CN114970371A - Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network - Google Patents

Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network Download PDF

Info

Publication number
CN114970371A
CN114970371A CN202210707959.3A CN202210707959A CN114970371A CN 114970371 A CN114970371 A CN 114970371A CN 202210707959 A CN202210707959 A CN 202210707959A CN 114970371 A CN114970371 A CN 114970371A
Authority
CN
China
Prior art keywords
neural network
deformation
wall part
slm
thin
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
CN202210707959.3A
Other languages
Chinese (zh)
Inventor
沈理达
张子凡
谢德巧
刘富玺
李哲晗
田宗军
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.)
Nanjing University of Aeronautics and Astronautics
Original Assignee
Nanjing University of Aeronautics and Astronautics
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 Nanjing University of Aeronautics and Astronautics filed Critical Nanjing University of Aeronautics and Astronautics
Priority to CN202210707959.3A priority Critical patent/CN114970371A/en
Publication of CN114970371A publication Critical patent/CN114970371A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/10—Additive manufacturing, e.g. 3D printing
    • 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
    • G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
    • 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
    • Y02P10/00—Technologies related to metal processing
    • Y02P10/25—Process efficiency

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Prostheses (AREA)

Abstract

The invention discloses a method for predicting deformation of an SLM (Selective laser melting) formed titanium alloy thin-wall part by applying a GA-BP (genetic algorithm-BP) neural network, belonging to the field of intelligent manufacturing, and comprising the following steps of: and designing a multi-curved-surface arc thin-wall model, carrying out SLM forming tests under different process parameters, and measuring to obtain deformation data of the formed arc thin-wall part. Normalizing the process parameters and the deformation data in the test and then respectively using the normalized process parameters and the normalized deformation data as the input and the output of the GA-BP neural network, designing and constructing a basic BP neural network, optimizing the basic BP neural network by using a genetic algorithm to obtain the optimal initial weight and threshold, and finally training the GA-BP neural network by combining test sample data. The trained GA-BP neural network model can accurately predict the deformation degree of the arc thin-wall part under a specific working condition, can effectively guide the research and development of a new process for forming the titanium alloy thin-wall part by the SLM, does not need to be deduced according to a deformation mechanism, can reduce the experiment cost and improve the forming precision of the part, is simple to use, and provides a new technical means for the design and manufacture of complex thin-wall parts.

Description

Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network
Technical Field
The invention belongs to the technical field of intelligent manufacturing, and particularly relates to a method for predicting deformation of an SLM (Selective laser melting) formed titanium alloy thin-wall part by applying a GA-BP (genetic algorithm-Back propagation) neural network.
Background
The additive manufacturing technology has wide application prospects in the fields of aerospace, biomedical treatment and the like, and the Selective Laser Melting (SLM) technology is used as one of metal additive manufacturing technologies, and compared with the traditional casting technology, the Selective Laser Melting (SLM) technology has the advantages of digital forming, high material utilization rate, short period, no limitation of part shapes and the like. The titanium alloy material has light weight, high strength, excellent biocompatibility and excellent corrosion resistance, the product with light weight and integrated integral structure and function becomes a development trend nowadays, and the shape and the structure of the accompanying metal part are increasingly special and complex, such as the structure with a thin wall, a curved surface and the like, so that the SLM-formed titanium alloy thin-wall part has better research prospect in the aspects of aerospace and biomedical application, and is particularly suitable for producing parts with complex structures.
However, for a multi-curved surface complex thin-wall part, in the selective laser melting and forming process, due to the extremely rapid melting and solidification of metal powder, the formation of a high temperature gradient can easily lead to the development of thermal stress in the printed sample part, cause the inconsistent shrinkage of a sample part layer, and finally lead to the deformation of the sample part after the continuous accumulation. The method not only reduces the size precision and mechanical property of the printed piece, but also directly causes the interruption of the additive manufacturing process to be serious, thereby limiting the development and application of the complex thin-walled piece.
The process of forming the multi-curved-surface complex thin-wall part by the SLM is complex, the affected factors are more, and the deformation degree of the formed part and the affected factors have a complex nonlinear relationship. At present, a large number of tests qualitatively summarize the influence rules of partial independent factors, but no model with a simple form and accurate prediction can be used for predicting the deformation degree of thin-wall parts under different process parameter combinations, so that guidance is provided for the optimization research and development of process parameters, a large number of tests and time cost are saved, and the forming precision and quality of formed parts are improved.
The artificial neural network has high nonlinear function approximation capability, strong fault-tolerant capability and self-adaptive learning capability in data tasks such as classification, regression and the like, can establish a high nonlinear relation between input and output, and provides a powerful tool for solving the nonlinear problems with complexity and many influencing factors. Therefore, the establishment of the SLM forming titanium alloy thin-wall part deformation prediction model which is simple in form and can accurately predict the deformation of the SLM forming titanium alloy thin-wall part and can be used for guiding engineering application is one of the problems to be solved in the field.
Disclosure of Invention
The invention aims to provide a method for predicting the deformation of an SLM (Selective laser melting) formed titanium alloy thin-wall part by applying a GA-BP (genetic algorithm-BP) neural network, and establishes a deformation prediction model with a simple form and high accuracy, so as to provide guidance for improving the forming precision of the SLM formed titanium alloy thin-wall part in engineering.
In order to realize the purpose, the invention provides the following technical scheme:
s1: according to the 3D printing application case of the human oral maxillary stent in the field of biological medical treatment, a simplified multi-curved-surface arc thin-wall model is designed.
S2: and obtaining deformation measurement values of the SLM forming circular arc thin-wall part under different process conditions.
S3: and taking the process parameters (including laser power, scanning speed, scanning distance and placing angle) in the S2 as input parameters of the GA-BP neural network, and taking the deformation measured value of the simple thin-wall bracket under the corresponding parameter condition as an output parameter. And all data are subjected to normalization preprocessing and are divided into training sample data and test sample data.
S4: and respectively determining the number of nodes of an input layer and an output layer of the neural network according to the input and output parameters in the S3, determining the number of hidden layers and the number of nodes of the hidden layers, and selecting the type of a transfer function between the layers to construct a basic BP neural network structure.
S5: based on the BP neural network structure of S4, the initial weight and the threshold value are searched by using a genetic algorithm, the fitness value is calculated, the optimal solution is updated according to the fitness value, and the optimization of the BP neural network is realized.
S6: and training the GA-BP network by using the data in the training sample set, continuously adjusting the network parameters of the GA-BP network, testing the GA-BP network by using the data in the testing sample set after the training termination condition is met, and comparing and analyzing the prediction result with the actual result to verify the prediction accuracy of the GA-BP network model on the deformation of the arc thin-walled workpiece.
The invention has the beneficial effects that:
the invention provides a method for predicting the deformation of an SLM (selective laser melting) forming titanium alloy thin-wall part by applying a GA-BP (genetic algorithm-Back propagation) neural network, which only needs to train the GA-BP neural network by using enough test data, and continuously optimizes by adjusting network training parameters, so that a neural network model completing a training target can effectively reflect the complex nonlinear mapping relation between each process parameter and the deformation value of the forming part, and can accurately predict the deformation condition of the SLM forming thin-wall part under the specified process condition.
Compared with a BP neural network model, the SLM forming titanium alloy thin-wall part deformation prediction model based on the GA-BP neural network has better robustness and prediction performance, solves the problem that the forming precision of the titanium alloy thin-wall part is difficult to monitor under different SLM process conditions, and can be used for guiding the complex thin-wall part to be optimized by utilizing the SLM forming process parameters in engineering application, thereby greatly saving the production and test cost, effectively improving the forming precision of the formed part, and having wide engineering significance and economic benefit.
Drawings
The drawings that accompany the detailed description can be briefly described as follows.
FIG. 1 is a schematic general flow diagram of the process of the present invention;
FIG. 2 is a schematic view of a circular arc thin wall model;
FIG. 3 is a graph showing the alignment effect of the circular arc thin wall initial model and the deformation model;
FIG. 4 is a schematic diagram of the deflection of a circular arc thin-walled part;
FIG. 5 is a schematic diagram of a BP neural network architecture;
FIG. 6 is a schematic diagram of a genetic algorithm optimized BP neural network model;
FIG. 7 is a schematic diagram comparing predicted values and actual measured values of a GA-BP neural network and a BP neural network.
Detailed Description
The technical solution of the present invention will be described in detail and clearly with reference to the drawings of the specification, taking a multi-curved arc thin-wall model as an example, and it is obvious that the described embodiment is a part of the embodiments of the present invention, but not all embodiments. Other embodiments, which can be derived by one of ordinary skill in the art from the embodiments given herein without making any creative effort, shall fall within the scope of the present invention.
FIG. 1 is a schematic flow chart of the implementation of the present invention, which mainly includes six steps. The following will describe each step in detail by taking a multi-surface circular arc thin-wall model as an example.
S1: according to the application case of 3D printing of the human oral maxillary stent in the biomedical field, a simplified multi-curved arc thin-wall model is designed, as shown in FIG. 2, specifically, the size parameters are as follows: the width is 40mm, the height is 15mm, the radius of the inner arc surface is 22mm, and the thickness is 1 mm.
S2: the method comprises the steps of carrying out part processing tests under different process parameter conditions by utilizing self-researched Selective Laser Melting (SLM) forming equipment, separating a formed part from a substrate by wire electrical discharge machining, and carrying out three-dimensional scanning on the formed arc thin-walled part by utilizing an optical three-dimensional scanner combining purple structure light and a photogrammetry operation principle to generate a complete STL deformation model. And finally, introducing an initial design model A of the arc thin-wall part and a deformation model B obtained through experimental measurement into Control X software, taking the model A as a reference geometry and the model B as a test geometry, and spatially aligning the models A and B by taking the maximum overlapping area as a target, as shown in FIG. 3. And calculating by 3D comparison to obtain the maximum deviation and the minimum deviation between the two models, and taking the average value of the maximum deviation and the minimum deviation in the middle of the outer edge of the arc thin-wall part as the final deformation measurement value of the arc thin-wall part shown in figure 4.
S3: and collecting and preprocessing multi-factor and multi-level test data in S2, taking 4 process parameters such as laser power, scanning speed, scanning interval, placing angle and the like as input parameters of the GA-BP neural network, and taking the deformation measurement value of the arc thin-wall part under the corresponding parameter condition as an output parameter of the GA-BP neural network. And all data are subjected to normalization preprocessing and are divided into training sample data and test sample data.
Mapping the acquired data to a (0, 1) interval through a normalization formula, wherein the normalization formula of the data is as follows:
Figure BSA0000275548580000021
s4: determining the number of nodes of the input layer and the output layer respectively according to the input and output parameters in S3, determining the number of hidden layers and the number of nodes of hidden layers, selecting the type of transfer function between layers of the neural network, and constructing a basic BP neural network structure, as shown in fig. 5, the specific implementation steps are as follows:
(a) defining the number of nodes of an input layer and an output layer of the BP neural network as m and n respectively, wherein the number of the nodes of the input layer and the number of the nodes of the output layer of the BP neural network are known as the dimensionality of an input variable and the dimensionality of an output variable respectively, namely m is 4, and n is 1;
(b) the number of hidden layers of the BP neural network is determined as one layer, and the number of nodes of the hidden layers meets the Kolmogorov theorem:
Y＝2X+1
in the formula, Y is the number of hidden layer nodes, X is the number of input layer nodes, namely the number of hidden layer nodes of the BP neural network is 9;
(c) an S-type tangent function tansig and a linear function purelin are respectively selected as transfer functions of an implicit layer and an output layer, and a training function tranlmm in a Levenberg-Marquardt algorithm is selected as a training method to reverse transfer errors.
Tansig transfer function:
Figure BSA0000275548580000031
purlin transfer function:
g(x)＝x (1.3)
s5: based on the BP neural network structure of S4, the initial weight and the threshold value are searched by using a genetic algorithm, the fitness value is calculated, the optimal solution is updated according to the fitness value, and the optimization of the BP neural network is realized. The process of optimizing the BP neural network by the genetic algorithm is shown in FIG. 6, and the specific steps are as follows:
(a) encoding an initial weight and a threshold in a binary real number form, and performing population initialization on the initial weight and the threshold;
(b) taking the training mean square error as a fitness function of the GA algorithm, calculating the fitness value of individuals in the population, and screening better initial weight and threshold after continuously undergoing selection, intersection and variation links in the maximum genetic algebra range if the fitness value at the moment is not optimal;
(c) when the fitness value or the genetic algebra meets the end condition, the BP neural network immediately obtains the optimal initial weight and threshold value, and completes the optimization, otherwise, the fitness value of the iteration individual is continuously updated.
And setting the values of relevant parameters of the genetic algorithm: the population scale is 40, the maximum genetic passage number is 50, the cross probability is 0.7, the mutation probability is 0.01, and the gully is 0.95.
S6: training the optimized GA-BP neural network by using data in the training sample set, and adjusting the network parameters to be as follows: the maximum learning frequency is 1000, the learning rate is 0.01, the convergence precision is 0.000001, and when the Mean Square Error (MSE) of the training set is less than 0.000001, or the obtained maximum value is 1000 epochs, or the MSE of the training set does not decrease in six iterations, the training of the GA-BP prediction model is completed. And after the training is finished, testing the test sample set by using the data of the test sample set, and comparing and analyzing the predicted result with the actual result.
The average relative error (MRE) of the prediction result output by the prediction model is selected to evaluate the prediction performance of the model.
Figure BSA0000275548580000032
In the formula, T is the data number of the test sample set, y' is a deformation prediction value after inverse normalization of the neural network, and y is an actual deformation measurement value, so that the prediction accuracy of the GA-BP network model on the deformation of the arc thin-wall part is verified.
Obtaining an optimized GA-BP neural network prediction model according to the method, inputting 4 sets of process parameter data in a test sample set to the network model after training is completed, and obtaining a predicted value of the deformation of the arc thin-wall part, thereby achieving the purpose of guiding the optimization of the process parameters. The prediction result of the BP neural network is added, and the effectiveness of the GA-BP prediction model is further verified. The result is shown in fig. 7, wherein MREs of the BP prediction model and the optimized GA-BP prediction model are respectively 8.9% and 1.2%, which indicates that the GA-BP model optimized by the genetic algorithm has higher prediction accuracy, better robustness and prediction performance, and is more favorable for optimizing SLM process parameters and improving deformation of thin-wall parts.
The above-described embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solutions of the present invention can be made by those skilled in the art without departing from the spirit of the present invention, and the technical solutions of the present invention are within the scope of the present invention defined by the claims.

Claims (8)

1. A method for predicting deformation of an SLM-formed titanium alloy thin-wall part by applying a GA-BP neural network is characterized by comprising the following steps of:
s1: according to the 3D printing application case of the human oral maxillary stent in the field of biological medical treatment, a multi-curved-surface arc thin-wall model is designed.
S2: and obtaining deformation measurement values of the SLM forming circular arc thin-wall part under different process conditions.
S3: and (4) taking the process parameters (including laser power, scanning speed, scanning interval and placing angle) in the S2 as input parameters of the GA-BP neural network, and taking the deformation measured value of the arc thin-wall part as an output parameter. And all data are subjected to normalization preprocessing and are divided into training sample data and test sample data.
S4: and respectively determining the number of nodes of an input layer and an output layer of the neural network according to the input and output parameters in the S3, determining the number of hidden layers and the number of nodes of the hidden layers, and selecting the type of a transfer function between the layers to construct a basic BP neural network structure.
S5: and based on the BP neural network of S4, searching the initial weight and the threshold value by using a genetic algorithm, calculating the fitness value, updating the optimal solution according to the fitness value, and realizing the optimization of the BP neural network.
S6: and (4) training the GA-BP network by using the training sample set data in the S3, continuously adjusting the network parameters of the GA-BP network, testing the GA-BP network by using the data in the test sample set after the training termination condition is met, and comparing and analyzing the prediction result with the actual result to verify the prediction accuracy of the GA-BP network model on the deformation of the arc thin-walled workpiece.
2. The method for predicting the deformation of the SLM-formed titanium alloy thin-wall part by using the GA-BP neural network as claimed in claim 1, wherein the designed multi-curved arc thin-wall model is 40mm in width, 15mm in height, 22mm in radius of the inner arc surface and 1mm in thickness, and belongs to a simplified version of a human oral maxillary stent model.
3. The method for predicting the deformation of the SLM-formed titanium alloy thin-wall part by applying the GA-BP neural network as claimed in claim 1, wherein an optical three-dimensional scanner and Control X software are used to obtain the deformation value of the SLM-formed part. The optical three-dimensional scanner combines purple structure light and a photogrammetry operation principle, can carry out three-dimensional scanning on a formed thin-walled workpiece, and generates a complete STL deformation model. And finally, introducing an initial design model A of the arc thin-wall part and a deformation model B obtained through experimental measurement into Control X software, taking the model A as a reference geometry and the model B as a test geometry, aligning the models A and B in space by taking the maximum overlapping area as a target, obtaining the maximum deviation and the minimum deviation between the two models through 3D comparison calculation, and taking the average value of the maximum deviation and the minimum deviation as the deformation measurement value of the arc thin-wall part.
4. The method for predicting the deformation of the SLM-formed titanium alloy thin-wall part by using the GA-BP neural network as claimed in claim 1, wherein 4 process parameters such as laser power, scanning speed, scanning interval and placing angle are selected as input of the GA-BP neural network in S3, namely the input dimension is 4, and the measured value of the deformation of the arc thin-wall part under different process parameter combinations is used as output of the GA-BP neural network, namely the output dimension is 1.
5. The method for predicting the deformation of the SLM-formed titanium alloy thin-wall part by applying the GA-BP neural network as claimed in claim 1, wherein the method comprises the steps of performing a multi-factor and multi-level SLM forming test on the arc thin-wall part by adjusting various horizontal values corresponding to different process parameters, performing normalization pretreatment on each group of test data, and dividing the test data into a training sample and a testing sample of the GA-BP neural network.
6. The method for predicting the deformation of the SLM-formed titanium alloy thin-wall part by applying the GA-BP neural network as claimed in claim 1, wherein the method for constructing the basic BP neural network structure in S4 is as follows:
(a) defining the number of nodes of an input layer and an output layer of the BP neural network as m and n respectively, wherein the number of the nodes of the input layer and the number of the nodes of the output layer of the BP neural network are known as the dimensionality of an input variable and the dimensionality of an output variable respectively, namely m is 4, and n is 1;
(b) the number of hidden layers of the BP neural network is determined as one layer, and the number of nodes of the hidden layers meets the Kolmogorov theorem: y is 2X +1, where Y is the number of hidden layer nodes, and X is the number of input layer nodes, that is, the number of hidden layer nodes of the BP neural network is 9;
(c) an S-type tangent function tansig and a linear function purelin are respectively selected as transfer functions of an implicit layer and an output layer, and a training function tranlmm in a Levenberg-Marquardt algorithm is selected as a training method to reverse transfer errors.
7. The method for predicting the deformation of the SLM-formed titanium alloy thin-wall part by using the GA-BP neural network as claimed in claim 1, wherein the genetic algorithm in S5 is optimized by the following main steps:
(a) encoding the initial weight and the threshold in a binary real number form, and performing population initialization on the initial weight and the threshold;
(b) taking the training mean square error as a fitness function of a genetic algorithm, calculating the fitness value of individuals in a population, and screening better initial weight and threshold after continuously undergoing selection, intersection and variation links in the maximum genetic algebra range if the fitness value at the moment is not optimal;
(c) when the fitness value or the genetic algebra meets the end condition, the BP neural network immediately obtains the optimal initial weight and threshold value, and completes the optimization, otherwise, the fitness value of the iteration individual is continuously updated.
8. The method of claim 1, wherein in step S6, the optimized GA-BP network is trained by using data in a training sample set, and when a Mean Square Error (MSE) of the training set is less than 0.000001, or a maximum value of 1000 epochs is obtained, or the MSE of the training set does not decrease in six iterations, the training of a GA-BP prediction model is completed, the prediction result is tested by using the data in the test sample set, and the prediction result is compared with an actual result.
CN202210707959.3A 2022-06-21 2022-06-21 Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network Pending CN114970371A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210707959.3A CN114970371A (en) 2022-06-21 2022-06-21 Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210707959.3A CN114970371A (en) 2022-06-21 2022-06-21 Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network

Publications (1)

Publication Number Publication Date
CN114970371A true CN114970371A (en) 2022-08-30

Family

ID=82964734

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210707959.3A Pending CN114970371A (en) 2022-06-21 2022-06-21 Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network

Country Status (1)

Country Link
CN (1) CN114970371A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116757534A (en) * 2023-06-15 2023-09-15 中国标准化研究院 Intelligent refrigerator reliability analysis method based on neural training network

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116757534A (en) * 2023-06-15 2023-09-15 中国标准化研究院 Intelligent refrigerator reliability analysis method based on neural training network
CN116757534B (en) * 2023-06-15 2024-03-15 中国标准化研究院 Intelligent refrigerator reliability analysis method based on neural training network

Similar Documents

Publication Publication Date Title
CN109508488B (en) Shot peening forming process parameter prediction method based on genetic algorithm optimization BP neural network
CN106020147A (en) Systematic analysis method for assembling precision of complex precise mechanical product
Fu et al. Springback prediction of high-strength sheet metal under air bending forming and tool design based on GA–BPNN
CN113204851B (en) Tolerance optimization method for all-welded assembly
CN109766597B (en) High-precision linear solving method for assembly pose by considering geometric errors
CN114970371A (en) Method for predicting deformation of SLM (Selective laser melting) formed titanium alloy thin-wall part by applying GA-BP (genetic algorithm-Back propagation) neural network
CN108108582A (en) A kind of method for numerical simulation of curved-surface piece flexible rolling forming process
CN115705443A (en) Method for manufacturing steel pipe cold-drawing die
CN110991605A (en) Low-pressure casting mold temperature prediction method of multivariable time series deep belief network
CN113722942A (en) Finite element calculation model considering 3D printing geometric defects
CN114274457B (en) Injection mold pouring system structural parameter decision method based on Kriging-GA
CN113705897A (en) Product quality prediction method and system for industrial copper foil production
Lin et al. Measurement point prediction of flatness geometric tolerance by using grey theory
Gao et al. A 6-sigma robust optimization method for stamping forming of automobile covering parts based on residual error and radial basis interpolation
CN111444619A (en) Online analysis method and equipment for injection mold cooling system
CN110909492A (en) Sewage treatment process soft measurement method based on extreme gradient lifting algorithm
CN116595827A (en) Infinite dimension strip shot peening process planning method and system
CN113221398B (en) Method for predicting L-shaped composite material part curing deformation rebound angle
CN112613146B (en) Straightening self-adaptive optimization method, straightening self-adaptive optimization system, storage medium and computing equipment
Narayanan Namboothiri Analysis and optimization of machining process using evolutionary algorithms
CN114863063B (en) Springback prediction method for single-point gradual change surface topography forming
Liu et al. Process optimal control of sheet metal forming springback based on evolutionary strategy
CN114880800B (en) Comprehensive gear shaping method for improving tooth surface unbalanced load
CN113221319B (en) Measurement and calculation method for C-shaped composite material part curing deformation resilience angle
CN114357875B (en) Intelligent data processing system based on machine learning

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