CN117373580A - Performance analysis method and system for realizing titanium alloy product based on time sequence network - Google Patents
Performance analysis method and system for realizing titanium alloy product based on time sequence network Download PDFInfo
- Publication number
- CN117373580A CN117373580A CN202311648729.5A CN202311648729A CN117373580A CN 117373580 A CN117373580 A CN 117373580A CN 202311648729 A CN202311648729 A CN 202311648729A CN 117373580 A CN117373580 A CN 117373580A
- Authority
- CN
- China
- Prior art keywords
- product
- time sequence
- titanium alloy
- data
- test data
- 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
Links
- 229910001069 Ti alloy Inorganic materials 0.000 title claims abstract description 190
- 238000004458 analytical method Methods 0.000 title claims abstract description 65
- 238000012360 testing method Methods 0.000 claims abstract description 150
- 230000003993 interaction Effects 0.000 claims abstract description 68
- 238000011156 evaluation Methods 0.000 claims abstract description 65
- 238000012512 characterization method Methods 0.000 claims abstract description 58
- 238000012545 processing Methods 0.000 claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 34
- 230000009467 reduction Effects 0.000 claims description 47
- 238000013528 artificial neural network Methods 0.000 claims description 43
- 238000012546 transfer Methods 0.000 claims description 32
- 239000011159 matrix material Substances 0.000 claims description 30
- 230000006870 function Effects 0.000 claims description 29
- 230000015654 memory Effects 0.000 claims description 20
- 125000004122 cyclic group Chemical group 0.000 claims description 17
- 230000001364 causal effect Effects 0.000 claims description 12
- 230000004913 activation Effects 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 9
- 230000000306 recurrent effect Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000001537 neural effect Effects 0.000 claims description 7
- 238000011176 pooling Methods 0.000 claims description 7
- 238000011282 treatment Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 5
- 230000002457 bidirectional effect Effects 0.000 claims description 4
- 230000008030 elimination Effects 0.000 claims description 2
- 238000003379 elimination reaction Methods 0.000 claims description 2
- 238000004891 communication Methods 0.000 description 8
- 238000013473 artificial intelligence Methods 0.000 description 5
- 238000007726 management method Methods 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000012216 screening Methods 0.000 description 3
- RTAQQCXQSZGOHL-UHFFFAOYSA-N Titanium Chemical compound [Ti] RTAQQCXQSZGOHL-UHFFFAOYSA-N 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 230000007797 corrosion Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012886 linear function Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 229910052719 titanium Inorganic materials 0.000 description 2
- 239000010936 titanium Substances 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000002441 X-ray diffraction Methods 0.000 description 1
- 229910045601 alloy Inorganic materials 0.000 description 1
- 239000000956 alloy Substances 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 238000012905 input function Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
- G06T7/41—Analysis of texture based on statistical description of texture
- G06T7/45—Analysis of texture based on statistical description of texture using co-occurrence matrix computation
-
- 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
-
- 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
- G06F2119/04—Ageing analysis or optimisation against ageing
-
- 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
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Databases & Information Systems (AREA)
- Chemical & Material Sciences (AREA)
- Crystallography & Structural Chemistry (AREA)
- Evolutionary Biology (AREA)
- Medical Informatics (AREA)
- Quality & Reliability (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of performance analysis of titanium alloy products, and provides a method and a system for realizing performance analysis of titanium alloy products based on a time sequence network, wherein the method comprises the following steps: determining performance evaluation indexes of titanium alloy products, and carrying out data classification processing on the product test data to obtain classification test data; inputting the classified test data as input data into a pre-trained time sequence network model, analyzing data time sequence characterization corresponding to the classified test data, determining time sequence interaction weights among the data time sequence characterization, and evaluating the performance of product indexes corresponding to the titanium alloy product by combining the classified test data, the time sequence interaction weights and the performance evaluation indexes; collecting a product appearance image of the titanium alloy product, identifying product appearance grains corresponding to the titanium alloy product, and calculating a product fatigue value of the titanium alloy product according to the product appearance grains; and generating a performance analysis report of the titanium alloy product. The invention aims to improve the accuracy of performance analysis of titanium alloy products.
Description
Technical Field
The invention relates to the technical field of performance analysis of titanium alloy products, in particular to a method and a system for realizing performance analysis of titanium alloy products based on a time sequence network.
Background
Titanium alloy is an alloy composed of titanium and other elements, and titanium has excellent characteristics of low density, high strength, corrosion resistance and the like, so that the titanium alloy is widely used in the fields of aerospace, ships, automobiles, medical appliances and the like, however, the titanium alloy is characterized by higher strength and rigidity and good corrosion resistance, so that performance analysis is required to be performed on titanium alloy products when the titanium alloy products are put into use.
The existing performance analysis method of the titanium alloy product mainly adopts an X-ray diffraction analysis method to emit X-rays onto the titanium alloy product, and the performance of the titanium alloy product is analyzed by recording and measuring the diffraction angle and diffraction intensity of rays, but partial performances of the titanium alloy product have interaction effects, so that the performance tested by the method cannot completely represent the performance of the titanium alloy product, thus the performance analysis of the titanium alloy product is inaccurate, and a method capable of improving the accuracy of the performance analysis of the titanium alloy product is needed.
Disclosure of Invention
The invention provides a method and a system for realizing performance analysis of a titanium alloy product based on a time sequence network, which mainly aim at improving the accuracy of the performance analysis of the titanium alloy product.
In order to achieve the above object, the present invention provides a method for analyzing performance of a titanium alloy product based on a time series network, comprising:
obtaining product test data of a titanium alloy product, determining performance evaluation indexes of the titanium alloy product, and carrying out data classification processing on the product test data according to the performance evaluation indexes to obtain classification test data;
inputting the classifying test data as input data into a pre-trained time sequence network model, analyzing data time sequence characterization corresponding to the classifying test data by using a time sequence convolution network in the time sequence network model, determining time sequence interaction weights among the data time sequence characterization by using a circulating neural network in the time sequence network model, and evaluating product index performance corresponding to the titanium alloy product by using a self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weights and the performance evaluation index;
collecting a product appearance image of the titanium alloy product, identifying product appearance grains corresponding to the titanium alloy product according to the product appearance image, and calculating a product fatigue value of the titanium alloy product according to the product appearance grains;
And generating a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
Optionally, the classifying the product test data according to the performance evaluation index to obtain classified test data includes:
performing dimension reduction processing on the product test data to obtain dimension reduction test data, and calculating a data center distance between the dimension reduction test data;
performing discrete data elimination processing on the dimension reduction test data according to the data center distance to obtain target test data;
extracting a data tag corresponding to the target test data, and calculating a confidence coefficient between the data tag and the performance evaluation index;
and carrying out data classification processing on the target test data according to the confidence coefficient to obtain classified test data.
Optionally, the calculating a confidence coefficient between the data tag and the performance evaluation index includes:
calculating a confidence coefficient between the data tag and the performance evaluation index by the following formula:
；
wherein B represents the confidence coefficient between the data label and the performance evaluation index, a and a+1 respectively represent the serial numbers corresponding to the data label and the performance evaluation index, Tag number representing data tag, +.>Representing an open square function>Representing the vector value corresponding to the a-th tag in the data tags,>the vector value corresponding to the (a+1) th index of the performance evaluation indexes is represented.
Optionally, the analyzing, by using a time sequence convolution network in the time sequence network model, a data time sequence representation corresponding to the categorizing test data includes:
identifying time sequence data corresponding to the classifying test data by using an input layer in the time sequence convolution network, and carrying out characterization extraction on the time sequence data by using the time sequence convolution layer in the time sequence convolution network to obtain a first time sequence characterization;
performing characterization extraction on the time sequence data by utilizing a flip convolution layer in the time sequence convolution network to obtain a second time sequence characterization;
performing dimension reduction processing on the first time sequence representation and the second time sequence representation by using a pooling layer in the time sequence convolution network to obtain a first dimension reduction time sequence representation and a second dimension reduction time sequence representation;
analyzing a characterization linear relationship between the first dimension reduction time sequence characterization and the second dimension reduction time sequence characterization by utilizing a linear analysis layer in the time sequence convolution network;
and outputting the data time sequence representation corresponding to the classifying test data from the first dimension reduction time sequence representation and the second dimension reduction time sequence representation by utilizing an output layer in the time sequence convolution network according to the representation linear relation.
Optionally, the determining the time sequence interaction weight between the data time sequence characterizations by using a cyclic neural network in the time sequence network model includes:
identifying a representation time step corresponding to the data time sequence representation by using a circulating attention layer in the circulating neural network, and analyzing a hidden state corresponding to the representation time step by using a bidirectional circulating layer in the circulating neural network;
extracting transfer information corresponding to the characterization time step by using a recurrent neural layer in the recurrent neural network according to the hidden state;
calculating an information interaction value between the transfer information by using an activation function in the cyclic neural network;
constructing an interaction value matrix corresponding to each piece of information in the transmitted information by utilizing a self-encoder in the cyclic neural network according to the information interaction value;
according to the interaction value matrix, distributing information interaction weights corresponding to each piece of information in the transmitted information by using a multi-layer perceptron in the cyclic neural network;
and outputting the information interaction weight by using an output function in the cyclic neural network to obtain the time sequence interaction weight between the data time sequence representations.
Optionally, the calculating the information interaction value between the transfer information by using an activation function in the recurrent neural network includes:
the specific calculation process of the activation function is as follows:
；
wherein,representing information interaction values between the transferred information, +.>The information entropy value corresponding to the d-th information in the transfer information, d represents the information serial number of the transfer information, t represents the information quantity corresponding to the transfer information, b represents the average information entropy corresponding to the transfer information,/and/or>Indicating the degree of embedding of the d-th information and the adjacent information in the transfer information.
Optionally, the evaluating the product index performance corresponding to the titanium alloy product by using the self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weight and the performance evaluation index includes:
extracting the characteristics of the classified test data by using a self-attention convolution layer in the self-attention network to obtain test data characteristics;
analyzing the dependency relationship between the test data features by using a causal expansion layer in the self-attention network according to the time sequence interaction weight;
and analyzing the product index performance of the titanium alloy product by utilizing a long-short-period memory network in the self-attention network according to the test data characteristics, the dependency relationship and the performance evaluation index.
Optionally, the identifying, according to the product appearance image, product appearance grains corresponding to the titanium alloy product includes:
denoising the product appearance image to obtain a denoising appearance image;
performing main body segmentation processing on the denoising appearance image to obtain a product main body image;
carrying out gray processing on the product main body image to obtain a gray product image;
constructing a product gray matrix corresponding to the gray product image, and calculating a matrix variance corresponding to the product gray matrix;
and extracting product appearance grains corresponding to the titanium alloy product from the grain region image according to the matrix variance.
Optionally, the calculating the product fatigue value of the titanium alloy product according to the product appearance texture includes:
inquiring an application environment corresponding to the titanium alloy product, and determining a product load and a load direction corresponding to the titanium alloy product according to the application environment;
according to the product load, calculating the load stress corresponding to the product appearance lines;
determining a stress direction corresponding to the load stress according to the load direction and the product appearance grain;
constructing a fatigue strength curve of the titanium alloy product according to the load stress and the stress direction;
And calculating the product fatigue value of the titanium alloy product according to the fatigue strength curve.
A performance analysis system for implementing a titanium alloy product based on a time series network, the system comprising:
the data classifying module is used for acquiring product test data of the titanium alloy product, determining performance evaluation indexes of the titanium alloy product, and carrying out data classifying treatment on the product test data according to the performance evaluation indexes to obtain classifying test data;
the performance analysis module is used for inputting the classifying test data into a pre-trained time sequence network model as input data, analyzing data time sequence characterization corresponding to the classifying test data by using a time sequence convolution network in the time sequence network model, determining time sequence interaction weights among the data time sequence characterization by using a circulating neural network in the time sequence network model, and evaluating the performance of the product index corresponding to the titanium alloy product by using a self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weights and the performance evaluation index;
the fatigue value calculation module is used for collecting product appearance images of the titanium alloy products, identifying product appearance lines corresponding to the titanium alloy products according to the product appearance images, and calculating product fatigue values of the titanium alloy products according to the product appearance lines;
And the report generation module is used for generating a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
According to the invention, specific data tested on each aspect of the titanium alloy product can be obtained by acquiring the product test data of the titanium alloy product, a basis is provided for performance analysis of the subsequent titanium alloy product, data screening and other treatments are conveniently carried out on the product test data by determining the performance evaluation index of the titanium alloy product, and the treatment efficiency of the subsequent data is improved. Therefore, the performance analysis method and system for realizing the titanium alloy product based on the time sequence network can improve the performance analysis accuracy of the titanium alloy product.
Drawings
FIG. 1 is a schematic flow chart of a method for implementing performance analysis of a titanium alloy product based on a time series network according to an embodiment of the present invention;
FIG. 2 is a functional block diagram of a performance analysis system for implementing a titanium alloy product based on a time series network according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an electronic device for implementing the performance analysis method for implementing a titanium alloy product based on a time sequence network according to an embodiment of the invention.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment of the application provides a performance analysis method for realizing a titanium alloy product based on a time sequence network. In the embodiment of the present application, the execution body of the performance analysis method for implementing the titanium alloy product based on the time sequence network includes, but is not limited to, at least one of a server, a terminal, and the like, which can be configured to execute the method provided in the embodiment of the present application. In other words, the performance analysis method for realizing the titanium alloy product based on the time sequence network can be executed by software or hardware installed on the terminal equipment or the service end equipment, wherein the software can be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.
Referring to fig. 1, a flow chart of a method for implementing performance analysis of a titanium alloy product based on a time sequence network according to an embodiment of the invention is shown. In this embodiment, the method for analyzing the performance of the titanium alloy product based on the time series network includes steps S1 to S4.
S1, obtaining product test data of a titanium alloy product, determining performance evaluation indexes of the titanium alloy product, and carrying out data classification processing on the product test data according to the performance evaluation indexes to obtain classification test data.
According to the invention, specific data of each aspect test of the titanium alloy product can be obtained by obtaining the product test data of the titanium alloy product, a basis is provided for the subsequent performance analysis of the titanium alloy product, the data screening and other treatments are conveniently carried out on the product test data by determining the performance evaluation index of the titanium alloy product, and the processing efficiency of the subsequent data is improved, wherein the product test data are multi-aspect test data of the titanium alloy product, such as product organization data or product physical property data, the performance evaluation index is a corresponding standard when the titanium alloy product performs the performance analysis, alternatively, the obtained product test data of the titanium alloy product can be obtained through a data collector, the data collector is compiled by a script language, and the performance evaluation index can be determined according to the functional range of the titanium alloy product in an application scene.
According to the performance evaluation index, the data classification processing is carried out on the product test data, so that data with obvious rules are obtained, irrelevant data can be removed, and the data quality of the product test data is improved, wherein the classification test data is obtained after the product test data is classified according to the performance evaluation index.
As an embodiment of the present invention, the classifying the product test data according to the performance evaluation index to obtain classified test data includes: performing dimension reduction processing on the product test data to obtain dimension reduction test data, calculating a data center distance between the dimension reduction test data, performing discrete data rejection processing on the dimension reduction test data according to the data center distance to obtain target test data, extracting a data tag corresponding to the target test data, calculating a confidence coefficient between the data tag and the performance evaluation index, and performing data classification processing on the target test data according to the confidence coefficient to obtain classified test data.
The dimension reduction test data are data obtained after the dimension reduction of the product test data from high dimension to low dimension, the distance between the data centers represents the distance between the dimension reduction test data, the target test data are data obtained after discrete data in the dimension reduction test data are removed, the data labels are data labels corresponding to the target test data, and the confidence coefficient represents the reliability degree between the data labels and the performance evaluation indexes.
Optionally, the dimension reduction processing of the product test data may be implemented by a principal component analysis method, calculating a data center distance between the dimension reduction test data may be implemented by an euclidean distance algorithm, performing discrete data rejection processing on the dimension reduction test data according to a value of the data center distance, extracting a data tag corresponding to the target test data may be implemented by a tag extraction tool, and performing data classification processing on the target test data according to a range of intervals in which the confidence coefficient is located.
Further, as an optional embodiment of the present invention, the calculating a confidence coefficient between the data tag and the performance evaluation index includes:
calculating a confidence coefficient between the data tag and the performance evaluation index by the following formula:
；
wherein B represents the confidence coefficient between the data label and the performance evaluation index, a and a+1 respectively represent the serial numbers corresponding to the data label and the performance evaluation index,tag number representing data tag, +.>Representing an open square function>Representing the vector value corresponding to the a-th tag in the data tags,>the vector value corresponding to the (a+1) th index of the performance evaluation indexes is represented.
S2, inputting the classifying test data into a pre-trained time sequence network model as input data, analyzing data time sequence characterization corresponding to the classifying test data by using a time sequence convolution network in the time sequence network model, determining time sequence interaction weights among the data time sequence characterization by using a circulating neural network in the time sequence network model, and evaluating product index performance corresponding to the titanium alloy product by using a self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weights and the performance evaluation index.
According to the invention, the time sequence convolution network in the time sequence network model is utilized to analyze the data time sequence representation corresponding to the classifying test data, so that the time sequence change characteristic of the classifying test data can be obtained, the subsequent time sequence interaction weight can be conveniently determined, and the interdependence relation among all performances can be found when the performance analysis is carried out on the titanium alloy product through the time sequence network model, so that the purpose of improving the performance analysis accuracy of the titanium alloy product is achieved, wherein the time sequence network model is a neural network model for processing the time sequence data, and the data time sequence representation is the time sequence change characteristic of the classifying test data.
As one embodiment of the present invention, the analyzing, by using a time sequence convolution network in the time sequence network model, the data time sequence characterization corresponding to the classifying test data includes: identifying the time sequence data corresponding to the classifying test data by using an input layer in the time sequence convolution network, carrying out representation extraction on the time sequence data by using a time sequence convolution layer in the time sequence convolution network to obtain a first time sequence representation, carrying out representation extraction on the time sequence data by using a flip convolution layer in the time sequence convolution network to obtain a second time sequence representation, carrying out dimension reduction processing on the first time sequence representation and the second time sequence representation by using a pooling layer in the time sequence convolution network to obtain a first dimension reduction time sequence representation and a second dimension reduction time sequence representation, analyzing a representation linear relation between the first dimension reduction time sequence representation and the second dimension reduction time sequence representation by using a linear analysis layer in the time sequence convolution network, and outputting the data representation corresponding to the classifying test data from the first dimension reduction time sequence representation and the second dimension reduction time sequence representation by using an output layer in the time sequence convolution network according to the representation linear relation.
The input layer is a first layer in the time sequence convolution network, the input layer is a neural network which converts received data into data which can be processed by the time sequence network model, the time sequence data are data with a relation with time in the classification test data, the time sequence convolution layer is a neural network which is used for extracting characterization in the time sequence convolution network, the first time sequence characterization is a representative characterization which is extracted from the time sequence data through the time sequence convolution layer, the flip convolution layer is a convolution network which is formed by the convolution cores in the time sequence convolution layer after being flipped or rotated by a certain angle, the pooling layer is a neural network which is used for reducing the dimension of the first time sequence characterization and the second time sequence characterization, and the linear analysis layer is a neural network which is used for analyzing the linear relation between the first dimension reduction time sequence characterization and the second dimension reduction time sequence characterization.
Optionally, the time series data corresponding to the classifying test data may be identified through an input neural unit in the input layer, the time series data may be subjected to characterization extraction through time series convolution check in the time series convolution layer, the time series data may be subjected to characterization extraction through flip convolution check in the flip convolution layer, the first time series characterization and the second time series characterization may be subjected to dimension reduction processing through a pooling function in the pooling layer, such as a maximum pooling function, a linear relationship, such as a linear function, between the first dimension reduction time series characterization and the second dimension reduction time series characterization may be analyzed through a linear function in the linear analysis layer, and the data time series characterization, such as an input function, corresponding to the classifying test data may be output from the first dimension reduction time series characterization and the second dimension reduction time series characterization through an output function in the output layer.
According to the invention, the time sequence interaction weight between the data time sequence characterization is determined by utilizing the cyclic neural network in the time sequence network model, so that the degree of interaction between the data time sequence characterization can be solved, and the accuracy of performance evaluation of the titanium alloy product is improved, wherein the time sequence interaction weight represents the importance of interaction between the data time sequence characterization.
As one embodiment of the present invention, the determining the time sequence interaction weight between the data time sequence characterizations by using the cyclic neural network in the time sequence network model includes: identifying the representation time steps corresponding to the data time sequence representation by using a circulation attention layer in the circulation neural network, analyzing the hidden states corresponding to the representation time steps by using a bidirectional circulation layer in the circulation neural network, extracting the transfer information corresponding to the representation time steps by using a recurrent neural layer in the circulation neural network according to the hidden states, calculating information interaction values among the transfer information by using an activation function in the circulation neural network, constructing an interaction value matrix corresponding to each information in the transfer information by using a self-encoder in the circulation neural network according to the information interaction values, distributing the information interaction weights corresponding to each information in the transfer information by using a multi-layer perceptron in the circulation neural network according to the interaction value matrix, and carrying out output processing on the information interaction weights by using an output function in the circulation neural network to obtain the time sequence interaction weights among the data time sequence representations.
The cyclic attention layer is a neural unit for identifying an input time point corresponding to the data time sequence representation, and is composed of a plurality of attention mechanisms, the representation time step is a time point corresponding to the data time sequence representation, the bidirectional cyclic layer is composed of a forward cyclic unit and a backward cyclic unit, the hidden state is an internal time state corresponding to the representation time step, the recurrent neural layer is a neural unit for extracting storage information in the representation time step, the transfer information is a related information stored in the representation time step and used for transfer, the information interaction value represents the probability of interaction between a certain information and all other information in the transfer information, the interaction value matrix is a square matrix composed of the information interaction values corresponding to each information in the transfer information, and the multi-layer perceptron is a neural network for distributing interaction weights corresponding to each information in the transfer information, wherein the information interaction weights represent the importance degree of each information in the transfer information on other information.
Optionally, as an optional embodiment of the present invention, the calculating the information interaction value between the transfer information by using an activation function in the recurrent neural network includes:
The specific calculation process of the activation function is as follows:
；
wherein,representing information interaction values between the transferred information, +.>The information entropy value corresponding to the d-th information in the transfer information, d represents the information serial number of the transfer information, t represents the information quantity corresponding to the transfer information, b represents the average information entropy corresponding to the transfer information,/and/or>Indicating the degree of embedding of the d-th information and the adjacent information in the transfer information.
According to the invention, the self-attention network in the time sequence network model is utilized to evaluate the product index performance corresponding to the titanium alloy product by combining the classification test data, the time sequence interaction weight and the performance evaluation index, so that the predicted performance condition of the corresponding index of the titanium alloy product can be obtained, and a basis is provided for determining the comprehensive performance of the subsequent product, wherein the product index performance is the predicted performance corresponding to the performance evaluation index of the titanium alloy product.
As one embodiment of the present invention, the evaluating the product index performance corresponding to the titanium alloy product by using the self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weight and the performance evaluation index includes: and extracting the characteristics of the classified test data by utilizing a self-attention convolution layer in the self-attention network to obtain test data characteristics, analyzing the dependency relationship among the test data characteristics by utilizing a causal expansion layer in the self-attention network according to the time sequence interaction weight, and analyzing the product index performance of the titanium alloy product by utilizing a long-period memory network in the self-attention network according to the test data characteristics, the dependency relationship and the performance evaluation index.
The self-attention convolution layer is a neural network for extracting data features of the classified test data and consists of a plurality of self-attention mechanisms, the test data features are representative features in the classified test data, the causal expansion layer is a neural network for analyzing the mutual dependence among the test data features, if the feature a changes, the feature b also changes, the causal expansion layer consists of a causal convolution layer and an expansion convolution layer, the dependence represents the causal relationship among the test data features, and the long-term memory network consists of an input gate, an output gate and a forgetting gate and is used for classifying and linearly processing the test data features according to the dependence so as to obtain the features of corresponding index categories.
Optionally, the feature weight of each data in the classified test data can be calculated through self-attention in a self-attention convolution layer, then the test data feature is extracted according to the feature weight, the causal relationship between the test data features can be analyzed by using a causal convolution layer in the causal expansion layer according to the time sequence interaction weight, the causal relationship is expanded and amplified by using an expansion convolution layer in the causal expansion layer, so that the dependency relationship between the test data features is analyzed, the index performance of the titanium alloy product can be analyzed by using a long-term and short-term memory network in the self-attention network according to the test data feature and the performance evaluation index, and the index performance is adjusted by combining the dependency relationship, so that the product index performance is obtained.
S3, acquiring a product appearance image of the titanium alloy product, identifying product appearance grains corresponding to the titanium alloy product according to the product appearance image, and calculating a product fatigue value of the titanium alloy product according to the product appearance grains.
According to the product appearance image, the product appearance texture corresponding to the titanium alloy product is identified, so that the product appearance texture condition of the titanium alloy product can be known, the appearance performance of the titanium alloy product is known, and a basis is provided for calculation of a subsequent product fatigue value, wherein the product appearance image is an image corresponding to the titanium alloy product, the product appearance texture is the appearance texture of the titanium alloy product, and optionally, acquisition of the product appearance image of the titanium alloy product can be realized through an image acquisition device.
As one embodiment of the present invention, the identifying product appearance lines corresponding to the titanium alloy product according to the product appearance image includes: denoising the product appearance image to obtain a denoised appearance image, carrying out main body segmentation processing on the denoised appearance image to obtain a product main body image, carrying out gray processing on the product main body image to obtain a gray product image, constructing a product gray matrix corresponding to the gray product image, calculating a matrix variance corresponding to the product gray matrix, and extracting product appearance grains corresponding to the titanium alloy product from the grain region image according to the matrix variance.
The product main image is a product image obtained by dividing and removing images irrelevant to products in the product appearance image, the gray product image is an image represented by the product main image through a single tone, the product gray matrix is a square matrix constructed by pixel gray values in the gray product image, and the matrix variance represents the discrete degree corresponding to the product gray matrix.
Optionally, denoising the product appearance image may be achieved by a low-pass filter, main body segmentation of the denoised appearance image may be achieved by a threshold segmentation method, an average value of three channels of each pixel in the product main body image may be calculated, replacement processing is performed on the pixels in the product main body image according to the average value, gray scale processing of the product main body image is completed, construction of a product gray scale matrix corresponding to the gray scale product image may be achieved by a matrix function, such as a zero matrix function, calculation of a matrix variance corresponding to the product gray scale matrix may be achieved by a variance calculator, and product appearance textures corresponding to the titanium alloy product may be extracted from the texture region image according to a value of the matrix variance.
According to the product appearance lines, the method calculates the product fatigue value of the titanium alloy product, and the structural durability of the titanium alloy product can be known through the fatigue value, so that the service life of the titanium alloy product can be predicted conveniently, wherein the product fatigue value represents the durability of the titanium alloy product in the use process.
As one embodiment of the present invention, the calculating the product fatigue value of the titanium alloy product according to the product appearance texture includes: inquiring an application environment corresponding to the titanium alloy product, determining a product load and a load direction corresponding to the titanium alloy product according to the application environment, calculating a load stress corresponding to the product appearance grain according to the product load, determining a stress direction corresponding to the load stress according to the load direction and the product appearance grain, constructing a fatigue strength curve of the titanium alloy product according to the load stress and the stress direction, and calculating a product fatigue value of the titanium alloy product according to the fatigue strength curve.
The application environment is an application scene of the titanium alloy product, the load of the product and the load direction are the load born by the titanium alloy product and the force direction generated by the corresponding load, the load stress is a specific bearing force corresponding to the product appearance grain, and the fatigue strength curve is a fatigue life description curve of the titanium alloy product under the load stress of different values.
Optionally, the service condition of the titanium alloy product may be determined according to the application environment, the product load and the load direction corresponding to the titanium alloy product may be determined according to the service condition, the ratio between the area of the product appearance grain and the area of the product load may be calculated, the product load may be multiplied by the ratio to obtain the load stress corresponding to the product appearance grain, the grain direction corresponding to the product appearance grain may be determined according to the load direction and the product appearance grain, and then the grain direction may be combined according to the load direction, so as to determine the stress direction corresponding to the load stress, the fatigue strength curve of the titanium alloy product may be constructed by combining a drawing tool and a finite element analysis method, and the product fatigue value of the titanium alloy product may be obtained by calculating the average value of the peak value and the valley value corresponding to the fatigue strength curve.
S4, generating a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
According to the fatigue value and the index performance of the product, the performance analysis report of the titanium alloy product is generated, so that the performance analysis result of the titanium alloy product is displayed, and the performance analysis data of the titanium alloy product can be intuitively known, wherein the performance analysis report is a visual result of the performance analysis of the titanium alloy product, optionally, the performance analysis report of the titanium alloy product can be realized through a report generator, and the report generator is compiled by Java language.
According to the invention, specific data tested on each aspect of the titanium alloy product can be obtained by acquiring the product test data of the titanium alloy product, a basis is provided for performance analysis of the subsequent titanium alloy product, data screening and other treatments are conveniently carried out on the product test data by determining the performance evaluation index of the titanium alloy product, and the treatment efficiency of the subsequent data is improved. Therefore, the performance analysis method for the titanium alloy product based on the time sequence network can improve the performance analysis accuracy of the titanium alloy product.
FIG. 2 is a functional block diagram of a performance analysis system for implementing a titanium alloy product based on a time series network according to an embodiment of the present invention.
The performance analysis system 100 for realizing the titanium alloy product based on the time sequence network can be installed in electronic equipment. Depending on the functions implemented, the performance analysis system 100 for implementing titanium alloy products based on a time series network may include a data categorization module 101, a performance analysis module 102, a fatigue value calculation module 103, and a report generation module 104. The module of the invention, which may also be referred to as a unit, refers to a series of computer program segments, which are stored in the memory of the electronic device, capable of being executed by the processor of the electronic device and of performing a fixed function.
In the present embodiment, the functions concerning the respective modules/units are as follows:
the data classifying module 101 is configured to obtain product test data of a titanium alloy product, determine a performance evaluation index of the titanium alloy product, and perform data classifying processing on the product test data according to the performance evaluation index to obtain classified test data;
the performance analysis module 102 is configured to input the categorizing test data as input data into a pre-trained time-series network model, analyze data time-series characterization corresponding to the categorizing test data by using a time-series convolution network in the time-series network model, determine time-series interaction weights between the data time-series characterization by using a cyclic neural network in the time-series network model, and evaluate product index performance corresponding to the titanium alloy product by using a self-attention network in the time-series network model in combination with the categorizing test data, the time-series interaction weights and the performance evaluation index;
The fatigue value calculating module 103 is configured to collect a product appearance image of the titanium alloy product, identify product appearance grains corresponding to the titanium alloy product according to the product appearance image, and calculate a product fatigue value of the titanium alloy product according to the product appearance grains;
the report generating module 104 is configured to generate a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
In detail, each module in the performance analysis system 100 for implementing a titanium alloy product based on a time sequence network in the embodiment of the present application adopts the same technical means as the performance analysis method for implementing a titanium alloy product based on a time sequence network in the above-mentioned fig. 1, and can produce the same technical effects, which are not described herein again.
Fig. 3 is a schematic structural diagram of an electronic device 1 according to an embodiment of the present invention for implementing a performance analysis method for a titanium alloy product based on a time series network.
The electronic device 1 may comprise a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may further comprise a computer program stored in the memory 11 and executable on the processor 10, such as a performance analysis method program for implementing a titanium alloy product based on a time series network.
The processor 10 may be formed by an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be formed by a plurality of integrated circuits packaged with the same function or different functions, including one or more central processing units (Central Processing Unit, CPU), a microprocessor, a digital processing chip, a graphics processor, a combination of various control chips, and so on. The processor 10 is a Control Unit (Control Unit) of the electronic device 1, connects respective parts of the entire electronic device using various interfaces and lines, executes or executes programs or modules stored in the memory 11 (for example, executes a performance analysis method program for realizing a titanium alloy product based on a time series network, etc.), and invokes data stored in the memory 11 to perform various functions of the electronic device and process the data.
The memory 11 includes at least one type of readable storage medium including flash memory, a removable hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. The memory 11 may in some embodiments be an internal storage unit of the electronic device, such as a mobile hard disk of the electronic device. The memory 11 may in other embodiments also be an external storage device of the electronic device, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the electronic device. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device. The memory 11 may be used not only for storing application software installed in an electronic device and various types of data, for example, a code for implementing a performance analysis method program of a titanium alloy product based on a time series network, etc., but also for temporarily storing data that has been output or is to be output.
The communication bus 12 may be a peripheral component interconnect standard (Peripheral Component Interconnect, PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, EISA) bus, among others. The bus may be classified as an address bus, a data bus, a control bus, etc. The bus is arranged to enable a connection communication between the memory 11 and at least one processor 10 etc.
The communication interface 13 is used for communication between the electronic device 1 and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (e.g., WI-FI interface, bluetooth interface, etc.), typically used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a Display (Display), an input unit such as a Keyboard (Keyboard), or alternatively a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch, or the like. The display may also be referred to as a display screen or display unit, as appropriate, for displaying information processed in the electronic device and for displaying a visual user interface.
Fig. 3 shows only an electronic device with components, it being understood by a person skilled in the art that the structure shown in fig. 3 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than shown, or may combine certain components, or may be arranged in different components.
For example, although not shown, the electronic device 1 may further include a power source (such as a battery) for supplying power to the respective components, and preferably, the power source may be logically connected to the at least one processor 10 through a power management system, so as to perform functions of charge management, discharge management, and power consumption management through the power management system. The power supply may also include one or more of any of a direct current or alternating current power supply, a recharging system, a power failure detection circuit, a power converter or inverter, a power status indicator, and the like. The electronic device 1 may further include various sensors, bluetooth modules, wi-Fi modules, etc., which will not be described herein.
It should be understood that the embodiments described are for illustrative purposes only and are not limited to this configuration in the scope of the patent application.
A performance analysis method program for implementing a titanium alloy product based on a time series network stored in the memory 11 of the electronic device 1 is a combination of a plurality of instructions, which when executed in the processor 10, can implement:
Obtaining product test data of a titanium alloy product, determining performance evaluation indexes of the titanium alloy product, and carrying out data classification processing on the product test data according to the performance evaluation indexes to obtain classification test data;
inputting the classifying test data as input data into a pre-trained time sequence network model, analyzing data time sequence characterization corresponding to the classifying test data by using a time sequence convolution network in the time sequence network model, determining time sequence interaction weights among the data time sequence characterization by using a circulating neural network in the time sequence network model, and evaluating product index performance corresponding to the titanium alloy product by using a self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weights and the performance evaluation index;
collecting a product appearance image of the titanium alloy product, identifying product appearance grains corresponding to the titanium alloy product according to the product appearance image, and calculating a product fatigue value of the titanium alloy product according to the product appearance grains;
and generating a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
In particular, the specific implementation method of the above instructions by the processor 10 may refer to the description of the relevant steps in the corresponding embodiment of fig. 1, which is not repeated herein.
Further, the modules/units integrated in the electronic device 1 may be stored in a computer readable storage medium if implemented in the form of software functional units and sold or used as separate products. The computer readable storage medium may be volatile or nonvolatile. For example, the computer readable medium may include: any entity or system capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM).
The present invention also provides a computer readable storage medium storing a computer program which, when executed by a processor of an electronic device, can implement:
obtaining product test data of a titanium alloy product, determining performance evaluation indexes of the titanium alloy product, and carrying out data classification processing on the product test data according to the performance evaluation indexes to obtain classification test data;
Inputting the classifying test data as input data into a pre-trained time sequence network model, analyzing data time sequence characterization corresponding to the classifying test data by using a time sequence convolution network in the time sequence network model, determining time sequence interaction weights among the data time sequence characterization by using a circulating neural network in the time sequence network model, and evaluating product index performance corresponding to the titanium alloy product by using a self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weights and the performance evaluation index;
collecting a product appearance image of the titanium alloy product, identifying product appearance grains corresponding to the titanium alloy product according to the product appearance image, and calculating a product fatigue value of the titanium alloy product according to the product appearance grains;
and generating a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
In several embodiments provided by the present invention, it should be understood that the provided apparatus, system, and method may be implemented in other ways. For example, the system embodiments described above are merely illustrative, e.g., the division of the modules is merely a logical function division, and other manners of division may be implemented in practice.
In addition, each functional module in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units can be realized in a form of hardware or a form of hardware and a form of software functional modules.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof.
The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Among these, artificial intelligence (Artificial Intelligence, AI) is a theory, method, technique, and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend, and extend artificial intelligence, sense the environment, acquire knowledge, and use knowledge to obtain optimal results.
Furthermore, it is evident that the word "comprising" does not exclude other elements or steps, and that the singular does not exclude a plurality. Multiple units or systems as set forth in the system claims may also be implemented by means of one unit or system in software or hardware. The terms first, second, etc. are used to denote a name, but not any particular order.
Finally, it should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.
Claims (10)
1. A method for implementing performance analysis of a titanium alloy product based on a time series network, the method comprising:
obtaining product test data of a titanium alloy product, determining performance evaluation indexes of the titanium alloy product, and carrying out data classification processing on the product test data according to the performance evaluation indexes to obtain classification test data;
inputting the classifying test data as input data into a pre-trained time sequence network model, analyzing data time sequence characterization corresponding to the classifying test data by using a time sequence convolution network in the time sequence network model, determining time sequence interaction weights among the data time sequence characterization by using a circulating neural network in the time sequence network model, and evaluating product index performance corresponding to the titanium alloy product by using a self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weights and the performance evaluation index;
Collecting a product appearance image of the titanium alloy product, identifying product appearance grains corresponding to the titanium alloy product according to the product appearance image, and calculating a product fatigue value of the titanium alloy product according to the product appearance grains;
and generating a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
2. The method for analyzing the performance of a titanium alloy product based on a time sequence network according to claim 1, wherein the step of classifying the product test data according to the performance evaluation index to obtain classified test data comprises the steps of:
performing dimension reduction processing on the product test data to obtain dimension reduction test data, and calculating a data center distance between the dimension reduction test data;
performing discrete data elimination processing on the dimension reduction test data according to the data center distance to obtain target test data;
extracting a data tag corresponding to the target test data, and calculating a confidence coefficient between the data tag and the performance evaluation index;
and carrying out data classification processing on the target test data according to the confidence coefficient to obtain classified test data.
3. The method for analyzing the performance of a titanium alloy product based on a time series network according to claim 2, wherein the calculating the confidence coefficient between the data tag and the performance evaluation index comprises:
calculating a confidence coefficient between the data tag and the performance evaluation index by the following formula:
；
wherein B represents the confidence coefficient between the data label and the performance evaluation index, a and a+1 respectively represent the serial numbers corresponding to the data label and the performance evaluation index,tag number representing data tag, +.>Representing an open square function>Representing the vector value corresponding to the a-th tag in the data tags,>the vector value corresponding to the (a+1) th index of the performance evaluation indexes is represented.
4. The method for analyzing performance of a titanium alloy product based on a time series network according to claim 1, wherein analyzing the data time series characterization corresponding to the classifying test data by using a time series convolution network in the time series network model comprises:
identifying time sequence data corresponding to the classifying test data by using an input layer in the time sequence convolution network, and carrying out characterization extraction on the time sequence data by using the time sequence convolution layer in the time sequence convolution network to obtain a first time sequence characterization;
Performing characterization extraction on the time sequence data by utilizing a flip convolution layer in the time sequence convolution network to obtain a second time sequence characterization;
performing dimension reduction processing on the first time sequence representation and the second time sequence representation by using a pooling layer in the time sequence convolution network to obtain a first dimension reduction time sequence representation and a second dimension reduction time sequence representation;
analyzing a characterization linear relationship between the first dimension reduction time sequence characterization and the second dimension reduction time sequence characterization by utilizing a linear analysis layer in the time sequence convolution network;
and outputting the data time sequence representation corresponding to the classifying test data from the first dimension reduction time sequence representation and the second dimension reduction time sequence representation by utilizing an output layer in the time sequence convolution network according to the representation linear relation.
5. The method for implementing performance analysis of titanium alloy products based on a time series network according to claim 1, wherein determining time series interaction weights between the data time series characterizations using a recurrent neural network in the time series network model comprises:
identifying a representation time step corresponding to the data time sequence representation by using a circulating attention layer in the circulating neural network, and analyzing a hidden state corresponding to the representation time step by using a bidirectional circulating layer in the circulating neural network;
Extracting transfer information corresponding to the characterization time step by using a recurrent neural layer in the recurrent neural network according to the hidden state;
calculating an information interaction value between the transfer information by using an activation function in the cyclic neural network;
constructing an interaction value matrix corresponding to each piece of information in the transmitted information by utilizing a self-encoder in the cyclic neural network according to the information interaction value;
according to the interaction value matrix, distributing information interaction weights corresponding to each piece of information in the transmitted information by using a multi-layer perceptron in the cyclic neural network;
and outputting the information interaction weight by using an output function in the cyclic neural network to obtain the time sequence interaction weight between the data time sequence representations.
6. The method for analyzing the performance of a titanium alloy product based on a time series network according to claim 5, wherein said calculating the information interaction value between the transfer information by using the activation function in the cyclic neural network comprises:
the specific calculation process of the activation function is as follows:
；
wherein,representing information interaction values between the transferred information, +.>The information entropy value corresponding to the d-th information in the transfer information, d represents the information serial number of the transfer information, t represents the information quantity corresponding to the transfer information, b represents the average information entropy corresponding to the transfer information,/and/or >Indicating the degree of embedding of the d-th information and the adjacent information in the transfer information.
7. The method for analyzing performance of a titanium alloy product based on a time series network according to claim 1, wherein the step of evaluating the performance of the product index corresponding to the titanium alloy product by using the self-attention network in the time series network model by combining the classifying test data, the time series interaction weight and the performance evaluation index comprises the steps of:
extracting the characteristics of the classified test data by using a self-attention convolution layer in the self-attention network to obtain test data characteristics;
analyzing the dependency relationship between the test data features by using a causal expansion layer in the self-attention network according to the time sequence interaction weight;
and analyzing the product index performance of the titanium alloy product by utilizing a long-short-period memory network in the self-attention network according to the test data characteristics, the dependency relationship and the performance evaluation index.
8. The method for analyzing the performance of a titanium alloy product based on a time sequence network according to claim 1, wherein the identifying product appearance lines corresponding to the titanium alloy product according to the product appearance image comprises the following steps:
Denoising the product appearance image to obtain a denoising appearance image;
performing main body segmentation processing on the denoising appearance image to obtain a product main body image;
carrying out gray processing on the product main body image to obtain a gray product image;
constructing a product gray matrix corresponding to the gray product image, and calculating a matrix variance corresponding to the product gray matrix;
and extracting product appearance grains corresponding to the titanium alloy product from the grain region image according to the matrix variance.
9. The method for analyzing the performance of a titanium alloy product based on a time sequence network according to claim 1, wherein the calculating the product fatigue value of the titanium alloy product according to the product appearance texture comprises:
inquiring an application environment corresponding to the titanium alloy product, and determining a product load and a load direction corresponding to the titanium alloy product according to the application environment;
according to the product load, calculating the load stress corresponding to the product appearance lines;
determining a stress direction corresponding to the load stress according to the load direction and the product appearance grain;
constructing a fatigue strength curve of the titanium alloy product according to the load stress and the stress direction;
And calculating the product fatigue value of the titanium alloy product according to the fatigue strength curve.
10. A performance analysis system for implementing a titanium alloy product based on a time series network, the system comprising:
the data classifying module is used for acquiring product test data of the titanium alloy product, determining performance evaluation indexes of the titanium alloy product, and carrying out data classifying treatment on the product test data according to the performance evaluation indexes to obtain classifying test data;
the performance analysis module is used for inputting the classifying test data into a pre-trained time sequence network model as input data, analyzing data time sequence characterization corresponding to the classifying test data by using a time sequence convolution network in the time sequence network model, determining time sequence interaction weights among the data time sequence characterization by using a circulating neural network in the time sequence network model, and evaluating the performance of the product index corresponding to the titanium alloy product by using a self-attention network in the time sequence network model in combination with the classifying test data, the time sequence interaction weights and the performance evaluation index;
the fatigue value calculation module is used for collecting product appearance images of the titanium alloy products, identifying product appearance lines corresponding to the titanium alloy products according to the product appearance images, and calculating product fatigue values of the titanium alloy products according to the product appearance lines;
And the report generation module is used for generating a performance analysis report of the titanium alloy product according to the product fatigue value and the product index performance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311648729.5A CN117373580B (en) | 2023-12-05 | 2023-12-05 | Performance analysis method and system for realizing titanium alloy product based on time sequence network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311648729.5A CN117373580B (en) | 2023-12-05 | 2023-12-05 | Performance analysis method and system for realizing titanium alloy product based on time sequence network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117373580A true CN117373580A (en) | 2024-01-09 |
| CN117373580B CN117373580B (en) | 2024-03-08 |
Family
ID=89389576
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311648729.5A Active CN117373580B (en) | 2023-12-05 | 2023-12-05 | Performance analysis method and system for realizing titanium alloy product based on time sequence network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117373580B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117969681A (en) * | 2024-03-28 | 2024-05-03 | 宝鸡富士特钛业（集团）有限公司 | Analysis method and system for internal damage and defects of titanium alloy pipe |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120084019A1 (en) * | 2009-10-05 | 2012-04-05 | Khonsari Michael M | Fracture Fatigue Entropy Determination |
| CN105320804A (en) * | 2014-08-01 | 2016-02-10 | 通用汽车环球科技运作有限责任公司 | Material property predictor for cast aluminum alloys |
| CN112289391A (en) * | 2020-10-28 | 2021-01-29 | 广西贺州市桂东电子科技有限责任公司 | Anode aluminum foil performance prediction system based on machine learning |
| MX2021003805A (en) * | 2018-10-05 | 2021-05-27 | Jfe Steel Corp | Automotive body part sensitivity analysis method and device, and automotive body part material property determination method. |
| KR20210128713A (en) * | 2020-04-17 | 2021-10-27 | 충북대학교 산학협력단 | Analysis and prediction method of manufacturing process quality using statistical analysis and deep learning, and recording medium thereof |
| CN113887328A (en) * | 2021-09-10 | 2022-01-04 | 天津理工大学 | Method for extracting space-time characteristics of photonic crystal space transmission spectrum in parallel by ECA-CNN fusion dual-channel RNN |
| CN114254533A (en) * | 2021-12-02 | 2022-03-29 | 中国兵器科学研究院宁波分院 | Method for examining influence and prediction of fatigue vibration on fixed angle of product group component |
| JP2022080367A (en) * | 2020-11-18 | 2022-05-30 | 日本電気株式会社 | Model evaluation device, model evaluation method, and program |
| CN115458079A (en) * | 2022-08-01 | 2022-12-09 | 西安航空制动科技有限公司 | Fatigue crack propagation life propagation finite element analysis method |
| CN115901502A (en) * | 2022-11-01 | 2023-04-04 | 中国航发沈阳黎明航空发动机有限责任公司 | Method for predicting relation between welding repair defects and fatigue life of TC17 titanium alloy blade |
| WO2023095008A1 (en) * | 2021-11-23 | 2023-06-01 | Bridgestone Europe Nv/Sa | Predictive method based upon machine learning for the development of composites for tyre tread compounds |
| CN116304587A (en) * | 2022-09-07 | 2023-06-23 | 南京工业大学 | Rolling bearing degradation trend prediction method based on CAE and AGRU |
| WO2023151233A1 (en) * | 2022-02-09 | 2023-08-17 | 南京航空航天大学 | Virtual strain energy-based metal material multi-axis fatigue life prediction method and system |
| CN116721724A (en) * | 2023-06-20 | 2023-09-08 | 小米汽车科技有限公司 | Alloy performance prediction method and device, storage medium, electronic equipment and chip |
| CN116913434A (en) * | 2023-07-27 | 2023-10-20 | 宝鸡拓普达钛业有限公司 | Method and device for improving mechanical properties of titanium alloy based on laser shock peening |
-
2023
- 2023-12-05 CN CN202311648729.5A patent/CN117373580B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120084019A1 (en) * | 2009-10-05 | 2012-04-05 | Khonsari Michael M | Fracture Fatigue Entropy Determination |
| CN105320804A (en) * | 2014-08-01 | 2016-02-10 | 通用汽车环球科技运作有限责任公司 | Material property predictor for cast aluminum alloys |
| MX2021003805A (en) * | 2018-10-05 | 2021-05-27 | Jfe Steel Corp | Automotive body part sensitivity analysis method and device, and automotive body part material property determination method. |
| KR20210128713A (en) * | 2020-04-17 | 2021-10-27 | 충북대학교 산학협력단 | Analysis and prediction method of manufacturing process quality using statistical analysis and deep learning, and recording medium thereof |
| CN112289391A (en) * | 2020-10-28 | 2021-01-29 | 广西贺州市桂东电子科技有限责任公司 | Anode aluminum foil performance prediction system based on machine learning |
| JP2022080367A (en) * | 2020-11-18 | 2022-05-30 | 日本電気株式会社 | Model evaluation device, model evaluation method, and program |
| CN113887328A (en) * | 2021-09-10 | 2022-01-04 | 天津理工大学 | Method for extracting space-time characteristics of photonic crystal space transmission spectrum in parallel by ECA-CNN fusion dual-channel RNN |
| WO2023095008A1 (en) * | 2021-11-23 | 2023-06-01 | Bridgestone Europe Nv/Sa | Predictive method based upon machine learning for the development of composites for tyre tread compounds |
| CN114254533A (en) * | 2021-12-02 | 2022-03-29 | 中国兵器科学研究院宁波分院 | Method for examining influence and prediction of fatigue vibration on fixed angle of product group component |
| WO2023151233A1 (en) * | 2022-02-09 | 2023-08-17 | 南京航空航天大学 | Virtual strain energy-based metal material multi-axis fatigue life prediction method and system |
| CN115458079A (en) * | 2022-08-01 | 2022-12-09 | 西安航空制动科技有限公司 | Fatigue crack propagation life propagation finite element analysis method |
| CN116304587A (en) * | 2022-09-07 | 2023-06-23 | 南京工业大学 | Rolling bearing degradation trend prediction method based on CAE and AGRU |
| CN115901502A (en) * | 2022-11-01 | 2023-04-04 | 中国航发沈阳黎明航空发动机有限责任公司 | Method for predicting relation between welding repair defects and fatigue life of TC17 titanium alloy blade |
| CN116721724A (en) * | 2023-06-20 | 2023-09-08 | 小米汽车科技有限公司 | Alloy performance prediction method and device, storage medium, electronic equipment and chip |
| CN116913434A (en) * | 2023-07-27 | 2023-10-20 | 宝鸡拓普达钛业有限公司 | Method and device for improving mechanical properties of titanium alloy based on laser shock peening |
Non-Patent Citations (4)
| Title |
|---|
| MASOUD SARRAF ET AL: "A state-of-the-art review of the fabrication and characteristics of titanium and its alloys for biomedical applications", 《BIO-DESIGN AND MANUFACTURING》, vol. 5, 26 October 2021 (2021-10-26), pages 371 * |
| SUMIT BAHL ET AL: "Comprehensive review on alloy design, processing, and performance of β Titanium alloys as biomedical materials", 《 INTERNATIONAL MATERIALS REVIEWS》, vol. 66, no. 2, 5 March 2020 (2020-03-05), pages 114 - 139 * |
| 尤琦;赵宇;马小兵;: "产品性能可靠性评估的时序分析方法", 北京航空航天大学学报, no. 05 * |
| 柳青秀;马红占;褚学宁;马斌彬;王峥;: "基于长短时记忆―自编码神经网络的风电机组性能评估及异常检测", 计算机集成制造系统, no. 12 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117969681A (en) * | 2024-03-28 | 2024-05-03 | 宝鸡富士特钛业（集团）有限公司 | Analysis method and system for internal damage and defects of titanium alloy pipe |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117373580B (en) | 2024-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113159147B (en) | Image recognition method and device based on neural network and electronic equipment | |
| CN113283446B (en) | Method and device for identifying object in image, electronic equipment and storage medium | |
| CN117373580B (en) | Performance analysis method and system for realizing titanium alloy product based on time sequence network | |
| CN111932547B (en) | Method and device for segmenting target object in image, electronic device and storage medium | |
| CN114496264B (en) | Health index analysis method, device, equipment and medium based on multidimensional data | |
| CN116186594B (en) | Method for realizing intelligent detection of environment change trend based on decision network combined with big data | |
| CN112306835A (en) | User data monitoring and analyzing method, device, equipment and medium | |
| CN113516417A (en) | Service evaluation method and device based on intelligent modeling, electronic equipment and medium | |
| CN116681923A (en) | Automatic ophthalmic disease classification method and system based on artificial intelligence | |
| CN115034315A (en) | Business processing method and device based on artificial intelligence, computer equipment and medium | |
| CN112101481B (en) | Method, device, equipment and storage medium for screening influence factors of target object | |
| CN117036732B (en) | Electromechanical equipment detection system, method and equipment based on fusion model | |
| CN111460293B (en) | Information pushing method and device and computer readable storage medium | |
| CN112052310A (en) | Information acquisition method, device, equipment and storage medium based on big data | |
| CN113705686B (en) | Image classification method, device, electronic equipment and readable storage medium | |
| CN116403693A (en) | Method, device, equipment and storage medium for dispatching questionnaire | |
| CN114049676A (en) | Fatigue state detection method, device, equipment and storage medium | |
| CN114187476A (en) | Vehicle insurance information checking method, device, equipment and medium based on image analysis | |
| CN112634938A (en) | Audio-based personnel positivity analysis method, device, equipment and storage medium | |
| CN113657488B (en) | User grading method, device, equipment and storage medium based on driving behaviors | |
| CN116522105B (en) | Method, device, equipment and medium for integrally constructing data based on cloud computing | |
| CN117611580B (en) | Flaw detection method, flaw detection device, computer equipment and storage medium | |
| CN116486097B (en) | Remote automatic feeding method and system applied to rodent feeding scene | |
| CN113723554B (en) | Model scheduling method, device, electronic equipment and computer readable storage medium | |
| CN117376977B (en) | Mobile phone 5G wireless signal testing system, method, equipment and medium |
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 |