Method for estimating fluidity of concrete, computer program product, and storage medium

Abstract

The invention discloses a concrete fluidity presumption method, a computer program product and a storage medium, and the method comprises the steps: building a concrete fluidity presumption model, taking the plastic viscosity data and dynamic yield stress data of concrete as the input variables of an artificial neural network; slump data and expansion data of the concrete serve as output variables of the artificial neural network; acquiring plastic viscosity data, dynamic yield stress data, slump data and expansion data corresponding to a plurality of parts of fresh concrete, constructing a data set, dividing the data set into a training set, a verification set and a test set, and learning the concrete fluidity presumption model; plastic viscosity data and dynamic yield stress data of to-be-detected fresh concrete are imported into the model, corresponding slump data and expansion data are presumed through the concrete fluidity presumption model, and the fluidity of the to-be-detected fresh concrete is judged through the obtained slump data and expansion data.

Description

Technical Field

[0001] The present invention relates to the technical field of material flow characteristic determination, and in particular to a method for estimating concrete fluidity, a computer program product and a storage medium. Background Art

[0002] With the rapid development of the production and construction industry, concrete, as an important building material, is widely used in bridges, railways, buildings, roads and other fields. There are many types of fresh concrete, including self-compacting concrete, high-performance concrete, superfluid concrete, etc. The quality of concrete directly affects the quality of commercial concrete construction projects.

[0003] Among them, the fluidity of concrete plays a key role in actual engineering. It is not only a characteristic of an engineering material, but also one of the important factors to ensure construction speed, engineering quality, safety and durability. In order to ensure that the quality of concrete meets the design requirements, prevent quality problems that may occur during construction, adjust the mix ratio, and improve construction efficiency and quality, a slump test must be carried out before construction. At present, the concrete fluidity test method mainly relies on traditional slump test and expansion test. The fluidity of concrete can be determined by measuring its slump and expansion. These two indicators reflect the flow characteristics and expansion range of concrete mixtures without external forces. However, there are problems such as the subjectivity of the test results, the extensive test methods, the cumbersome test process, and the high requirements for test personnel, which make it difficult to accurately express the fluidity of concrete. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by the present invention is: how to provide a method for estimating concrete fluidity with a simple operation process and stable and accurate test results.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A method for estimating concrete fluidity comprises the following steps:

[0007] (1) A concrete fluidity estimation model was established, with the plastic viscosity data and dynamic yield stress data of concrete as input variables of the artificial neural network, and the slump data and expansion data of concrete as output variables of the artificial neural network;

[0008] (2) obtaining rheological test data and fluidity test data of several fresh concretes, obtaining corresponding plastic viscosity data, dynamic yield stress data, slump data, and expansion data, constructing a data set, randomly dividing the data set into a training set, a validation set, and a test set in proportion, using the training set to train, the validation set to evaluate, and the test set to test the concrete fluidity estimation model, and learning the concrete fluidity estimation model;

[0009] (3) The plastic viscosity data and dynamic yield stress data of the fresh concrete to be tested are imported into the concrete fluidity estimation model, and the corresponding slump data and spread data are estimated by the concrete fluidity estimation model. The fluidity of the fresh concrete to be tested is determined by the obtained slump data and spread data.

[0010] As an optimization, before the concrete fluidity estimation model is learned, the input variables and output variables in the data set are normalized respectively, and the normalized threshold is [-1, 1]; the hyperbolic tangent function is used as the activation function of the hidden layer in the concrete fluidity estimation model, and the linear function is used as the activation function of the output layer in the concrete fluidity estimation model; the output value in the concrete fluidity estimation model is denormalized to obtain the concrete fluidity estimation value, the learning rate is set to (0.1, 0.01, 0.001, 0.0001), and the momentum factor is set to (0.5-0.95).

[0011] As an optimization, the division ratio of the training set, validation set and test set is (60-80):(10-20):(10-20).

[0012] As an optimization, when training the concrete fluidity estimation model, the Levenberg-Marquardt algorithm is used to train the concrete fluidity estimation model.

[0013] A computer program product comprises a computer program, and when the computer program is executed by a computer, the method described above is implemented.

[0014] A computer-readable storage medium stores a computer program, which, when executed by a computer, implements the method described above.

[0015] Compared with the prior art, the present invention has the following advantages: compared with the existing concrete fluidity testing technology, the present invention adopts the concrete rheology testing method to infer the concrete fluidity, which can directly study the performance of concrete materials instead of judging the concrete fluidity by empirical methods. At the same time, the rheology testing method is more convenient, intelligent, requires fewer personnel, saves time and effort, and has accurate and stable results compared with the slump and expansion tests. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flow chart for establishing and using the model in the present invention;

[0017] Figure 2 It is a heat map of the correlation between input variables and output variables in the present invention;

[0018] Figure 3 It is an analysis diagram of the estimated value and the measured value of the concrete slump in the present invention;

[0019] Figure 4 This is an analysis chart of the estimated and measured values ​​of concrete expansion in the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0021] like Figure 1 As shown, the method for estimating concrete fluidity in this specific embodiment includes the following steps:

[0022] (1) Establishing a concrete fluidity estimation model, using the plastic viscosity data and dynamic yield stress data of concrete as input variables of the artificial neural network, and using the slump data and expansion data of concrete as output variables of the artificial neural network; (2) Obtaining a number of rheological test and fluidity test data of freshly mixed concrete, obtaining the corresponding plastic viscosity data, dynamic yield stress data, slump data and expansion data, constructing a data set, randomly dividing the data set into a training set, a validation set and a test set in proportion, using the training set to train, the validation set to evaluate and the test set to test the concrete fluidity estimation model, and learning the concrete fluidity estimation model;

[0023] (3) The plastic viscosity data and dynamic yield stress data of the fresh concrete to be tested are imported into the concrete fluidity estimation model, and the corresponding slump data and spread data are estimated by the concrete fluidity estimation model. The fluidity of the fresh concrete to be tested is determined by the obtained slump data and spread data.

[0024] like Figure 2As shown, the correlation verification of model parameters: Shapiro-Wilk test variable normality is used, and the model input and output variables are in line with normal distribution. Pearson correlation coefficient is used to analyze the correlation between variables. The input variables and output variables have a strong and significant correlation, which can indicate that the slump and expansion of concrete will be affected by the changes in the plastic viscosity and dynamic yield stress of concrete. Therefore, it is feasible to establish a concrete fluidity estimation model based on the plastic viscosity and dynamic yield stress of concrete. The slump estimation structure in the concrete fluidity estimation model is determined to be 2-6-1 by the empirical method and trial and error method, indicating that the input parameter is 2, with a hidden layer structure, the number of hidden layer neurons is 6, and the final output parameter is 1; the concrete expansion estimation structure is 2-3-1.

[0025] In this specific implementation, before the concrete fluidity estimation model is learned, the input variables and output variables in the data set are normalized respectively, and the threshold after normalization is [-1,1]; the hyperbolic tangent function is used as the activation function of the hidden layer in the concrete fluidity estimation model, so that the neural network can learn and represent complex nonlinear relationships, and the linear function is used as the activation function of the output layer in the concrete fluidity estimation model, which can improve the expression ability of the model. This function has the characteristics of simplicity, rapidity, stability, etc., and conforms to the regression model of input variables and output variables; the output value in the concrete fluidity estimation model is denormalized to obtain the concrete fluidity estimation value, and the learning rate is set to 0.01 and the momentum factor is set to 0.9. The input parameters and output parameters are normalized to avoid deviations in the weight distribution of the model to the features due to different dimensions. Normalizing the data before modeling can speed up the calculation speed, avoid excessive sensitivity of the model to the data, and improve the performance and accuracy of the model.

[0026] In this specific implementation, the division ratio of the training set, the validation set and the test set is 70:15:15.

[0027] In this specific implementation, when training the concrete fluidity estimation model, the Levenberg-Marquardt algorithm is used to train the concrete fluidity estimation model, which has the characteristics of fast convergence to the local optimal solution, better description of the characteristics of the observed data, better interpretability, etc.

[0028] A computer program product comprises a computer program, and when the computer program is executed by a computer, the method described above is implemented.

[0029] A computer-readable storage medium stores a computer program, which, when executed by a computer, implements the method described above.

[0030] Eight groups of concrete tests were designed to carry out concrete rheological tests and concrete slump and expansion tests at the same time. The slump and expansion tests were conducted by professionals, and the concrete plastic viscosity and dynamic yield stress were tested using a concrete rheometer. The obtained concrete plastic viscosity and dynamic yield stress were used as input parameters, and the fluidity (slump and expansion) test results measured by the test personnel were used as true values. The plastic viscosity and dynamic yield stress measured by the rheological experiment were imported as independent variables into the constructed concrete fluidity estimation model. The plastic viscosity value was stored in the first row of the numerical matrix X, and the dynamic yield stress value was stored in the second row of the numerical matrix X. The command sim_y=sim(net,X) was used to call the trained network to estimate the slump and expansion of concrete. The estimated value of slump was stored in the numerical matrix H_y; the estimated value of expansion was stored in the numerical matrix L_y.

[0031]

[0032] H_y=[257 256 216 161 213 211 204 207]

[0033] L_y=[646 633 520 300 477 474 444 482]

[0034] Comparative analysis of the estimated and measured values ​​of concrete slump and expansion Figure 3 and Figure 4 The error comparison is shown in Table 1:

[0035]

[0036] Table 1

[0037] The mean absolute error MAE, root mean square error MSE, and determination coefficient R 2 Evaluate the test results.

[0038]

[0039] The above formula is used to calculate the average error MAE of the 8 groups of fresh concrete slump estimation results, which is 9.5, the root mean square error MSE is 113.5, and the determination coefficient R 2 The average error MAE of the estimated results of the expansion degree of 8 groups of fresh concrete is 113.75, the root mean square error MSE is 179.375, and the determination coefficient R 2 The estimated results of concrete slump and expansion have good determination coefficients. Comparative tests show that the present invention has practical significance and can be used to estimate the slump and expansion of fresh concrete and determine the fluidity of concrete.

[0040] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described with reference to the preferred embodiments of the present invention, it should be understood by those skilled in the art that various changes may be made in form and details without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A method for estimating concrete fluidity, characterized in that: The following steps are involved: (1) A concrete fluidity estimation model was established, with the plastic viscosity data and dynamic yield stress data of concrete as input variables of the artificial neural network, and the slump data and expansion data of concrete as output variables of the artificial neural network; (2) Obtaining rheological test data and fluidity test data of several fresh concretes, obtaining corresponding plastic viscosity data, dynamic yield stress data, slump data, and expansion data, constructing a data set, and randomly dividing the data set into a training set, a validation set, and a test set in proportion. Using the training set to train, the validation set to evaluate, and the test set to test the concrete fluidity estimation model, the concrete fluidity estimation model is learned; (3) The plastic viscosity data and dynamic yield stress data of the fresh concrete to be tested are imported into the concrete fluidity estimation model, and the corresponding slump data and spread data are estimated by the concrete fluidity estimation model. The fluidity of the fresh concrete to be tested is determined by the obtained slump data and spread data.

2. The method for estimating concrete fluidity according to claim 1, characterized in that: Before learning the concrete fluidity estimation model, the input variables and output variables in the data set are normalized, and the normalized threshold is [-1, 1]; The hyperbolic tangent function is used as the activation function of the hidden layer in the concrete fluidity estimation model, and the linear function is used as the activation function of the output layer in the concrete fluidity estimation model. The output value in the concrete fluidity estimation model is denormalized to obtain the concrete fluidity estimation value. The learning rate is set to (0.1, 0.01, 0.001, 0.0001), and the momentum factor is set to (0.5-0.95).

3. The method for estimating concrete fluidity according to claim 1, characterized in that: The ratio of training set, validation set and test set is (60-80): (10-20): (10-20).

4. The method for estimating concrete fluidity according to claim 1, characterized in that: When training the concrete fluidity estimation model, the Levenberg-Marquardt algorithm is used to train the concrete fluidity estimation model.

5. A computer program product, characterized in that: The method comprises a computer program, which, when executed by a computer, implements the method according to any one of claims 1 to 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a computer, the method according to any one of claims 1 to 4 is implemented.

Images