CN116759031B - Design method of sludge ash concrete material mixing ratio based on ANN - Google Patents

Abstract

The invention discloses a design method of sludge ash concrete material mixing ratio based on ANN, which comprises the following steps: collecting and constructing an initial data set of raw material mixing proportion and mechanical property parameters of sludge ash concrete; randomly dividing an initial data set into a training set and a testing set; performing iterative training construction based on the training set to obtain an ANN model, verifying the ANN model according to the testing set, outputting the constructed ANN model if verification is passed, and re-executing the training process to optimize the ANN model if verification is not passed until the ANN model is passed; and inputting the performance index of the sludge ash concrete material to be prepared into an ANN model which is based on verification, and outputting the mixing ratio of the sludge ash concrete raw materials through the ANN model. The invention can solve the problems of long time consumption, large workload and the like of the existing concrete mixing proportion design method.

Description

An ANN-based design method for sludge ash concrete material mix ratio

Technical field

The invention relates to the technical field of concrete mix proportion design. Specifically, it is a design method of sludge ash concrete material mix ratio based on ANN.

Background technique

Sludge ash refers to the ash obtained after drying, burning and other treatments of sludge produced during sewage treatment. Sludge ash has high activity, small particles, and certain cement activity. It can replace part of the cement or aggregate in concrete to achieve the purpose of resource reuse and reducing environmental pollution. It is of great significance to realize the utilization of urban solid waste, reduce environmental pollution, reduce the use of cement, and help achieve the "double carbon" goal.

However, the properties of sludge ash are complex, and its impact on concrete strength, durability and other properties is significantly different from traditional concrete materials. In order to ensure the stability and reliability of sludge ash concrete, reasonable mix ratio design must be carried out. The current mix proportion design methods are mostly mass method or volume method, but these methods are time-consuming and require a large workload, and cannot meet the mix design requirements of sludge ash concrete. In this context, it is necessary to explore new, fast, and efficient design methods.

Contents of the invention

To this end, the technical problem to be solved by the present invention is to provide a design method of sludge ash concrete material mix ratio based on ANN (Artificial Neural Network, artificial neural network), so as to solve the problem of the existing sludge ash concrete mix ratio design method that consumes Issues such as time duration and heavy workload.

In order to solve the above technical problems, the present invention provides the following technical solutions:

An ANN-based design method for the material mix of sludge ash concrete, including the following steps:

Step (1): Collect and construct an initial data set of sludge ash concrete raw material mix ratio and mechanical property parameters;

Step (2): Randomly divide the initial data set into a training set and a test set;

Step (3): Use the training set to perform iterative training to construct an ANN model, and verify the trained ANN model based on the test set to test the fit between the mix ratio output by the ANN model and the actual mix ratio; if the verification passes Output the constructed ANN model. If the verification fails, re-execute the training process to optimize the ANN model until the ANN model passes the verification;

Step (4): Input the performance indicators of the sludge ash concrete material to be prepared into the verified ANN model, and output the mix ratio of the sludge ash concrete raw materials through the ANN model.

In the above ANN-based design method of sludge ash concrete material mix ratio, in step (3), the ANN model construction method is:

Step (3-1), use the sludge ash concrete raw material mix ratio as the output variable and the mechanical property parameters as the input variables, and normalize the input variables and output variables respectively;

Step (3-2): Determine the loss function of the ANN model, adjust the internal parameters of the ANN model based on the gradient descent optimization algorithm, and use the Bayesian optimization algorithm to adaptively adjust and select the hyperparameters of the ANN model so that the training iterations are The loss function of the ANN model is reduced to an acceptable level;

The weight w and bias b are set in each layer of neurons in the ANN; for the known input x and output y, the neural network continuously adjusts w and b during training to fit the relationship between the existing y and x , the simplified relational expression y=wx+b is obtained at the end of the training, w and b are the huge parameter matrices trained by the ann model; the gradient descent algorithm used in the present invention is an algorithm for optimizing the internal parameters of the model, which belongs to the model parameters Optimization algorithm, the adjusted model parameter is w in the model expression y=wx+b. The Bayesian hyperparameter optimization algorithm is an algorithm for selecting and setting hyperparameters of ANN models. It is an automatic parameter adjustment algorithm and can replace manual parameter adjustment, grid search and other methods. It is convenient, short-time consuming, and can determine the optimal hyperparameter. Parameters can solve the problem that it is time-consuming and laborious to adjust parameters based on empirical methods alone, and it is impossible to obtain hyper-parameters that optimize model performance; hyper-parameters include: the number of iterations, the number of network layers, and other hyper-parameters that need to be set during modeling;

Step (3-3), use the test set to verify the trained ANN model, input the input variables of the test set into the trained ANN model, and use the Min-Max denormalization algorithm to map the output results to the real interval. And compare it with the output variables of the test set;

Step (3-4), use the root mean square error and coefficient of determination to evaluate the ANN model training results. If the evaluation objectives are met, the training model will be output as the ANN model. If the evaluation objectives are not met, the training process will be re-executed until the evaluation is met. Goal requirements.

In the above ANN-based design method of sludge ash concrete material mix ratio, in step (3-1), the Min-Max normalization method is used to normalize the input variables and output variables, and the calculation formula of the normalization process is for:

(1);

In formula (1), x represents an unnormalized value, represents the value after normalization; max represents the maximum value in the same batch of data, and min represents the minimum value in the same batch of data;

In step (3-2), the loss function of the ANN model is:

(2);

In formula (2), n represents the total number of samples in the training set, Indicates the predicted value of the mix ratio given by the model,/> Indicates the actual mix ratio after normalization; when the loss value is in the range of 0 to 0.05, it indicates that the model loss function has reduced to an acceptable level;

In step (3-3), the calculation formula of the Min-Max denormalization algorithm is:

(3);

In formula (3), x represents an unnormalized value, represents the value after normalization; max represents the maximum value in the same batch of data, and min represents the minimum value in the same batch of data;

In step (3-4), the calculation formula of ^the root mean square error RMSE and the coefficient of determination R2 is:

(4);

(5);

In Formula (4) and Formula (5), n represents the total number of samples in the test set, Represents the predicted value after denormalization,/> Represents the actual value,/> Represents the mean of the actual values of n test set samples; when the root mean square error RMSE is less than or equal to 0.01, and the coefficient of determination R2 is greater than or equal to 0.95, it indicates that the training model meets the requirements; the closer the value of ^the root mean square error RMSE is to 0 , indicating that the better the fitting effect of the model is, the closer the value of the determination coefficient R ² is to 1, indicating the better the fitting effect of the model. Therefore, if the value of the root mean square error RMSE is large or the value of the coefficient of determination R2 is small, the training process should be adjusted and re-executed until the ^root mean square error RMSE and the coefficient of determination R2 meet ^the requirements.

In the above ANN-based design method of sludge ash concrete material mix ratio, in step (3-2), the Bayesian hyperparameter optimization algorithm is used to adaptively adjust and select the hyperparameters of the ANN model to improve the construction of the ANN model. mold accuracy, and finally obtain a highly accurate ANN design model of sewage ash concrete mix ratio; the specific tuning method is:

First, design or adjust the optimization space Θ of each hyperparameter in the ANN model based on experience or model testing performance to reduce computational complexity; the optimization space can be adjusted based on the prediction performance of subsequent models;

Secondly, set the objective function h ( θ ) for hyperparameter tuning:

(6);

In formula (6), n represents the total number of samples, Represents the predicted value,/> represents the actual value, θ represents the hyperparameter;

Then, the Bayesian hyperparameter optimization method is used to solve to obtain the hyperparameter that minimizes the h ( θ ) value; the solution process is expressed as:

(7);

In formula (7), θ * is the optimal hyperparameter that the Bayesian hyperparameter optimization algorithm needs to find, θ is the input hyperparameter, Θ is the set parameter space, n represents the total number of samples, Represents the predicted value,/> represents the actual value.

The above-mentioned ANN-based design method of sludge ash concrete material mix ratio and the Bayesian hyperparameter optimization algorithm to find the optimal hyperparameters are:

The objective function h ( θ ) obeys Gaussian distribution, that is:

(8);

In formula (8), μ ( θ ) is the mean value of h ( θ ), O ( θ , θ ´) is the covariance matrix of h ( θ ), and the initial value of O ( θ , θ' ) is expressed as:

(9);

In the Bayesian hyperparameter optimization process, the covariance matrix O ( θ , θ' ) will change with the iterations of training. As the training iterations proceed, assume that the hyperparameter input at step t+1 is θ _{t+ 1} , then the value of O ( θ , θ' ) is expressed as:

(10);

Therefore, the calculation formula for the posterior probability of the objective function h ( θ ) is:

(11);

In equation (11), D is the observation value, μ _{t+ 1} ( θ ) is the mean value of h ( θ ) at step t+1, and σ ² _{t+ 1} ( θ ) is the variance of h ( θ ) at step t+1;

After obtaining the posterior probability, the Bayesian hyperparameter optimization algorithm searches for the space near the last hyperparameter according to the posterior distribution. The search method is:

(12);

In formula (12), ζ _t+1 is a constant, set to 0.01; θ _t+1 is the selected hyperparameter of step t+1;

Through continuous iterative search, the Bayesian hyperparameter optimization algorithm can be used to determine the optimal hyperparameters in a given hyperparameter optimization space Θ. This set of hyperparameters can enable the ANN model to achieve optimal training results.

The above-mentioned ANN-based design method of sludge ash concrete material mix ratio, the ANN model hyperparameters tuned by the Bayesian hyperparameter optimization algorithm include the number of hidden layers, the number of training iterations, the optimizer, the batch sample size, the learning rate, and the activation function. and discard ratio.

In the above ANN-based design method of sludge ash concrete material mix ratio, in step (1), the initial data set of sludge ash concrete raw material mix ratio and mechanical properties is obtained by collecting public data and/or experiments.

In the above ANN-based design method of sludge ash concrete material mix ratio, in step (2), the data in the initial data set are randomly divided into a training set and a test set in a ratio of 7:3.

In the above ANN-based design method of sludge ash concrete material mix ratio, in step (1), the raw materials of sludge ash concrete include sludge ash, cement, gravel, sand, water reducing agent, thickener, expansion agent and water. ; Mechanical performance parameters include compressive strength, flexural strength, slump and durability.

For the above-mentioned ANN-based design method of sludge ash concrete material mix ratio, the durability is one or two or more of the chloride ion permeability coefficient, frost resistance and alkali aggregate reaction.

The technical solution of the present invention achieves the following beneficial technical effects:

1. Since the influence of sludge ash on the performance of concrete is relatively complex, using existing mix proportion design methods is time-consuming and requires a large workload, and it is difficult to meet the complex mix design requirements of sludge ash concrete. The mix ratio design model based on ANN and Bayesian hyperparameter optimization proposed by this invention not only has a much higher prediction accuracy than the existing technology, but also has a higher level of intelligence and automation, making it more suitable for engineering practice.

2. The present invention is designed based on the performance indicators of sludge ash concrete, uses the ANN artificial neural network method to model the performance parameters such as the mix ratio and compressive strength of various additives of sludge ash concrete, and uses Bayesian super The parameter optimization algorithm tunes multiple hyperparameters of the ANN model. Compared with the manual tuning method based on experience, it can fully utilize the fitting ability of the ANN model and obtain the optimal design model.

3. The present invention uses ANN to construct a regression prediction model of the mix ratio of each raw material of sludge ash concrete, replacing the traditional mass method, volume method and mathematical relational expressions. It can model up to 14 variables and better simulate Combining the complex nonlinear relationship between the performance indicators and mix ratio of sludge ash concrete, we can obtain a mix ratio design that is closer to the predetermined performance indicators.

Description of the drawings

Figure 1 is a schematic diagram of the model training process when constructing an ANN model in the embodiment of the present invention;

Figure 2 is a schematic diagram of the model testing process when constructing the ANN model in the embodiment of the present invention;

Figure 3 is a comparison chart between model prediction values and actual values based on the validation set in the embodiment of the present invention.

Detailed ways

The ANN-based sludge ash concrete material mix design method in this embodiment includes the following steps:

Step (1): Collect and construct an initial data set of sludge ash concrete raw material mix ratio and mechanical property parameters; in this embodiment, the sludge ash concrete raw material mix ratio and mechanical property data are obtained by collecting online public data and experimental collection. Build the initial data set.

Step (2): Randomly divide the initial data set into a training set and a test set; the compressive strength, flexural strength, slump, and durability in the initial data set (durability includes chloride ion permeability coefficient, frost resistance, alkali resistance, etc. The value of aggregate reaction) is used as the input variable, and the mixing ratio of sludge ash, cement, gravel, sand, water, water reducing agent, thickener and expanding agent is used as the output variable; the initial data set is based on the ratio of 7:3 It is randomly divided into a training set and a test set. The training set is used to train the model to obtain the optimal fitting effect. The test set is used to verify and evaluate the calculation effect of the training model on the sludge ash concrete mix ratio.

Step (3): Use the training set to perform iterative training to build an ANN model, and verify the trained ANN model based on the test set. If the verification passes, the constructed ANN model will be output. If the verification fails, the training process will be re-executed to optimize the ANN. model until the ANN model passes the verification; that is, construct the ANN model and iteratively train the model based on the training set so that the model can output results that are closer and closer to the output variable based on the values of the input variables in the training set. In order to obtain a hyperparameter combination that improves the performance of the ANN model, this embodiment uses the Bayesian hyperparameter optimization algorithm to adaptively tune model hyperparameters such as the number of iterations and the learning rate. Use the test set to verify the ANN model based on the optimal parameter combination, and use the root mean square error and coefficient of determination as evaluation indicators to evaluate the training effect of the model.

Specifically, the construction method of the ANN model is:

Step (3-1): Use the sludge ash concrete raw material mix ratio as the output variable and the mechanical property parameters as the input variables. Normalize the input variables and output variables respectively; use the Min-Max normalization method to normalize the input variables. And the output variables are normalized. The calculation formula of normalization is:

(1);

In formula (1), x represents an unnormalized value, represents the normalized value; max represents the maximum value in the same batch of data, and min represents the minimum value in the same batch of data.

Step (3-2): Determine the loss function of the ANN model, adjust the internal parameters of the ANN model based on the gradient descent optimization algorithm, and use the Bayesian optimization algorithm to adaptively adjust and select the hyperparameters of the ANN model so that the training iterations are The loss function of the ANN model is reduced to an acceptable level; the loss function of the ANN model is:

(2);

In formula (2), n represents the total number of samples in the training set, Indicates the predicted value of the mix ratio given by the model,/> Indicates the actual mix ratio after normalization; when the loss value is in the range of 0 to 0.05, it indicates that the model loss function has reduced to an acceptable level;

The ANN model hyperparameters tuned by the Bayesian hyperparameter optimization algorithm include the number of hidden layers, the number of training iterations, the optimizer, the batch sample size, the learning rate, the activation function and the dropout ratio; the Bayesian hyperparameter optimization algorithm has a positive impact on the ANN model. The method for tuning the internal hyperparameters is:

First, design or adjust the optimization space Θ of each hyperparameter in the ANN model based on experience or model test performance;

Secondly, set the objective function h ( θ ) for hyperparameter tuning:

(6);

In formula (6), n represents the total number of samples, Represents the predicted value,/> represents the actual value, θ represents the hyperparameter;

Then, the Bayesian hyperparameter optimization method is used to solve to obtain the hyperparameter that minimizes the h ( θ ) value; the solution process is expressed as:

(7);

In formula (7), θ * is the optimal hyperparameter that the Bayesian hyperparameter optimization algorithm needs to find, θ is the input hyperparameter, Θ is the set parameter space, n represents the total number of samples, Represents the predicted value,/> represents the actual value.

Step (3-3), use the test set to verify the trained ANN model, input the input variables of the test set into the trained ANN model, and use the Min-Max denormalization algorithm to map the output results to the real interval. And compare it with the output variables of the test set; the calculation formula of the Min-Max denormalization algorithm is:

(3);

In formula (3), x represents an unnormalized value, represents the value after normalization; max represents the maximum value in the same batch of data, and min represents the minimum value in the same batch of data;

Step (3-4), use the root mean square error and coefficient of determination to evaluate the ANN model training results. If the evaluation objectives are met, the training model will be output as the ANN model. If the evaluation objectives are not met, the training process will be re-executed until the evaluation is met. Goal requirements. The calculation formula of ^the root mean square error RMSE and the coefficient of determination R2 is:

(4);

(5);

In Formula (4) and Formula (5), n represents the total number of samples in the test set, Represents the predicted value after denormalization,/> Represents the actual value,/> Represents the mean of the actual values of n test set samples; when ^the root mean square error RMSE is less than or equal to 0.01, and the coefficient of determination R2 is greater than or equal to 0.95, it indicates that the training model meets the requirements.

The method for finding optimal hyperparameters using the Bayesian hyperparameter optimization algorithm is:

The objective function h ( θ ) obeys Gaussian distribution, that is:

(8);

In formula (8), μ ( θ ) is the mean value of h ( θ ), O ( θ , θ ´) is the covariance matrix of h ( θ ), and the initial value of O ( θ , θ' ) is expressed as:

(9);

As the training iteration proceeds, assuming that the hyperparameter input at step t+1 is θ _{t+ 1} , the value of O ( θ , θ' ) is expressed as:

(10);

Therefore, the calculation formula for the posterior probability of the objective function h ( θ ) is:

(11);

In equation (11), D is the observation value, μ _{t+ 1} ( θ ) is the mean value of h ( θ ) at step t+1, and σ ² _{t+ 1} ( θ ) is the variance of h ( θ ) at step t+1;

After obtaining the posterior probability, the Bayesian hyperparameter optimization algorithm searches for the space near the last hyperparameter according to the posterior distribution. The search method is:

(12);

In formula (12), ζ _t+1 is a constant, set to 0.01; θ _t+1 is the selected hyperparameter of step t+1;

Through continuous iterative search, the optimal hyperparameters in the given hyperparameter optimization space Θ are determined.

In the process of building an ANN model, the number of hidden layers, number of iterations, learning rate and other hyperparameters of the ANN model do not need to be set in advance. They are adaptively determined by the Bayesian hyperparameter optimization algorithm during the training process and only need to be given in advance. Optimization space of hyperparameters; the optimization space given in this embodiment is shown in Table 1.

Table 1 Hyperparameter optimization space

Before the data is input into the model, it must first undergo Min-Max normalization processing; after the model outputs the results, it must also undergo denormalization calculation processing to obtain the predicted value within the true value range.

There are 84 sets of training data in the initial data set collected in this embodiment, and the remaining 36 sets are used as test sets to verify the performance of the model after training (due to the large amount of data, this embodiment does not list each data). During _the model training _process , the compressive strength ( X ₁ ), flexural strength ( X ₂ ), _slump ( , alkali _aggregate reaction _{( _} _ _{_} _ _{_} ( Y ₅ ), water reducing agent ( Y ₆ ), thickener ( Y ₇ ), and expansion agent ( Y ₈ ), a total of eight groups of variables.

After training, you can view the hyperparameter group obtained by Bayesian hyperparameter optimization in this training. The hyperparameter set in this embodiment is shown in Table 2.

Table 2 Hyperparameter groups obtained by Bayesian hyperparameter optimization

Finally, 36 sets of test data were used to verify the performance of the model, and two evaluation indicators, root mean square error ( RMSE ) and coefficient of determination ( R ² ), were used to evaluate the model fitting effect. After calculation, it is found that RMSE =0.00263 and R ² =0.98221. The comparison results between the predicted values given by the model and the actual values are shown in Figure 3.

The sludge ash concrete mix proportion modeling in this embodiment achieved test results of RMSE =0.00263 and R ² =0.98221 in the test set. According to the formulas and meanings of RMSE and R ² , it can be seen that the closer the value of RMSE is to 0, the closer the value of R ² is to 1, indicating that the predicted value given by the model is closer to the actual value. Therefore, it can be seen from the above evaluation indicators that the ANN model constructed in this embodiment fits the performance index and mix ratio of sludge ash concrete very well, indicating that the ANN model can accurately predict the relationship between the performance index and mix ratio of sludge ash concrete. With sufficient learning, the required mix ratio can be given based on the expected performance index values. It can also be concluded from the comparison between the predicted values and actual values in Figure 2: ANN combined with the Bayesian hyperparameter optimization method gives mix ratio regression prediction results that are very close to the actual values.

Step (4): Save the verified ANN model. During engineering application, first enter the expected sludge ash concrete performance index value, and then the ANN model will be calculated based on the trained parameters to output the matching Compare. The sludge ash concrete is prepared according to this mix ratio.

In this embodiment, the sludge ash concrete prepared according to the mix ratio output by the ANN model was experimentally tested on the performance indicators of the sludge ash concrete. The results found that the performance of the sludge ash concrete configured according to the mix ratio output by the ANN model was in line with the expected requirements. The achieved performance indicators are basically consistent and fully meet the needs of engineering applications.

In some other embodiments, the type of sludge ash concrete raw material can be adjusted as an output variable according to the actual situation, or the index type of mechanical property data can be adjusted as an input variable according to the application indicators of sludge ash concrete to construct a new ANN model.

Obviously, the above-mentioned embodiments are only examples for clear explanation and are not intended to limit the implementation. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. An exhaustive list of all implementations is neither necessary nor possible. The obvious changes or modifications derived therefrom are still within the protection scope of the claims of this patent application.

Claims (4)

1. The design method of the sludge ash concrete material mixing ratio based on the ANN is characterized by comprising the following steps of:

step (1), collecting and constructing an initial data set of raw material mixing proportion and mechanical property parameters of sludge ash concrete; the sludge ash concrete raw materials comprise sludge ash, cement, broken stone, sand, a water reducing agent, a thickening agent, an expanding agent and water; mechanical properties include compressive strength, flexural strength, slump and durability;

step (2), randomly dividing the initial data set into a training set and a testing set;

performing iterative training by using a training set to construct an ANN model, verifying the ANN model obtained by training according to a test set, outputting the constructed ANN model if verification is passed, and re-executing the training process to optimize the ANN model if verification is not passed until the ANN model is passed;

step (4), inputting performance indexes of the sludge ash concrete material to be prepared into an ANN model passing verification, and outputting the mixing ratio of the sludge ash concrete raw materials through the ANN model;

in the step (3), the construction method of the ANN model comprises the following steps:

step (3-1), taking the mixing ratio of sludge ash concrete raw materials as an output variable and taking mechanical property parameters as an input variable, and respectively carrying out normalization treatment on the input variable and the output variable;

in the step (3-1), a Min-Max normalization method is adopted to normalize the input variable and the output variable, and a calculation formula of the normalization is as follows:

（1）；

in the formula (1), the components are as follows,indicating non-normalized values, ++>Representing the value after normalization; />Represents the maximum value in the same batch of data, +.>Representing the minimum value in the same batch of data;

step (3-2), determining a loss function of the ANN model, adjusting internal parameters of the ANN model based on a gradient descent optimization algorithm, and adopting a Bayesian optimization algorithm to adaptively adjust and select the super parameters of the ANN model so as to reduce the loss function of the ANN model in training iteration to an acceptable level;

in step (3-2), the loss function of the ANN model is:

（2）；

in the formula (2), the amino acid sequence of the compound,representing the total number of training set samples, +.>Representing the predicted value of the mix given by the model, < + >>Representing the normalized actual mix ratio;lossvalues in the range of 0 to 0.05 indicate that the model loss function is reduced to an acceptable level;

step (3-3), verifying the trained ANN model by using the test set, inputting the input variable of the test set into the trained ANN model, mapping the output result to a real interval by using a Min-Max inverse normalization algorithm, and comparing the output result with the output variable of the test set;

in the step (3-3), the Min-Max inverse normalization algorithm has a calculation formula:

（3）；

in the formula (3), the amino acid sequence of the compound,indicating non-normalized values, ++>Representing the value after normalization; />Represents the maximum value in the same batch of data, +.>Representing the minimum value in the same batch of data;

step (3-4), evaluating the training result of the ANN model by using root mean square error and a decision coefficient, outputting the training model as the ANN model if the evaluation target is met, and re-executing the training process if the evaluation target is not met until the requirement of the evaluation target is met;

in step (3-4), root mean square errorRMSEDetermining coefficientsR ² The calculation formula of (2) is as follows:

（4）；

（5）；

in the formulas (4) and (5),representing the total number of test set samples, +.>Representing the inverse normalized predicted value,/->Representing the actual value +.>Representation->An average of actual values of the individual test set samples; when root mean square errorRMSELess than or equal to 0.01 and determining coefficientsR ² When the training model is more than or equal to 0.95, the training model is shown to meet the requirements;

in the step (3-2), adopting a Bayes super-parameter optimization algorithm to carry out self-adaptive adjustment and selection on the super parameters of the ANN model; the ANN model super parameters optimized by the Bayesian super parameter optimization algorithm comprise the number of hidden layers, the training iteration times, an optimizer, batch sample size, learning rate, activation function and discarding method ratio; the specific tuning method comprises the following steps:

firstly, designing or adjusting an optimization space Θ of each super parameter in an ANN model according to the performance of experience or model test;

secondly, setting an objective function of super parameter tuningh(θ)：

（6）；

In the formula (6), the amino acid sequence of the compound,representing the total number of samples->Representing predicted values +.>The actual value is represented by a value that is,θrepresenting the super-parameters;

then, solving by using a Bayes super-parameter optimization method to obtain the leadh(θ) Super-parameters with minimum values; the solving process is expressed as:

（7）；

in the formula (7), the amino acid sequence of the compound,θ* The optimal super-parameters to be searched for by the Bayes super-parameter optimization algorithm,θfor the input hyper-parameters, Θ is the set parameter space,representing the total number of samples->Representing predicted values +.>Representing the actual value;

the method for searching the optimal super-parameters by the Bayes super-parameter optimization algorithm comprises the following steps:

objective functionh(θ) Obeys gaussian distribution, i.e.:

（8）；

in the formula (8), the amino acid sequence of the compound,μ(θ) Is thath(θ) Is used for the average value of (a),O(θ,θ(v) ish(θ) Is used for the co-variance matrix of (a),O(θ,θ’) The initial value of (2) is expressed as:

（9）；

as training iterations progress, assume that the super-parameters entered at step t+1 areθ _t+1 ThenO(θ,θ’) The values of (2) are expressed as:

（10）；

thus, the objective functionh(θ) The posterior probability of (2) is calculated by the following formula:

（11）；

in the formula (11), the amino acid sequence of the compound,Din order to observe the value of the value,μ _t+1 (θ) Is t+1st steph(θ) Mean, sigma of ² _t+1 (θ) Is t+1st steph(θ) Is a variance of (2);

after posterior probability is obtained, searching the space near the last super parameter according to posterior distribution by a Bayes super parameter optimization algorithm, wherein the searching method comprises the following steps:

（12）；

zeta in (12) _t+1 Is a constant, set to 0.01;θ _t+1 is the super parameter of the selected t+1 step;

by constant iterative search, the optimal superparameter in a given superparameter optimization space Θ is determined.

2. The method for designing the mix proportion of sludge ash concrete materials based on ANN according to claim 1, wherein in the step (1), the initial data set of the mix proportion of sludge ash concrete raw materials and mechanical properties is obtained by collecting public data and/or experimental means.

3. The method for designing a mix ratio of sludge ash concrete materials based on ANN according to claim 1, wherein in step (2), the data in the initial data set is randomly divided into a training set and a test set according to a ratio of 7:3.

4. The method for designing the mix ratio of the ANN-based sludge ash concrete material according to claim 1, wherein the durability is one or two or more of chloride ion permeability coefficient, freezing resistance and alkali aggregate reaction.