation mix ratio used for the low shrinkage concrete during the drying shrinkage test and the compressive strength test, and combining all the preparation mix ratios to obtain a mix ratio set;

[0019] S22, setting the mixing ratio set as the independent variable, the drying shrinkage value and the mechanical property value as the dependent variables, and performing threshold screening on the independent variables, and selecting the dominant secondary variable based on the screening results;

[0020] S23, analyzing the fluctuation between the dominant secondary variable and the independent variable, and calculating the correlation between the dependent variable and the fluctuation, to determine the correlation between the changing trends of the independent variable and the dependent variable;

[0021] S24. Calculate the responsibility coefficient for the correlation of the changing trend, and define the responsibility value between the independent variable and the dependent variable according to the responsibility coefficient result to obtain the correlation result.

[0022] Preferably, the expression for correlation calculation is:

[0023]

[0024] In the formula, u represents the correlation calculation result, a represents the total number of tests, represents the bth dependent variable in the cth test, represents the average value of the dependent variable at the cth test, represents the bth fluctuation during the cth test, Represents the average value of volatility during the cth test.

[0025] Preferably, a shrinkage performance prediction model of low shrinkage concrete is established based on the correlation result and the mixing ratio set, and the shrinkage performance prediction model is used to output the shrinkage performance of low shrinkage concrete under different mixing ratios, including:

[0026] S31, constructing an independent variable data matrix according to the proportion parameters related to the mixed proportion set and the correlation results, and calculating the variance and standard deviation of the proportion parameters, and obtaining a target matrix based on the independent variable data matrix, the variance and the standard deviation;

[0027] S32, obtaining positive eigenvalues ​​and eigenvectors of the target matrix, determining a first target ratio parameter based on the positive eigenvalues ​​and the eigenvectors, and constructing a first prediction model according to the first target ratio parameter and the Solow model;

[0028] S33, screening the first target ratio parameter, obtaining a second target ratio parameter based on the screening result, and combining the second target ratio parameter with the first prediction model to generate a shrinkage performance prediction model;

[0029] S34, defining data sets with different mixing ratios, inputting the data sets into a shrinkage performance prediction model, and obtaining the shrinkage performance of low shrinkage concrete under different mixing ratios based on the output results.

[0030] Preferably, according to the ratio parameters related to the mixed ratio set and the correlation results, an independent variable data matrix is ​​constructed, and the variance and standard deviation of the ratio parameters are calculated. The target matrix is ​​obtained based on the independent variable data matrix, the variance and the standard deviation, including:

[0031] S311, extracting a mixing ratio parameter as an independent variable according to the mixing ratio set, and extracting a mix ratio as a related ratio parameter by combining the independent variable with the correlation result;

[0032] S312, defining a matrix structure in which the mix ratio represents rows and the independent variables represent columns, and using the mix ratio set and the matrix structure to arrange the independent variables in order into a matrix form, to obtain an independent variable data matrix;

[0033] S313, using variance and standard deviation calculation technology to obtain the variance and standard deviation of the proportion parameter, and standardizing the independent variable data matrix according to the obtained results and the normalization formula to obtain a standard data matrix;

[0034] S314, using matrix target normalization technology to perform secondary standard processing on the standard data matrix, and obtain the target matrix based on the processing results.

[0035] Preferably, obtaining the positive eigenvalues ​​and eigenvectors of the target matrix, determining the first target ratio parameter based on the positive eigenvalues ​​and the eigenvectors, and constructing the first prediction model according to the first target ratio parameter and the Solow model includes:

[0036] S321, obtaining all eigenvalues ​​of the target matrix to obtain an eigenvalue set, extracting eigenvalues ​​greater than zero from the eigenvalue set as positive eigenvalues, and obtaining a corresponding eigenvector for each positive eigenvalue;

[0037] S322, analyzing the contribution rate of the mix ratio in the target matrix according to the positive eigenvalue and contribution rate calculation technology, and analyzing the cumulative contribution rate of the positive eigenvalue based on the contribution rate;

[0038] S323, performing a difference comparison between the cumulative contribution rate and a preset value, selecting a principal component with the largest contribution rate according to the comparison result, and determining a first target ratio parameter based on the principal component and the eigenvector;

[0039] S324. Set the prediction framework of the Solow model, take the first target proportion parameter as the input, perform iterative training operations, and obtain a first prediction model with the output being the shrinkage performance of low-shrinkage concrete based on the training results.

[0040] Preferably, combine the shrinkage performance with the optimization objective and use the particle swarm optimization algorithm to construct a strong classifier. The optimal preparation ratio of low-shrinkage concrete output based on the strong classifier includes:

[0041] S41. Generate a sample data set based on the shrinkage performance value, mechanical property value, and mixing ratio set, divide the sample data set into training samples, initialize a particle swarm with a size of the threshold value, and set a particle to represent a training sample.

[0042] S42. Construct several weak learners, and sequentially train the weak learners through the particle swarm to obtain weak classifiers. Construct a final strong classifier according to the weak classifiers, and use the final classifier to output the objective function value corresponding to the particle with a classification result of one.

[0043] S43. Input the objective function value into the shrinkage performance prediction model, predict the shrinkage performance of the corresponding training sample, compare the prediction result with the target value, and screen the objective function value based on the comparison result to obtain the fitness value.

[0044] S44. Calculate the global optimal solution based on the fitness value result to determine the optimal particle position of the training sample, and select the mixing ratio set to determine the optimal preparation ratio of low-shrinkage concrete.

[0045] Preferably, constructing several weak learners, and sequentially training the weak learners through the particle swarm to obtain weak classifiers. Construct a final strong classifier according to the weak classifiers, and using the final classifier to output the objective function value corresponding to the particle with a classification result of one includes:

[0046] S421. Construct several weak learners, initialize the weight distribution of the particle swarm in each weak learner, and perform classification training on each weak learner based on the particle swarm after the weight distribution is initialized to obtain weak classifiers.

[0047] S422. Calculate the classification error rate of the weak classifier, analyze the weight coefficient of the weak classifier according to the classification error rate, and update the weight distribution of the particle swarm using the weight coefficient.

[0048] S423. Train the weak classifier with the particle swarm after the weight distribution is updated to obtain a final strong classifier, and use the final classifier to output the objective function value corresponding to the particle with a classification result of one.

[0049] Preferably, the expression of the classification error rate is:

[0050]

[0051] In the formula, E m represents the classification error rate, N represents the total number of particles, x represents the xth particle, and p mi represents the xth particle weight of the mth weak classifier, R[G m (y x )≠z x ] represents the classification result of the mth weak classifier, z x represents the classification result of the xth particle, y x Represents the training sample corresponding to the xth particle.

[0052] According to another aspect of the present invention, there is also provided a low shrinkage concrete, which is composed of the following raw materials in parts by weight:

[0053] 100-150 parts of low-heat cement, 50-90 parts of fly ash, 500-900 parts of sand, 600-900 parts of crushed stone, 50-100 parts of water and 1.0-1.5 parts of shrinkage reducer.

[0054] (III) Beneficial effects

[0055] Compared with the prior art, the present invention provides a method for detecting shrinkage performance of low shrinkage concrete and low shrinkage concrete, which have the following beneficial effects:

[0056] (1) The present invention comprehensively evaluates the physical properties of low shrinkage concrete through drying shrinkage test and compressive strength test, and provides reliable basic data for subsequent analysis and model construction. At the same time, by obtaining the mixing ratio set and analyzing the correlation, the relationship between drying shrinkage value, mechanical properties and mix ratio is clarified, and the scientific nature of concrete mix design is improved. The shrinkage performance prediction model established based on the correlation and mixing ratio set can quickly predict the concrete performance under different mixes, reduce the number of tests and R&D costs, and finally, through the combination of strong classifier and optimization algorithm, it can automatically explore and output the optimal concrete mix that meets the target performance, which can significantly improve the efficiency, scientific nature and practical effect of low shrinkage concrete design, and provide a better solution for concrete engineering.

[0057] (2) The present invention constructs an independent variable data matrix by correcting the variance and standard deviation of the proportion parameters, reduces the impact of data noise, ensures the quality and consistency of the input data, and extracts the first target proportion parameter based on the positive eigenvalue and eigenvector, avoiding the overfitting and computational complexity problems that may be caused by the input of all variables. At the same time, the first target proportion parameter is logically linked to the actual concrete performance through the Solow model, which strengthens the theoretical basis of the prediction model, improves the interpretability of the model, and further enhances the accuracy and applicability of the shrinkage performance prediction, providing a reliable tool for the subsequent optimization of the mixing ratio performance.

[0058] (3) The present invention generates a sample data set using shrinkage performance values, mechanical property values ​​and mixing ratio sets, converts experimental data into inputs that can be used for optimization, provides rich initial information for the particle swarm algorithm, constructs multiple weak learners, trains the particle swarm in stages, and uses its gradually improved characteristics to reduce the overfitting and underfitting problems of the model. The particles whose classification results are marked as one correspond to the optimal objective function value, providing an accurate basis for subsequent performance prediction and fitness calculation. At the same time, the objective function value is input into the shrinkage performance prediction model to predict the performance of the corresponding mixing ratio, reducing the time and resources required for actual experiments, so that the mixing ratio set can quickly determine the preparation ratio that meets the requirements of low shrinkage performance and high mechanical properties after the optimal particle position is selected, thereby improving the scientificity and efficiency of concrete design. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0060] Figure 1 4 is a flow chart of a method for detecting shrinkage performance of low shrinkage concrete according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0062] According to an embodiment of the present invention, a method for detecting shrinkage performance of low-shrinkage concrete and low-shrinkage concrete are provided.

[0063] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the method for detecting shrinkage performance of low shrinkage concrete according to an embodiment of the present invention, the performance detection method includes:

[0064] S1. Use the drying shrinkage test method and pressure testing machine to carry out drying shrinkage test and compressive strength test on low shrinkage concrete, and obtain the drying shrinkage value and mechanical property value based on the test results.

[0065] In specific applications, each component is weighed by weight, and 100-150 parts of low-heat cement, 50-90 parts of fly ash, 500-900 parts of sand, 600-900 parts of crushed stone, 50-100 parts of water, and 1.0-1.5 parts of shrinkage reducer are added to a concrete mixer to obtain low-shrinkage concrete materials for complex plateau environments. After the concrete is cast and molded and demolded, a water-replenishing type moisturizing curing film with an inner water-replenishing layer of highly absorbent resin and an outer water-retaining layer of plastic film is used for film curing without watering. The mechanical properties and shrinkage properties of the prepared concrete are tested with reference to GB / T50107-2010 "Concrete Strength Test and Evaluation Standard" and GB / T 50082-2009 "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete". The specific preparation process is as follows:

[0066] The corresponding weight portions of the above-mentioned low-heat cement, sand and crushed stone are added into a mixer in sequence, stirred for 1-2 minutes, and corresponding weight portions of water and shrinkage reducing agent are added. After stirring for 10 minutes, corresponding weight portions of fly ash are added, and stirring is continued for 20-30 minutes to obtain the low shrinkage concrete material of the present invention.

[0067] After the material preparation is completed, it can be poured into the mold. After forming and demolding, a high-efficiency moisturizing curing film must be used for curing according to the following steps: Before removing the concrete formwork, use 10%-20% concentration of silicate to replenish water on one side of the curing membrane water storage layer until it is saturated; the concrete specimen is demolded 24 hours after forming, and one side of the curing membrane water storage layer is wrapped tightly against the concrete surface, and the overlap is sealed with transparent tape.

[0068] S2. Obtain a set of mixing ratios corresponding to drying shrinkage values ​​and mechanical property values, and analyze the correlation between drying shrinkage performance, mechanical properties and mixing ratio based on a correlation analysis method.

[0069] In one embodiment, obtaining a set of mixing ratios corresponding to drying shrinkage values ​​and mechanical property values, and analyzing the correlation between drying shrinkage performance, mechanical properties and mixing ratios based on a correlation analysis method includes:

[0070] S21, obtaining the preparation mix ratio used for the low shrinkage concrete during the drying shrinkage test and the compressive strength test, and combining all the preparation mix ratios to obtain a mix ratio set;

[0071] S22, setting the mixing ratio set as the independent variable, the drying shrinkage value and the mechanical property value as the dependent variables, and performing threshold screening on the independent variables, and selecting the dominant secondary variable based on the screening results;

[0072] S23, analyzing the fluctuation between the dominant secondary variable and the independent variable, and calculating the correlation between the dependent variable and the fluctuation, to determine the correlation between the changing trends of the independent variable and the dependent variable;

[0073] S24. Calculate the responsibility coefficient for the correlation of the changing trend, and define the responsibility value between the independent variable and the dependent variable according to the responsibility coefficient result to obtain the correlation result.

[0074] In one embodiment, the expression for correlation calculation is:

[0075]

[0076] In the formula, u represents the correlation calculation result, a represents the total number of tests, represents the bth dependent variable in the cth test, represents the average value of the dependent variable at the cth test, represents the bth fluctuation during the cth test, Represents the average value of volatility during the cth test.

[0077] S3. A shrinkage performance prediction model for low shrinkage concrete is established based on the correlation results and the mixing ratio set, and the shrinkage performance prediction model is used to output the shrinkage performance of low shrinkage concrete under different mixing ratios.

[0078] In one embodiment, a shrinkage performance prediction model of low shrinkage concrete is established based on the correlation result and the mixing ratio set, and the shrinkage performance prediction model is used to output the shrinkage performance of low shrinkage concrete under different mixing ratios, including:

[0079] S31, constructing an independent variable data matrix according to the proportion parameters related to the mixed proportion set and the correlation results, and calculating the variance and standard deviation of the proportion parameters, and obtaining a target matrix based on the independent variable data matrix, the variance and the standard deviation;

[0080] S32, obtaining positive eigenvalues ​​and eigenvectors of the target matrix, determining a first target ratio parameter based on the positive eigenvalues ​​and the eigenvectors, and constructing a first prediction model according to the first target ratio parameter and the Solow model;

[0081] S33, screening the first target ratio parameter, obtaining a second target ratio parameter based on the screening result, and combining the second target ratio parameter with the first prediction model to generate a shrinkage performance prediction model;

[0082] S34, defining data sets with different mixing ratios, inputting the data sets into a shrinkage performance prediction model, and obtaining the shrinkage performance of low shrinkage concrete under different mixing ratios based on the output results.

[0083] In one embodiment, according to the ratio parameters related to the mixed ratio set and the correlation results, an independent variable data matrix is ​​constructed, and the variance and standard deviation of the ratio parameters are calculated. The target matrix is ​​obtained based on the independent variable data matrix, the variance and the standard deviation, including:

[0084] S311, extracting a mixing ratio parameter as an independent variable according to the mixing ratio set, and extracting a mix ratio as a related ratio parameter by combining the independent variable with the correlation result;

[0085] S312, defining a matrix structure in which the mix ratio represents rows and the independent variables represent columns, and using the mix ratio set and the matrix structure to arrange the independent variables in order into a matrix form, to obtain an independent variable data matrix;

[0086] S313, using variance and standard deviation calculation technology to obtain the variance and standard deviation of the proportion parameter, and standardizing the independent variable data matrix according to the obtained results and the normalization formula to obtain a standard data matrix;

[0087] S314, using matrix target normalization technology to perform secondary standard processing on the standard data matrix, and obtain the target matrix based on the processing results.

[0088] In one embodiment, obtaining positive eigenvalues ​​and eigenvectors of a target matrix, determining a first target ratio parameter based on the positive eigenvalues ​​and the eigenvectors, and constructing a first prediction model according to the first target ratio parameter and the Solow model includes:

[0089] S321, obtaining all eigenvalues ​​of the target matrix to obtain an eigenvalue set, extracting eigenvalues ​​greater than zero from the eigenvalue set as positive eigenvalues, and obtaining a corresponding eigenvector for each positive eigenvalue;

[0090] S322, analyzing the contribution rate of the mix ratio in the target matrix according to the positive eigenvalue and contribution rate calculation technology, and analyzing the cumulative contribution rate of the positive eigenvalue based on the contribution rate;

[0091] S323, performing a difference comparison between the cumulative contribution rate and a preset value, selecting a principal component with the largest contribution rate according to the comparison result, and determining a first target ratio parameter based on the principal component and the eigenvector;

[0092] S324, setting a prediction framework of the Solow model, and taking the first target ratio parameter as input, performing an iterative training operation, and obtaining a first prediction model whose output is the shrinkage performance of low shrinkage concrete based on the training result.

[0093] S4. Combining shrinkage performance with optimization objectives, a strong classifier is constructed using the particle swarm optimization algorithm, and the optimal preparation ratio of low shrinkage concrete is output based on the strong classifier.

[0094] In one embodiment, shrinkage performance is combined with optimization objectives and a strong classifier is constructed using a particle swarm optimization algorithm. The optimal preparation ratio of low shrinkage concrete outputted based on the strong classifier includes:

[0095] S41, generating a sample data set based on the contraction performance value, the mechanical performance value and the mixed ratio set, and dividing the sample data set into training samples, initializing a particle group with a size of a threshold, and setting one particle to represent one training sample;

[0096] S42, constructing several weak learners, and training the weak learners in turn through the particle swarm to obtain weak classifiers, constructing a final strong classifier based on the weak classifiers, and using the final classifier to output the objective function value corresponding to the particle whose classification result is one;

[0097] S43, inputting the objective function value into the shrinkage performance prediction model, predicting the shrinkage performance of the corresponding training sample, and comparing the prediction result with the target value, screening the objective function value based on the comparison result, and obtaining the fitness value;

[0098] S44. Calculate the global optimal solution based on the fitness value result to determine the optimal particle position of the training sample, and select a set of mixing ratios to determine the optimal preparation ratio of the low shrinkage concrete.

[0099] In one embodiment, several weak learners are constructed, and the weak learners are trained in sequence by a particle swarm to obtain weak classifiers. A final strong classifier is constructed based on the weak classifiers. The objective function value corresponding to the particle whose classification result is one output by the final classifier includes:

[0100] S421, constructing several weak learners, and initializing the weight distribution of the particle swarm in each weak learner, and performing classification training on each weak learner based on the particle swarm after the weight distribution is initialized to obtain a weak classifier;

[0101] S422, calculating the classification error rate of the weak classifier, analyzing the weight coefficient of the weak classifier according to the classification error rate, and using the weight coefficient to update the weight distribution of the particle swarm;

[0102] S423, training the weak classifier according to the particle swarm after weight distribution update to obtain a final strong classifier, and using the final classifier to output the objective function value corresponding to the particle whose classification result is one.

[0103] It should be explained that, taking the shrinkage and crack resistance of concrete as the optimization target, the particle swarm optimization algorithm is used to establish a concrete mix intelligent optimization model (strong classifier), and the feasibility of the model in optimizing the mix is ​​verified by the test results of the following embodiments. The mix optimization process is as follows:

[0104] According to the requirements for concrete performance, the target mechanical properties and shrinkage performance parameters of the model are input into the optimization model, and the concrete mix ratio is randomly represented as the position of the PSO algorithm particles. Each particle represents a possible mix ratio scheme to form an initialized solution space. The objective function values ​​of all particles in the solution space are calculated and input into the established prediction model of mechanical properties and shrinkage properties. The shrinkage performance is predicted and compared with the target value. The calculated objective function value is used as the fitness value of the entire optimization model to evaluate the feasibility of a set of mix ratios. According to the constraints and fitness values, the individual extreme values ​​and group extreme values ​​after this iteration are screened, and according to the update strategy of the particle swarm algorithm, the speed and position of each particle are updated to judge the convergence. If it does not converge, iterate until convergence to obtain the optimal particles in the group and their corresponding mix ratios, that is, the optimized mix ratios.

[0105] In one embodiment, the classification error rate is expressed as:

[0106]

[0107] In the formula, E m represents the classification error rate, N represents the total number of particles, x represents the xth particle, p mi represents the xth particle weight of the mth weak classifier, R[G m (y x )≠z x ] represents the classification result of the mth weak classifier, z x represents the classification result of the xth particle, y x Represents the training sample corresponding to the xth particle.

[0108] According to another embodiment of the present invention, a low shrinkage concrete is provided, which is composed of the following raw materials in parts by weight:

[0109] 100-150 parts of low-heat cement, 50-90 parts of fly ash, 500-900 parts of sand, 600-900 parts of crushed stone, 50-100 parts of water and 1.0-1.5 parts of shrinkage reducer.

[0110] The following are some embodiments of the present invention. The mix ratios of low shrinkage concrete for bridge piers and towers under complex environments prepared in different embodiments are shown in Table 1. The concrete mechanical properties are tested with reference to GB / T 50107-2010 "Concrete Strength Test and Evaluation Standard", and the concrete shrinkage performance is tested with reference to GB / T 50082-2009 "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete".

[0111] Table 1 Example concrete mix ratio

[0112] Low heat cement Fly ash sand stone water Shrinkage Reducing Agent 100 52 595 643 54 1.07 110 61 517 523 62 1.29 120 55 640 685 68 1.14 130 88 586 723 76 1.37 140 79 759 782 82 1.48 150 86 867 870 90 1.44

[0113] Example 1

[0114] The low shrinkage concrete for bridge piers and towers under complex environments in this embodiment comprises the following components, measured in parts by weight: 100 parts of low-heat cement, 52 parts of fly ash, 595 parts of sand, 643 parts of crushed stone, 54 parts of water, and 1.07 parts of shrinkage reducer. Each component is weighed according to weight, and the low-heat cement, fly ash, sand, crushed stone, water, and shrinkage reducer are added to a concrete mixer to obtain low-shrinkage concrete for bridge piers and towers under complex environments. Subsequently, a 12% concentration of sodium silicate solution curing film is used to replenish water and then the film is covered for curing. After that, the test is carried out in accordance with the standards "Concrete Strength Inspection and Evaluation Standard" and "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete". The results are shown in Table 2.

[0115] Example 2

[0116] The low shrinkage concrete for bridge piers and towers under complex environments in this embodiment comprises the following components, measured in parts by weight: 110 parts of low-heat cement, 61 parts of fly ash, 517 parts of sand, 523 parts of crushed stone, 62 parts of water, and 1.29 parts of shrinkage reducer. Each component is weighed according to weight, and low-heat cement, fly ash, sand, crushed stone, water, and shrinkage reducer are added to a concrete mixer to obtain low-shrinkage concrete for bridge piers and towers under complex environments. Subsequently, a curing film is cured with 18% sodium silicate solution and then covered with film for curing after hydration. Thereafter, tests are carried out in accordance with the standards "Concrete Strength Inspection and Evaluation Standard" and "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete". The results are shown in Table 2.

[0117] Example 3

[0118] The low shrinkage concrete for bridge piers and towers under complex environments in this embodiment comprises the following components, measured in parts by weight: 120 parts of low-heat cement, 55 parts of fly ash, 640 parts of sand, 685 parts of crushed stone, 68 parts of water, and 1.14 parts of shrinkage reducer. Each component is weighed by weight, and the low-heat cement, fly ash, sand, crushed stone, water, and shrinkage reducer are added to a concrete mixer to obtain low-shrinkage concrete for bridge piers and towers under complex environments. Subsequently, a 16% concentration of sodium silicate solution curing film is used to replenish water and then the film is covered for curing. After that, the test is carried out in accordance with the standards "Concrete Strength Inspection and Evaluation Standard" and "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete". The results are shown in Table 2.

[0119] Example 4

[0120] The low shrinkage concrete for bridge piers and towers under complex environments in this embodiment comprises the following components, measured in parts by weight: 130 parts of low-heat cement, 88 parts of fly ash, 586 parts of sand, 723 parts of crushed stone, 76 parts of water, and 1.37 parts of shrinkage reducer. Each component is weighed by weight, and the low-heat cement, fly ash, sand, crushed stone, water, and shrinkage reducer are added to a concrete mixer to obtain low-shrinkage concrete for bridge piers and towers under complex environments. Subsequently, a 15% concentration of sodium silicate solution curing film is used to replenish water and then the film is covered for curing. After that, the test is carried out in accordance with the standards "Concrete Strength Inspection and Evaluation Standard" and "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete". The results are shown in Table 2.

[0121] Example 5

[0122] The low shrinkage concrete for bridge piers and towers under complex environments in this embodiment comprises the following components, measured in parts by weight: 140 parts of low-heat cement, 79 parts of fly ash, 759 parts of sand, 782 parts of crushed stone, 82 parts of water, and 1.48 parts of shrinkage reducer. Each component is weighed according to weight, and low-heat cement, fly ash, sand, crushed stone, water, and shrinkage reducer are added to a concrete mixer to obtain low-shrinkage concrete for bridge piers and towers under complex environments. Subsequently, a 13% concentration of sodium silicate solution curing film is used to replenish water and then the film is covered for curing. After that, the film is tested in accordance with the standards "Concrete Strength Inspection and Evaluation Standard" and "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete". The results are shown in Table 2.

[0123] Example 6

[0124] The low shrinkage concrete for bridge piers and towers under complex environments in this embodiment comprises the following components, measured in parts by weight: 150 parts of low-heat cement, 86 parts of fly ash, 867 parts of sand, 870 parts of crushed stone, 90 parts of water, and 1.44 parts of shrinkage reducer. Each component is weighed according to weight, and low-heat cement, fly ash, sand, crushed stone, water, and shrinkage reducer are added to a concrete mixer to obtain low-shrinkage concrete for bridge piers and towers under complex environments. Subsequently, a 17% concentration of sodium silicate solution curing film is used to replenish water and then the film is covered for curing. After that, the test is carried out in accordance with the standards "Concrete Strength Inspection and Evaluation Standard" and "Standard for Test Methods for Long-term Performance and Durability of Ordinary Concrete". The results are shown in Table 2.

[0125] Table 2 Example concrete mix ratio

[0126]

[0127]

[0128] It can be seen from Table 2 that the compressive strength of the low shrinkage concrete for bridge piers and towers in complex environments increases steadily with the increase of age, and the shrinkage rate is maintained below 0.030%, indicating that the low shrinkage concrete for bridge piers and towers in complex environments prepared by the present invention has stable strength development, good shrinkage reduction performance, improved crack resistance, effectively controls the risk of shrinkage cracking of concrete structures, and greatly improves durability.

[0129] To sum up, with the help of the above-mentioned technical scheme of the present invention, the present invention combines low-heat cement with special matrix shrinkage-reducing materials, and adopts a curing method of covering with a water-replenishing and moisturizing film containing a silicate solution to regulate the shrinkage and cracking problem of concrete, and designs and prepares large-volume concrete materials suitable for bridge piers and towers under complex environments. Compared with the single regulation method of materials or curing methods, the shrinkage of concrete structural components under complex environments is greatly reduced, the risk of early and long-term cracking is suppressed, and the cracking problem of concrete of bridge piers and towers of the Sichuan-Tibet Railway under complex environments is controlled. At the same time, compared with the existing technical methods, the concrete prepared by the present invention has the characteristics of small shrinkage, high strength, good durability, etc., and is suitable for concrete in complex environments such as large temperature difference, extreme dryness, and strong radiation in the Sichuan-Tibet region, and is well applied in the concrete engineering of the Sichuan-Tibet Railway.

[0130] The present invention comprehensively evaluates the physical properties of low shrinkage concrete through drying shrinkage test and compressive strength test, provides reliable basic data for subsequent analysis and model construction, and at the same time, clarifies the relationship between drying shrinkage value, mechanical properties and mix ratio by obtaining a mixing ratio set and analyzing the correlation, thereby improving the scientific nature of concrete mix design, and a shrinkage performance prediction model established based on the correlation and mixing ratio set can quickly predict the concrete performance under different mix ratios, reduce the number of tests and R&D costs, and finally, through a strong classifier combined with an optimization algorithm, can automatically explore and output the optimal concrete mix that meets the target performance, which can significantly improve the efficiency, scientific nature and practical effect of low shrinkage concrete design, and provide a better solution for concrete engineering. The present invention constructs an independent variable data matrix through variance and standard deviation correction of proportion parameters, reduces the influence of data noise, ensures the quality and consistency of input data, and extracts the first target proportion parameter based on positive eigenvalues ​​and eigenvectors, avoiding overfitting and calculation complexity problems that may be caused by full variable input. At the same time, the first target proportion parameter is logically linked with the actual concrete performance through the Solow model, which strengthens the theoretical basis of the prediction model, improves the interpretability of the model, further improves the accuracy and applicability of shrinkage performance prediction, and provides a reliable tool for subsequent optimization of mixing ratio performance.

[0131] The present invention generates a sample data set by using shrinkage performance values, mechanical property values ​​and a mixture ratio set, converts experimental data into input that can be used for optimization, provides rich initial information for a particle swarm algorithm, constructs multiple weak learners, trains a particle swarm in stages, and uses its gradually improved characteristics to reduce overfitting and underfitting problems of the model. The particle corresponding to a classification result marked as one corresponds to an optimal objective function value, providing an accurate basis for subsequent performance prediction and fitness calculation. At the same time, the objective function value is input into a shrinkage performance prediction model to predict the performance of the corresponding mixture ratio, thereby reducing the time and resources required for actual experiments, so that the mixture ratio set can quickly determine a preparation ratio that meets the requirements of low shrinkage performance and high mechanical performance through the selection of the optimal particle position, thereby improving the scientificity and efficiency of concrete design.

[0132] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting shrinkage performance of low shrinkage concrete, characterized in that: The performance testing method includes: S1. Use drying shrinkage test method and pressure testing machine to perform drying shrinkage test and compressive strength test on low shrinkage concrete, and obtain drying shrinkage value and mechanical property value based on the test results; S2. Obtaining a set of mixing ratios corresponding to drying shrinkage values ​​and mechanical property values, and analyzing the correlation between drying shrinkage performance, mechanical properties and mixing ratio based on a correlation analysis method; S3. establishing a shrinkage performance prediction model for low shrinkage concrete based on the correlation results and the mix ratio set, and using the shrinkage performance prediction model to output the shrinkage performance of low shrinkage concrete under different mix ratios; S4. Combining shrinkage performance with optimization objectives, a strong classifier is constructed using the particle swarm optimization algorithm, and the optimal preparation ratio of low shrinkage concrete is output based on the strong classifier.

2. A method for detecting shrinkage performance of low shrinkage concrete according to claim 1, characterized in that: The step of obtaining a set of mixing ratios corresponding to drying shrinkage values ​​and mechanical property values, and analyzing the correlation between drying shrinkage performance, mechanical properties and mixing ratios based on a correlation analysis method includes: S21, obtaining the preparation mix ratio used for the low shrinkage concrete during the drying shrinkage test and the compressive strength test, and combining all the preparation mix ratios to obtain a mix ratio set; S22, setting the mixing ratio set as the independent variable, the drying shrinkage value and the mechanical property value as the dependent variables, and performing threshold screening on the independent variables, and selecting the dominant secondary variable based on the screening results; S23, analyzing the fluctuation between the dominant secondary variable and the independent variable, and calculating the correlation between the dependent variable and the fluctuation, to determine the correlation between the changing trends of the independent variable and the dependent variable; S24. Calculate the responsibility coefficient for the correlation of the changing trend, and define the responsibility value between the independent variable and the dependent variable according to the responsibility coefficient result to obtain the correlation result.

3. A method for detecting shrinkage performance of low shrinkage concrete according to claim 2, characterized in that: The expression for calculating the correlation is: In the formula, u represents the correlation calculation result, a represents the total number of tests, represents the bth dependent variable in the cth test, represents the average value of the dependent variable at the cth test, represents the bth fluctuation during the cth test, Represents the average value of volatility during the cth test.

4. A method for detecting shrinkage performance of low shrinkage concrete according to claim 3, characterized in that: The shrinkage performance prediction model of low shrinkage concrete is established based on the correlation result and the mixing ratio set, and the shrinkage performance prediction model is used to output the shrinkage performance of low shrinkage concrete under different mixing ratios, including: S31, constructing an independent variable data matrix according to the proportion parameters related to the mixed proportion set and the correlation results, and calculating the variance and standard deviation of the proportion parameters, and obtaining a target matrix based on the independent variable data matrix, the variance and the standard deviation; S32, obtaining positive eigenvalues ​​and eigenvectors of the target matrix, determining a first target ratio parameter based on the positive eigenvalues ​​and the eigenvectors, and constructing a first prediction model according to the first target ratio parameter and the Solow model; S33, screening the first target ratio parameter, obtaining a second target ratio parameter based on the screening result, and combining the second target ratio parameter with the first prediction model to generate a shrinkage performance prediction model; S34, defining data sets with different mixing ratios, inputting the data sets into a shrinkage performance prediction model, and obtaining the shrinkage performance of low shrinkage concrete under different mixing ratios based on the output results.

5. A method for detecting shrinkage performance of low shrinkage concrete according to claim 4, characterized in that: The method of constructing an independent variable data matrix based on the proportion parameters related to the mixed proportion set and the correlation results, and calculating the variance and standard deviation of the proportion parameters, and obtaining a target matrix based on the independent variable data matrix, the variance and the standard deviation includes: S311, extracting a mixing ratio parameter as an independent variable according to the mixing ratio set, and extracting a mix ratio as a related ratio parameter by combining the independent variable with the correlation result; S312, defining a matrix structure in which the mix ratio represents rows and the independent variables represent columns, and using the mix ratio set and the matrix structure to arrange the independent variables in order into a matrix form, to obtain an independent variable data matrix; S313, using variance and standard deviation calculation technology to obtain the variance and standard deviation of the proportion parameter, and standardizing the independent variable data matrix according to the obtained results and the normalization formula to obtain a standard data matrix; S314, using matrix target normalization technology to perform secondary standard processing on the standard data matrix, and obtain the target matrix based on the processing results.

6. A method for detecting shrinkage performance of low shrinkage concrete according to claim 5, characterized in that: The step of obtaining positive eigenvalues ​​and eigenvectors of the target matrix, determining a first target ratio parameter based on the positive eigenvalues ​​and the eigenvectors, and constructing a first prediction model according to the first target ratio parameter and the Solow model includes: S321, obtaining all eigenvalues ​​of the target matrix to obtain an eigenvalue set, extracting eigenvalues ​​greater than zero from the eigenvalue set as positive eigenvalues, and obtaining a corresponding eigenvector for each positive eigenvalue; S322, analyzing the contribution rate of the mix ratio in the target matrix according to the positive eigenvalue and contribution rate calculation technology, and analyzing the cumulative contribution rate of the positive eigenvalue based on the contribution rate; S323, performing a difference comparison between the cumulative contribution rate and a preset value, selecting a principal component with the largest contribution rate according to the comparison result, and determining a first target ratio parameter based on the principal component and the eigenvector; S324, setting a prediction framework of the Solow model, and taking the first target ratio parameter as input, performing an iterative training operation, and obtaining a first prediction model whose output is the shrinkage performance of low shrinkage concrete based on the training result.

7. A method for detecting shrinkage performance of low shrinkage concrete according to claim 6, characterized in that: The method of combining shrinkage performance with optimization objectives and using a particle swarm optimization algorithm to construct a strong classifier and outputting an optimal preparation ratio of low shrinkage concrete based on the strong classifier includes: S41, generating a sample data set based on the contraction performance value, the mechanical performance value and the mixed ratio set, and dividing the sample data set into training samples, initializing a particle group with a size of a threshold, and setting one particle to represent one training sample; S42, constructing several weak learners, and training the weak learners in turn through the particle swarm to obtain weak classifiers, constructing a final strong classifier based on the weak classifiers, and using the final classifier to output the objective function value corresponding to the particle whose classification result is one; S43, inputting the objective function value into the shrinkage performance prediction model, predicting the shrinkage performance of the corresponding training sample, and comparing the prediction result with the target value, screening the objective function value based on the comparison result, and obtaining the fitness value; S44. Calculate the global optimal solution based on the fitness value result to determine the optimal particle position of the training sample, and select a set of mixing ratios to determine the optimal preparation ratio of the low shrinkage concrete.

8. A method for detecting shrinkage performance of low shrinkage concrete according to claim 7, characterized in that: The method constructs several weak learners, trains the weak learners in sequence through a particle swarm, obtains a weak classifier, constructs a final strong classifier based on the weak classifier, and uses the final classifier to output a classification result of one corresponding to the particle objective function value, which includes: S421, constructing several weak learners, and initializing the weight distribution of the particle swarm in each weak learner, and performing classification training on each weak learner based on the particle swarm after the weight distribution is initialized to obtain a weak classifier; S422, calculating the classification error rate of the weak classifier, analyzing the weight coefficient of the weak classifier according to the classification error rate, and using the weight coefficient to update the weight distribution of the particle swarm; S423, training the weak classifier according to the particle swarm after weight distribution update to obtain a final strong classifier, and using the final classifier to output the objective function value corresponding to the particle whose classification result is one.

9. A method for detecting shrinkage performance of low shrinkage concrete according to claim 8, characterized in that: The expression of the classification error rate is: In the formula, E m represents the classification error rate, N represents the total number of particles, x represents the xth particle, and p mi represents the xth particle weight of the mth weak classifier, R[G m (y x )≠z x ] represents the classification result of the mth weak classifier, z x represents the classification result of the xth particle, y x Represents the training sample corresponding to the xth particle.

10. A low shrinkage concrete, used to implement the detection of shrinkage performance of low shrinkage concrete according to any one of claims 1 to 9, characterized in that: The low shrinkage concrete is composed of the following raw materials in parts by weight: 100-150 parts of low-heat cement, 50-90 parts of fly ash, 500-900 parts of sand, 600-900 parts of crushed stone, 50-100 parts of water and 1.0-1.5 parts of shrinkage reducer.

Images