CN113793653A - High arch dam model test similar material mixing ratio determination method based on neural network - Google Patents

High arch dam model test similar material mixing ratio determination method based on neural network Download PDF

Info

Publication number
CN113793653A
CN113793653A CN202110934011.7A CN202110934011A CN113793653A CN 113793653 A CN113793653 A CN 113793653A CN 202110934011 A CN202110934011 A CN 202110934011A CN 113793653 A CN113793653 A CN 113793653A
Authority
CN
China
Prior art keywords
neural network
mechanical property
artificial neural
similar material
mixing ratio
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
Application number
CN202110934011.7A
Other languages
Chinese (zh)
Inventor
魏庆阳
曹茂森
付荣华
韦黎
席一粟
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.)
Hohai University HHU
Original Assignee
Hohai University HHU
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 Hohai University HHU filed Critical Hohai University HHU
Priority to CN202110934011.7A priority Critical patent/CN113793653A/en
Publication of CN113793653A publication Critical patent/CN113793653A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C60/00—Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
    • 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
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/048—Activation functions
    • 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
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/70—Machine learning, data mining or chemometrics

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Geometry (AREA)
  • Biophysics (AREA)
  • Mathematical Physics (AREA)
  • Biomedical Technology (AREA)
  • Computer Hardware Design (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Chemical & Material Sciences (AREA)
  • Architecture (AREA)
  • Civil Engineering (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Databases & Information Systems (AREA)
  • Structural Engineering (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)

Abstract

The invention provides a method for determining the mixing ratio of similar materials in a high arch dam model test based on a neural network, and belongs to the field of dam model tests. The method comprises the following steps of S1, establishing a training sample set of similar materials according to a preliminarily designed orthogonal experimental scheme and a mechanical property test result thereof; s2, establishing an artificial neural network for indicating the relation between the similar material mix proportion parameter and the similar material mechanical property, and training the artificial neural network by using the training sample set to obtain a target artificial neural network; s3, calculating to obtain a direct mapping set of similar material mixing ratio parameters and similar material mechanical properties by using the target artificial neural network and a further designed detailed experimental scheme; and S4, selecting similar material mixing ratio parameters corresponding to the mechanical property requirements in the direct mapping set by combining fitting error indexes according to the mechanical property requirements of the high arch dam model test. The method saves a great deal of time, cost and resources for design.

Description

High arch dam model test similar material mixing ratio determination method based on neural network
Technical Field
The invention relates to the field of dam model tests, in particular to a method for determining the mix proportion of similar materials in a high arch dam model test based on a neural network.
Background
The southwest area of China contains abundant water energy resources, and is the main position of water conservancy construction in China at the present stage. With the increase of social demands and the development of construction technology, many projects in the arch dam projects planned and built in the region are all over 200 m-300 m. However, in these areas, the geological conditions are complex, the seismic intensity is high, and the seismic acceleration is high, so that the seismic performance research of the arch dam is a main problem in engineering construction and scientific research tasks. At present, two main methods of numerical calculation and model test are mainly adopted for the research on the seismic performance problem of the high arch dam. In the former, due to the difference of the selected constitutive relation, the loading method and the integration means, results obtained by different researches are often greatly different. The model test can intuitively express the mechanical behavior of the arch dam under the action of earthquake, truly reflect the earthquake-resistant performance of the arch dam, and is necessary supplement and verification for numerical calculation. Therefore, in the design stage of the arch dam, it is necessary to research the dynamic response characteristics of the arch dam during earthquake, the crack propagation rule and determine the design earthquake acceleration through model tests.
For a special structure of the high arch dam, the stress level caused by the self weight of the high arch dam is much higher than that of a common structure, in addition, the inertia force generated under the action of an earthquake is not negligible, and therefore, the model test of the high arch dam under the action of the earthquake must meet the gravity-inertia force similarity criterion as far as possible. The actual scale of the project and laboratory conditions for different arch dam projects have determined the geometric scale of the model test, and therefore the weight and strength requirements that similar materials used to make physical models need to meet are stringent. In order to meet the requirements of both heavy weight and strength, the raw materials selected for preparing similar materials are often not common materials, and the mechanical properties are not easy and complete to grasp. In addition, unlike ordinary concrete materials and the like, similar materials used to prepare physical models have different requirements in different arch dam projects. Both of the above aspects bring great difficulty to the determination of the similar materials of the model test aiming at a specific high arch dam project.
Traditionally, the formulation of similar materials has relied on extensive formulation attempts and experiential experience by laboratory personnel. The method comprises the steps of determining a plurality of raw materials according to the needs of a specific model test, further carrying out a large amount of mixing ratio design and testing the performance of the raw materials, thereby judging the influence of the content of each raw material on the mechanical property of similar materials, and finally gradually adjusting the mixing ratio design to determine the materials meeting the requirements. The method has the problem of low efficiency and can bring huge resource waste. More importantly, when the mechanical properties to be considered are not unique, the ratio determined empirically is often only satisfactory to a certain extent and is not optimal. In addition, some mathematical models are used to fit the relationship between the mix ratio parameter and the property, but when similar materials are determined, the relationship between the mix ratio parameter and the mechanical property is highly non-linear, and it is difficult to fit with a certain mathematical model. Therefore, it is very urgent and practical to determine the similar material ratio needed efficiently and accurately with a small amount of experimental results of ratio trial.
Disclosure of Invention
Based on the defects of the prior art, the invention aims to provide a method for determining the mixing ratio of similar materials in a high arch dam model test based on a neural network so as to meet the requirements of the high arch dam dynamic model test.
In order to achieve the purpose, the invention provides the following technical scheme: a method for determining a high arch dam model test similar material based on a neural network comprises the following steps:
s1, establishing a training sample set of similar materials according to the preliminarily designed orthogonal experimental scheme and the mechanical property test result thereof;
s2, establishing an artificial neural network for indicating the relation between the similar material mixing ratio parameters and the similar material mechanical properties, and training the artificial neural network by using the obtained training sample set to obtain a target artificial neural network;
and S3, calculating to obtain a direct mapping set of the mix proportion parameters and the mechanical properties of the similar materials by using the target artificial neural network and a further designed detailed experimental scheme.
And S4, selecting similar material mixing ratio parameters corresponding to the mechanical property requirements in the direct mapping set by combining the fitting error indexes according to the mechanical property requirements of the high arch dam model test to be developed.
In one embodiment, the design factors of the orthogonal experimental scheme preliminarily designed in step S1 include: based on the kind of raw materials proposed for formulating similar materials, the design level of the preliminarily designed orthogonal experimental protocol includes: 4-6 grades are divided according to the mass ratio of the raw materials.
In one embodiment, the mechanical properties required to be measured in step S1 include at least one of the following parameters: the weight, compressive strength, split strength, elastic modulus, etc. of similar materials.
In an embodiment, the mechanical property test result is obtained by testing and calculating a set of 6 test pieces, and the calculation method is as follows:
Figure BDA0003210607630000021
wherein, yijThe marked value of the mechanical property parameter j under the mix proportion i is shown,
Figure BDA0003210607630000022
1 group of 6 test pieces are sorted from small to large, and k is more than or equal to 1 and less than or equal to 6.
In one embodiment, the specific steps of step S2 are as follows:
s21: input feature vectors of an input training sample set
Figure BDA0003210607630000031
And outputting the feature vector
Figure BDA0003210607630000032
Figure BDA0003210607630000033
And training the artificial neural network, and storing a training result.
Wherein x isimDenotes the content of the raw material m at the mixing ratio i, yijThe mechanical property parameter j at the mixing ratio i is shown.
S22: a network accuracy error index is established and the accuracy of the artificial neural network trained in S21 is calculated.
Figure BDA0003210607630000034
Where n denotes the number of test samples for testing the accuracy of the artificial neural network, yijIs the labeled value of the mechanical property parameter j under the mix proportion i, yfijIs the predicted value of the mechanical property parameter j under the mixing proportion i. a is1，a2，...，ajRepresenting the weight of the factor 1, 2, j in the accuracy error index, the weight taking a value between 1% and 100% according to the importance degree of each index in the studied model test problem.
S23: and repeating the steps of S21 and S22 1000 times, training 1000 artificial neural networks, calculating to obtain corresponding 1000 accuracy error indexes, and selecting the most accurate artificial neural network according to the minimum value of the accuracy error indexes to determine the target artificial neural network.
In one embodiment, the selected neural network is a BP neural network, the hidden layer is selected from 1-2 layers, the number of nodes is selected from 5-15 layers, the transfer function is selected from a Sigmoid function, and the training method uses a steepest descent method.
In one embodiment, the refined protocol is generated by:
generating several columns of random numbers on the formulated interval, the number of columns is identical to the number of the factors m, the number of rows is set to 10000 rows, and forming a fine design scheme of 10000 mix proportions
Figure BDA0003210607630000035
And calculating the relation between the mix proportion parameter and the mechanical property through the target artificial neural network to obtain a direct mapping set capable of comprehensively reflecting the relation between the mix proportion parameter and the mechanical property. r represents a random number and Z represents a formed random number matrix.
In an embodiment, the establishing of the fitting error index in step S4 specifically includes:
the fitting error index is calculated by the following formula:
Figure BDA0003210607630000041
where ω is the fitting error index, ytijIs the target value of the mechanical property parameter j at the mix proportion i, yfijIs the predicted value of the mechanical property parameter j under the mixing proportion i. a is1，a2，...，ajAnd representing the weight of the design factors 1, 2, the weight of j in the fitting error index, taking a value between 1% and 100% according to the importance degree of each index in the researched model test problem, finally confirming that the performance is closer to the required performance, and meeting the similar material mixing ratio parameters of the model test on the requirements of similar materials.
Compared with the prior art, the invention provides a method for determining the mix proportion of similar materials in a high arch dam model test based on a neural network, which establishes a neural network training sample set and calculates a direct mapping set reflecting the mix proportion and mechanical property parameters of the materials. The most accurate neural network is selected by using an accuracy error index in the sample training process, and the training accuracy is guaranteed in probability. With the aid of the fitting error index, the established direct mapping set can provide choices for tests with different performance requirements. Compared with the prior art, the method has the advantages that the determined mix proportion design saves a great deal of time, cost and resources for developing material proportion design.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention;
FIG. 2 is a diagram of a neural network architecture for use with the present invention;
FIG. 3 is a graph of the accuracy calculation of 1000 neural networks in an embodiment of the present invention;
FIG. 4 is a four-column random number used to design a refined mix-proportion scheme.
Detailed Description
In order to make the objects and technical solutions of the present invention more apparent, the present invention will be further described with reference to the following embodiments and the accompanying drawings.
Example (b): the dam height of a certain arch dam is 240 meters, C30 concrete is adopted for pouring, the dam height of a physical model is designed to be 1 meter, and the geometric scale of the dam height is 240 meters. Referring to the flowchart shown in fig. 1, the present embodiment includes the following steps;
step S1: and establishing a training sample set of similar materials according to the preliminarily designed orthogonal experimental scheme and the mechanical property test result thereof. The raw materials to be drawn up comprise barite sand, barite powder, fly ash, cement and water, and the design factor of the preliminarily designed orthogonal experimental scheme is x1Water/flyash + cement, x2Cement, flyash, recrystallized powder, recrystallized sand, x3Cement/fly ash, x4The design level of the recrystallized powder/recrystallized sand is that 4 grades are divided according to the mass ratio and are shown in table 1, and 16 mixing ratio schemes are designed and are shown in table 2.
TABLE 1
Figure BDA0003210607630000042
Figure BDA0003210607630000051
The mechanical properties to be measured in step S1 include y for similar materials1Severe, y2Compressive strength, y3Cleavage Strength and y4Modulus of elasticity. The test of any mechanical property is obtained by testing and calculating a group of 6 poured test pieces, and the calculation mode is as follows:
Figure BDA0003210607630000052
wherein, yijThe mechanical property parameter j under the mixing proportion i is shown,
Figure BDA0003210607630000053
the 6 test pieces in 1 group are sorted from small to large, and the measurement results of the mechanical property parameters in the embodiment are summarized in table 2.
TABLE 2
Figure BDA0003210607630000054
Figure BDA0003210607630000061
Step S2: and establishing an artificial neural network for indicating the relation between the similar material mixing ratio parameters and the similar material mechanical properties, and training the artificial neural network by using the obtained training sample set. The specific steps of step S2 are as follows:
s21: input feature vectors of an input training sample set
Figure BDA0003210607630000062
And outputting the feature vector
Figure BDA0003210607630000063
Figure BDA0003210607630000064
And training the artificial neural network, and storing a training result.
Wherein x isimDenotes the content of the raw material m at the mixing ratio i, yijThe mechanical property parameter j at the mixing ratio i is shown. Still further, the neural network selected in step S22 is a BP neural network, the hidden layer selects 1 layer, the number of nodes selects 8, the transfer function selects a Sigmoid function, and the training method uses a steepest descent method. The structure of the neural network used is shown in figure 2.
S22: a network accuracy error index is established and the accuracy of the artificial neural network trained in S21 is calculated.
Figure BDA0003210607630000065
Where n denotes the number of test samples used to test the accuracy of the artificial neural network, yijIs the labeled value of the mechanical property parameter j under the mix proportion i, yfijIs the predicted value of the mechanical property parameter j under the mixing proportion i. a is1＝50％，a2＝70％，a3＝ 50％，a4100% represents the weight of the design factor 1, 2.., j in the accuracy error index, which takes values between 1% and 100% depending on the degree of importance of the respective index in the model test problem under study.
S23: repeating the steps of S21 and S22 1000 times, training 1000 artificial neural networks and calculating to obtain corresponding 1000 accuracy error indexes, and selecting 896 th neural network according to the minimum value of the accuracy error indexes, as shown in figure 3, so as to determine the target artificial neural network.
Step S3: further refinement of the mix proportion design scheme yields 4 columns of random numbers over a proposed interval, the number of random numbers in each column being 10000, the intervals being [0.516, 0.727], [0.176, 0.214], [0.1, 0.7], [0.5, 1], respectively, as shown in FIG. 4. And calculating the relation between the mix proportion parameter and the mechanical property through the target artificial neural network determined in the step S23 to obtain a direct mapping set r representing random numbers capable of comprehensively reflecting the relation between the mix proportion parameter and the mechanical property, wherein Z represents a formed random number matrix.
Step S4: and selecting corresponding similar material mixing ratio parameters in the calculated direct mapping set according to the mechanical property requirement of the high arch dam model test to be developed. As the elasticity problem is researched, the elastic modulus needs to be ensured to be completely similar and certain strength is ensured, and the target mechanical property requirements are 2550kg/m3 with heavy weight, 2MPa compressive strength, 150Kpa splitting strength and 130MPa elastic modulus. Establishing a fitting error index, calculating the fitting error index in a direct mapping set by combining with mechanical property requirements, and selecting a mix proportion parameter with the highest fitting as a mix proportion finally used for configuring the high arch dam model test similar material, wherein the fitting error index is as follows:
Figure BDA0003210607630000071
where ω is the fitting error index, ytijIs the target value of the mechanical property parameter j at the mix proportion i, yfijThe predicted value of the mechanical property parameter j under the mixing proportion i is shown, and the mechanical property parameter j in the embodiment comprises 4 types, namely, the weight, the compressive strength, the splitting strength and the elastic modulus. a is1＝50％，a2＝70％，a3＝50％，a4100% represents the weight of the factor 1, 2.. j in the accuracy error index, taking values between 1% and 100%, depending on how important each index is in the model test problem under study. The finally determined similar material mixing ratio parameters are as follows: x is the number of1＝0.647，x2＝0.198，x3＝0.697，x40.501, its predicted performance is y1＝2555.70kg/m3，y2＝2.09Mpa，y3＝160.96Kpa，y4130.14Mpa, which is close to the required performance, can meet the requirement of model test for similar materials.

Claims (8)

1. A high arch dam model test similar material mixing ratio determination method based on a neural network is characterized by comprising the following steps:
s1, establishing a training sample set of similar materials according to the designed orthogonal experimental scheme and the mechanical property test result thereof;
s2, establishing an artificial neural network for indicating the relation between the similar material mixing ratio parameter and the similar material mechanical property, and training the artificial neural network by using the training sample set to obtain a target artificial neural network;
s3, calculating to obtain a direct mapping set of similar material mixing ratio parameters and similar material mechanical properties by using the target artificial neural network and the detailed experimental scheme;
s4, establishing a fitting error index, and selecting a similar material mixing ratio parameter with the highest fitting degree corresponding to the mechanical property requirement in the direct mapping set according to the mechanical property requirement of the high arch dam model test by combining the fitting error index, so as to obtain the finally determined similar material mixing ratio parameter.
2. The method of claim 1, wherein the design factors of the orthogonal experimental protocol designed in step S1 include: based on the kind of raw materials proposed for preparing similar materials; the design levels of the preliminarily designed orthogonal experimental scheme include: 4-6 grades are divided according to the mass ratio of the raw materials.
3. The method according to claim 1, wherein the mechanical properties required to be tested in step S1 include at least one of the following parameters: the gravity, compressive strength, split strength, and elastic modulus of similar materials.
4. The method according to claim 1, wherein in the step S1, the mechanical property test result is obtained by testing and calculating a set of 6 test pieces to be poured, and the calculation is as follows:
Figure FDA0003210607620000011
wherein, yijThe mechanical property parameters at the mixing ratio ijThe value of the label of (a) is,
Figure FDA0003210607620000012
1 group of 6 test pieces are sorted from small to large, and k is more than or equal to 1 and less than or equal to 6.
5. The method according to claim 1, wherein step S2 includes:
s21: input feature vectors of an input training sample set
Figure FDA0003210607620000013
And outputting the feature vector
Figure FDA0003210607620000014
Figure FDA0003210607620000021
Training an artificial neural network, and storing a training result;
wherein x isimDenotes the content of the raw material m at the mixing ratio i, yijThe value of the mechanical property parameter j under the mix proportion i is expressed;
s22: establishing a network accuracy error index, and calculating the accuracy of the artificial neural network trained in the S21:
Figure FDA0003210607620000022
where n denotes the number of test samples for testing the accuracy of the artificial neural network, yijIs the labeled value of the mechanical property parameter j under the mix proportion i, yfijIs the predicted value of the mechanical property parameter j under the mix proportion i; a is1，a2，...，ajRepresenting the weight of the design factor 1, 2.. times, j in the accuracy error index, in accordance with the respective index in the model test problem under studyThe importance degree is between 1% and 100%;
s23: and repeating the steps of S21 and S22 1000 times, training 1000 artificial neural networks, calculating to obtain corresponding 1000 accuracy error indexes, and selecting the most accurate artificial neural network according to the minimum value of the accuracy error indexes to determine the target artificial neural network.
6. The method of claim 5, wherein the artificial neural network is a BP neural network, the hidden layer is selected from 1-2 layers, the number of nodes is selected from 5-15, the transfer function is selected from a Sigmoid function, and the training method uses a steepest descent method.
7. The method of claim 1, wherein the refined protocol is generated by:
generating a plurality of columns of random numbers in a formulated interval, wherein the number of the columns is consistent with the number m of raw materials, the number of the rows is set to 10000 rows, and a refined design scheme with 10000 mix proportions is formed:
Figure FDA0003210607620000023
and calculating the relation between the mix proportion parameter and the mechanical property through the target artificial neural network to obtain a direct mapping set capable of comprehensively reflecting the relation between the mix proportion parameter and the mechanical property, wherein r represents a random number, and Z represents a formed random number matrix.
8. The method according to claim 1, wherein in the step S4
The establishment of the fitting error index specifically comprises the following steps:
the fitting error index is calculated by the following formula:
Figure FDA0003210607620000024
wherein ω isFitting error index, ytijIs the target value of the mechanical property parameter j at the mix proportion i, yfijIs the predicted value of the mechanical property parameter j under the mix proportion i; a is1，a2，...，ajAnd representing the weight of the design factors 1, 2, the weight of j in the fitting error index, taking a value between 1% and 100% according to the importance degree of each index in the researched model test problem, finally confirming that the performance is closer to the required performance, and meeting the similar material mixing ratio parameters of the model test on the requirements of similar materials.
CN202110934011.7A 2021-08-13 2021-08-13 High arch dam model test similar material mixing ratio determination method based on neural network Pending CN113793653A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110934011.7A CN113793653A (en) 2021-08-13 2021-08-13 High arch dam model test similar material mixing ratio determination method based on neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110934011.7A CN113793653A (en) 2021-08-13 2021-08-13 High arch dam model test similar material mixing ratio determination method based on neural network

Publications (1)

Publication Number Publication Date
CN113793653A true CN113793653A (en) 2021-12-14

Family

ID=79181703

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110934011.7A Pending CN113793653A (en) 2021-08-13 2021-08-13 High arch dam model test similar material mixing ratio determination method based on neural network

Country Status (1)

Country Link
CN (1) CN113793653A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114496125A (en) * 2022-02-10 2022-05-13 西南交通大学 Preparation method, device and equipment of similar material and readable storage medium
CN115964931A (en) * 2022-11-04 2023-04-14 广西大学 Theory, method and formula for preparing materials with similar dynamic and static characteristics of rock high-strength Gao Cui

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104991051A (en) * 2015-06-30 2015-10-21 华侨大学 Method for predicting concrete strength based on hybrid model
CN109696355A (en) * 2019-01-21 2019-04-30 中国林业科学研究院木材工业研究所 A kind of measuring method for the Long-Term Tensile Strength recombinating composite construction bamboo wood
CN110262219A (en) * 2019-06-14 2019-09-20 广东工业大学 A kind of motor PID automatic setting method based on BP neural network
AU2020101453A4 (en) * 2020-07-23 2020-08-27 China Communications Construction Co., Ltd. An Intelligent Optimization Method of Durable Concrete Mix Proportion Based on Data mining
CN112149797A (en) * 2020-08-18 2020-12-29 Oppo（重庆）智能科技有限公司 Neural network structure optimization method and device and electronic equipment

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104991051A (en) * 2015-06-30 2015-10-21 华侨大学 Method for predicting concrete strength based on hybrid model
CN109696355A (en) * 2019-01-21 2019-04-30 中国林业科学研究院木材工业研究所 A kind of measuring method for the Long-Term Tensile Strength recombinating composite construction bamboo wood
CN110262219A (en) * 2019-06-14 2019-09-20 广东工业大学 A kind of motor PID automatic setting method based on BP neural network
AU2020101453A4 (en) * 2020-07-23 2020-08-27 China Communications Construction Co., Ltd. An Intelligent Optimization Method of Durable Concrete Mix Proportion Based on Data mining
CN112149797A (en) * 2020-08-18 2020-12-29 Oppo（重庆）智能科技有限公司 Neural network structure optimization method and device and electronic equipment

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
陈智勇 等: "基于前馈网络的柔性材料配比模型的研究", 《长江科学院院报》, vol. 1, no. 4, pages 2 *
顾清恒: "似膏体充填材料配比的BP网络优化方法", 《金属矿山》, vol. 1, no. 3, pages 1 - 3 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114496125A (en) * 2022-02-10 2022-05-13 西南交通大学 Preparation method, device and equipment of similar material and readable storage medium
CN114496125B (en) * 2022-02-10 2023-05-02 西南交通大学 Preparation method, device and equipment of similar materials and readable storage medium
CN115964931A (en) * 2022-11-04 2023-04-14 广西大学 Theory, method and formula for preparing materials with similar dynamic and static characteristics of rock high-strength Gao Cui
CN115964931B (en) * 2022-11-04 2023-09-05 广西大学 Theory, method and formula for preparing rock materials with high strength and high brittleness and similar dynamic and static characteristics

Similar Documents

Publication Publication Date Title
Dingqiang et al. A novel approach for developing a green Ultra-High Performance Concrete (UHPC) with advanced particles packing meso-structure
Behnood et al. Prediction of the compressive strength of normal and high-performance concretes using M5P model tree algorithm
CN113793653A (en) High arch dam model test similar material mixing ratio determination method based on neural network
Esposito et al. Literature review of modelling approaches for ASR in concrete: a new perspective
CN109522577B (en) Concrete fatigue life prediction method and device based on Weibull equation and maximum fatigue deformation
CN101976221B (en) Particle swarm taboo combination-based parallel test task dispatching method and platform
CN112113875B (en) Intelligent gradient temperature control method, system, equipment and readable storage medium
CN111310390B (en) Intelligent prediction method for concrete pumping performance
CN102425148A (en) Rapid sub-grade settlement predicting method based on static sounding and BP (Back Propagation) neural network
Eskandari et al. Effect of 32.5 and 42.5 cement grades on ANN prediction of fibrocement compressive strength
CN112016663A (en) Polymer slurry parameter identification method based on group intelligent optimization algorithm
CN110046427B (en) Orthogonal design and normal cloud model based sand T-beam concrete mixing proportion method
Zhou et al. Creep parameter inversion for high CFRDs based on improved BP neural network response surface method
Chaube et al. Modelling of concrete performance: Hydration, microstructure and mass transport
CN109238950A (en) Atmospheric corrosion of metal materials prediction technique based on qualitative analysis and quantitative forecast
Hong et al. Dynamic evaluation for compaction quality of roller compacted concrete based on reliability metrics
Zhao et al. Research on the application of computer intelligent detection in civil engineering technology
Ren et al. The application of BIM technology in the research on seismic performance of shear wall structure of prefabricated residential buildings
CN110516405B (en) Construction method of hydration heat presumption-free prediction model of portland cement-based cementing material system
CN111504779B (en) Method and device for determining rock softening curve by using brittleness index
Liu et al. Cement-based grouting material development and prediction of material properties using PSO-RBF machine learning
Jia et al. Optimization inversion of material parameters of arch dam based on PSO-LSTM
CN116959636A (en) Deep coal series stratum grouting material design method and grouting material
da Silva et al. Fuzzy affinity hydration model
CN109522568B (en) Concrete fatigue life prediction method and device based on index Weibull equation and maximum fatigue deformation

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