CN115796346A - Yield optimization method and system and non-transitory computer readable storage medium - Google Patents
Yield optimization method and system and non-transitory computer readable storage medium Download PDFInfo
- Publication number
- CN115796346A CN115796346A CN202211465631.1A CN202211465631A CN115796346A CN 115796346 A CN115796346 A CN 115796346A CN 202211465631 A CN202211465631 A CN 202211465631A CN 115796346 A CN115796346 A CN 115796346A
- Authority
- CN
- China
- Prior art keywords
- yield
- reaction
- numerical
- parameter
- 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
- 238000005457 optimization Methods 0.000 title claims abstract description 76
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000006243 chemical reaction Methods 0.000 claims abstract description 124
- 238000002474 experimental method Methods 0.000 claims abstract description 48
- 239000000126 substance Substances 0.000 claims abstract description 19
- 230000003993 interaction Effects 0.000 claims abstract description 8
- 238000012795 verification Methods 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 23
- 230000008569 process Effects 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 12
- 238000003775 Density Functional Theory Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 8
- 208000025174 PANDAS Diseases 0.000 claims description 3
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 claims description 3
- 235000016496 Panda oleosa Nutrition 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 240000000220 Panda oleosa Species 0.000 claims 1
- 238000012827 research and development Methods 0.000 abstract description 3
- 239000000758 substrate Substances 0.000 description 8
- 239000003446 ligand Substances 0.000 description 6
- 239000002904 solvent Substances 0.000 description 5
- 230000006872 improvement Effects 0.000 description 4
- -1 concentrations Substances 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 240000004718 Panda Species 0.000 description 2
- 239000003153 chemical reaction reagent Substances 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000012803 optimization experiment Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003054 catalyst Substances 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000013401 experimental design Methods 0.000 description 1
- 238000003050 experimental design method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000376 reactant Substances 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Images
Classifications
-
- 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
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a yield optimization method, a system and a non-transitory computer readable storage medium, wherein the method comprises the following steps: s1, selecting reaction parameters influencing the chemical reaction yield, and respectively coding numerical reaction parameters and non-numerical reaction parameters; s2, carrying out Cartesian product combination operation on values corresponding to all reaction parameters to obtain all sets, namely a reaction parameter space; s3, selecting data in a reaction parameter space, carrying out a chemical reaction experiment to obtain yield data, constructing a model by using a Bayesian optimization algorithm, and recommending an experiment parameter combination of the next round; and S4, carrying out specific experimental verification according to the experimental parameter combination recommended in the step S3, and carrying out multiple rounds of interaction with the algorithm in the step S3 to finally complete yield optimization of the whole chemical reaction so as to obtain a satisfactory yield. The method can effectively improve the experimental efficiency of chemical reaction yield optimization, and brings great convenience to the research and development work of chemical experiments.
Description
Technical Field
The invention relates to a yield optimization method, a yield optimization system and a non-transitory computer readable storage medium, and belongs to the technical field of yield optimization methods.
Background
Currently, optimization of chemical reaction yield is a very challenging task, requiring experts in the chemical field to evaluate various reaction parameters, such as substrates, additives, solvents, concentrations, catalysts, temperature, etc. Due to time and experience-accumulating limitations, experts can only simply evaluate a small fraction of these conditions during a standard optimization procedure. It is also very challenging to achieve a higher yield. Although thousands of experiments can be completed in a short time by means of high-throughput experiments, the cost consumption is very large. In addition to this, an empirical accumulation of similar reactions and an understanding of the chemical mechanisms of the reactants can play an important role in the optimization of the yield by consulting the relevant literature, but this also places very high professional demands on the relevant experimenters.
In order to solve these problems, a chemical expert usually performs a control experiment by using some experimental design methods, for example, designing an experimental scheme by using DOE experimental design ideas, but there are some problems of optimization design and experience priority selection. Yield optimization essentially belongs to parameter optimization, so a Bayesian optimization algorithm in machine learning can be adopted. The algorithm aims to balance the exploration of uncertainty fields and the utilization of available information, thereby achieving high quality configuration in fewer evaluations. Meanwhile, the algorithm supports parallel calculation of a plurality of experiments, which means that a plurality of groups of experiments can be performed in one round of optimization experiment, one parallel requirement of chemical reaction is met, and meanwhile, the requirement on professional literacy of testers and the cost required by the experiments are reduced. Therefore, the Bayesian optimization algorithm is applied to the project of chemical reaction yield optimization, and the problems of the traditional optimization method are solved.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides an interactive yield optimization method based on a Bayesian optimization algorithm, which can effectively improve the experimental efficiency of chemical reaction yield optimization and bring great convenience to the research and development work of chemical experiments.
The technical scheme for solving the technical problems is as follows: a method for yield optimization, said method comprising the steps of:
s1, pretreatment of experimental data:
selecting reaction parameters (including substrate, solvent, ligand, temperature, reagent and the like) influencing the chemical reaction yield, dividing the reaction parameters into a numerical type and a non-numerical type, and then respectively coding the numerical type and the non-numerical type reaction parameters; for example, temperature, humidity, etc. are numerical reaction parameters, and the solvent type is non-numerical reaction parameters.
S2, constructing a reaction parameter space:
carrying out Cartesian product combination operation on values corresponding to all reaction parameters to obtain all sets, namely a reaction parameter space; for example: the substrates are a, b and c; the ligand is A, B, C, and the parameter space is that all combinations can be obtained by carrying out Cartesian product combination operation on the substrate and the ligand.
S3, constructing and optimizing model
Selecting data in a reaction parameter space, carrying out a chemical reaction experiment to obtain yield data, constructing a model by using a Bayesian optimization algorithm, recommending a new round of experiment parameter combination, and constructing an optimization model by using an experiment result obtained by the previous round of experiment parameter combination as prior knowledge and recommending the next round of experiment parameter combination;
and S4, carrying out specific experimental verification according to the experimental parameter combination recommended in the step S3, and carrying out multiple rounds of interaction with the algorithm in the step S3 to finally complete yield optimization of the whole chemical reaction so as to obtain a satisfactory yield.
Further, in step S1, the numerical reaction parameter encoding method is: the experimenter sets the upper and lower limits of the numerical range according to the numerical parameters, and simultaneously defines the segmentation scale of each numerical reaction parameter.
Further, in step S1, the non-numerical reaction parameter encoding manner is: and inputting SMILES codes corresponding to each non-numerical reaction parameter, and combining one or more algorithms of one-hot algorithm, density functional theory algorithm (DFT) and mordred descriptor to obtain the coded value of the non-numerical reaction parameter.
Further, in step S2, when constructing the reaction parameter space, firstly, redundant information generated when encoding non-numerical reaction parameters, such as file names, mean values, variances, and other redundant information generated by DFT algorithm, is deleted;
then carrying out Cartesian product combination operation; the reaction parameter names in the reaction parameter space are then aligned with the encoded values using the pandas data processing tool.
Further, in step S3, a bayesian optimization algorithm includes a gaussian process and a selection function, and the bayesian optimization algorithm constructs a chemical reaction parameter model;
constructing a prior knowledge model for the Bayesian optimization algorithm through a Gaussian process, updating an original data set by using the combination of the reaction parameters recommended in the previous round as new data in the Bayesian optimization algorithm in the algorithm interaction process, and updating the prior knowledge model in the Gaussian process;
selecting a new round of sampling points by a selection function through a priori knowledge model, and calculating posterior distribution through the sampling points; at the same time, the selection of the optimal sampling point in the current region and the sampling point of the unknown region is balanced.
Further, the gaussian process formula is as follows:
f(x)～gp(m(x),k(x，x′))
wherein m (x) is a mean function, k (x, x ') is a covariance function, and x' are two sets of input parameters;
the selection function in the Bayesian optimization algorithm comprises the following steps: upper Confidence Bound (UCB) and Expected Improvement (EI).
Further, the Expected improvement selection function is:
wherein,
phi (z) and phi (z) are respectively a Gaussian distribution cumulative probability function and a probability density function, f (x +) represents the existing maximum value, and x + enables f (x +) to obtain the maximum value as an input parameter
The Upper confidence bound selection function is:
UCB(x)＝μ(x)+kσ(x)
wherein μ (x) is the mean; σ (x) is the standard deviation; k is an adjustment coefficient.
Further, the Bayesian optimization algorithm flow is as follows:
firstly, initializing the chemical reaction parameter model, and randomly selecting five groups of reaction parameter data from the reaction parameter space obtained in the step S2 as Gaussian process regression training data;
carrying out specific chemical experiments by experimenters according to the initialized reaction parameter data, aligning the yield with each group of reaction parameter data after obtaining the yield, updating a data set, and establishing a priori knowledge model according to the updated data set in the Gaussian process;
then, selecting a function to calculate a new batch of 4-6 groups of reaction parameter combinations through a priori knowledge model, and carrying out specific chemical experiments by experimenters to obtain the yield of each group of reaction parameter experiments;
finally, the experimenter judges whether the current requirement is met according to the obtained yield data, and if the current requirement is met, the experiment is stopped; if the yield does not meet the requirement of the experimenter, updating the reaction parameters and the yield data of the specific chemical experiment into a data set, carrying out model optimization and calculation again, recommending a new batch of reaction parameter combinations, and repeating the experiment steps until the yield meets the requirement of the experimenter.
The invention also discloses a yield optimization system, which at least comprises a central processing unit (cpu) and a memory in communication connection with the cpu, wherein the cpu can execute the program of the yield optimization method, and the memory can store the program instruction called and executed by the cpu and the relevant parameter model.
Also disclosed is a non-transitory computer readable storage medium storing computer instructions that cause a computer to perform the yield optimization method.
The invention has the beneficial effects that:
the invention provides a yield optimization method, which is characterized in that by means of a Bayesian optimization algorithm and an interactive experimental means, a satisfactory reaction yield is obtained for experimenters finally. The experimental efficiency of chemical reaction yield optimization can be effectively improved, and great convenience is brought to the research and development work of chemical experiments.
The method is an interactive yield optimization method based on a Bayesian optimization algorithm, supports parallel calculation of a plurality of experiments, can perform a plurality of groups of experiments in one round of optimization experiment, meets a parallel requirement of chemical reaction, reduces professional literacy requirements on testers, and reduces the cost required by the experiments.
The yield optimization method can realize the processing calculation of numerical reaction parameters, can also realize the processing calculation of non-numerical reaction parameters, and realizes the comprehensive processing of the numerical reaction parameters and the non-numerical reaction parameters, thereby providing comprehensive and comprehensive yield optimization conditions for experimenters, greatly improving the yield optimization efficiency of chemical experiments, and reducing the labor intensity of the experimenters.
Drawings
FIG. 1 is a detailed flowchart of the Bayesian optimization algorithm in an embodiment;
FIG. 2 is a flow chart of experimental interactions in an example.
Detailed Description
The present invention will be described in detail with reference to the following embodiments in order to make the aforementioned objects, features and advantages of the invention more comprehensible. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
A method for yield optimization, said method comprising the steps of:
s1, pretreatment of experimental data:
reaction parameters (including substrates, solvents, ligands, temperature, reagents and the like) influencing the yield of the chemical reaction are selected, the reaction parameters are divided into a numerical type and a non-numerical type, and then the numerical type and the non-numerical type are respectively coded. For example, temperature, humidity, etc. are numerical reaction parameters, and the solvent type is non-numerical reaction parameters.
The numerical reaction parameter coding mode is as follows: the experiment personnel sets the upper limit and the lower limit of a numerical range according to the numerical parameters and simultaneously demarcates the division scale of each numerical reaction parameter. For example, temperature, humidity and the like belong to numerical reaction parameters, each parameter of temperature and humidity needs to define a division scale, for example, the upper and lower limits of temperature define a range of 10-20 ℃, the temperature can be divided into [10, 12, 14, 16, 18, 20] by experimenters, aiming at finding out the specific temperature scale when the optimal state is reached in the experiment, and the numerical reaction parameters (temperature) are coded as follows:
temperature [10, 12, 14, 16, 18, 20]
The non-numerical reaction parameter coding mode is as follows: and inputting SMILES codes corresponding to each non-numerical reaction parameter, and combining one or more algorithms of one-hot algorithm, density functional theory algorithm (DFT) and mordred descriptor to obtain the coded value of the non-numerical reaction parameter.
The one-hot algorithm, the density functional theory algorithm (DFT) and the mordred descriptor may be used in combination or separately, and the manner of use depends on an empirical evaluation of the experiment by an experimenter and which encoding the compound represented by the SMILES code can be performed.
When the calculation force is enough, a DFT coding mode can be selected preferably, and enough chemical mechanisms can be obtained as characteristics; the mordred algorithm may extract chemical mechanisms from SMILES as features; the one-hot algorithm extracts features from a purely mathematical perspective and these features are used for modeling.
S2, constructing a reaction parameter space:
carrying out Cartesian product combination operation on values corresponding to all reaction parameters to obtain all sets, namely a reaction parameter space; for example: the substrates are a, b and c; the ligand is A, B, C, and the parameter space is that all combinations can be obtained by carrying out Cartesian product combination operation on the substrate and the ligand.
Such as non-numerical reaction parameters: the substrates are a, b and c; numerical reaction parameters: the temperature is 15 to 20; the parameter space contains the combinations: [a 15] And [ a 20], [ b 15], [ b 20], [ c 15], [ c 20], that is, a cartesian combination.
When a reaction parameter space is constructed, firstly, redundant information generated during non-numerical reaction parameter coding, such as file names, mean values, variances and other redundant information generated by a DFT algorithm, is deleted; then carrying out Cartesian product combination operation; the reaction parameter names in the reaction parameter space are then aligned with the encoded values using the pandas data processing tool.
S3, constructing and optimizing model
Constructing a model by using a Bayesian optimization algorithm, and then constructing an optimization model by using an experimental result obtained by the previous round of experimental parameter combination as prior knowledge and recommending the next round of experimental parameter combination;
the Bayesian optimization algorithm comprises a Gaussian process and a selection function, and the Bayesian optimization algorithm constructs a chemical reaction parameter model f (x) x → R, wherein x is a group of reaction parameter combinations in a reaction parameter space;
constructing a prior knowledge model for the Bayesian optimization algorithm through a Gaussian process, updating an original data set by using the combination of the reaction parameters recommended in the previous round as new data in the Bayesian optimization algorithm in the algorithm interaction process, and updating the prior knowledge model in the Gaussian process;
selecting a new round of sampling points by a selection function through a priori knowledge model, and calculating posterior distribution through the sampling points; at the same time, the selection of the optimal sampling point in the current region and the sampling point of the unknown region is balanced.
Further, the gaussian process equation is as follows:
f(x)～gp(m(x),k(x,x′))
wherein m (x) is a mean function and k (x, x') is a covariance function;
the selection function in the Bayesian optimization algorithm comprises the following steps: upper Confidence Bound (UCB) and Expected Improvement (EI).
Further, the Expected improvement selection function is:
wherein,
phi (z) and phi (z) are respectively a Gaussian distribution cumulative probability function and a probability density function, f (x +) represents the existing maximum value, and x + enables f (x +) to obtain the maximum value as an input parameter
The Upper confidence bound selection function is:
UCB(x)＝μ(x)+kσ(x)
wherein μ (x) is the mean; σ (x) is the standard deviation; k is an adjustment parameter.
And S4, carrying out specific experimental verification according to the experimental parameter combination recommended in the step S3, and carrying out multiple rounds of interaction with the algorithm in the step S3 to finally complete yield optimization of the whole chemical reaction so as to obtain a satisfactory yield.
In this embodiment, the bayesian optimization algorithm flow is as follows:
s3-1, initializing the chemical reaction parameter model, and randomly selecting five groups of reaction parameter data from the reaction parameter space obtained in the step S2 as Gaussian process regression training data;
s3-2, carrying out a specific chemical experiment by an experimenter according to the initialized reaction parameter data, aligning the yield with each group of reaction parameter data after obtaining the yield, updating a data set, and establishing a priori knowledge model according to the updated data set in the Gaussian process;
s3-3, calculating a new batch of five groups of reaction parameter combinations by a selection function through a priori knowledge model, deleting the calculated five groups of reaction parameter combinations by experimenters according to experience knowledge in the field, or adding a plurality of groups of reaction parameter combinations according to experience to determine the reaction parameter combinations which finally need to carry out specific chemical experiments, and then carrying out specific chemical experiments by the experimenters to obtain the yield of each group of reaction parameter experiments;
s3-4, judging whether the current requirement is met or not by experimenters according to the obtained yield data, and stopping the experiment if the current requirement is met; if the yield does not meet the requirement, the reaction parameters and the yield data of the specific chemical experiment performed in the S3-3 are updated to the data set, model optimization and calculation are performed again, a new batch of five groups of reaction parameter combinations are calculated, and the experiment steps can be repeated by combining the experience of professional experimenters until the yield meets the requirement of the experimenters, as shown in figure 2.
A yield optimization system, the system at least has a Central Processing Unit (CPU), and a memory connected with the CPU for communication, the CPU can execute the program of the yield optimization method, the memory can store the program instructions called by the CPU to execute and the relevant parameter models.
A non-transitory computer readable storage medium storing computer instructions that cause a computer to perform the yield optimization method.
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is specific and detailed, but not to be understood as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (8)
1. A method for optimizing yield, comprising the steps of:
s1, pretreatment of experimental data:
selecting reaction parameters influencing the chemical reaction yield, dividing the reaction parameters into a numerical type and a non-numerical type, and then respectively coding the numerical type reaction parameters and the non-numerical type reaction parameters;
s2, constructing a reaction parameter space:
carrying out Cartesian product combination operation on values corresponding to all reaction parameters to obtain all sets, namely a reaction parameter space;
s3, constructing and optimizing model
Selecting data in a reaction parameter space, carrying out a chemical reaction experiment to obtain yield data, constructing a model by using a Bayesian optimization algorithm, recommending a new round of experiment parameter combination, and constructing an optimization model by using an experiment result obtained by the previous round of experiment parameter combination as prior knowledge and recommending the next round of experiment parameter combination;
and S4, carrying out specific experimental verification according to the experimental parameter combination recommended in the step S3, and carrying out multiple rounds of interaction with the algorithm in the step S3 to finally complete yield optimization of the whole chemical reaction so as to obtain a satisfactory yield.
2. The yield optimization method according to claim 1, wherein in step S1, the numerical reaction parameters are encoded by: the experiment personnel sets the upper limit and the lower limit of a numerical range according to the numerical parameters and simultaneously demarcates the division scale of each numerical reaction parameter.
3. The yield optimization method according to claim 1, wherein in step S1, the non-numerical reaction parameters are encoded by: and inputting the SMILES code corresponding to each non-numerical reaction parameter, and combining one-hot algorithm, density functional theory algorithm and mordred descriptor to obtain the coded value of the non-numerical reaction parameter.
4. The yield optimization method according to claim 3, wherein in step S2, when constructing the reaction parameter space, redundant information generated during encoding of non-numerical reaction parameters is first deleted; then carrying out Cartesian product combination operation; the reaction parameter names in the reaction parameter space are then aligned with the encoded values using the pandas data processing tool.
5. The yield optimization method according to claim 1, wherein in step S3, the bayesian optimization algorithm comprises a gaussian process and a selection function, and the bayesian optimization algorithm constructs a chemical reaction parameter model;
constructing a prior knowledge model for the Bayesian optimization algorithm through a Gaussian process, updating an original data set by using the combination of the reaction parameters recommended in the previous round as new data in the Bayesian optimization algorithm in the algorithm interaction process, and updating the prior knowledge model in the Gaussian process;
selecting a new round of sampling points by the selection function through a priori knowledge model, and calculating posterior distribution through the sampling points; at the same time, the selection of the optimal sampling point in the current region and the sampling point of the unknown region is balanced.
6. The yield optimization method according to claim 5, wherein the Bayesian optimization algorithm flow is as follows:
firstly, initializing the chemical reaction parameter model, and randomly selecting five groups of reaction parameter data from the reaction parameter space obtained in the step S2 as Gaussian process regression training data;
carrying out specific chemical experiments by experimenters according to the initialized reaction parameter data, aligning the yield with each group of reaction parameter data after obtaining the yield, updating a data set, and establishing a priori knowledge model according to the updated data set in the Gaussian process;
then, selecting a function, calculating a new batch of 4-6 groups of reaction parameter combinations through a priori knowledge model, and carrying out specific chemical experiments by experimenters to obtain the yield of each group of reaction parameter experiments;
finally, the experimenter judges whether the current requirement is met according to the obtained yield data, and if the current requirement is met, the experiment is stopped; if the yield does not meet the requirement of the experimenter, updating the reaction parameters and the yield data of the specific chemical experiment into a data set, carrying out model optimization and calculation again, recommending a new batch of reaction parameter combinations, and repeating the experiment steps until the yield meets the requirement of the experimenter.
7. A yield optimization system, characterized in that the system comprises at least a central processing unit, and a memory communicatively connected to the central processing unit, the central processing unit being capable of executing a program of the yield optimization method according to any one of claims 1 to 6, the memory being capable of storing program instructions to be invoked and executed by the central processing unit and associated parametric models.
8. A non-transitory computer readable storage medium storing computer instructions for causing a computer to perform the yield optimization method of any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211465631.1A CN115796346B (en) | 2022-11-22 | 2022-11-22 | Yield optimization method, system and non-transitory computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211465631.1A CN115796346B (en) | 2022-11-22 | 2022-11-22 | Yield optimization method, system and non-transitory computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115796346A true CN115796346A (en) | 2023-03-14 |
| CN115796346B CN115796346B (en) | 2023-07-21 |
Family
ID=85440012
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211465631.1A Active CN115796346B (en) | 2022-11-22 | 2022-11-22 | Yield optimization method, system and non-transitory computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115796346B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020087281A1 (en) * | 2018-10-30 | 2020-05-07 | 深圳市大疆创新科技有限公司 | Hyper-parameter optimization method and apparatus |
| CN111898212A (en) * | 2020-06-17 | 2020-11-06 | 西安交通大学 | Impeller mechanical profile design optimization method based on BezierGAN and Bayesian optimization |
| WO2021007812A1 (en) * | 2019-07-17 | 2021-01-21 | 深圳大学 | Deep neural network hyperparameter optimization method, electronic device and storage medium |
| GB202100758D0 (en) * | 2021-01-20 | 2021-03-03 | Jems Energetska Druzba D O O | Systems and methods for plant process optimisation |
| WO2022058113A1 (en) * | 2020-09-18 | 2022-03-24 | Bystronic Laser Ag | Computer implemented method of and optimisation tool for refinement of laser cutting process parameters by means of an optimization tool |
| CN114781280A (en) * | 2022-06-21 | 2022-07-22 | 烟台国工智能科技有限公司 | Interactive yield optimization method and system based on recurrent neural network |
| CN114912342A (en) * | 2022-03-28 | 2022-08-16 | 南京邮电大学 | Packaging lead bonding process parameter optimization method based on multiple quality parameters |
| CN115129982A (en) * | 2022-06-22 | 2022-09-30 | 电计科技发展(上海)有限公司 | Experiment parameter recommendation method, device, terminal and medium based on improved Bayesian optimization |
-
2022
- 2022-11-22 CN CN202211465631.1A patent/CN115796346B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020087281A1 (en) * | 2018-10-30 | 2020-05-07 | 深圳市大疆创新科技有限公司 | Hyper-parameter optimization method and apparatus |
| WO2021007812A1 (en) * | 2019-07-17 | 2021-01-21 | 深圳大学 | Deep neural network hyperparameter optimization method, electronic device and storage medium |
| CN111898212A (en) * | 2020-06-17 | 2020-11-06 | 西安交通大学 | Impeller mechanical profile design optimization method based on BezierGAN and Bayesian optimization |
| WO2022058113A1 (en) * | 2020-09-18 | 2022-03-24 | Bystronic Laser Ag | Computer implemented method of and optimisation tool for refinement of laser cutting process parameters by means of an optimization tool |
| GB202100758D0 (en) * | 2021-01-20 | 2021-03-03 | Jems Energetska Druzba D O O | Systems and methods for plant process optimisation |
| CN114912342A (en) * | 2022-03-28 | 2022-08-16 | 南京邮电大学 | Packaging lead bonding process parameter optimization method based on multiple quality parameters |
| CN114781280A (en) * | 2022-06-21 | 2022-07-22 | 烟台国工智能科技有限公司 | Interactive yield optimization method and system based on recurrent neural network |
| CN115129982A (en) * | 2022-06-22 | 2022-09-30 | 电计科技发展(上海)有限公司 | Experiment parameter recommendation method, device, terminal and medium based on improved Bayesian optimization |
Non-Patent Citations (1)
| Title |
|---|
| 何龙: "《深入理解XGBoost 高效机器学习算法与进阶》", 机械工业出版社, pages: 313 - 314 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115796346B (en) | 2023-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Astakhova et al. | Forecasting method for grouped time series with the use of k-means algorithm | |
| CN113361680A (en) | Neural network architecture searching method, device, equipment and medium | |
| CN103077253A (en) | High-dimensional mass data GMM (Gaussian Mixture Model) clustering method under Hadoop framework | |
| CN114186749A (en) | Flexible workshop scheduling method and model based on reinforcement learning and genetic algorithm | |
| CN113343427B (en) | Structural topology configuration prediction method based on convolutional neural network | |
| CN107516148B (en) | System modeling optimization method and storage medium | |
| CN112990515A (en) | Workshop resource scheduling method based on heuristic optimization algorithm | |
| CN114580678A (en) | Product maintenance resource scheduling method and system | |
| Chen et al. | A GA-based search method for the tolerance allocation problem | |
| CN114519429B (en) | Method, device and medium for obtaining observability quantity of target system | |
| Zhang et al. | A robust MILP and gene expression programming based on heuristic rules for mixed-model multi-manned assembly line balancing | |
| CN113821983A (en) | Engineering design optimization method and device based on proxy model and electronic equipment | |
| CN111785326B (en) | Gene expression profile prediction method after drug action based on generation of antagonism network | |
| CN113935235A (en) | Engineering design optimization method and device based on genetic algorithm and agent model | |
| CN112669899A (en) | 16S and metagenome sequencing data correlation analysis method, system and equipment | |
| CN118136141A (en) | Synthetic route development method and device for comprehensive inverse synthesis, condition and reaction prediction | |
| CN118070609A (en) | Simulation prediction method, simulation prediction system, simulation prediction device and computer readable storage medium based on finite element parametric modeling and point cloud deep learning | |
| CN115796346B (en) | Yield optimization method, system and non-transitory computer readable storage medium | |
| CN116520787A (en) | Production line optimization method and system and electronic equipment | |
| Chang et al. | An experimental study of approximation algorithms for the joint spectral radius | |
| JP2012093976A (en) | Clustering system, method and program | |
| CN115344386A (en) | Method, device and equipment for predicting cloud simulation computing resources based on sequencing learning | |
| CN115391156A (en) | Cross-domain collaborative training task calculation efficiency simulation method and related equipment | |
| CN114926075A (en) | Mechanical part production scheduling method based on man-hour prediction | |
| Serdyukov et al. | Using genetic algorithm for generating optimal data sets to automatic testing the program code |
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 |