CN116090049A - Method for correcting deformation value of containment structure integrity test based on BP neural network - Google Patents

Method for correcting deformation value of containment structure integrity test based on BP neural network Download PDF

Info

Publication number
CN116090049A
CN116090049A CN202211646049.5A CN202211646049A CN116090049A CN 116090049 A CN116090049 A CN 116090049A CN 202211646049 A CN202211646049 A CN 202211646049A CN 116090049 A CN116090049 A CN 116090049A
Authority
CN
China
Prior art keywords
data
deformation
neural network
test
containment structure
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
CN202211646049.5A
Other languages
Chinese (zh)
Inventor
杨昕光
汪全
宋翔
郑磊
单强
刘涵婧
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.)
China Nuclear Power Operation Co Ltd
MCC Inspection and Certification Co Ltd
Original Assignee
China Nuclear Power Operation Co Ltd
MCC Inspection and Certification 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 China Nuclear Power Operation Co Ltd, MCC Inspection and Certification Co Ltd filed Critical China Nuclear Power Operation Co Ltd
Priority to CN202211646049.5A priority Critical patent/CN116090049A/en
Publication of CN116090049A publication Critical patent/CN116090049A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
    • G01N3/08—Investigating strength properties of solid materials by application of mechanical stress by applying steady tensile or compressive forces
    • 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
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
    • G01N2203/0014—Type of force applied
    • G01N2203/0016—Tensile or compressive
    • G01N2203/0019—Compressive
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2203/00—Investigating strength properties of solid materials by application of mechanical stress
    • G01N2203/02—Details not specific for a particular testing method
    • G01N2203/06—Indicating or recording means; Sensing means
    • G01N2203/067—Parameter measured for estimating the property
    • G01N2203/0682—Spatial dimension, e.g. length, area, angle
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • 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
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E30/00—Energy generation of nuclear origin
    • Y02E30/30—Nuclear fission reactors

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Pathology (AREA)
  • Mathematical Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Architecture (AREA)
  • Civil Engineering (AREA)
  • Immunology (AREA)
  • Chemical & Material Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computational Mathematics (AREA)
  • Biochemistry (AREA)
  • Analytical Chemistry (AREA)
  • Artificial Intelligence (AREA)
  • Structural Engineering (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)

Abstract

The invention discloses a containment structure integrity test deformation measurement value correction method, device and equipment based on a BP neural network algorithm, and relates to the technical field of prestress concrete structure safety evaluation and machine learning, wherein the method comprises the steps of sampling data in a pre-test process and performing first data processing to obtain sample data containing first environment temperature data, first solar radiation intensity and first deformation measurement data; training a BP neural network model by using sample data to obtain an optimal BP neural network model; inputting the actual measurement data in the formal test period after normalization processing into an optimal BP neural network model for calculation to obtain second deformation prediction data in the formal test period; and correcting the comprehensive deformation measured value of the containment structure during the formal test by using the second deformation prediction data to obtain a deformation measurement corrected value of the containment structure under the action of the test pressure. The invention has the advantages of high measurement precision of the deformation of the containment structure and high practical engineering application value.

Description

Method for correcting deformation value of containment structure integrity test based on BP neural network
Technical Field
The invention relates to the technical field of prestress concrete structure safety evaluation and machine learning, in particular to a containment structure integrity test deformation measurement value correction method, device and equipment based on BP neural network algorithm.
Background
The containment vessel serves as the last barrier to leakage of fission products to the environment, ensuring its integrity is particularly important. In order to verify whether the overall tightness, structural performance and strength of the containment meet the requirements, a containment structural integrity test must be performed every ten years after the first refueling overhaul of the unit.
In the test, the prestressed concrete containment structure mainly bears the test pressure effect, and measures deformation, strain, temperature, prestress steel beam force value and the like, so that the prestressed concrete containment structure is a typical low-load and small-range precision test, and has high measurement accuracy requirement. For example, under maximum test internal pressure, the vertical displacement of the dome is only about one ten thousandth of the height of the containment, and the deformation measurement accuracy is required to be hundreds of thousands. Therefore, the influence of comprehensive environmental factors such as ambient temperature and solar radiation on the integrity test of the containment structure needs to be considered, and the deformation of the containment structure is reasonably corrected to obtain a measurement result of the load precision requirement.
Most of the existing structural deformation measurement correction methods are based on linear fitting methods to obtain an approximate relation curve of the ambient temperature and the structural deformation, so that the fitting relation of the ambient temperature and the structural deformation is required to be determined in advance, the interference of human factors is large, the calculation is complicated, the influence of solar radiation on the structural deformation is not considered in the existing correction methods, the error is large, and a scientific and accurate safety shell structural integrity test deformation measurement value correction method is difficult to form.
Disclosure of Invention
Therefore, in order to overcome the defects in the prior art, the invention provides a method, a device and equipment for correcting the structural integrity test deformation measured value of the containment structure based on the BP neural network algorithm, wherein an intelligent algorithm is applied to actual engineering, the influence of ambient temperature and solar radiation on the structural deformation measured value is comprehensively considered, the relationship among the ambient temperature, the solar radiation intensity and the structural deformation can be automatically judged, the interference of human factors is eliminated, and the structural deformation measuring precision of the containment structure is improved.
Therefore, the method for correcting the deformation measured value of the containment structure integrity test based on the BP neural network algorithm comprises the following steps:
s1, sampling data in a continuous time period similar to weather parameters in a formal test period in a pre-test process and performing first data processing to obtain sample data; the sample data includes first ambient temperature data, first solar radiation intensity, and first deformation measurement data caused by the first ambient temperature data and the first solar radiation intensity; the weather parameters comprise ambient temperature, humidity, wind direction, wind power, solar radiation intensity and the like; the duration of the time period is not less than 72h;
s2, training the BP neural network model on the basis of the sample data to obtain an optimal BP neural network model; the BP neural network model is input into first environment temperature data and first solar radiation intensity, and is output into first deformation prediction data;
s3, acquiring data in a formal test period and performing second data processing to obtain actual measurement data; the measured data includes second ambient temperature data and second solar radiation intensity;
s4, inputting the actually measured data into the optimal BP neural network model for calculation after normalization processing, and obtaining second deformation prediction data caused by second environmental temperature data and second solar radiation intensity during a formal test;
s5, acquiring a comprehensive deformation measured value of the containment structure during a formal test;
and S6, correcting the comprehensive deformation measured value by using the second deformation prediction data to obtain a deformation measurement correction value of the containment structure under the action of the test pressure.
Preferably, the data during the sampling pre-test includes a first ambient temperature, a first solar radiation intensity and a first deformation measurement.
Preferably, the first data processing includes a first outlier processing, a deformation hysteresis processing, and a first difference processing. The first outlier processing comprises interpolation substitution of outliers with data on the left and right sides of the outliers if the outliers suddenly increase or decrease in the data in the pre-test process. The deformation hysteresis process includes adjusting a time course curve of the first deformation measurement to eliminate hysteresis of the deformation. The first difference processing includes subtracting the first initial ambient temperature from the first ambient temperature at each time to obtain first ambient temperature data and subtracting the initial measured value from the first deformation measured value at each time to obtain first deformation measured data.
Preferably, the step of training the BP neural network model based on the sample data to obtain an optimal BP neural network model includes:
s21, carrying out normalization processing on the sample data and dividing the sample data into a training sample and a test sample;
s22, constructing a BP neural network model with an input layer, an implicit layer and an output layer which are sequentially connected, wherein the output of each layer serves as the input of the next layer. The input layer is provided with 2 nodes and is used for receiving the first environmental temperature data and the normalized value of the first solar radiation intensity in a one-to-one correspondence manner; the output layer has 1 node for outputting first deformation prediction data. The activation function from the input layer to the hidden layer is a tangent S-shaped function, the activation function from the hidden layer to the output layer is a linear function, and the training function adopts a Levenberg-Marquardt optimization algorithm. The number of nodes of the hidden layer is according to an empirical formula
Figure BDA0004009574040000031
Determining, wherein m is the number of nodes of an input layer, n is the number of nodes of an output layer, and a is [1,10 ]]An integer therebetween.
S23, inputting a training sample into the BP neural network model, performing BP neural network training, verifying a training effect by using a test sample, and obtaining a trained optimal BP neural network model through error analysis and determination of first deformation prediction data and first deformation measurement data output by the model.
Preferably, the data collected during the formal test includes a second ambient temperature and a second solar radiation intensity.
Preferably, the second data processing includes a second outlier processing and a second difference processing. The first outlier processing includes interpolating the outlier with data on both its left and right sides if there is an outlier that suddenly increases or decreases during the formal test. The second difference processing includes subtracting the second initial ambient temperature from the second ambient temperature at each time to obtain second ambient temperature data.
Preferably, the step of correcting the integrated deformation measurement value by using the second deformation prediction data to obtain a deformation measurement correction value of the containment structure includes:
and S61, after the second deformation prediction data is subjected to inverse normalization processing, removing the second deformation prediction data from the comprehensive deformation measurement value to obtain a deformation measurement correction value of the containment structure under the action of test pressure.
The invention relates to a containment structure integrity test deformation measurement value correction device based on BP neural network algorithm, which comprises:
the pre-test sampling unit is used for carrying out a pre-test of the containment structure integrity test in a time period similar to weather parameters during a formal test of the containment structure integrity test, sampling data in the pre-test process and carrying out first data processing to obtain sample data; the sample data includes first ambient temperature data, first solar radiation intensity, and first deformation measurement data caused by the first ambient temperature data and the first solar radiation intensity; the weather parameters include a second ambient temperature, humidity, wind direction, wind force, and a second solar radiation intensity; the duration of the time period is not less than 72h;
the BP neural network model training unit is used for training the BP neural network model on the basis of the sample data to obtain an optimal BP neural network model; the BP neural network model is input into first environment temperature data and first solar radiation intensity, and is output into first deformation prediction data;
the formal test sampling unit is used for collecting data in a formal test period and performing second data processing to obtain actual measurement data; the measured data includes second ambient temperature data and second solar radiation intensity;
the deformation prediction unit is used for inputting the actually measured data into the optimal BP neural network model for calculation after normalization processing to obtain second deformation prediction data caused by second environmental temperature data and second solar radiation intensity during a formal test;
the comprehensive deformation measurement value acquisition unit is used for acquiring the comprehensive deformation measurement value of the containment structure during the formal test;
and the deformation measurement value correction unit is used for correcting the comprehensive deformation measurement value by using the second deformation prediction data to obtain a deformation measurement correction value of the containment structure under the action of the test pressure.
Preferably, the data during the sampling pre-test includes a first ambient temperature, a first solar radiation intensity and a first deformation measurement.
Preferably, the first data processing includes a first outlier processing, a deformation hysteresis processing, and a first difference processing. The first outlier processing comprises interpolation substitution of outliers with data on the left and right sides of the outliers if the outliers suddenly increase or decrease in the data in the pre-test process. The deformation hysteresis process includes adjusting a time course curve of the first deformation measurement to eliminate hysteresis of the deformation. The first difference processing includes subtracting the first initial ambient temperature from the first ambient temperature at each time to obtain first ambient temperature data and subtracting the initial measured value from the first deformation measured value at each time to obtain first deformation measured data.
Preferably, the BP neural network model training unit includes:
the normalization processing unit is used for performing normalization processing on the sample data and dividing the sample data into a training sample and a test sample;
and the BP neural network model construction unit is used for constructing the BP neural network model with an input layer, an hidden layer and an output layer which are sequentially connected, and the output of each layer is used as the input of the next layer. The input layer is provided with 2 nodes and is used for receiving the first environmental temperature data and the normalized value of the first solar radiation intensity in a one-to-one correspondence manner; the output layer has 1 node for outputting first deformation prediction data. The activation function from the input layer to the hidden layer is a tangent S-shaped function, the activation function from the hidden layer to the output layer is a linear function, and the training function adopts a Levenberg-Marquardt optimization algorithm. The number of nodes of the hidden layer is according to an empirical formula
Figure BDA0004009574040000051
Determining, wherein m is the number of nodes of an input layer, n is the number of nodes of an output layer, and a is [1,10 ]]An integer therebetween.
The optimal BP neural network model determining unit is used for inputting a training sample into the BP neural network model, performing BP neural network training, verifying the training effect by using a test sample, and obtaining a trained optimal BP neural network model through error analysis and determination of first deformation prediction data and first deformation measurement data output by the model.
Preferably, the data collected during the formal test includes a second ambient temperature and a second solar radiation intensity.
Preferably, the second data processing includes a second outlier processing and a second difference processing. The first outlier processing includes interpolating the outlier with data on both its left and right sides if there is an outlier that suddenly increases or decreases during the formal test. The second difference processing includes subtracting the second initial ambient temperature from the second ambient temperature at each time to obtain second ambient temperature data.
Preferably, the deformation measurement correction unit includes:
and the rejecting unit is used for rejecting the second deformation prediction data from the comprehensive deformation measured value after performing inverse normalization processing to obtain a deformation measurement correction value of the containment structure.
The invention relates to a containment structure integrity test deformation measured value correction device based on BP neural network algorithm, which comprises:
one or more processors; and
a storage device for storing one or more programs; the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the above-described containment structure integrity test deformation measurement correction method based on the BP neural network algorithm.
The method, the device and the equipment for correcting the deformation measured value of the containment structure integrity test based on the BP neural network algorithm have the following advantages:
1. the BP neural network is used for establishing a mapping relation among the environmental temperature, the solar radiation intensity and the structural deformation measured value, and the deformation measured value is corrected during the structural integrity test of the prestressed concrete containment structure, so that the influence of the environmental temperature and the solar radiation on the structural deformation measured value is comprehensively considered, and the functional relation among the environmental temperature and the solar radiation is not required to be determined in advance, thereby improving the structural deformation measurement precision of the containment, eliminating the interference of human factors and having higher innovation and practical engineering application value.
2. The machine learning algorithm is applied to actual engineering, the multi-parameter input of the environmental temperature and the solar radiation intensity can be realized, the corresponding relation between the comprehensive environmental factors and the structural deformation can be automatically judged, the complicated calculation is given to a computer, and the calculation efficiency and the accuracy are scientifically and effectively improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a specific example of a method for correcting a deformation measurement value of a containment structure integrity test based on a BP neural network algorithm in an embodiment of the present invention;
FIG. 2 is a graph of the ambient temperature and first deformation measurements for a containment vessel station pre-test;
FIG. 3 is a graph of the measurement of the first solar radiation intensity for a certain containment point prediction test;
FIG. 4 is a graph of predicted and expected values for a test sample at a test point of a containment vessel;
FIG. 5 is a graph of the measured results of the second solar radiation intensity, ambient temperature during a point of care formal test of a containment;
FIG. 6 is a graph of the structural displacement variation time course predicted by the BP neural network during a certain measurement point formal test of a certain containment;
fig. 7 is a comparison of the deformation correction of a certain measuring point of a certain containment vessel before and after.
Detailed Description
The following description of the embodiments of the present invention will be made apparent and fully in view of the accompanying drawings, in which some, but not all embodiments of the invention are shown. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In describing the present invention, it should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The use of the terms "comprises" and/or "comprising," when used in this specification, are intended to specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. The terms "first," "second," and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
Furthermore, some of the figures in this specification are flowcharts for illustrating methods. It will be understood that each block of the flowchart illustrations, and combinations of blocks in the flowchart illustrations, can be implemented by computer program instructions. These computer program instructions may be loaded onto a computer or other programmable apparatus to produce a machine, such that the instructions which execute on the computer or other programmable apparatus create means for implementing the functions specified in the flowchart block or blocks. These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart block or blocks. The computer program instructions may also be loaded onto a computer or other programmable apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks.
Accordingly, blocks of the flowchart support combinations of means for performing the specified functions and combinations of steps for performing the specified functions. It will also be understood that each block of the flowchart, and combinations of blocks in the flowchart, can be implemented by special purpose hardware-based computer systems which perform the specified functions or steps, or combinations of special purpose hardware and computer instructions.
In addition, the technical features of the different embodiments of the present invention described below may be combined with each other as long as they do not collide with each other.
The embodiment adopts a BP neural network algorithm to correct the deformation measured value of the integrity test of the containment structure, and forms a method for correcting the deformation measured value of the integrity test of the containment structure based on the BP neural network algorithm, which comprises the following specific steps as shown in figure 1:
s1, sampling data (such as a first ambient temperature, a first solar radiation intensity, a first deformation measured value and the like) in a continuous time period similar to weather parameters in a formal test process, and performing first data processing to obtain sample data. The sample data includes first ambient temperature data, first solar radiation intensity, and first deformation measurement data caused by the first ambient temperature data and the first solar radiation intensity.
The data sampled during the pre-test is selected from a period of time similar to weather parameters (e.g., ambient temperature, humidity, wind direction, wind force, solar radiation intensity, etc.) during the formal test, for not less than 72 hours. As shown in fig. 2, the measurement of the first deformation measurement (structural horizontal radial displacement) is shown at the ambient temperature of the pilot test. As shown in fig. 3, the measurement of the first solar radiation intensity of the pre-test is shown.
Preferably, the first data processing includes a first outlier processing, a deformation hysteresis processing, and a first difference processing. Since there are suddenly increasing or decreasing outliers in the pre-test and formal test data, the first outlier processing includes interpolating the outliers with the data on the left and right sides if there are suddenly increasing or decreasing outliers in the data during the pre-test. Because the concrete has poor thermal conductivity, the structural displacement change has certain hysteresis compared with the environmental temperature and the solar radiation intensity change, and the environmental temperature and the displacement curve wave peak value and wave trough value can be approximately overlapped by adjusting the hysteresis time of the structural displacement change, the deformation hysteresis treatment comprises adjusting the time course curve of the first deformation measured value to eliminate the hysteresis of deformation. The first difference processing includes subtracting the first initial ambient temperature from the first ambient temperature at each time to obtain first ambient temperature data (ambient temperature difference) and subtracting the initial measured value from the first deformation measured value at each time to obtain first deformation measured data (structural displacement measured value difference).
S2, training the BP neural network model on the basis of the sample data to obtain an optimal BP neural network model. The BP neural network model comprises 2 nodes, wherein the input layer is used for inputting first environment temperature data and first solar radiation intensity, and the output layer is used for outputting first deformation prediction data, and the output layer is 1 node. The method specifically comprises the following steps:
s21, carrying out normalization processing on the sample data and dividing the sample data into training samples and test samples. For example, the training sample may sample the first 70% of the normalized sample data and the test sample may sample the last 30% of the normalized sample data.
S22, constructing a BP neural network model with an input layer, an implicit layer and an output layer which are sequentially connected, wherein the output of each layer serves as the input of the next layer. The input layer is provided with 2 nodes and is used for receiving the first environmental temperature data and the normalized value of the first solar radiation intensity in a one-to-one correspondence manner; the output layer has 1 node for outputting first deformation prediction data. The activation function from the input layer to the hidden layer is a tangent S-shaped function, the activation function from the hidden layer to the output layer is a linear function, and the training function adopts a Levenberg-Marquardt optimization algorithm. The number of nodes of the hidden layer is according to an empirical formula
Figure BDA0004009574040000081
Determining, wherein m is the number of nodes of an input layer, n is the number of nodes of an output layer, and a is [1,10 ]]An integer therebetween. Therefore, the number of the optimal hidden layer nodes is calculated to be between 2 and 11, and the optimal hidden layer nodes are determined through the established BP neural network.
S23, inputting a training sample into the BP neural network model, performing BP neural network training, verifying a training effect by using a test sample, and obtaining a trained optimal BP neural network model through error analysis and determination of first deformation prediction data and first deformation measurement data output by the model. For example, the training cycle number of the BP neural network model is 1000 at maximum, the learning rate is 0.01, and the training minimum error is 0.000001. In a specific example, 202 sets of data of a certain measuring point of a certain containment are selected, wherein 143 sets of training samples and 60 sets of test samples are used for obtaining first parallel prediction data (predicted value) output by a model after the test samples are input into an optimal BP neural network model, the comparison situation of the predicted value and an expected value (first deformation measurement data corresponding to the test samples) is shown in fig. 4, and errors between the predicted value and the expected value are found to meet the requirements.
The following table shows the comparison of five errors, namely the square sum of errors, the average absolute error, the mean square error, the root mean square error and the average percentage error, of the linear fitting, the single factor BP neural network of the ambient temperature and the multi-factor BP neural network of the ambient temperature and the solar radiation intensity.
Figure BDA0004009574040000091
From the above table, it can be seen that each error of the linear fitting is larger than the BP neural network considering only the ambient temperature, and the BP neural network considering only the ambient temperature is larger than the BP neural network considering the ambient temperature and the solar radiation intensity, which indicates that the error of the structural deformation correction method of the BP neural network considering the ambient temperature and the solar radiation intensity is much lower than that of the previous two correction methods, so that the correction method of the embodiment significantly improves the measurement accuracy of the deformation (structural displacement).
And S3, acquiring data (such as a second ambient temperature, a second solar radiation intensity and the like) during the formal test period, and performing second data processing to obtain actual measurement data. The measured data includes second ambient temperature data and second solar radiation intensity. As shown in fig. 5, the measurement of the ambient temperature, the second solar radiation intensity during the formal test is shown.
Preferably, the second data processing includes a second outlier processing and a second difference processing. The first outlier processing includes interpolating the outlier with data on both its left and right sides if there is an outlier that suddenly increases or decreases during the formal test. The second difference processing includes subtracting the second initial ambient temperature from the second ambient temperature at each time to obtain second ambient temperature data.
S4, inputting the actually measured data into the optimal BP neural network model for calculation after normalization processing, and obtaining second deformation prediction data caused by second environmental temperature data and second solar radiation intensity during a formal test. As shown in fig. 6, BP predicted displacement (obtained after the second deformation prediction data inverse normalization process) during the formal test is shown.
S5, acquiring comprehensive deformation measurement values (caused by the second ambient temperature, the second solar radiation intensity, the test pressure and the like) of the containment structure during the formal test.
S6, correcting the comprehensive deformation measured value by utilizing the second deformation prediction data, namely subtracting structural displacement change caused by the ambient temperature and the solar radiation intensity during the formal test on the basis of the comprehensive deformation measured value, specifically subtracting the second deformation prediction data from the comprehensive deformation measured value and performing inverse normalization processing to obtain a deformation measurement correction value of the containment structure under the effect of test pressure, wherein the effects before and after correction are shown in figure 7.
According to the method for correcting the deformation measurement value of the structural integrity test of the containment structure based on the BP neural network algorithm, the influence of the ambient temperature and solar radiation on the structural deformation measurement value is comprehensively considered, and the functional relation between the ambient temperature and the solar radiation is not required to be determined in advance, so that the deformation measurement precision of the containment structure is improved, the interference of human factors is eliminated, and the method has higher innovation and practical engineering application value.
It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations or modifications of the above teachings will be apparent to those of ordinary skill in the art. It is not necessary here nor is it exhaustive of all embodiments. While still being apparent from variations or modifications that may be made by those skilled in the art are within the scope of the invention.

Claims (10)

1. The containment structure integrity test deformation measurement value correction method based on the BP neural network algorithm is characterized by comprising the following steps of:
sampling data in a continuous time period similar to weather parameters in a formal test process and performing first data processing to obtain sample data; the sample data includes first ambient temperature data, first solar radiation intensity, and first deformation measurement data caused by the first ambient temperature data and the first solar radiation intensity;
training the BP neural network model on the basis of the sample data to obtain an optimal BP neural network model; the BP neural network model is input into first environment temperature data and first solar radiation intensity, and is output into first deformation prediction data;
collecting data in a formal test period and performing second data processing to obtain actual measurement data; the measured data includes second ambient temperature data and second solar radiation intensity;
inputting the normalized measured data into the optimal BP neural network model for calculation to obtain second deformation prediction data caused by second environmental temperature data and second solar radiation intensity during a formal test;
acquiring a comprehensive deformation measured value of the containment structure during a formal test;
and correcting the comprehensive deformation measured value by using the second deformation prediction data to obtain a deformation measurement correction value of the containment structure under the action of the test pressure.
2. The method of claim 1, wherein the time period is not less than 72 hours long.
3. The method according to claim 1 or 2, wherein the first data processing includes a first outlier processing, a deformation hysteresis processing, and a first difference processing;
the first abnormal value processing comprises the step of interpolating and replacing the abnormal value by the data on the left side and the right side if the abnormal value suddenly increases or decreases in the data in the pre-test process;
the deformation hysteresis process includes adjusting a time course curve of the first deformation measurement to eliminate hysteresis of the deformation;
the first difference processing includes subtracting the first initial ambient temperature from the first ambient temperature at each time to obtain first ambient temperature data and subtracting the initial measured value from the first deformation measured value at each time to obtain first deformation measured data.
4. A method according to any one of claims 1-3, wherein the step of training the BP neural network model on the basis of the sample data to obtain an optimal BP neural network model comprises:
normalizing the sample data and dividing the sample data into a training sample and a test sample;
constructing a BP neural network model with an input layer, an hidden layer and an output layer which are sequentially connected, wherein the output of each layer is used as the input of the next layer; the input layer is provided with 2 nodes and is used for receiving the first environmental temperature data and the normalized value of the first solar radiation intensity in a one-to-one correspondence manner; the output layer is provided with 1 node and is used for outputting first deformation prediction data;
and inputting a training sample into the BP neural network model, performing BP neural network training, verifying a training effect by using a test sample, and obtaining a trained optimal BP neural network model through error analysis and determination of first deformation prediction data and first deformation measurement data output by the model.
5. The method according to claim 4, wherein the activation function from the input layer to the hidden layer is a tangent S-type function, the activation function from the hidden layer to the output layer is a linear type function, and the training function adopts a Levenberg-Marquardt optimization algorithm; the number of nodes of the hidden layer is according to an empirical formula
Figure FDA0004009574030000021
Determining, wherein m is the number of nodes of an input layer, n is the number of nodes of an output layer, and a is[1,10]An integer therebetween.
6. The method of any of claims 1-5, wherein the second data processing comprises a second outlier processing and a second difference processing;
the first outlier processing comprises the step of interpolating and replacing the outlier with data on the left side and the right side of the outlier if the outlier suddenly increases or decreases in the data during the formal test;
the second difference processing includes subtracting the second initial ambient temperature from the second ambient temperature at each time to obtain second ambient temperature data.
7. The method of any one of claims 1-6, wherein the step of correcting the integrated deformation measurement using the second deformation prediction data to obtain a deformation measurement correction value for the containment structure comprises:
and after carrying out inverse normalization processing on the second deformation prediction data, removing the second deformation prediction data from the comprehensive deformation measurement value to obtain a deformation measurement correction value of the containment structure under the action of test pressure.
8. The utility model provides a containment structure integrity test deformation measured value correcting unit based on BP neural network algorithm which characterized in that includes:
the pre-test sampling unit is used for carrying out a pre-test of the containment structure integrity test in a time period similar to weather parameters during a formal test of the containment structure integrity test, sampling data in the pre-test process and carrying out first data processing to obtain sample data; the sample data includes first ambient temperature data, first solar radiation intensity, and first deformation measurement data caused by the first ambient temperature data and the first solar radiation intensity;
the BP neural network model training unit is used for training the BP neural network model on the basis of the sample data to obtain an optimal BP neural network model; the BP neural network model is input into first environment temperature data and first solar radiation intensity, and is output into first deformation prediction data;
the formal test sampling unit is used for collecting data in a formal test period and performing second data processing to obtain actual measurement data; the measured data includes second ambient temperature data and second solar radiation intensity;
the deformation prediction unit is used for inputting the actually measured data into the optimal BP neural network model for calculation after normalization processing to obtain second deformation prediction data caused by second environmental temperature data and second solar radiation intensity during a formal test;
the comprehensive deformation measurement value acquisition unit is used for acquiring the comprehensive deformation measurement value of the containment structure during the formal test;
and the deformation measurement value correction unit is used for correcting the comprehensive deformation measurement value by using the second deformation prediction data to obtain a deformation measurement correction value of the containment structure under the action of the test pressure.
9. A computer readable storage medium having instructions stored thereon, which when executed by a processor, implement the containment structure integrity test deformation measurement correction method based on the BP neural network algorithm of any one of claims 1-7.
10. The containment structure integrity test deformation measurement value correction device based on the BP neural network algorithm is characterized by comprising:
one or more processors; and
a storage device for storing one or more programs; the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the containment structure integrity test deformation measurement correction method based on the BP neural network algorithm of any one of claims 1-7.
CN202211646049.5A 2022-12-21 2022-12-21 Method for correcting deformation value of containment structure integrity test based on BP neural network Pending CN116090049A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211646049.5A CN116090049A (en) 2022-12-21 2022-12-21 Method for correcting deformation value of containment structure integrity test based on BP neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211646049.5A CN116090049A (en) 2022-12-21 2022-12-21 Method for correcting deformation value of containment structure integrity test based on BP neural network

Publications (1)

Publication Number Publication Date
CN116090049A true CN116090049A (en) 2023-05-09

Family

ID=86211275

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211646049.5A Pending CN116090049A (en) 2022-12-21 2022-12-21 Method for correcting deformation value of containment structure integrity test based on BP neural network

Country Status (1)

Country Link
CN (1) CN116090049A (en)

Similar Documents

Publication Publication Date Title
CN110443405A (en) A kind of built photovoltaic power station power generation amount forecasting system and method
CN107229824B (en) Photovoltaic power station power generation unit power curve modeling method and device
CN112100574B (en) AAKR model uncertainty calculation method and system based on resampling
CN113126019A (en) Intelligent ammeter error remote estimation method, system, terminal and storage medium
CN110390343B (en) Correction method and system for space meteorological data
CN117591964B (en) Electric power intelligent analysis method based on artificial intelligence
CN114897331A (en) Power transformer risk assessment method based on three-parameter interval gray number decision
CN117273195A (en) Steam heating system demand load prediction method
CN116070068B (en) Stay cable damage identification method, device and terminal based on first-order derivative wavelet transform of main beam deflection difference
CN111578984B (en) System for monitoring stress state of steel structure in full life cycle of station house in severe cold region
CN109861293B (en) Method for evaluating influence of photovoltaic uncertainty on small signal stability of power system
CN116090049A (en) Method for correcting deformation value of containment structure integrity test based on BP neural network
CN113052380A (en) Method, device, equipment and medium for detecting operation and maintenance cleaning effect of photovoltaic power station
CN114965794B (en) Performance evaluation method of natural gas analysis system
CN114062998B (en) Verification device running state monitoring method, electronic device and storage medium
CN111833011B (en) Wind resistance evaluation method and device applied to engineering construction
CN110555220A (en) Calibration method and system of photoelectric conversion model
CN114282761A (en) Hydroelectric generating set health evaluation method and system based on characteristic statistical analysis
CN118826643B (en) Intelligent detection method and device for photovoltaic equipment
CN118348341B (en) Data model for intelligent monitoring of operation state of substation lightning arrester
CN114004045B (en) Y-type pipe node stress concentration coefficient calculation method, device and storable medium
CN118071178B (en) Method and device for evaluating abnormal state of power station, electronic equipment and storage medium
CN105806943B (en) Composite material beam construction damage positioning method and system based on discrete wavelet
US12044601B2 (en) Displacement reconstruction method for lattice tower structure based on improved mode superposition
CN112098711B (en) Moon rover sailboard power attenuation determining method and device based on sparse telemetry

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