CN103745271A - Forecasting method for induction thermal deposition calcium-phosphate coating process on basis of neural network - Google Patents
Forecasting method for induction thermal deposition calcium-phosphate coating process on basis of neural network Download PDFInfo
- Publication number
- CN103745271A CN103745271A CN201410001043.1A CN201410001043A CN103745271A CN 103745271 A CN103745271 A CN 103745271A CN 201410001043 A CN201410001043 A CN 201410001043A CN 103745271 A CN103745271 A CN 103745271A
- Authority
- CN
- China
- Prior art keywords
- coating
- neural network
- frequency
- network
- mentioned
- 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
Images
Landscapes
- Application Of Or Painting With Fluid Materials (AREA)
- Investigating Or Analyzing Materials Using Thermal Means (AREA)
Abstract
The invention discloses a forecasting method for an induction thermal deposition calcium-phosphate coating process on the basis of a neural network, which comprises the following steps of preparing a calcium-phosphate coating by utilizing an induction thermal deposition method and collecting a test sample; fitting out a training sample by the test sample; establishing the BP (Back Propagation) neural network, training the network by adopting a Levenberg-Marquardt algorithm and utilizing the training sample, and testing the generalization ability of the network by the test sample; designing orthogonal horizontal process parameters, using the orthogonal horizontal process parameters as input parameters, simulating output parameters by utilizing the trained network and constructing an orthogonal sample; and carrying out analyzing calculation on the obtained orthogonal sample and forecasting influence of the process parameters on the deposition process. According to the invention, the data mining ability of the neural network is successfully utilized and the relation between three process parameters of time, frequency and temperature of induction thermal deposition and the deposition rate is analyzed, so that design of the calcium-phosphate coating preparation process is effectively guided.
Description
Technical field
The present invention relates to the method for testing of a coating performance, the Forecasting Methodology that particularly relates to a kind of coating process, also relate to a kind of nerve network system analysis and application method, be applied to set up technological parameter knowledge base and the analysis process parameter influence technique field to deposition process, experimental design and commercial Application have been had to good technological guidance's meaning.
Background technology
HAp coating C/C compound substance is that potential bone of new generation substitutes and repair materials, has broad application prospects.The method of preparing at present C/C composite material surface HAp coating mainly contains: plasma spraying method, coating sintering method, electrochemical process, alkali heat treating process, biomimetic method and induction heat sedimentation.
Test confirmation, induction heat sedimentation is simple, and gentle, the applicable complicated substrate surface deposition of process conditions, overflows from carbon/carbon composite material base surface without gas again.Therefore,, by controlling process conditions, can realize at the surface of carbon/carbon composite of modification and preparing by the closelypacked fine and close calcium phosphor coating of granular crystal grain, and the intensity of coating can reach the standard-required of clinical use.In view of above advantage, induction heat sedimentation is a kind of new technology of worth continuation research and development.
Yet, induction heat sedimentation, be the new technology of the preparation HAp coating C/C compound substance that grows up for 2009, current research is confined to the preparation of coating mostly, coating pattern, component, the sign of structure, and the test of coating performance, the research to its technological parameter action rule, be limited by experiment condition more, fail to carry out effectively rapidly.
Error back propagation neural network, BP network, has stronger associative memory and generalization ability, is more ripe in theory neural network, is also the network of succeeding and applying in many science and technology field.Its self study and adaptive ability become carries out to various types of signal a kind of natural instrument that multi-usage processing is processed, can to challenge, make rational judgement decision-making according to the experience of the knowledge of having learned and the problem of processing, provide more satisfied answer, or following process is made and effectively predicted and estimate.The patent that relevant neural network is applied in practice and paper are existing a lot, for example, Jiang Xiuhua etc. have invented a kind of method for evaluating video quality based on artificial neural network, Han Yang etc. have invented a kind of CT perfusion medical image intelligent method for fusing based on neural network model, and Fei Chunguo etc. have invented a kind of neural network of solving-optimizing problem.The report that neural network is prepared aspect calcium phosphor coating Forecast on Process in induction heat sedimentation is more rare, but theoretically, BP neural network can, in preparation technology of coating research, still also not have at present about BP Application of Neural Network is deposited to the relevant report in calcium phosphor coating technology assessment and prediction field in induction heat completely.
Summary of the invention
In order to solve prior art problem, the research that the object of the invention is to overcome induction heat sedimentation technological parameter relies on the deficiency that laboratory facilities exist more, and development of research thinking provides a kind of induction heat deposition calcium phosphor coating Forecast on Process method based on neural network.Self-learning capability and the adaptive ability of BP neural network have been utilized, on the basis of the finite data obtaining in experiment, adopt Levenberg-Marquardt Algorithm for Training network, meeting, network error performance function value is less, and under the good prerequisite of network generalization, determine the structure and parameter of network, then design orthogonal test sample, utilize the network having trained to carry out simulation data, obtain quadrature samples collection, enrich technological parameter knowledge base, on this basis, analysis process parameter is on being coated with the impact of layer deposition process, experiment is had to good directive significance.
For reaching foregoing invention, create object, the present invention adopts following technical proposals:
An induction heat deposition calcium phosphor coating Forecast on Process method based on neural network, comprises the following steps:
A. under laboratory condition, adopt induction heat sedimentation, at surface of carbon/carbon composite, prepare calcium phosphor coating, by data collector, record time, frequency and three technological parameters of temperature in experimentation and corresponding coating deposition weight, arrange experimental data, structure test sample book, the test sample book that is organized into 3 input 1 outputs, wherein time, frequency and temperature are as input parameter, and deposition is as output parameter;
B. the test sample book data of the data collector collection in above-mentioned a step arrangement are inputted to data analysis system, the test sample book data that utilization constructs in above-mentioned a step, by data analysis system, draw the time, the graph of a relation of three technological parameters of frequency and temperature and coating deposition weight, then utilize data analysis system to get regularly a little in the drawings, read point coordinate data, with the time, frequency, 4 groups of data that temperature and coating deposition weight form are as a row sample array, using and get a quantity as another row sample array, matching arranges out a dimension and is 4 * get the training sample of 3 input 1 outputs of a quantity, wherein choosing of number of training will be in conjunction with complicacy and the input parameter of induction heat depositing operation, the number of output parameter decides,
C. by the data analysis system in above-mentioned b step, set up BP nerve network system, the training sample of utilization matching in above-mentioned b step, adopt Levenberg-Marquardt Algorithm for Training network system, and be used in the generalization ability of the test sample book test network of constructing in above-mentioned a step, first by the time, three technological parameters of frequency and temperature are normalized, result is as the input parameter of BP neural network, coating deposition weight is as output parameter, at the error function of taking into account training sample, be less than under the preferential prerequisite of the generalization ability of goal-setting value and test sample book, determine and preserve the structure and parameters of this BP neural network, the structure of BP nerve network system adopts hit-and-miss method to determine the number of plies of network hidden layer of BP network system and the hidden unit number of every layer, decision condition is the desired value that the error performance index of training sample is less than setting, and the generalization ability that meets test sample book is more excellent, the structure of preferably determining and preserving this BP neural network is pair hidden layer configurations, and hidden unit number is respectively 8 and 3, constructs the BP network of [3,8,3, a 1] structure,
D. utilize the coating deposition weight parameter of the BP nerve network system output training in above-mentioned c step, make coating deposition weight parameter correspondence in setting-up time region, regular value in setpoint frequency scope and design temperature interval, structure is set the quadrature level input sample of quantity, then quadrature level input sample is processed according to the method for normalizing in above-mentioned c step, result data are re-used as the input parameter input data analysis system of BP neural network, the network that utilization trains in above-mentioned c step carries out emulation, obtain corresponding coating deposition weight output parameter, then disposal data, construct the quadrature samples of complete setting dimension, the coating deposition weight parameter of the BP nerve network system output that utilization trains in above-mentioned c step, preferably make coating deposition weight parameter correspondence even value in 2 ~ 25min time zone, 289 ~ 381kHz frequency range and 378 ~ 408K temperature range, structure is set the quadrature level input sample of quantity,
E. in the data analysis system in above-mentioned d step, continue the quadrature samples data that analytical calculation obtains in above-mentioned d step, analysis time, frequency and three technological parameters of temperature relation that affects on coating deposition weight parameter, deposits the action rule of weight thereby obtain time, frequency and three technological parameters of temperature to coating respectively; When preferably analysis time, frequency and three technological parameters of temperature are related to affecting of coating deposition weight parameter respectively, preferably by the mapping software in data analysis system, draw and show, at least draw and show that any one technological parameter in output time, frequency and three technological parameters of temperature deposits the trend map that affects of weight on coating; Particularly preferably draw and show that any two technological parameters in output time, frequency and three technological parameters of temperature deposit the trend map that affects of weight on coating.
The present invention compared with prior art, has following apparent outstanding substantive distinguishing features and remarkable advantage:
1. under the limited experimental data condition that the present invention gives user, utilize BP neural network, learning ability, adaptive ability, the fault-tolerant ability of neural network have been brought into play, effectively experimental data is learnt, foundation has the BP network of certain generalization ability, on this basis, utilize the quadrature level input sample of design, simulation data parameter, a series of quadrature samples have been constructed, enrich induction heat deposition process parameters knowledge base, for analysis process parameter provides a kind of new method to the impact of deposition process, expanded research thinking;
2. the present invention is incorporated into induction heat sedimentation by BP neural network and prepares in the Forecast on Process method of calcium phosphor coating, on limited experimental data basis, analog computation has obtained abundant technological parameter knowledge base, utilize knowledge base, by the data processing in later stage, can infer the impact of technological parameter on deposition process, thereby effectively instruct calcium phosphor coating preparation technology design, experimental design and commercial Application are had to good technological guidance's meaning.
Accompanying drawing explanation
Fig. 1 is the theory diagram of the embodiment of the present invention one induction heat deposition calcium phosphor coating Forecast on Process method.
Fig. 2 is the BP artificial neural network structural drawing of the embodiment of the present invention one.
Fig. 3 is the algorithm flow chart of the BP neural network of the embodiment of the present invention one.
Fig. 4 is the time m-coating heavy deposition magnitude relation X-Y scheme of the quadrature samples of the embodiment of the present invention one making.
Fig. 5 is temperature-coating heavy deposition magnitude relation X-Y scheme of the quadrature samples of the embodiment of the present invention one making.
Fig. 6 is frequency-coating heavy deposition magnitude relation X-Y scheme of the quadrature samples of the embodiment of the present invention one making.
Fig. 7 is T/F-coating heavy deposition magnitude relation three-dimensional plot of the quadrature samples of the embodiment of the present invention two making.
Fig. 8 is frequency-temperature-coating heavy deposition magnitude relation three-dimensional plot of the quadrature samples of the embodiment of the present invention two making.
Fig. 9 is temperature-time m-coating heavy deposition magnitude relation three-dimensional plot of the quadrature samples of the embodiment of the present invention two making.
Embodiment
Details are as follows for the preferred embodiments of the present invention:
embodiment mono-:
In the present embodiment, referring to Fig. 1～Fig. 3, the induction heat deposition calcium phosphor coating Forecast on Process method based on neural network, specifically can be divided into following five steps and realize:
A. utilize induction heat sedimentation to prepare calcium phosphor coating, collecting test sample: adopt induction heat sedimentation under laboratory condition, at surface of carbon/carbon composite, prepare calcium phosphor coating, by data collector, record the time in experimentation, three technological parameters of frequency and temperature and corresponding coating deposition weight, arrange experimental data, form the test sample book of 3 input 1 outputs, construct 4 * 48 two-dimensional array as test sample book, 48 represent that one has 48 groups of data, 4 represent that every group of data comprise three input parameters and an output parameter, time wherein, frequency and temperature are as input parameter, the coating deposition weight that corresponding experiment records is as output parameter,
B. the test sample book of utilizing the first step to gather, matching training sample: the test sample book data of the data collector collection in above-mentioned a step arrangement are inputted to data analysis system, 48 groups of test sample book data that utilization constructs in above-mentioned a step, while making frequency f=289kHz, temperature T=378K, 388K, 398K, time m-heavy deposition magnitude relation figure during 408K, the like, the time m-heavy deposition spirogram at four temperature spot places while remaking out f=319kHz and 381kHz, then utilize data analysis system to get regularly a little in the drawings, read point coordinate data, when done each in m-heavy deposition spirogram, evenly get a little 240, record the time of each some representative, frequency, temperature and coating deposition weight, arrange out the training sample that a dimension is 4 * 240,
C. set up BP neural network, utilize training sample training network, and with the generalization ability of test sample book test network: set up BP nerve network system by the data analysis system in above-mentioned b step, first the first three rows to 240 groups of training samples and 48 groups of test sample books, be that input parameter is normalized, normalization formula is: X=(X-Xmin)/(Xmax-Xmin), Xmax refers to the maximal value of technological parameter, Xmin refers to the minimum value of technological parameter, after normalization, each technological parameter changes between 0 ~ 1; Utilize 240 groups of training sample training networks after normalization, the generalization ability of 48 groups of test sample book test networks; By the BP network creation function newff in matlab, adopt Levenberg-Marquardt Algorithm for Training network, network hidden layer adopts tansig function, and output layer adopts purelin function, adopts hit-and-miss method to determine hidden layer configuration; Design the BP network of different hidden layer configurations, by the training sample after normalization and test sample book input BP network, training network is to convergence, be less than under the desired value 0.0001 of setting and the good prerequisite of generalization ability of test sample book considering training sample error performance index, determine the structure of network, preserve weights and the threshold value of network.Final this experiment determines that network be two hidden layer configurations, and hidden nodes is followed successively by 8,3, and we have just obtained one reliable [3 like this, 8,3,1] type BP network, Fig. 2 is the BP artificial neural network structural drawing of the present embodiment, and Fig. 3 is the algorithm flow chart of the BP neural network of the present embodiment;
D. design the technological parameter of quadrature level as input parameter, utilize the network simulation output parameter training, structure quadrature samples: the design technology parameter time is even value in 2 ~ 25min, and step-length is 1, gets 24 numbers; Frequency is even value in 289 ~ 381kHz, and step-length is 4, gets 24 numbers; Temperature is even value between 378 ~ 408, and step-length is 1, gets 31 numbers.Then within the scope of this, construct 2064 quadrature input samples; 2064 quadrature input samples are normalized, then input [3,8,3,1] network of the 3rd step gained, the corresponding output parameter of simulation data, deposits weight.It is 4 * 2064 quadrature samples that disposal data obtains dimension;
E. analyze quadrature samples, the impact of predict process parameter on deposition process: in the data analysis system in above-mentioned d step, continue 2064 groups of quadrature samples that analytical calculation obtains in above-mentioned d step, distinguish analysis time, when three technological parameters of frequency and temperature are related to affecting of coating deposition weight parameter, by the mapping software in data analysis system, draw and show, at least draw and show output time, the affect trend X-Y scheme of any one technological parameter in three technological parameters of frequency and temperature on coating deposition weight, referring to Fig. 4～Fig. 6, Fig. 4 is when frequency f=289kHz and temperature T=378K, the time m-coating heavy deposition magnitude relation X-Y scheme of quadrature samples, observe Fig. 4 and can find out the variation along with the time, deposition substantial linear increases, can infer that deposition process is an at the uniform velocity process, Fig. 5 is when frequency f=289kHz and time t=10s, temperature-coating heavy deposition magnitude relation X-Y scheme of quadrature samples, observes Fig. 5 and can find out the rising along with temperature, and deposition increases gradually, by further computational reasoning, can infer that deposition process may be subject to surface reaction controlling, Fig. 6 is when temperature T=378K and time t=10s, and frequency-coating heavy deposition magnitude relation X-Y scheme of quadrature samples is observed Fig. 6 and can be seen in figure, there is a minimal point, can infer in deposition process, an interval may be there is, when frequency changes in this interval, lower deposition will be obtained.The present embodiment has utilized the data mining ability of neural network, has analyzed three technological parameter times of induction heat deposition, frequency, and the relation of temperature and rate of sedimentation, thus effectively instruct calcium phosphor coating preparation technology design.
embodiment bis-:
The present embodiment and embodiment mono-are basic identical, and special feature is:
In the present embodiment, referring to Fig. 7～Fig. 9, the induction heat deposition calcium phosphor coating Forecast on Process method based on neural network, specifically can be divided into following five steps and realize:
A. utilize induction heat sedimentation to prepare calcium phosphor coating, collecting test sample: identical with embodiment mono-;
B. the test sample book of utilizing the first step to gather, matching training sample: identical with embodiment mono-;
C. set up BP neural network, utilize training sample training network, and with the generalization ability of test sample book test network: identical with embodiment mono-;
D. design the technological parameter of quadrature level as input parameter, utilize the network simulation output parameter training, structure quadrature samples: identical with embodiment mono-;
E. analyze quadrature samples, the impact of predict process parameter on deposition process: in the data analysis system in above-mentioned d step, continue 2064 groups of quadrature samples that analytical calculation obtains in above-mentioned d step, when analysis time, frequency and three technological parameters of temperature are related to affecting of coating deposition weight parameter respectively, in above-mentioned steps e, draw and show that any two technological parameters in output time, frequency and three technological parameters of temperature deposit the trend map that affects of weight on coating, referring to Fig. 7～Fig. 9
Fig. 7 is when frequency temperature T=378K, and T/F-coating heavy deposition magnitude relation three-dimensional plot of quadrature samples is observed Fig. 7 and can be found out the variation along with time and frequency, and deposition evenly increases substantially; In different frequency ranges, there is the region that deposition is lower, can infer that deposition process is an at the uniform velocity process, and spray more than minimum deposition, there is the effect of surface tension of surface coating; Fig. 8 is when time t=10s, and the frequency-temperature of quadrature samples-coating heavy deposition magnitude relation three-dimensional plot is observed Fig. 8 and can be found out the rising along with temperature, and deposition increases gradually; In different frequency ranges, there is the region that deposition is lower, by further computational reasoning, can infer that deposition process may be subject to surface reaction controlling, and spray more than minimum deposition, there is the effect of surface tension of surface coating; Fig. 9 is when frequency f=289kHz, temperature-time m-coating heavy deposition magnitude relation three-dimensional plot of quadrature samples, observe Fig. 9 and can see that in figure, three-dimension curved surface approaches clinoplane, illustrate by controlling temperature and evenly heat up or uniform decrease in temperature according to time shaft, can realize the accurate control of coating thickness.The present embodiment can more clearly be observed the affect trend of technological parameter on deposition weight at three dimensions, the information containing in analysis chart 7～Fig. 9, the action rule of tentative prediction technological parameter to deposition process more exactly, find preferably technological parameter, thus guiding experiment design and commercial production.
By reference to the accompanying drawings the embodiment of the present invention is illustrated above; but the invention is not restricted to above-described embodiment; can also make multiple variation according to the object of innovation and creation of the present invention; the change of making under all Spirit Essences according to technical solution of the present invention and principle, modification, substitute, combination, simplify; all should be equivalent substitute mode; as long as goal of the invention according to the invention; only otherwise deviate from know-why and the inventive concept of the induction heat deposition calcium phosphor coating Forecast on Process method that the present invention is based on neural network, all belong to protection scope of the present invention.
Claims (5)
1. the deposition of the induction heat based on a neural network calcium phosphor coating Forecast on Process method, is characterized in that, comprises the following steps:
A. under laboratory condition, adopt induction heat sedimentation, at surface of carbon/carbon composite, prepare calcium phosphor coating, by data collector, record time, frequency and three technological parameters of temperature in experimentation and corresponding coating deposition weight, arrange experimental data, structure test sample book, the test sample book that is organized into 3 input 1 outputs, wherein time, frequency and temperature are as input parameter, and deposition is as output parameter;
B. the test sample book data of the data collector collection in above-mentioned a step arrangement are inputted to data analysis system, the test sample book data that utilization constructs in above-mentioned a step, by data analysis system, draw the time, the graph of a relation of three technological parameters of frequency and temperature and coating deposition weight, then utilize data analysis system to get regularly a little in the drawings, read point coordinate data, with the time, frequency, 4 groups of data that temperature and coating deposition weight form are as a row sample array, using and get a quantity as another row sample array, matching arranges out a dimension and is 4 * get the training sample of 3 input 1 outputs of a quantity,
C. by the data analysis system in above-mentioned b step, set up BP nerve network system, the training sample of utilization matching in above-mentioned b step, adopt Levenberg-Marquardt Algorithm for Training network system, and be used in the generalization ability of the test sample book test network of constructing in above-mentioned a step, first by the time, three technological parameters of frequency and temperature are normalized, result is as the input parameter of BP neural network, coating deposition weight is as output parameter, at the error function of taking into account training sample, be less than under the preferential prerequisite of the generalization ability of goal-setting value and test sample book, determine and preserve the structure and parameters of this BP neural network, the structure of BP nerve network system adopts hit-and-miss method to determine the number of plies of network hidden layer of BP network system and the hidden unit number of every layer, decision condition is the desired value that the error performance index of training sample is less than setting, and the generalization ability that meets test sample book is more excellent,
D. utilize the coating deposition weight parameter of the BP nerve network system output training in above-mentioned c step, make coating deposition weight parameter correspondence in setting-up time region, regular value in setpoint frequency scope and design temperature interval, structure is set the quadrature level input sample of quantity, then quadrature level input sample is processed according to the method for normalizing in above-mentioned c step, result data are re-used as the input parameter input data analysis system of BP neural network, the network that utilization trains in above-mentioned c step carries out emulation, obtain corresponding coating deposition weight output parameter, then disposal data, construct the quadrature samples of complete setting dimension,
E. in the data analysis system in above-mentioned d step, continue the quadrature samples data that analytical calculation obtains in above-mentioned d step, analysis time, frequency and three technological parameters of temperature relation that affects on coating deposition weight parameter, deposits the action rule of weight thereby obtain time, frequency and three technological parameters of temperature to coating respectively.
2. the induction heat based on neural network deposits calcium phosphor coating Forecast on Process method according to claim 1, it is characterized in that: in above-mentioned steps e, when analysis time, frequency and three technological parameters of temperature are related to affecting of coating deposition weight parameter respectively, by the mapping software in data analysis system, draw and show, at least draw and show that any one technological parameter in output time, frequency and three technological parameters of temperature deposits the trend map that affects of weight on coating.
3. the induction heat based on neural network deposits calcium phosphor coating Forecast on Process method according to claim 2, it is characterized in that: in above-mentioned steps e, draw and show that any two technological parameters in output time, frequency and three technological parameters of temperature deposit the trend map that affects of weight on coating.
4. according to the deposition of the induction heat based on neural network calcium phosphor coating Forecast on Process method described in any one in claim 1～3, it is characterized in that: in above-mentioned c step, the structure of determining and preserving this BP neural network is two hidden layer configurations, hidden unit number is respectively 8 and 3, construct one [3,8,3,1] the BP network of structure.
5. according to the deposition of the induction heat based on neural network calcium phosphor coating Forecast on Process method described in any one in claim 1～3, it is characterized in that: in above-mentioned d step, the coating deposition weight parameter of the BP nerve network system output that utilization trains in above-mentioned c step, make coating deposition weight parameter correspondence even value in 2 ~ 25min time zone, 289 ~ 381kHz frequency range and 378 ~ 408K temperature range, structure is set the quadrature level input sample of quantity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410001043.1A CN103745271B (en) | 2014-01-02 | 2014-01-02 | Forecasting method for induction thermal deposition calcium-phosphate coating process on basis of neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410001043.1A CN103745271B (en) | 2014-01-02 | 2014-01-02 | Forecasting method for induction thermal deposition calcium-phosphate coating process on basis of neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103745271A true CN103745271A (en) | 2014-04-23 |
| CN103745271B CN103745271B (en) | 2017-05-10 |
Family
ID=50502288
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410001043.1A Expired - Fee Related CN103745271B (en) | 2014-01-02 | 2014-01-02 | Forecasting method for induction thermal deposition calcium-phosphate coating process on basis of neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103745271B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109234661A (en) * | 2018-09-21 | 2019-01-18 | 北京联合涂层技术有限公司 | A kind of heat spraying method and system based on artificial neural network |
| CN110047247A (en) * | 2019-05-21 | 2019-07-23 | 武汉理工大学 | A kind of smart home device accurately identifying Falls in Old People |
| CN110165297A (en) * | 2019-05-31 | 2019-08-23 | 重庆市科学技术研究院 | A kind of procedure parameter selection method of film lithium cell solid electrolyte |
| CN110756803A (en) * | 2019-10-27 | 2020-02-07 | 浙江亚通焊材有限公司 | Preparation method of die steel powder material for computer-formatted 3D printing |
| CN110814342A (en) * | 2019-10-26 | 2020-02-21 | 浙江亚通焊材有限公司 | Preparation method of computer-formalized 3D printing metal material |
| CN111596527A (en) * | 2020-06-19 | 2020-08-28 | 成都路维光电有限公司 | Method for optimizing rotation parameters in mask plate rotation gluing |
| CN112053019A (en) * | 2019-06-06 | 2020-12-08 | 湖南师范大学 | Optical fiber preform deposition process intelligent method based on big data model predictive control framework |
| WO2022141751A1 (en) * | 2020-12-31 | 2022-07-07 | 杭州富加镓业科技有限公司 | High-resistance gallium oxide prediction method and preparation method based on deep learning and bridgman-stockbarger method |
| CN115294671A (en) * | 2022-08-08 | 2022-11-04 | 杭州哲达科技股份有限公司 | Air compressor outlet pressure prediction method and prediction system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103439413B (en) * | 2013-08-07 | 2015-11-18 | 湘潭大学 | A kind of Analysis of Acoustic Emission Signal method that heat barrier coatings damage pattern identifies automatically |
-
2014
- 2014-01-02 CN CN201410001043.1A patent/CN103745271B/en not_active Expired - Fee Related
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109234661A (en) * | 2018-09-21 | 2019-01-18 | 北京联合涂层技术有限公司 | A kind of heat spraying method and system based on artificial neural network |
| CN110047247A (en) * | 2019-05-21 | 2019-07-23 | 武汉理工大学 | A kind of smart home device accurately identifying Falls in Old People |
| CN110165297A (en) * | 2019-05-31 | 2019-08-23 | 重庆市科学技术研究院 | A kind of procedure parameter selection method of film lithium cell solid electrolyte |
| CN112053019B (en) * | 2019-06-06 | 2022-10-25 | 湖南师范大学 | Method for realizing intellectualization of optical fiber preform deposition process |
| CN112053019A (en) * | 2019-06-06 | 2020-12-08 | 湖南师范大学 | Optical fiber preform deposition process intelligent method based on big data model predictive control framework |
| CN110814342B (en) * | 2019-10-26 | 2021-10-29 | 浙江亚通焊材有限公司 | Preparation method of computer-formalized 3D printing metal material |
| CN110814342A (en) * | 2019-10-26 | 2020-02-21 | 浙江亚通焊材有限公司 | Preparation method of computer-formalized 3D printing metal material |
| CN110756803A (en) * | 2019-10-27 | 2020-02-07 | 浙江亚通焊材有限公司 | Preparation method of die steel powder material for computer-formatted 3D printing |
| CN110756803B (en) * | 2019-10-27 | 2021-10-26 | 浙江亚通焊材有限公司 | Preparation method of die steel powder material for computer-formatted 3D printing |
| CN111596527A (en) * | 2020-06-19 | 2020-08-28 | 成都路维光电有限公司 | Method for optimizing rotation parameters in mask plate rotation gluing |
| CN111596527B (en) * | 2020-06-19 | 2023-09-05 | 成都路维光电有限公司 | Rotation parameter optimization method in mask rotary gluing |
| WO2022141751A1 (en) * | 2020-12-31 | 2022-07-07 | 杭州富加镓业科技有限公司 | High-resistance gallium oxide prediction method and preparation method based on deep learning and bridgman-stockbarger method |
| US12026616B2 (en) | 2020-12-31 | 2024-07-02 | Hangzhou Fujia Gallium Technology Co. Ltd. | Preparation method of high resistance gallium oxide based on deep learning and vertical bridgman growth method |
| CN115294671A (en) * | 2022-08-08 | 2022-11-04 | 杭州哲达科技股份有限公司 | Air compressor outlet pressure prediction method and prediction system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103745271B (en) | 2017-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103745271A (en) | Forecasting method for induction thermal deposition calcium-phosphate coating process on basis of neural network | |
| CN105788249B (en) | A kind of traffic flow forecasting method, prediction model generation method and device | |
| CN107909227A (en) | Ultra-short term predicts the method, apparatus and wind power generating set of wind power | |
| CN103942422B (en) | Granular-computation-based long-term prediction method for converter gas holder positions in metallurgy industry | |
| CN108090500A (en) | The flexible measurement method of edible fungus fermented production process key parameters | |
| CN109934422A (en) | Neural network wind speed prediction method based on time series data analysis | |
| CN107782857A (en) | Flexible measurement method based on the edible fungus fermented process key parameter for improving CS BPNN | |
| Wang et al. | Modeling of soft sensor based on DBN-ELM and its application in measurement of nutrient solution composition for soilless culture | |
| CN107341359A (en) | Stalk fermentation produces the flexible measurement method of ethanol process key parameters | |
| CN110414115A (en) | A kind of wavelet neural network tomato yield prediction technique based on genetic algorithm | |
| CN106980874A (en) | A kind of multi-time Scales dimension data fusion method towards distribution big data | |
| CN100470427C (en) | Industrial soft measuring instrument based on bionic intelligence and soft measuring method therefor | |
| CN105568732A (en) | Disc mill control method | |
| CN109408896B (en) | Multi-element intelligent real-time monitoring method for anaerobic sewage treatment gas production | |
| Xu et al. | Soft sensor for ammonia concentration at the ammonia converter outlet based on an improved particle swarm optimization and BP neural network | |
| CN107703760A (en) | Absorbing natural gas tower sweetening process control method based on RBF and GDHP | |
| CN104182854A (en) | Mixed energy consumption measuring method for enterprise energy management system | |
| CN111126827A (en) | Input-output accounting model construction method based on BP artificial neural network | |
| CN107944552A (en) | A kind of industrial Internet of Things parameter prediction method based on Elman neutral nets | |
| CN113536685B (en) | Neural network extrapolation-based wind speed probability model modeling method | |
| CN110348576A (en) | A kind of decision-making technique for improving multiple target bacterial chemotaxis algorithm based on index | |
| Venkateswarlu | Advances in monitoring and state estimation of bioreactors | |
| Hattori et al. | Theoretical and evolutionary parameter tuning of neural oscillators with a double-chain structure for generating rhythmic signals | |
| Shi et al. | Research on monthly flow uncertain reasoning model based on cloud theory | |
| CN103675012A (en) | Industrial melt index soft measurement instrument and method based on BP particle swarm optimization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20160927 Address after: 200444 Baoshan District Road, Shanghai, No. 99 Applicant after: Shanghai University Applicant after: AVIC Commercial Aircraft Engine Co.,Ltd. Address before: 200444 Baoshan District Road, Shanghai, No. 99 Applicant before: Shanghai University |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170510 Termination date: 20200102 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |