CN111488713A - Method, system and storage medium for predicting early carbonization of concrete - Google Patents
Method, system and storage medium for predicting early carbonization of concrete Download PDFInfo
- Publication number
- CN111488713A CN111488713A CN202010290133.2A CN202010290133A CN111488713A CN 111488713 A CN111488713 A CN 111488713A CN 202010290133 A CN202010290133 A CN 202010290133A CN 111488713 A CN111488713 A CN 111488713A
- Authority
- CN
- China
- Prior art keywords
- concrete
- random forest
- early
- carbonization
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003763 carbonization Methods 0.000 title claims abstract description 103
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000007637 random forest analysis Methods 0.000 claims abstract description 86
- 238000012549 training Methods 0.000 claims abstract description 42
- 238000013528 artificial neural network Methods 0.000 claims abstract description 18
- 238000002790 cross-validation Methods 0.000 claims abstract description 14
- 238000004364 calculation method Methods 0.000 claims abstract description 10
- 239000004568 cement Substances 0.000 claims description 19
- 238000011156 evaluation Methods 0.000 claims description 11
- 238000010835 comparative analysis Methods 0.000 claims description 6
- 239000000463 material Substances 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 239000010881 fly ash Substances 0.000 claims description 4
- 229910021487 silica fume Inorganic materials 0.000 claims description 4
- 238000003066 decision tree Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 238000013179 statistical model Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 12
- 230000000694 effects Effects 0.000 description 10
- 238000011160 research Methods 0.000 description 9
- 230000009471 action Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 3
- 230000006872 improvement Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 239000002994 raw material Substances 0.000 description 2
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 230000002301 combined effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of early concrete carbonization prediction, and discloses a method, a prediction system and a storage medium for early concrete carbonization prediction, wherein an initial concrete carbonization index system and an original training set are constructed; carrying out sample diversity, establishing parameters of a random forest model, and simultaneously carrying out cross validation on each index; calculating the importance degree of each index by using a random forest model, calculating the correlation coefficient of each index by using a Pearson function model, and comparing the calculation result of the random forest model with the calculation result of the Pearson function model; and judging the error sizes of the random forest model, the BP artificial neural network and the wavelet neural network by using the two parameters of the goodness-of-fit and the root-mean-square error. The method can effectively predict the change result of the early carbonization of the concrete, and provides a feasible method for realizing the prediction of the early carbonization of the concrete.
Description
Technical Field
The invention belongs to the technical field of concrete early carbonization prediction, and particularly relates to a method, a prediction system and a storage medium for predicting concrete early carbonization.
Background
Currently, carbonization of concrete is a chemical attack to which concrete is subjected. The properties of cement, the strength of concrete and the components of concrete are closely related to the carbonization of concrete. In recent years, improvement of the durability of concrete structures has been a problem of great concern, early carbonization of concrete has been one of factors affecting the durability of concrete, and has been the focus of research personnel, and many researchers are troubling the problems of rapid carbonization of concrete, serious consequences of early carbonization of concrete, how to reduce carbonization of concrete, and the like.
The prior art has developed relevant research around concrete carbonation analysis. Some researchers reveal from the microscopic aspect that the essence of the early carbonization of the concrete has the preliminary knowledge of the early carbonization phenomenon of the concrete, and most researches are to carry out deep excavation on the early carbonization characteristic based on the existing experimental data, which indicates that a great amount of experimental data support is needed if a further conclusion is to be obtained, thereby causing great labor and time waste and greatly increasing the research cost. Meanwhile, most researches mainly aim at the influence of a certain specific influence factor on early carbonization, and the action and effect of various influence factors are not comprehensively considered. In addition to the above two points, there are few studies considering which factors are prominent for early carbonization under the combined action of different influencing factors. Therefore, predicting the early carbonization property through the data of a plurality of influence factors accumulated in the early stage becomes a better research mode.
Through the above analysis, the problems and defects of the prior art are as follows: (1) in the prior art, a large amount of experimental data is required for supporting in concrete carbonization analysis, so that a large amount of labor and time are wasted, and the cost is also greatly increased.
(2) Most researches mainly relate to the influence of a certain specific influence factor on early carbonization, and the action and effect of various influence factors are not comprehensively considered.
(3) Few studies have considered ranking the importance of different influencing factors to early carbonization under their combined effect.
The difficulty in solving the above problems and defects is: in order to research the effect of different influencing factors on early carbonization of concrete, data information needs to be collected in various aspects. In order to explore the importance degree of the influence of different influencing factors on early carbonization, a machine learning model needs to be introduced to rank the influencing factors.
The significance of solving the problems and the defects is as follows: on the basis of saving a large amount of cost and time, the deep research on the early carbonization of the concrete can be realized. The comprehensive influence effect on the early carbonization of the concrete under the combined action of various influence factors can be researched. The importance degree of the influencing factors is analyzed, and the early carbonization of the concrete can be more effectively controlled and managed.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a method, a prediction system and a storage medium for predicting early carbonization of concrete. The influence of the influence factors on the early carbonization depth of the concrete is predicted through relatively few and relatively large influence factor data, and the influence of each influence factor on the early carbonization depth of the concrete is predicted and analyzed.
The invention is realized in such a way that a method for predicting early carbonization of concrete comprises the following steps:
the method comprises the following steps: constructing an initial index system and an initial training set for early carbonization of concrete;
step two: carrying out sample diversity on the constructed training set, establishing parameters of a random forest model, and simultaneously carrying out cross validation on each index;
step three: calculating the importance degree of each index by using a random forest model, calculating the correlation coefficient of each index by using a Pearson function model, and comparing the calculation result of the random forest model with the calculation result of the Pearson function model;
step four: and judging the error sizes of the random forest model, the BP artificial neural network and the wavelet neural network by using two parameters of goodness of fit and root mean square error, wherein the greater the goodness of fit is, the smaller the root mean square error is, the better the effect of the corresponding prediction model is, and otherwise, the worse the prediction effect is.
Further, the concrete initial index system in the step one comprises an early carbonization influence factor and an evaluation index, wherein the influence factor comprises 9 factors including cement strength, cement dosage, fly ash dosage, fine aggregate dosage, coarse aggregate dosage, concrete strength, silica fume dosage, water-cement ratio and average grain size; the early carbonization evaluation index is the carbonization depth.
Further, the original training set in the step one is: and taking different influence factors in the index system as variables of the random forest, collecting statistical relevant data, and taking corresponding data as an original training set.
Further, the step two of performing reasonable sample diversity on the created training set, establishing appropriate random forest model parameters, and performing cross validation on each index simultaneously includes the following steps:
the first step is as follows: randomly dividing a selected monitoring sample into 5 equal parts, optionally selecting 4 parts as a sample training set, wherein the sample training set is used for determining random forest parameters and constructing a random forest model, and the other 1 part is used as a sample test set and used for evaluating the performance of model prediction;
the second step is that: two important parameters of the random forest model: mtry is the number of influencing factors, and ntree is a tree of the decision tree; and (3) establishing a regression model by setting the variable number in the data set as P, wherein mtry is P or mtry is P/3 and ntree is 500.
Further, the third step of calculating the importance degree of each index by using a random forest model, and calculating the correlation coefficient of each index by using a Pearson function model, includes the following steps:
step 1: the method comprises the steps of calculating out-of-bag data errors of each tree in a random forest according to the OOB (calculated out and recorded as errOOB1, adding noise interference into all samples of the out-of-bag data OOB randomly, calculating out-of-bag data errors repeatedly and recorded as errOOB2, scoring importance of different influence factors, visually drawing the importance scores, and calculating the importance of a certain variable characteristic according to the following formula:
step 2: analyzing the correlation degree of different influencing factors and the early carbonization depth by utilizing a Pearson function model, wherein the Pearson sample correlation coefficient r is as follows:
wherein
Respectively representing the mean values of the samples of X, Y;
and step 3: and comparing and analyzing the calculation result of the importance of the random forest model and the calculation result of the correlation of the Pearson function model, and if the importance sequence in the random forest and the Pearson function correlation sequence have consistency, indicating that the prediction of the random forest has reliability.
Further, in the fourth step, the two parameters of root mean square error and goodness of fit are selected for evaluating the prediction result as the evaluation basis of the prediction accuracy of the model; the mean square error and goodness of fit expression is:
wherein, yobsIs the actual observed value, ypredIs a predicted value; r2The ratio of variability in a data set interpreted by a statistical model is used for measuring the fitting degree of a regression straight line to an observed value, the range of R2 is between 0 and 1, and the closer to 1, the more accurate the observed data is; the RMSE value is the sum of the individual differences between the estimated value of the estimator and the actual observed value, the RMSE value is equal to or greater than 0,closer to 0 indicates that the observed data is statistically more perfect.
Another object of the present invention is to provide a concrete early carbonation prediction system comprising:
the system comprises a concrete early carbonization initial index system and an original training set construction module, wherein the concrete early carbonization initial index system and the original training set construction module are used for obtaining relevant influence factors through engineering practice materials, collecting relevant influence factor data and creating a concrete early carbonization initial index system and an original training set;
the cross validation module is used for carrying out reasonable sample diversity on the initial index system of the early carbonization of the concrete and the training set established by the original training set establishing module, establishing proper random forest model parameters and simultaneously carrying out cross validation on each index;
the reliability establishing module of the random forest is used for calculating the importance of the related factors by using a random forest model to obtain the importance degree sequence of different factors, simultaneously performing the relevance degree between each factor and the early concrete carbonization factor by using a Pearson function model, and performing comparative analysis on the results of the two models to further establish the reliability of the random forest;
and the random forest model feasibility analysis module is used for judging the error sizes of the random forest model, the BP artificial neural network and the wavelet neural network by using two parameters of goodness of fit R2 and root mean square error RMSE, comparing and analyzing the error sizes, and proving the feasibility of the random forest model by using the fact that the error of the random forest model is far smaller than the error of the other two models.
It is another object of the present invention to provide a program storage medium for receiving a user input, the stored computer program causing an electronic device to execute the method for early concrete carbonation prediction.
It is a further object of the present invention to provide a computer program product stored on a computer readable medium, comprising a computer readable program for providing a user input interface for implementing said method of early concrete carbonation prediction when executed on an electronic device.
Another object of the present invention is to provide a terminal having a processor for executing the method for early concrete carbonation prediction.
By combining all the technical schemes, the invention has the advantages and positive effects that: the method for predicting the early concrete carbonization based on the random forest comprises the steps of firstly, obtaining relevant influence factors by referring to a large amount of relevant documents and engineering practice materials, and collecting data of the relevant influence factors, so that a set of initial concrete carbonization index system and an original training set are created; performing reasonable sample diversity on the created training set, determining proper random forest model parameters, and performing cross validation on each index; thirdly, calculating the importance of the relevant factors by using a random forest model to obtain the importance degree sequence of different factors, simultaneously performing the relevance degree between each factor and the early concrete carbonization factor by using a Pearson function model, and performing comparative analysis on the results of the two models to further determine the reliability of the random forest; and step four, judging the error sizes of the random forest model, the BP artificial neural network and the wavelet neural network by using two parameters of goodness of fit R2 and root mean square error RMSE, comparing and analyzing the error sizes, and finally finding that the error of the random forest model is far smaller than the error of the other two models, thereby confirming the feasibility of the random forest model. The method for predicting early carbonization of concrete is based on the conventional experimental method, introduces a random forest algorithm to establish an initial sample set for the influence factors of early carbonization of concrete and carries out importance evaluation, and then compares the correlation result obtained by using a Pearson function model with the correlation result to prove the accuracy of the random forest model, so that the change result of early carbonization of concrete can be effectively predicted, and a feasible method is provided for realizing prediction of early carbonization of concrete.
The effects and advantages obtained by combining experimental or experimental data with the prior art are: the invention realizes the deep exploration of the early carbonization of the concrete on the basis of saving a large amount of cost and time.
The comprehensive influence effect on the early carbonization of the concrete under the combined action of all the influence factors is analyzed. The invention can be used for controlling and managing the importance degree of the influencing factors so as to more effectively control and manage the early carbonization of the concrete.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained from the drawings without creative efforts.
FIG. 1 is a flow chart of a method for predicting early carbonation of concrete according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a method for predicting early carbonization of concrete according to an embodiment of the present invention.
FIG. 3 is a chart of an importance ranking of influencing factors provided by embodiments of the present invention.
Fig. 4 is a third diagram provided in the embodiment of the present invention, which is a node purity importance score of the influencing factor in the embodiment of the present invention.
FIG. 5 is a graph showing the correlation between the influence factors of early carbonization of concrete according to the embodiment of the present invention.
Fig. 6 is a fitting graph and a prediction graph of the random forest model of the influencing factors provided by the embodiment of the invention.
Fig. 7 is a schematic diagram of a concrete early carbonization prediction system according to an embodiment of the present invention.
In the figure: 1. an initial index system for early concrete carbonization and an original training set building module; 2. a cross validation module; 3. a reliability establishing module of the random forest; 4. and a random forest model feasibility analysis module.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
In the prior art, a large amount of experimental data is required for supporting in concrete carbonization analysis, so that a large amount of labor and time are wasted, and the cost is also greatly increased.
In view of the problems in the prior art, the present invention provides a method, a prediction system and a storage medium for early carbonization prediction of concrete, which will be described in detail below with reference to the accompanying drawings.
As shown in fig. 1, the method for predicting early carbonization of concrete provided by the embodiment of the invention comprises the following steps:
s101, establishing an initial index system and an original training set: based on a large amount of engineering practices and documents, it can be known that through query analysis of a large amount of documents and summary of practical engineering experience, the early carbonization property of concrete is related to raw material components and proportion of the components, so that the influence factors influencing the early carbonization of concrete are mainly selected from a material level, and the specific factors are as follows: cement strength, cement dosage, fly ash dosage, fine aggregate dosage, coarse aggregate dosage, concrete strength, silica fume dosage, water-cement ratio, average grain size and the like. Evaluation of the index for evaluation of carbonization was determined as the carbonization depth.
S102, random forest parameter setting: in order to meet the operating conditions of the random forest model, the related parameters of the model need to be correspondingly set.
And S103, calculating the importance and the relevance.
And S104, predicting and comparing by regression analysis.
Step S102 includes the following steps:
the first step is as follows: sample diversity. The selected monitoring samples are randomly divided into 5 equal parts, 4 parts of the samples are selected as a sample training set, the sample training set is used for determining random forest parameters and constructing a random forest model, the other 1 part of the samples is used as a sample test set and used for evaluating the performance of model prediction, and meanwhile, the original training set and the original test set which appear in the following text are emphasized to be the training set and the test set mentioned herein.
The second step is that: and (5) cross validation. Two important parameters of the random forest model: mtry is the number of influencing factors, and ntree is the tree of the decision tree. Setting the number of variables in the data set as P, and setting mtry to be P (classification model) or mtry to be P/3 (regression model) under the default condition; ntree is 500. Breiman's study indicated that: when the parameters of the random forest are set to be default values, the model can often achieve a relatively ideal effect. Therefore, the invention establishes a regression model by taking mtry as P/3 and ntree as 500, thereby ensuring the accuracy.
Step S103 specifically includes:
step 1: and (5) calculating the importance. The importance degree of each influencing factor on the early carbonization depth is analyzed through a random forest regression model, meanwhile, the importance degree is arranged according to a descending mode, and the importance arrangement distribution is shown in figure 3. Fig. 3 is an importance measure of each influencing factor in the random forest training model, and the larger the variation amplitude of the node purity (inodepurity), the more important the influencing factor is. As can be seen from the importance ranking and scores in FIG. 3, the importance measures of the variables such as the cement amount, the average particle size, the concrete strength, the water-cement ratio, the fine aggregate amount and the like are relatively large.
Step 2: and (5) calculating the correlation. The correlation between the various influence factors and the correlation between the influence factors and the early carbonization depth can be analyzed by utilizing the Pearson function, the correlation between the influence factors and the early carbonization depth can be used as a verification means for the prediction result of the random forest regression model, and as can be seen from the graph in FIG. 4, the correlation between the average particle size, the water-cement ratio, the cement dosage, the concrete strength, the fine aggregate dosage and the carbonization depth is obviously higher than that of other influence factors and is generally consistent with an importance ranking diagram, and the influence factors are further explained to have great influence on the carbonization depth property.
And step 3: the fitting result of the random forest regression model on the concrete early carbonization related data training sample is shown in fig. 5, and the prediction result of the test sample is shown in fig. 6.
Step S104 specifically includes:
and selecting two parameters of Root Mean Square Error (RMSE) and goodness of fit (R2) as evaluation basis of prediction accuracy of the model. The mean square error and goodness of fit expression is as follows:
wherein, yobsIs the actual observed value, ypredIs a predicted value. R2The proportion of variability in a data set explained by a statistical model can measure the fitting degree of a regression straight line to an observed value, the range of R2 is between 0 and 1, and the closer to 1, the more accurate the observed data is; the RMSE value is the sum of the individual differences between the estimated and actual observed values, with a RMSE value equal to or greater than 0, with closer to 0 indicating that the observed data is statistically more perfect.
According to the method for predicting the early carbonization of the concrete based on the random forest, disclosed by the invention, the importance of the influence factors is evaluated by the random forest by utilizing a random forest regression prediction model aiming at the characteristics of more influence factors and complex noise interference of the early carbonization of the concrete to obtain a better index set, the correlation among all the factors is calculated by utilizing a Pearson function model, and the result is compared with the importance, so that the accuracy of the random forest model is verified.
Fig. 2 is a schematic diagram of a method for predicting early carbonization of concrete according to an embodiment of the present invention.
As shown in fig. 7, the present invention provides a concrete early stage carbonization prediction system including:
the concrete early carbonization initial index system and original training set building module 1 is used for obtaining relevant influence factors through engineering practice materials, collecting relevant influence factor data, and creating the concrete early carbonization initial index system and an original training set.
And the cross validation module 2 is used for carrying out reasonable sample diversity on the training set created by the initial index system of early concrete carbonization and the original training set construction module, establishing proper random forest model parameters and simultaneously carrying out cross validation on each index.
And the random forest reliability establishing module 3 is used for calculating the importance of the related factors by using a random forest model to obtain the importance degree sequence of different factors, simultaneously performing the relevance degree between each factor and the concrete early carbonization factor by using a Pearson function model, and performing comparative analysis on the results of the two models to further establish the reliability of the random forest.
And the random forest model feasibility analysis module 4 is used for judging the error sizes of the random forest model, the BP artificial neural network and the wavelet neural network by using two parameters of goodness of fit R2 and root mean square error RMSE, comparing and analyzing the error sizes, and proving the feasibility of the random forest model by using the fact that the error of the random forest model is far smaller than the error of the other two models.
The invention is further described with reference to specific examples.
Examples
The method for predicting the early carbonization of the concrete provided by the embodiment of the invention comprises the following steps:
(1) establishing an initial index system and an original training set
Based on a large number of engineering practices and literatures, the early carbonization property of concrete is related to the raw material components and the proportion of the components, so the invention mainly selects the influencing factors influencing the early carbonization of the concrete from the material level, and the specific factors are as follows: cement strength, cement dosage, fly ash dosage, fine aggregate dosage, coarse aggregate dosage, concrete strength, silica fume dosage, water-cement ratio, average grain size and the like. The evaluation index for the carbonization was defined as the carbonization depth. And taking the concrete carbonization depth of the civil engineering of the Songtong project as an output variable of the concrete carbonization property. Selecting 14 monitored groups of data as an original training set, wherein the data is as shown in a table I:
And analyzing the importance degree of each influence factor on the early carbonization depth through a random forest regression model, and meanwhile, arranging the importance degrees in a descending manner, wherein the importance arrangement distribution is shown as a second figure. And the third graph is the importance measurement of each influence factor in the random forest training model, and the larger the variation amplitude of the node purity (InNodePurity), the more important the influence factor is. According to the importance ranking and scores of the second graph and the third graph, the importance measurement values of the variables such as the cement dosage, the average grain diameter, the concrete strength, the water-cement ratio, the fine aggregate dosage and the like are relatively larger
In order to further check the superiority of predicting early carbonization based on a random forest model (RF), a BP artificial neural network and a wavelet neural network are selected for comparative analysis, and the accuracy is determined by comparing two coefficients obtained by the three prediction models. Error results were obtained for the ratio table two:
table two error comparison
The root mean square errors of the random forest prediction model, the BP artificial neural network prediction model and the wavelet neural network analysis are respectively 0.00057, 0.016 and 0.0138, and the certainty coefficients are respectively 0.694, 0.734 and 0.5433, so that on the premise that the certainty coefficients are very similar, the root mean square error obtained by the random forest model is far smaller than the root mean square errors obtained by the artificial neural model and the wavelet neural network model, and the prediction result of the random forest model is most close to the actual value, and is higher in precision and better in effect.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A method for predicting early carbonization of concrete, which is characterized in that the method for predicting early carbonization of concrete comprises the following steps:
the method comprises the following steps: constructing an initial index system and an initial training set for early carbonization of concrete;
step two: carrying out sample diversity on the constructed training set, establishing parameters of a random forest model, and simultaneously carrying out cross validation on each index;
step three: calculating the importance degree of each index by using a random forest model, calculating the correlation coefficient of each index by using a Pearson function model, and comparing the calculation result of the random forest model with the calculation result of the Pearson function model;
step four: and judging the error sizes of the random forest model, the BP artificial neural network and the wavelet neural network by using two parameters of goodness-of-fit and root-mean-square error, and performing comparative analysis.
2. The method for predicting early carbonization of concrete according to claim 1, wherein the concrete initial index system in the first step comprises two parts of early carbonization influencing factors and evaluation indexes, wherein the influencing factors comprise 9 factors of cement strength, cement dosage, fly ash dosage, fine aggregate dosage, coarse aggregate dosage, concrete strength, silica fume dosage, water-cement ratio and average particle size; the early carbonization evaluation index is the carbonization depth.
3. The method for predicting early carbonation of concrete according to claim 1, wherein said original training set in step one is: and taking different influence factors in the index system as variables of the random forest, collecting statistical relevant data, and taking corresponding data as an original training set.
4. The method for early concrete carbonation prediction according to claim 1, wherein said step two of performing reasonable sample diversity on the created training set and establishing appropriate random forest model parameters, and performing cross validation on each index comprises the following steps:
the first step is as follows: randomly dividing a selected monitoring sample into 5 equal parts, optionally selecting 4 parts as a sample training set, wherein the sample training set is used for determining random forest parameters and constructing a random forest model, and the other 1 part is used as a sample test set and used for evaluating the performance of model prediction;
the second step is that: two important parameters of the random forest model: mtry is the number of influencing factors, and ntree is a tree of the decision tree; and (3) establishing a regression model by setting the variable number in the data set as P, wherein mtry is P or mtry is P/3 and ntree is 500.
5. The method for predicting early carbonization of concrete according to claim 1, wherein the step three comprises calculating importance of each index by using a random forest model, and calculating correlation coefficient of each index by using a Pearson function model, and comprises the following steps:
step 1: the method comprises the steps of calculating out-of-bag data errors of each tree in a random forest according to the OOB (calculated out and recorded as errOOB1, adding noise interference into all samples of the out-of-bag data OOB randomly, calculating out-of-bag data errors repeatedly and recorded as errOOB2, scoring importance of different influence factors, visually drawing the importance scores, and calculating the importance of a certain variable characteristic according to the following formula:
step 2: analyzing the correlation degree of different influencing factors and the early carbonization depth by utilizing a Pearson function model, wherein the Pearson sample correlation coefficient r is as follows:
wherein
Respectively representing the mean values of the samples of X, Y;
and step 3: and comparing and analyzing the calculation result of the importance of the random forest model and the calculation result of the correlation of the Pearson function model.
6. The method for predicting early carbonization of concrete according to claim 1, wherein in the fourth step, the two parameters of root mean square error and goodness of fit are selected as the evaluation basis of the prediction accuracy of the model; the mean square error and goodness of fit expression is:
wherein, yobsIs the actual observed value, ypredIs a predicted value; r2Is a proportion of variability in the data set that is accounted for by the statistical model, measuring the degree of fit, R, of the regression line to the observed value2Ranges from 0 to 1, with closer to 1 indicating more accurate observed data; the RMSE value is the sum of the individual differences between the estimated and actual observed values, with a RMSE value equal to or greater than 0, with closer to 0 indicating that the observed data is statistically more perfect.
7. A concrete early carbonation prediction system for implementing the method for concrete early carbonation prediction according to claims 1 to 6, wherein the concrete early carbonation prediction system comprises:
the system comprises a concrete early carbonization initial index system and an original training set construction module, wherein the concrete early carbonization initial index system and the original training set construction module are used for obtaining relevant influence factors through engineering practice materials, collecting relevant influence factor data and creating a concrete early carbonization initial index system and an original training set;
the cross validation module is used for carrying out reasonable sample diversity on the initial index system of the early carbonization of the concrete and the training set established by the original training set establishing module, establishing proper random forest model parameters and simultaneously carrying out cross validation on each index;
the reliability establishing module of the random forest is used for calculating the importance of the related factors by using a random forest model to obtain the importance degree sequence of different factors, simultaneously performing the relevance degree between each factor and the early concrete carbonization factor by using a Pearson function model, and performing comparative analysis on the results of the two models to further establish the reliability of the random forest;
and the random forest model feasibility analysis module is used for judging the error sizes of the random forest model, the BP artificial neural network and the wavelet neural network by using two parameters of goodness of fit R2 and root mean square error RMSE, comparing and analyzing the error sizes, and proving the feasibility of the random forest model by using the fact that the error of the random forest model is far smaller than the error of the other two models.
8. A program storage medium for receiving user input, the stored computer program causing an electronic device to perform the method of early concrete carbonation prediction according to any one of claims 1 to 6.
9. A computer program product stored on a computer readable medium, comprising a computer readable program for providing a user input interface for implementing a method for early concrete carbonation prediction according to any one of claims 1 to 6 when executed on an electronic device.
10. A terminal, characterized in that the terminal is provided with a processor for operating the method for predicting early carbonization of concrete according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010290133.2A CN111488713A (en) | 2020-04-14 | 2020-04-14 | Method, system and storage medium for predicting early carbonization of concrete |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010290133.2A CN111488713A (en) | 2020-04-14 | 2020-04-14 | Method, system and storage medium for predicting early carbonization of concrete |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111488713A true CN111488713A (en) | 2020-08-04 |
Family
ID=71795199
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010290133.2A Pending CN111488713A (en) | 2020-04-14 | 2020-04-14 | Method, system and storage medium for predicting early carbonization of concrete |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111488713A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111986737A (en) * | 2020-08-07 | 2020-11-24 | 华中科技大学 | Durable concrete mixing proportion optimization method based on RF-NSGA-II |
| CN111985796A (en) * | 2020-08-07 | 2020-11-24 | 华中科技大学 | Method for predicting concrete structure durability based on random forest and intelligent algorithm |
| CN112016244A (en) * | 2020-08-07 | 2020-12-01 | 中国交通建设股份有限公司吉林省分公司 | Durable concrete multi-target mix proportion optimization method based on SVM and intelligent algorithm |
| CN112070356A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting anti-carbonization performance of concrete based on RF-LSSVM model |
| CN112069567A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting compressive strength of concrete based on random forest and intelligent algorithm |
| CN112488191A (en) * | 2020-11-30 | 2021-03-12 | 海南电网有限责任公司电力科学研究院 | Metal corrosion distribution diagram drawing method based on KNN intelligent algorithm |
| CN112505494A (en) * | 2020-10-30 | 2021-03-16 | 西安交通大学 | Method and device for evaluating insulation water content of oiled paper |
| CN112530528A (en) * | 2020-11-27 | 2021-03-19 | 华能西藏雅鲁藏布江水电开发投资有限公司 | Concrete carbonization parameter prediction method, device and experimental system |
| CN112668802A (en) * | 2021-01-05 | 2021-04-16 | 广东工业大学 | Construction carbon emission prediction method based on design parameters |
| CN113868960A (en) * | 2021-10-18 | 2021-12-31 | 青岛农业大学 | Soil heavy metal characteristic selection method and system based on typical relevant forest |
| CN116108998A (en) * | 2023-02-22 | 2023-05-12 | 葛洲坝集团交通投资有限公司 | Expressway construction project carbon emission prediction method and system |
| CN117113291A (en) * | 2023-10-23 | 2023-11-24 | 合肥综合性国家科学中心人工智能研究院(安徽省人工智能实验室) | Analysis method for importance of production parameters in semiconductor manufacturing |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108153989A (en) * | 2018-01-09 | 2018-06-12 | 天津大学 | Consider the concrete dam compaction quality method for quick predicting that parameter uncertainty influences |
| CN108334668A (en) * | 2018-01-09 | 2018-07-27 | 天津大学 | Consider the earth and rockfill dam compaction quality method for quick predicting that parameter uncertainty influences |
-
2020
- 2020-04-14 CN CN202010290133.2A patent/CN111488713A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108153989A (en) * | 2018-01-09 | 2018-06-12 | 天津大学 | Consider the concrete dam compaction quality method for quick predicting that parameter uncertainty influences |
| CN108334668A (en) * | 2018-01-09 | 2018-07-27 | 天津大学 | Consider the earth and rockfill dam compaction quality method for quick predicting that parameter uncertainty influences |
Non-Patent Citations (2)
| Title |
|---|
| 林开春 等: ""基于随机森林和神经网络的空气质量预测研究"", vol. 33, no. 2, pages 33 * |
| 齐小华 等: "《混凝土结构加固设计与施工》", 西安电子科技大学出版社, pages: 206 - 208 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112069567B (en) * | 2020-08-07 | 2024-01-12 | 湖北交投十巫高速公路有限公司 | Method for predicting compressive strength of concrete based on random forest and intelligent algorithm |
| CN111985796A (en) * | 2020-08-07 | 2020-11-24 | 华中科技大学 | Method for predicting concrete structure durability based on random forest and intelligent algorithm |
| CN112016244A (en) * | 2020-08-07 | 2020-12-01 | 中国交通建设股份有限公司吉林省分公司 | Durable concrete multi-target mix proportion optimization method based on SVM and intelligent algorithm |
| CN112070356A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting anti-carbonization performance of concrete based on RF-LSSVM model |
| CN112069567A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting compressive strength of concrete based on random forest and intelligent algorithm |
| CN111986737A (en) * | 2020-08-07 | 2020-11-24 | 华中科技大学 | Durable concrete mixing proportion optimization method based on RF-NSGA-II |
| CN112070356B (en) * | 2020-08-07 | 2024-05-14 | 湖北交投十巫高速公路有限公司 | Method for predicting carbonization resistance of concrete based on RF-LSSVM model |
| CN112505494A (en) * | 2020-10-30 | 2021-03-16 | 西安交通大学 | Method and device for evaluating insulation water content of oiled paper |
| CN112505494B (en) * | 2020-10-30 | 2022-05-03 | 西安交通大学 | Method and device for evaluating insulation water content of oiled paper |
| CN112530528A (en) * | 2020-11-27 | 2021-03-19 | 华能西藏雅鲁藏布江水电开发投资有限公司 | Concrete carbonization parameter prediction method, device and experimental system |
| CN112488191A (en) * | 2020-11-30 | 2021-03-12 | 海南电网有限责任公司电力科学研究院 | Metal corrosion distribution diagram drawing method based on KNN intelligent algorithm |
| CN112488191B (en) * | 2020-11-30 | 2022-11-25 | 海南电网有限责任公司电力科学研究院 | Metal corrosion distribution map drawing method based on KNN intelligent algorithm |
| CN112668802A (en) * | 2021-01-05 | 2021-04-16 | 广东工业大学 | Construction carbon emission prediction method based on design parameters |
| CN113868960B (en) * | 2021-10-18 | 2024-04-16 | 青岛农业大学 | Soil heavy metal characteristic selection method and system based on typical related forests |
| CN113868960A (en) * | 2021-10-18 | 2021-12-31 | 青岛农业大学 | Soil heavy metal characteristic selection method and system based on typical relevant forest |
| CN116108998B (en) * | 2023-02-22 | 2023-12-15 | 葛洲坝集团交通投资有限公司 | Expressway construction project carbon emission prediction method and system |
| CN116108998A (en) * | 2023-02-22 | 2023-05-12 | 葛洲坝集团交通投资有限公司 | Expressway construction project carbon emission prediction method and system |
| CN117113291A (en) * | 2023-10-23 | 2023-11-24 | 合肥综合性国家科学中心人工智能研究院(安徽省人工智能实验室) | Analysis method for importance of production parameters in semiconductor manufacturing |
| CN117113291B (en) * | 2023-10-23 | 2024-02-09 | 合肥综合性国家科学中心人工智能研究院(安徽省人工智能实验室) | Analysis method for importance of production parameters in semiconductor manufacturing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111488713A (en) | Method, system and storage medium for predicting early carbonization of concrete | |
| CN112069567B (en) | Method for predicting compressive strength of concrete based on random forest and intelligent algorithm | |
| CN111985796B (en) | Method for predicting concrete structure durability based on random forest and intelligent algorithm | |
| CN109685289B (en) | Method, device and system for forward prediction of blast furnace conditions | |
| CN111507518A (en) | Wavelet neural network concrete impermeability prediction method based on random forest | |
| CN112070356B (en) | Method for predicting carbonization resistance of concrete based on RF-LSSVM model | |
| CN112329262A (en) | Residential building gas carbon emission prediction method | |
| CN101976307A (en) | Printing and dyeing process sewage monitoring index time constraint associated rule mining algorithm | |
| CN110751176A (en) | Lake water quality prediction method based on decision tree algorithm | |
| CN112016815A (en) | User side comprehensive energy efficiency evaluation method based on neural network | |
| CN114331238B (en) | Intelligent model algorithm optimization method, system, storage medium and computer equipment | |
| CN112200459A (en) | Power distribution network data quality analysis and evaluation method and system | |
| CN113962477A (en) | Industrial electric quantity association aggregation prediction method, device, equipment and storage medium | |
| CN111861264A (en) | Method for predicting concrete durability based on data mining and intelligent algorithm | |
| CN114626692A (en) | Method and system for optimizing town scale structure, computer equipment and storage medium | |
| CN114065616A (en) | Long-term aging viscosity prediction method based on neural network and grey correlation | |
| CN112116139A (en) | Power demand prediction method and system | |
| TWI695285B (en) | Regression method and system based on system program infrastructure | |
| Fu et al. | Discovering admissible Web services with uncertain QoS | |
| CN116720662B (en) | Distributed energy system applicability evaluation method based on set pair analysis | |
| CN117764454B (en) | Method for evaluating development degree of site flaky stripping | |
| CN117575106B (en) | Method, system, electronic equipment and medium for predicting gas production profile of coal-bed gas well | |
| CN114818990B (en) | Method and system for grading quality of maintenance effect of aero-engine | |
| CN114139995B (en) | Test area monitoring and evaluating method and device, electronic equipment and storage medium | |
| CN111079344B (en) | Method and system for measuring softness of paper |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200804 |