CN111444629A - Reinforcing steel bar corrosion parameter prediction method based on support vector machine - Google Patents

Reinforcing steel bar corrosion parameter prediction method based on support vector machine Download PDF

Info

Publication number
CN111444629A
CN111444629A CN202010296670.8A CN202010296670A CN111444629A CN 111444629 A CN111444629 A CN 111444629A CN 202010296670 A CN202010296670 A CN 202010296670A CN 111444629 A CN111444629 A CN 111444629A
Authority
CN
China
Prior art keywords
steel bar
support vector
vector machine
section
corrosion
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.)
Granted
Application number
CN202010296670.8A
Other languages
Chinese (zh)
Other versions
CN111444629B (en
Inventor
张四化
吕亚军
李大望
武轶彬
刘云龙
王俊伟
邹亮
熊程
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Henan Zhangbotuan Construction Technology Co ltd
Zhengzhou University Multi Functional Design And Research Academy Ltd
China Second Metallurgy Group Co Ltd
Original Assignee
Henan Zhangbotuan Construction Technology Co ltd
Zhengzhou University Multi Functional Design And Research Academy Ltd
China Second Metallurgy Group Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Henan Zhangbotuan Construction Technology Co ltd, Zhengzhou University Multi Functional Design And Research Academy Ltd, China Second Metallurgy Group Co Ltd filed Critical Henan Zhangbotuan Construction Technology Co ltd
Priority to CN202010296670.8A priority Critical patent/CN111444629B/en
Publication of CN111444629A publication Critical patent/CN111444629A/en
Application granted granted Critical
Publication of CN111444629B publication Critical patent/CN111444629B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N17/00—Investigating resistance of materials to the weather, to corrosion, or to light
    • G01N17/006—Investigating resistance of materials to the weather, to corrosion, or to light of metals
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Geometry (AREA)
  • Computer Graphics (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Ecology (AREA)
  • Environmental & Geological Engineering (AREA)
  • Environmental Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Testing Resistance To Weather, Investigating Materials By Mechanical Methods (AREA)

Abstract

The invention relates to the technical fields of reinforced concrete, machine learning technology and big data, in particular to a method for predicting steel bar corrosion parameters based on a support vector machine. The steel bar corrosion parameter prediction method based on the support vector machine comprises the following steps: the method comprises the following steps: manufacturing a test piece of the corrosion reinforcing steel bar; step two: 3D scanning is carried out on the test piece, and a 3D image model is generated; step three: according to the 3D image model, calculating the specific characteristic parameters of the section of the test piece and the steel bar corrosion rate corresponding to the specific characteristic parameters of the section; step four: setting specific characteristic parameters of the cross section and the corrosion rate of the steel bar corresponding to the specific characteristic parameters as a group of basic data, inputting the multiple groups of basic data into a support vector machine for learning and training to obtain a prediction model for predicting the corrosion rate of the steel bar to be measured. The method can realize the prediction of the corrosion rate of the steel bar only by knowing the specific characteristic parameters of the corroded steel bar, and has high accuracy and convenient use.

Description

Reinforcing steel bar corrosion parameter prediction method based on support vector machine
Technical Field
The invention relates to the technical fields of reinforced concrete, machine learning technology and big data, in particular to a method for predicting steel bar corrosion parameters based on a support vector machine.
Background
After the reinforcing steel bar in the concrete is corroded, the apparent appearance becomes uneven, the mechanical property of the reinforcing steel bar is reduced due to the reduction of the section, and the strength and the ductility of the reinforcing steel bar are also reduced. Thereby affecting the load bearing capacity and durability of the concrete.
In order to evaluate the form, the bearing capacity and the durability of the steel bar after rusting and reflect the real corrosion condition of the steel bar, researchers carry out related researches on quantitative evaluation of the corrosion condition of the steel bar. The quality rust rate is a parameter frequently adopted by scholars, and the scholars frequently use a weighing method, a vernier caliper method, a drainage method and the like. However, these analysis methods still have some defects, certain errors exist in data measurement, and accurate analysis of the cross-sectional shape characteristics of the corroded steel bar is difficult to perform.
With the progress of the technology, scholars measure the form of the corroded steel bar by using methods such as X-ray transmission computed tomography (XCT) and 3D scanning, and the measuring methods can accurately reflect the corrosion condition of the steel bar.
The detection method is very accurate, can extract the cross-sectional shape of the corroded steel bar, and can digitize the cross-sectional shape, so as to realize accurate extraction of corroded steel bar data, but it needs to be pointed out that the analysis method after data extraction has some defects, for example, after the corroded form of the steel bar is accurately obtained, only the preliminary analysis is performed on the corroded condition distribution of the steel bar, but the relation between the corroded form of the steel bar and the corroded rate of the steel bar does not exist.
That is to say, can only be after the reinforcing bar corrosion, measure the corrosion rate mechanical energy of reinforcing bar, but when the reinforcing bar was not dismantled, the corrosion rate of accurate prediction reinforcing bar can not be accomplished for in some occasions, especially in the building field, must dismantle just to know the corrosion condition of reinforcing bar, and dismantle the reinforcing bar and can exert an influence to the building body, and then comparatively inconvenient.
Disclosure of Invention
The invention aims to provide a steel bar corrosion parameter prediction method based on a support vector machine, which can accurately obtain the relationship between the steel bar corrosion form and the steel bar corrosion rate, and further can realize the prediction of the steel bar corrosion rate.
The embodiment of the invention is realized by the following steps:
a steel bar corrosion parameter prediction method based on a support vector machine comprises the following steps:
the method comprises the following steps: manufacturing a test piece of the corrosion reinforcing steel bar;
step two: 3D scanning is carried out on the test piece, and a 3D image model is generated;
step three: according to the 3D image model, calculating the specific characteristic parameters of the section of the test piece and the steel bar corrosion rate corresponding to the specific characteristic parameters of the section;
step four: setting specific characteristic parameters of the section and the corrosion rate of the steel bar corresponding to the specific characteristic parameters as a group of basic data, inputting a plurality of groups of basic data into a support vector machine for learning and training to obtain a prediction model for predicting the corrosion rate of the steel bar to be measured;
the specific characteristic parameters of the cross section of the test piece comprise:
the ratio of the minimum inscribed circle radius to the fitted circle radius η, the ratio of the maximum circumscribed circle radius to the fitted circle radius, the ratio of the minimum inscribed circle radius to the maximum circumscribed circle radius upsilon, the eccentricity e, the ratio of the short side of the fitted ellipse to the long side of the fitted ellipse, the roundness chi and the section roughness gamma.
Preferably, in the step one, the corrosion steel bar test piece is manufactured by adopting an electrifying method.
Preferably, before the 3D scanning of the test piece, the test piece is cleaned to remove a rusted portion.
Preferably, the calculation formulas of the specific characteristic parameters are respectively as follows:
η＝r1/r0(3)
＝r2/r0(4)
ν＝r1/r2(5)
Figure BDA0002452444160000031
＝a/b (7)
χ＝p2/(4πA1) (8)
γ＝A1/(πr0 2) (9)
wherein r is1The minimum inscribed circle radius of the cross section profile; r is2The maximum circumscribed circle radius of the cross section profile; r is0Fitting a circle radius for the profile of the section of the steel bar; a is the fitting ellipse short side of the section outline of the steel bar(ii) a b is a long edge of a section contour fitting ellipse of the steel bar; y is0，z0Fitting ellipse center coordinates; and p is the perimeter of the residual cross section area after the steel bar is corroded.
Preferably, the specific characteristic parameters of the section of the test piece are calculated by programming according to a 3D image model by using MAT L AB.
Preferably, when a plurality of groups of basic data are input into the support vector machine for learning training, a PSO particle swarm optimization method is used for optimizing an optimal penalty factor C and a kernel function parameter g, and the optimal penalty factor C and the kernel function parameter g are used for establishing an optimal regression model as a prediction model.
Preferably, when a plurality of groups of basic data are input into the support vector machine for learning and training, a GS grid optimization method is used for establishing an optimal regression model as a prediction model.
Preferably, after the prediction model is obtained, inputting a plurality of groups of base data which are not trained by learning into the prediction model, and verifying the prediction model;
when the final error is within an allowable range, the prediction model is a final prediction model;
and when the final error is not in the allowable range, the prediction model is inaccurate, the step three is returned, the specific characteristic parameters of the section of the test piece are recalculated, and then a new prediction model is established again.
Preferably, the calculated final error comprises a mean relative error, a mean absolute error, a root mean square error, a correlation coefficient and/or a mean square error.
Preferably, the average relative error MRE is calculated by:
Figure BDA0002452444160000041
the mean absolute error MAE is calculated as:
Figure BDA0002452444160000042
the root mean square error RMSE is calculated as:
Figure BDA0002452444160000043
coefficient of correlation R2The calculation method is as follows:
Figure BDA0002452444160000044
the mean square error MSE is calculated as follows:
Figure BDA0002452444160000045
in the formula: y'iTo predict value, yiIn order to be the true value of the value,
Figure BDA0002452444160000046
the average of the true values, i ═ 1,2, …, n.
The embodiment of the invention has the beneficial effects that:
and carrying out corrosion simulation on the prefabricated corrosion steel bar test piece, then carrying out learning training on the simulated corrosion steel bar by using a support vector machine to obtain a matching function, forming a prediction model, and further realizing prediction on the corrosion rate of the unknown steel bar by using the prediction model.
The method can realize the prediction of the corrosion rate of the steel bar only by knowing the specific characteristic parameters of the corroded steel bar, and has high accuracy and convenient use.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a flowchart of a method for predicting a steel bar corrosion rate parameter based on a support vector machine according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a cross-sectional corrosion condition of a test piece in the steel bar corrosion rate parameter prediction method based on a support vector machine according to the embodiment of the present invention;
fig. 3 is a schematic characteristic cross-sectional view of a test piece in the method for predicting corrosion rate parameters of steel bars based on a support vector machine according to the embodiment of the invention;
fig. 4 is a schematic diagram of a method for extracting characteristic parameters of a characteristic cross section of a corroded steel bar in a steel bar corrosion rate parameter prediction method based on a support vector machine according to an embodiment of the present invention;
fig. 5 is a schematic diagram of an extraction method of a characteristic parameter η of a characteristic cross section of a corroded steel bar in the steel bar corrosion rate parameter prediction method based on a support vector machine according to the embodiment of the present invention;
fig. 6 is a schematic diagram of a method for extracting a characteristic parameter υ of a characteristic cross section of a corroded steel bar in a steel bar corrosion rate parameter prediction method based on a support vector machine according to an embodiment of the present invention;
fig. 7 is a schematic diagram of a method for extracting characteristic parameters of a characteristic cross section of a corroded steel bar in a steel bar corrosion rate parameter prediction method based on a support vector machine according to an embodiment of the present invention;
fig. 8 is a schematic diagram of an extraction method of a characteristic parameter e of a characteristic cross section of a corroded steel bar in the steel bar corrosion rate parameter prediction method based on a support vector machine according to the embodiment of the present invention;
fig. 9 is a schematic structural diagram of a support vector machine in the method for predicting corrosion rate of steel bars based on the support vector machine according to the embodiment of the present invention;
fig. 10 is a flow chart of PSO optimization in the method for predicting steel bar corrosion rate parameters based on a support vector machine according to the embodiment of the present invention;
fig. 11 is a schematic diagram of a PSO optimization result in the steel bar corrosion rate parameter prediction method based on a support vector machine according to the embodiment of the present invention;
fig. 12 is a comparison graph of the predicted result and the actual result of the steel bar corrosion rate of the PSO optimization method and the GS grid optimization method in the steel bar corrosion rate parameter prediction method based on the support vector machine according to the embodiment of the present invention;
fig. 13 is a schematic diagram illustrating the correlation between the predicted value and the true value of the PSO optimization method in the steel bar corrosion rate parameter prediction method based on the support vector machine according to the embodiment of the present invention;
fig. 14 is a schematic diagram of correlation between a predicted value and a true value of a GS grid optimization method in the steel bar corrosion rate parameter prediction method based on a support vector machine according to the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings or the orientations or positional relationships that the products of the present invention are conventionally placed in use, and are only used for convenience in describing the present invention and simplifying the description, but do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," "third," and the like are used solely to distinguish one from another and are not to be construed as indicating or implying relative importance.
Furthermore, the terms "horizontal", "vertical", "overhang" and the like do not imply that the components are required to be absolutely horizontal or overhang, but may be slightly inclined. For example, "horizontal" merely means that the direction is more horizontal than "vertical" and does not mean that the structure must be perfectly horizontal, but may be slightly inclined.
In the description of the present invention, it should also be noted that, unless otherwise explicitly specified or limited, the terms "disposed," "mounted," "connected," and "connected" are to be construed broadly and may, for example, be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Some embodiments of the present invention will be described in detail below with reference to fig. 1 to 14. The embodiments described below and the features of the embodiments can be combined with each other without conflict.
Machine learning is increasingly involved in various fields due to the development of big data. The BP neural network, the support vector machine, the random forest method and other algorithms train data and establish a prediction model, and the method has the characteristics of strong operability and accurate prediction. For example, Xue and other methods based on BP neural network restore the image, the image display effect obtained by the method is obviously improved, the detail and edge identification degree of the image are improved, and the image quality is greatly improved. Gong et al estimate the roughness index of a flexible road surface by using a random forest model, the adopted prediction parameters comprise the traffic volume, the maintenance condition, the structural form, the local climate characteristics and the like of the road, 11000 data samples are collected, 80% of the samples are used for training the model, 20% of the samples are used for testing the prediction model, and the obtained model has higher accuracy. Therefore, the intelligent algorithm can be used for accurately analyzing and predicting the data.
Based on the above situation, the present invention provides a steel bar corrosion parameter prediction method based on a support vector machine, as shown in fig. 1, which includes the following steps:
the method comprises the following steps: manufacturing a test piece of the corrosion reinforcing steel bar;
step two: 3D scanning is carried out on the test piece, and a 3D image model is generated;
step three: according to the 3D image model, calculating the specific characteristic parameters of the section of the test piece and the steel bar corrosion rate corresponding to the specific characteristic parameters of the section;
step four: setting specific characteristic parameters of the cross section and the corrosion rate of the steel bar corresponding to the specific characteristic parameters as a group of basic data, inputting the multiple groups of basic data into a support vector machine for learning and training to obtain a prediction model for predicting the corrosion rate of the steel bar to be measured.
Specifically, in the first step of this example, the fabricated test piece of rusted steel bars was a reinforced concrete test piece 400mm × 350mm × 250mm, the composition ratio of the concrete is shown in table 1, the concrete test piece contains 3 steel bars, the diameter of each steel bar is 14mm, the type of the steel bar is HPB300, the chemical composition list of the steel bars is shown in table 2, and the reinforced concrete test piece is first cured in air at room temperature of about 25 ℃ for 28 days.
TABLE 1 concrete composition
Figure BDA0002452444160000081
TABLE 2HPB300 Reinforcement Bar Components
Figure BDA0002452444160000082
Before the reinforced concrete test pieces were produced, each of the 400 mm-long reinforcing bars was first subjected to sand blasting to remove rust and oxidation substances on the surface thereof. The end parts of the two sides of the steel bar, which are 75mm long, are covered by epoxy resin and then by an isolation strip, so that the manufactured reinforced concrete test piece is ensured to be the same as the using environment of the steel bar, and the part of the middle part of the steel bar, which is 250mm, is corroded.
In the second step of this embodiment, the surface morphology of the corroded steel bar is measured by using the 3D optical scanner, the scanning precision is 0.05mm, the scanning process provides the three-dimensional coordinates of each point on the surface of the steel bar, and according to the three-dimensional coordinates of each point, a 3D image of the corrosion of the steel bar can be accurately obtained.
The cross section of the corrosion-resistant steel bar is shown in figure 3.
In the third step of this embodiment, the 3D image data obtained by 3D scanning is put into the calculation software to be read out, and then a plurality of groups of specific characteristic parameters and the steel bar corrosion rate corresponding to the characteristic parameters are obtained through processing.
Specifically, on each section, specific characteristic parameters and the corresponding steel bar corrosion rate can be obtained. At this time, specific characteristic parameters of a plurality of sections and the corrosion rates of the steel bars corresponding to the specific characteristic parameters are calculated.
More specifically, in this embodiment, the calculation method of the corrosion rate of the steel bar is the ratio of the difference between the area of the cross section of the corroded steel bar and the area of the cross section of the original steel bar to the cross section of the original steel bar.
The specific calculation formula is as shown in formula (1):
Figure BDA0002452444160000091
A0the original cross-sectional areas are the same as the cross-sectional areas of the steel bars before corrosion; a. the1Calculating the residual cross-sectional area of the steel bar after being corroded by an integral method, as shown in formula (2):
Figure BDA0002452444160000092
in the formula: y isi，yi+1Is a y-axis coordinate, zi、zi+1Is the z-axis coordinate.
The calculation method of seven parameters adopts the formulas (3) - (9)
η＝r1/r0(3)
＝r2/r0(4)
ν＝r1/r2(5)
Figure BDA0002452444160000093
＝a/b (7)
χ＝p2/(4πA1) (8)
γ＝A1/(πr0 2) (9)
Wherein r is1The minimum inscribed circle radius of the cross section profile; r is2The maximum circumscribed circle radius of the cross section profile; r is0Fitting a circle radius for the profile of the section of the steel bar; a is a short side of a fitting ellipse of the section profile of the steel bar; b is a long edge of a section contour fitting ellipse of the steel bar; y is0，z0Fitting ellipse center coordinates; and p is the perimeter of the residual cross section area after the steel bar is corroded.
Specific characteristic parameter factors of the section of the steel bar are obtained through the characteristic parameters, namely the ratio η of the minimum inscribed circle radius to the radius of the fitting circle, the ratio of the maximum circumscribed circle radius to the radius of the fitting circle, the ratio upsilon of the minimum inscribed circle radius to the maximum circumscribed circle radius, the eccentricity e, the ratio of the short side of the fitting ellipse to the long side of the fitting ellipse, the roundness chi and the section roughness gamma.
Preferably, the specific characteristic parameters of the cross section of the test piece are calculated by programming with MAT L AB according to the 3D image model.
It should be noted that the calculation method of the above-mentioned 7 specific characteristic parameters of the cross section of the test piece may be a calculation programmed by using MAT L AB, but is not limited to the calculation method of MAT L AB, as long as the above-mentioned 7 specific characteristic parameters can be obtained from the 3D image model.
The extraction of the characteristic parameters of the corrosion steel bar characteristic section is shown in figures 4-8.
In the fourth step of this embodiment, the specific characteristic parameters on each cross section and the steel bar corrosion rate corresponding to the specific characteristic parameters are set as a set of basic data, and a plurality of sets of basic data are input into the support vector machine for learning and training, so that after a certain rule range can be obtained, a preliminary prediction model is obtained for predicting the corrosion rate of the steel bar to be measured.
Specifically, a linear regression function is established in a support vector machine by using the principle of minimizing the structural risk:
y′＝Asc(x)＝w·φ(x)+b (10)
in the formula, w is a weight vector, w ∈ F, b is an offset vector, b ∈ R, y' is a predicted value, and y is an actual measured value.
The structural diagram of the regression-type support vector machine is shown in FIG. 9.
In FIG. 9, the corrosion rate A of the steel bar is outputtedscIs a linear combination of intermediate nodes, each of which corresponds to a support vector, x1,x2,x3…,xnAs an input variable, ci-diIs the network weight.
Preferably, in the step one, the corrosion steel bar test piece is manufactured by adopting an electrifying method.
In the manufacturing process of the steel bar test piece, corrosion is accelerated by adopting a power-on method so as to reduce the manufacturing time of the steel bar test piece.
Specifically, the method for accelerating corrosion by electrifying is a steel bar anodic corrosion test method, in which steel bars are used as anodes and an external metal (usually stainless steel or copper) is used as a cathode. External power having a constant current or a constant voltage is supplied between the anode and the cathode to introduce chloride ions from an external solution into the concrete, thereby causing corrosion of the reinforcing steel.
The corrosion state of the section of the corroded steel bar is shown in figure 2.
It should be noted that the corrosion acceleration may be realized by accelerating through energization, or may be realized by other methods, such as by a chemical method, that is, it is only necessary to realize accelerated corrosion of the steel bar in the same environment.
Preferably, before the 3D scanning of the test piece, the test piece is cleaned for removing the rusted parts.
After the accelerated corrosion test, the corroded steel reinforcement is carefully removed from the sample by crushing the concrete and then the reinforcement is cleaned by the specified standard method.
After the steel bars are cleaned, the accuracy of data in 3D scanning of corroded steel bars can be guaranteed, and the accuracy of calculation and prediction of the final corrosion rate is further guaranteed.
Preferably, when a plurality of groups of basic data are used as training sets and input into a support vector machine for learning training, a PSO particle swarm optimization method is used for optimizing an optimal penalty factor C and a kernel function parameter g, and the optimal penalty factor C and the kernel function parameter g are used for establishing an optimal regression model as a prediction model.
The method of PSO particle swarm is used to find the optimal combination of penalty parameter c and kernel function parameter g, and the specific process is shown in fig. 10.
The first step is initialization. Randomly generating the speed and position of particles, and setting learning factor c1And c2Evolution algebra E penalty factor C and kernel parameter g.
And secondly, evaluating the fitness. A fitness function value (fitness) of each particle is calculated, and an individual optimal value and a global optimal value are initialized.
The third step is an update process. And updating the speed and the position of the particles to generate a new population, respectively comparing the adaptive value of the current parameter C with the self historical optimal value and the population optimal value, and updating the global optimal values of the population parameters C and g.
The fourth step is a stop condition. Optimizing to reach the maximum evolution algebra, finishing optimizing and outputting the optimal C and g.
The structure diagram of the regression type support vector machine is shown in FIG. 9.
Preferably, when a plurality of groups of basic data are used as a training set and input into a support vector machine for learning and training, a GS grid optimization method is used for establishing an optimal regression model as a prediction model.
In order to show the accuracy of the PSO particle swarm optimization method, a group of control groups is added in the experiment, namely a GS grid optimization method is adopted for comparison.
The principle of the GS grid optimization method is that optimization is firstly carried out in a large range, then detail optimization is carried out by narrowing the range, C and g are enabled to divide the grid in a certain range and traverse all points in the grid to carry out value taking, for the determined C and g, the verification and classification accuracy of the training set under the group of C and g is obtained by using a K-CV cross verification method, and finally the group of C which enables the verification and classification accuracy of the training set to be the highest and the best parameters are obtained.
Preferably, after the prediction model is obtained, inputting a plurality of groups of basic data which are not trained by learning into the prediction model as a prediction set, and verifying the prediction model;
when the final error is within the allowable range, the prediction model is the final prediction model;
and when the final error is not in the allowable range, the prediction model is inaccurate, the step three is returned, the specific characteristic parameters of the section of the test piece are recalculated, and then a new prediction model is established again.
The method comprises the steps of taking seven characteristic parameters as input values, taking the condition of the steel bar corrosion rate as an output value, selecting the front 950 groups of data from 1000 groups of data as a training set, training by using a support vector machine, constructing a prediction model of the steel bar corrosion rate parameters, training the training set by using an L ibsvm toolbox in the training process, taking the remaining 50 groups of data as a prediction set, inputting the seven characteristic parameters of the predicted corrosion steel bar into the obtained prediction model, obtaining a prediction result, obtaining the result by software operation, and drawing a relevant curve image of the true value and the predicted value of the steel bar corrosion rate under a PSO particle swarm optimization method and a GS grid optimization method after the corresponding code operation is finished.
The PSO optimizing result is shown in figure 11, the comparison graph 12 of the predicted result and the actual result of the corrosion rate of the reinforcing steel bars by the two optimizing methods is shown, and the correlation between the predicted value and the actual value is shown in figures 13 and 14.
Preferably, the error calculation method includes an average relative error, an average absolute error, a root mean square error, a correlation coefficient and/or a mean square error.
And finally, evaluating the performance and the prediction effect of the SVM model by selecting the average relative error (MRE), the average absolute error (MAE), the Root Mean Square Error (RMSE), a correlation coefficient (R2) and the Mean Square Error (MSE). MRE reflects the degree of deviation of the predicted value from the true value, MAE reflects the actual situation of the error of the predicted value, RMSE measures the deviation between the predicted value and the true value, MSE reflects the difference degree between the true value and the predicted value, the closer the value to 0, the smaller the error of the prediction model is, R2 can measure the fitting degree of the model, and the closer the value to 1, the better the fitting effect of the model is.
The specific calculation formula is as follows:
Figure BDA0002452444160000131
Figure BDA0002452444160000132
Figure BDA0002452444160000141
Figure BDA0002452444160000142
Figure BDA0002452444160000143
in the formula: y'iTo predict value, yiIn order to be the true value of the value,
Figure BDA0002452444160000144
the average of the true values, i ═ 1,2, …, n.
The comparison results are shown in FIG. 8.
The embodiment of the invention has the beneficial effects that:
and carrying out corrosion simulation on the prefabricated corrosion steel bar test piece, then carrying out learning training on the simulated corrosion steel bar by using a support vector machine to obtain a matching function, forming a prediction model, and further realizing prediction on the corrosion rate of the unknown steel bar by using the prediction model.
The method can realize the prediction of the corrosion rate of the steel bar only by knowing the specific characteristic parameters of the corroded steel bar, and has high accuracy and convenient use.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A steel bar corrosion parameter prediction method based on a support vector machine is characterized by comprising the following steps:
the method comprises the following steps: manufacturing a test piece of the corrosion reinforcing steel bar;
step two: 3D scanning is carried out on the test piece, and a 3D image model is generated;
step three: according to the 3D image model, calculating the specific characteristic parameters of the section of the test piece and the steel bar corrosion rate corresponding to the specific characteristic parameters of the section;
step four: setting specific characteristic parameters of the section and the corrosion rate of the steel bar corresponding to the specific characteristic parameters as a group of basic data, inputting a plurality of groups of basic data into a support vector machine for learning and training to obtain a prediction model for predicting the corrosion rate of the steel bar to be measured;
the specific characteristic parameters of the cross section of the test piece comprise:
the ratio of the minimum inscribed circle radius to the fitted circle radius η, the ratio of the maximum circumscribed circle radius to the fitted circle radius, the ratio of the minimum inscribed circle radius to the maximum circumscribed circle radius upsilon, the eccentricity e, the ratio of the short side of the fitted ellipse to the long side of the fitted ellipse, the roundness chi and the section roughness gamma.
2. The method for predicting the steel bar corrosion parameters based on the support vector machine according to claim 1, wherein in the step one, a corroded steel bar test piece is manufactured by adopting a power-on method.
3. The support vector machine-based steel bar corrosion parameter prediction method according to claim 1, wherein before 3D scanning of the test piece, the test piece is cleaned to remove a corroded part.
4. The method for predicting steel bar corrosion parameters based on the support vector machine according to claim 1, wherein the calculation formulas of the specific characteristic parameters are respectively as follows:
η＝r1/r0；
＝r2/r0；
ν＝r1/r2；
Figure FDA0002452444150000011
＝a/b；
χ＝p2/(4πA1)；
γ＝A1/(πr0 2)；
wherein r is1The minimum inscribed circle radius of the cross section profile; r is2The maximum circumscribed circle radius of the cross section profile; r is0Fitting a circle radius for the profile of the section of the steel bar; a is a short side of a fitting ellipse of the section profile of the steel bar; b is a long edge of a section contour fitting ellipse of the steel bar; y is0，z0Fitting ellipse center coordinates; and p is the perimeter of the residual cross section area after the steel bar is corroded.
5. The method for predicting the steel bar corrosion parameters based on the support vector machine according to claim 1, wherein the specific characteristic parameters of the cross section of the test piece are calculated by programming MAT L AB according to a 3D image model.
6. The support vector machine-based steel bar corrosion parameter prediction method according to claim 1, wherein when a plurality of groups of basic data are input into the support vector machine for learning training, a PSO particle swarm optimization method is used for optimizing an optimal penalty factor C and a kernel function parameter g for establishing an optimal regression model as a prediction model.
7. The support vector machine-based steel bar corrosion parameter prediction method according to claim 1, wherein when a plurality of groups of basic data are input into the support vector machine for learning training, a GS grid optimization method is used for establishing an optimal regression model as a prediction model.
8. The method for predicting the steel bar corrosion parameters based on the support vector machine according to claim 1, wherein after the prediction model is obtained, a plurality of groups of base data which are not trained in learning are input into the prediction model, and the prediction model is verified;
when the final error is within an allowable range, the prediction model is a final prediction model;
and when the final error is not in the allowable range, the prediction model is inaccurate, the step three is returned, the specific characteristic parameters of the section of the test piece are recalculated, and then a new prediction model is established again.
9. The support vector machine-based steel bar corrosion parameter prediction method according to claim 8, wherein the calculated final error comprises an average relative error, an average absolute error, a root mean square error, a correlation coefficient and/or a mean square error.
10. The method for predicting the steel bar corrosion parameter based on the support vector machine of claim 9, wherein the average relative error MRE is calculated by:
Figure FDA0002452444150000031
the mean absolute error MAE is calculated as:
Figure FDA0002452444150000032
the root mean square error RMSE is calculated as:
Figure FDA0002452444150000033
coefficient of correlation R2The calculation method is as follows:
Figure FDA0002452444150000034
the mean square error MSE is calculated as follows:
Figure FDA0002452444150000035
in the formula: y'iTo predict value, yiIn order to be the true value of the value,
Figure FDA0002452444150000036
the average of the true values, i ═ 1,2, …, n.
CN202010296670.8A 2020-04-15 2020-04-15 Reinforcing steel bar corrosion parameter prediction method based on support vector machine Active CN111444629B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010296670.8A CN111444629B (en) 2020-04-15 2020-04-15 Reinforcing steel bar corrosion parameter prediction method based on support vector machine

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010296670.8A CN111444629B (en) 2020-04-15 2020-04-15 Reinforcing steel bar corrosion parameter prediction method based on support vector machine

Publications (2)

Publication Number Publication Date
CN111444629A true CN111444629A (en) 2020-07-24
CN111444629B CN111444629B (en) 2023-07-18

Family

ID=71654325

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010296670.8A Active CN111444629B (en) 2020-04-15 2020-04-15 Reinforcing steel bar corrosion parameter prediction method based on support vector machine

Country Status (1)

Country Link
CN (1) CN111444629B (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112345382A (en) * 2020-11-03 2021-02-09 西北农林科技大学 Method for detecting mechanical strength of heat-treated wood
CN112668078A (en) * 2020-12-24 2021-04-16 青岛理工大学 Method for identifying damage of rusted reinforced concrete beam after fire disaster
CN113052207A (en) * 2021-03-10 2021-06-29 中国特种设备检测研究院 Stainless steel environment cracking judgment method and system based on Hybrid idea
CN113610160A (en) * 2021-08-09 2021-11-05 广州市稳建工程检测有限公司 Reinforcing steel bar detection and classification method and system based on neural network and storage medium
EP3945301A1 (en) * 2020-07-29 2022-02-02 ABB Schweiz AG Technologies for producing training data for identifying degradation of physical components
CN114904195A (en) * 2022-05-13 2022-08-16 常州机电职业技术学院 Fire early warning fire extinguishing system based on large-space warehouse fire early warning model
CN115112854A (en) * 2022-05-26 2022-09-27 北京理工大学 Ground penetrating radar corrosion steel bar identification method based on polarization characteristics and machine learning

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007149055A (en) * 2005-05-19 2007-06-14 Nano Geometry Kenkyusho:Kk Pattern inspection apparatus and method
CN102928580A (en) * 2012-11-20 2013-02-13 郑州大学 Corrosion monitoring device and method for reinforcement bar in concrete structure
JP2014074474A (en) * 2012-10-05 2014-04-24 Sumitomo Wiring Syst Ltd Fitting structure of fastening band
CN106769602A (en) * 2016-12-26 2017-05-31 深圳大学 Automatic testing equipment and method that a kind of Corrosion Reinforcement section is axially distributed
CN107729671A (en) * 2017-10-25 2018-02-23 西安科技大学 A kind of mine laneway coefficient of frictional resistance Forecasting Methodology based on SVMs
CN108204944A (en) * 2018-01-13 2018-06-26 福州大学 The Buried Pipeline rate prediction method of LSSVM based on APSO optimizations
CN108805355A (en) * 2018-06-09 2018-11-13 贵州电网有限责任公司 A kind of ground net corrosion prediction technique based on least square method supporting vector machine
CN109883940A (en) * 2019-02-28 2019-06-14 北京石油化工学院 A kind of construction method of degree of corrosion based on steel bar stress electrochemical signals and mechanical property degradation rule model
CN109948712A (en) * 2019-03-20 2019-06-28 天津工业大学 A kind of nanoparticle size measurement method based on improved Mask R-CNN
CN110838128A (en) * 2019-11-07 2020-02-25 华侨大学 Image method aggregate stacking void ratio prediction method and system

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007149055A (en) * 2005-05-19 2007-06-14 Nano Geometry Kenkyusho:Kk Pattern inspection apparatus and method
JP2014074474A (en) * 2012-10-05 2014-04-24 Sumitomo Wiring Syst Ltd Fitting structure of fastening band
CN102928580A (en) * 2012-11-20 2013-02-13 郑州大学 Corrosion monitoring device and method for reinforcement bar in concrete structure
CN106769602A (en) * 2016-12-26 2017-05-31 深圳大学 Automatic testing equipment and method that a kind of Corrosion Reinforcement section is axially distributed
CN107729671A (en) * 2017-10-25 2018-02-23 西安科技大学 A kind of mine laneway coefficient of frictional resistance Forecasting Methodology based on SVMs
CN108204944A (en) * 2018-01-13 2018-06-26 福州大学 The Buried Pipeline rate prediction method of LSSVM based on APSO optimizations
CN108805355A (en) * 2018-06-09 2018-11-13 贵州电网有限责任公司 A kind of ground net corrosion prediction technique based on least square method supporting vector machine
CN109883940A (en) * 2019-02-28 2019-06-14 北京石油化工学院 A kind of construction method of degree of corrosion based on steel bar stress electrochemical signals and mechanical property degradation rule model
CN109948712A (en) * 2019-03-20 2019-06-28 天津工业大学 A kind of nanoparticle size measurement method based on improved Mask R-CNN
CN110838128A (en) * 2019-11-07 2020-02-25 华侨大学 Image method aggregate stacking void ratio prediction method and system

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
JIN XIA等: "Numerical simulation of steel corrosion in chloride contaminated concrete", 《CONSTRUCTION AND BUILDING MATERIALS》, vol. 228 *
YA-JUN LV等: "Steel corrosion prediction based on support vector machines", 《CHAOS, SOLITONS AND FRACTALS》, no. 136 *
张家发等: "碎石颗粒形状测量与评定的初步研究", 《岩土力学》, pages 343 - 349 *
毕傲睿等: "基于主成分和粒子群优化支持向量机的管道内腐蚀预测", 《表面技术》, pages 133 - 140 *
王巧华等: "基于机器视觉的产地脏污鸭蛋外形扁平度在线检测", 《中国食品学报》 *
甘海龙等: "基于神经网络技术预测锈胀开裂后混凝土中钢筋锈蚀量", 《科技通报》, pages 133 - 140 *
雷大鹏;陈文彬;朱成魁;朱小娇;: "基于勾心标准样品定形尺寸的纵向刚度数学模型", 西部皮革, no. 03 *
魏仁: "锈蚀钢筋表观形貌及本构特征统计分析", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, pages 12 - 30 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3945301A1 (en) * 2020-07-29 2022-02-02 ABB Schweiz AG Technologies for producing training data for identifying degradation of physical components
CN112345382A (en) * 2020-11-03 2021-02-09 西北农林科技大学 Method for detecting mechanical strength of heat-treated wood
CN112668078A (en) * 2020-12-24 2021-04-16 青岛理工大学 Method for identifying damage of rusted reinforced concrete beam after fire disaster
CN113052207A (en) * 2021-03-10 2021-06-29 中国特种设备检测研究院 Stainless steel environment cracking judgment method and system based on Hybrid idea
CN113610160A (en) * 2021-08-09 2021-11-05 广州市稳建工程检测有限公司 Reinforcing steel bar detection and classification method and system based on neural network and storage medium
CN113610160B (en) * 2021-08-09 2024-03-15 广州市稳建工程检测有限公司 Neural network-based reinforcement detection classification method, system and storage medium
CN114904195A (en) * 2022-05-13 2022-08-16 常州机电职业技术学院 Fire early warning fire extinguishing system based on large-space warehouse fire early warning model
CN115112854A (en) * 2022-05-26 2022-09-27 北京理工大学 Ground penetrating radar corrosion steel bar identification method based on polarization characteristics and machine learning

Also Published As

Publication number Publication date
CN111444629B (en) 2023-07-18

Similar Documents

Publication Publication Date Title
CN111444629A (en) Reinforcing steel bar corrosion parameter prediction method based on support vector machine
CN114511203A (en) Method and system for rapidly outputting project amount list based on BIM (building information modeling)
CN110555247A (en) structure damage early warning method based on multipoint sensor data and BilSTM
CN112069656B (en) LSSVM-NSGAII durable concrete mixing ratio multi-objective optimization method
CN113722942A (en) Finite element calculation model considering 3D printing geometric defects
CN115688564A (en) Method for predicting mechanical property of twisted steel under rolling coupling effect
CN117162225B (en) Demoulding and forming method and system for concrete prefabricated part
CN114722478A (en) Clastic rock contact model parameter acquisition method and slope stability analysis method
CN116858665B (en) Analysis method for outline of round bar sample during uniaxial stretching necking deformation
CN113312812B (en) Method for generating initial geometric defects of cold-formed steel member
CN117313471A (en) Door machine stress field inversion method and system and electronic equipment
CN112784218A (en) Kurtosis estimation method for non-Gaussian random vibration signal
CN108304649B (en) High-rise building deformation prediction method
CN108665090B (en) Urban power grid saturation load prediction method based on principal component analysis and Verhulst model
Alyami et al. Estimation methods based on ranked set sampling for the arctan uniform distribution with application
CN106934729B (en) Building detection and identification method and device
CN109101759A (en) A kind of parameter identification method based on forward and reverse response phase method
JP7089808B2 (en) Information processing equipment, methods and programs for estimating measurement accuracy
CN114092545A (en) Self-adaptive grid searching method suitable for spherical target fitting
Mahammad Using AI in Dimensional Metrology
CN109858699B (en) Water quality quantitative simulation method and device, electronic equipment and storage medium
CN113361209A (en) Quantitative analysis method for magnetic anomaly of surface defects of high-temperature alloy
CN113283137A (en) Structural member corrosion effect control simulation method under organic protective coating system
CN118378048B (en) Method for identifying damage of rod system structure based on data-driven damage file
CN110009112B (en) Rock joint three-dimensional joint surface roughness measuring method

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
GR01 Patent grant
GR01 Patent grant