CN117236198A - Machine learning solving method of flame propagation model of blasting under sparse barrier - Google Patents
Machine learning solving method of flame propagation model of blasting under sparse barrier Download PDFInfo
- Publication number
- CN117236198A CN117236198A CN202311507623.3A CN202311507623A CN117236198A CN 117236198 A CN117236198 A CN 117236198A CN 202311507623 A CN202311507623 A CN 202311507623A CN 117236198 A CN117236198 A CN 117236198A
- Authority
- CN
- China
- Prior art keywords
- explosion
- model
- flame propagation
- propagation model
- parameters
- 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
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000010801 machine learning Methods 0.000 title claims abstract description 18
- 230000004888 barrier function Effects 0.000 title claims abstract description 14
- 238000005422 blasting Methods 0.000 title description 2
- 238000004880 explosion Methods 0.000 claims abstract description 66
- 238000002474 experimental method Methods 0.000 claims abstract description 28
- 239000012530 fluid Substances 0.000 claims abstract description 27
- 230000000694 effects Effects 0.000 claims abstract description 25
- 238000004088 simulation Methods 0.000 claims abstract description 15
- 238000012544 monitoring process Methods 0.000 claims abstract description 12
- 238000005070 sampling Methods 0.000 claims abstract description 12
- 230000007246 mechanism Effects 0.000 claims abstract description 11
- 238000009826 distribution Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims abstract description 6
- 230000003449 preventive effect Effects 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 14
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000009828 non-uniform distribution Methods 0.000 claims description 8
- 238000010206 sensitivity analysis Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 4
- 238000000540 analysis of variance Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 5
- 230000009286 beneficial effect Effects 0.000 abstract description 4
- 230000002265 prevention Effects 0.000 abstract description 2
- 230000006698 induction Effects 0.000 abstract 1
- 239000007789 gas Substances 0.000 description 29
- 239000013598 vector Substances 0.000 description 20
- 239000000126 substance Substances 0.000 description 5
- 238000002485 combustion reaction Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a machine learning solving method of a flame propagation model of a flame explosion under a sparse barrier, and belongs to the technical field of flame explosion accident prevention and control. The technical proposal is as follows: firstly, constructing a corresponding gas explosion experiment system and acquiring explosion pressure monitoring point experiment data; then, an OpenFOAM is utilized to construct an explosion model, latin sampling is carried out on turbulence parameters representing fluid instability and obstacle induction in the k-epsilon turbulence model, and simulation is carried out to form a simulation training data set; then, a turbulent flame propagation model is obtained and embedded into a control equation and the turbulent flame propagation model; finally, under sparse distribution of obstacles, the representation of a multi-space-time evolution mechanism and overpressure characteristics of oil gas explosion is realized, the turbulence effect jointly induced by fluid instability and the obstacles is analyzed, and the establishment of preventive measures is guided. The beneficial effects are that: the method can effectively solve the evolution process and the size range of parameters such as pressure, flame and the like of gas explosion, and ensure the safe operation of petrochemical enterprises.
Description
Technical Field
The invention belongs to the technical field of explosion accident prevention and control under sparse distribution of obstacles, and particularly relates to a machine learning solving method of an explosion flame propagation model under sparse obstacles.
Background
In industrial processes, particularly in chemical and petrochemical production, combustible gases are widely used. The gas explosion accident can occur when the ignition source is encountered because the device equipment is in an explosion limit due to the defect or human factor that the device is mixed with air to form a combustible gas mixture. Once such an explosion accident occurs, serious property loss and casualties may result. In addition, for chemical industry and petrochemical enterprises, due to the fact that the equipment is numerous, the distribution condition is complex, and after accidents occur, the results are more difficult to analyze.
Modeling the target scenario combustion explosion given scenario-related variables (e.g., ignition location, gas concentration, gas volume, etc.) is critical to supporting real-time emergency response planning and decision-making processes. Ideally, comprehensive experimental testing is the most reliable method, however, this method is difficult to achieve due to its high cost and high risk. Computational Fluid Dynamics (CFD) has been widely used to accurately simulate overpressure conditions generated after a combustion explosion of various gases in different spaces. But for the scene of space, complex structure and uneven equipment layout of chemical and petrochemical enterprises, the evolution process of flame and pressure in the explosion accident needs to be further analyzed.
In view of the dangerousness and environmental destructiveness of gas explosion accidents of chemical and petrochemical enterprises, the traditional experimental method has the defects of high cost and high dangerousness. At present, chemical and petrochemical enterprises are urgent to develop researches on the numerical solution of the spatial and temporal evolution of the explosion of dangerous gas, realize the advanced simulation prediction of the overpressure and flame evolution of the explosion, and provide technical support for the emergency treatment and disaster reduction and explosion resistance of the explosion.
Disclosure of Invention
In order to solve the technical problems, the invention provides a machine learning solving method of a flame propagation model of a flame explosion under a sparse barrier.
The machine learning solving method of the flame propagation model of the explosion under the sparse barrier is characterized by comprising the following steps of:
s1, constructing a gas explosion experiment system under non-uniform distribution of obstacles, and acquiring experiment data of explosion pressure monitoring points;
s2, constructing an explosion model identical to the gas explosion experimental system by using OpenFOAM, sampling Latin for parameters of fluid instability and turbulence effect, and performing simulation to form a simulation training data set;
s3, utilizing machine learning to carry out inverse solution on parameters of fluid instability and turbulence effect, thereby solving a turbulent flame propagation model, and embedding the turbulent flame propagation model into a control equation and the turbulent model;
and S4, under the sparse distribution of the obstacles, the representation of a multi-space-time evolution mechanism and overpressure characteristics of oil gas explosion is realized, the turbulence effect jointly induced by fluid instability and the obstacles is analyzed, and the establishment of preventive measures is guided.
Further, the step of acquiring the experimental data of the explosion pressure monitoring point in the step S1 is as follows:
s11: the gas explosion experiment system under the non-uniform distribution of the barriers is formed by an experiment model, a dangerous gas charging system, an ignition system and a data collection system, and a plurality of pressure detection points are arranged at different positions in the gas explosion experiment system;
s12: and carrying out an explosion experiment on pressure monitoring points which are arranged in advance, and acquiring pressure data of different positions of the model, namely the experimental data.
Further, in the step S2, an OpenFOAM is used to construct an explosion model identical to the gas explosion experiment system, and the specific steps of latin sampling the parameters of the fluid instability and the turbulence effect are as follows:
s21: carrying out model construction and grid division by using a snappyhexmesh function in CAD and OpenFOAM, and carrying out grid sensitivity analysis;
s22: performing ANOVA analysis on parameters of fluid instability and turbulence effect in a turbulence model to obtain parameters with high importance, wherein the turbulence model is expressed as follows:
wherein,is an empirical parameter->Is the Reynolds average density, +.>Is time, & lt>Is->Mass average speed of component, +.>Is turbulent energy, +.>Is the dissipation of turbulent energy, +.>Is->Turbulent planter of>Is->Is used for the number of plannches,is turbulent viscosity->Is Reynolds stress->Is a friction factor;
s23: and defining upper and lower range boundaries for the parameters of the fluid instability and the turbulence effect, and sampling Latin from the boundary ranges to form a simulation combination.
Further, the specific steps of S3 are as follows:
s31: automatic differential variation reasoning is implemented using python, as follows:
posterior density of Bayesian reasoningParameterizing and taking the difference between the approximate class distribution and the true posterior +.>Minimization, can be expressed as:
in the method, in the process of the invention,is called the variation density, is a parameterized density, consisting of +.>Parameterization, should be easy to sample and evaluate;KLthe (Kullback-Leibler) divergence is a measure of the "distance" between two densities.
Subsequently, it will(lower evidence limit) maximization to replace +.>Can be expressed as:
then the expression (3) can be expressed as:
subsequently, a fixed variational approximation is introduced, defining a one-to-one, differentiable function:
and identifies the converted variable asζ=T(θ)；
In automatic differential variational, it is assumed that all model parameters are continuous and willGradient to obtain:
placing the gradient within the desired value, yielding:
wherein the method comprises the steps ofExtraction from standard normal mapz，（μ，σ) Is the variation mean and standard deviation.
Finally, some gradient descent is performed to obtain a solution to the pair. Automatic differentiation can be used to calculate the gradient of the desired interior. Random gradient ascent was used to automatically change the score reasoning.
S32: attention mechanisms are introduced into fully connected neural networks, as follows:
for input vectorsx=[x 1 ,x 2 ,…,x n ] T It passes through the hidden layer to obtain linear output vectorz=[z 1 ,z 2 ,…,z n ] T Mainly by weight vectorsWAnd offset vectorTo determine, expressed as:
after obtaining the linear output vector, converting the output vector by using a nonlinear function to obtain the output vector of the hidden layerOr the output vector of the output layer->In neural networks, this nonlinear function is also called an activation function, which takes the form:
for all input vectorsPerforming similarity operation to obtain attention valueH=/>The attention value is normalized by using a SoftMax function to obtain an attention weight coefficient, and the formula is as follows:
and carrying out weighted summation on all the input Attention values and the corresponding weight coefficients to obtain an output vector Attention (h), wherein the formula is as follows:
s33: an inverse parametric solution model of fluid instability and turbulence effects was constructed using fully connected neural networks that draw attention mechanisms and automatic differential variational reasoning.
The beneficial effects are that: the invention provides a machine learning solving method of a flame propagation model of a flame explosion under a sparse barrier. The method can simulate the explosion evolution of the dangerous gas of the offshore platform, has good generalization capability, and is beneficial to more comprehensively and reasonably predicting the dangerous area covered by the gas explosion of the platform; meanwhile, on the basis of the traditional CFD model, the inverse solving work is carried out, so that the evolution condition of flame and pressure can be accurately solved, and the method has good application potential. In general, the method for constructing the combustion model under the non-uniform obstacle by machine learning is a suitable alternative scheme for constructing an emergency management digital twin system, is beneficial to safe operation of an offshore platform, and effectively guarantees safe operation of petrochemical enterprises.
Detailed Description
The present invention will be described in further detail with reference to the following examples in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the particular embodiments described herein are illustrative only and are not intended to limit the invention, i.e., the embodiments described are merely some, but not all, of the embodiments of the invention.
Example 1
The machine learning solving method of the flame propagation model of the explosion under the sparse barrier is characterized by comprising the following steps of:
s1, constructing a gas explosion experiment system under non-uniform distribution of obstacles, and acquiring experiment data of explosion pressure monitoring points;
s2, constructing an explosion model identical to the gas explosion experimental system by using OpenFOAM, sampling Latin for parameters of fluid instability and turbulence effect, and performing simulation to form a simulation training data set;
s3, utilizing machine learning to carry out inverse solution on parameters of fluid instability and turbulence effect, thereby solving a turbulent flame propagation model, and embedding the turbulent flame propagation model into a control equation and the turbulent model;
and S4, under the sparse distribution of the obstacles, the representation of a multi-space-time evolution mechanism and overpressure characteristics of oil gas explosion is realized, the turbulence effect jointly induced by fluid instability and the obstacles is analyzed, and the establishment of preventive measures is guided.
Further, the step of acquiring the experimental data of the explosion pressure monitoring point in the step S1 is as follows:
s11: the gas explosion experiment system under the non-uniform distribution of the barriers is formed by an experiment model, a dangerous gas charging system, an ignition system and a data collection system, and a plurality of pressure detection points are arranged at different positions in the gas explosion experiment system;
s12: and carrying out an explosion experiment on pressure monitoring points which are arranged in advance, and acquiring pressure data of different positions of the model, namely the experimental data.
Further, in the step S2, an OpenFOAM is used to construct an explosion model identical to the gas explosion experiment system, and the specific steps of latin sampling the parameters of the fluid instability and the turbulence effect are as follows:
s21: carrying out model construction and grid division by using a snappyhexmesh function in CAD and OpenFOAM, and carrying out grid sensitivity analysis;
s22: performing ANOVA analysis on parameters of fluid instability and turbulence effect in a turbulence model to obtain parameters with high importance, wherein the turbulence model is expressed as follows:
wherein,is an empirical parameter->Is the Reynolds average density, +.>Is time, & lt>Is->Mass average speed of component, +.>Is turbulent energy, +.>Is the dissipation of turbulent energy, +.>Is->Turbulent planter of>Is->Is used for the number of plannches,is the viscosity of the turbulent flow,/>is Reynolds stress->Is a friction factor;
s23: and defining upper and lower range boundaries for the parameters of the fluid instability and the turbulence effect, and sampling Latin from the boundary ranges to form a simulation combination.
Further, the specific steps of S3 are as follows:
s31: automatic differential variation reasoning is implemented using python, as follows:
posterior density of Bayesian reasoningParameterizing and taking the difference between the approximate class distribution and the true posterior +.>Minimization, can be expressed as:
in the method, in the process of the invention,is called the variation density, is a parameterized density, consisting of +.>Parameterization, should be easy to sample and evaluate;KLthe (Kullback-Leibler) divergence is a measure of the "distance" between two densities.
Subsequently, it will(lower evidence limit) maximization to replace +.>Can be expressed as:
then the expression (3) can be expressed as:
subsequently, a fixed variational approximation is introduced, defining a one-to-one, differentiable function:
and identifies the converted variable asζ=T(θ)；
In automatic differential variational, it is assumed that all model parameters are continuous and willGradient to obtain:
placing the gradient within the desired value, yielding:
wherein the method comprises the steps ofExtraction from standard normal mapz，（μ，σ) Is the variation mean and standard deviation.
Finally, some gradient descent is performed to obtain a solution to the pair. Automatic differentiation can be used to calculate the gradient of the desired interior. Random gradient ascent was used to automatically change the score reasoning.
S32: attention mechanisms are introduced into fully connected neural networks, as follows:
for input vectorsx=[x 1 ,x 2 ,…,x n ] T It passes through the hidden layer to obtain linear output vectorz=[z 1 ,z 2 ,…,z n ] T Mainly by weight vectorsWAnd offset vectorTo determine, expressed as:
after obtaining the linear output vector, converting the output vector by using a nonlinear function to obtain the output vector of the hidden layerOr the output vector of the output layer->In neural networks, this nonlinear function is also called an activation function, which takes the form:
for all input vectorsPerforming similarity operation to obtain attention valueH=/>The attention value is normalized by using a SoftMax function to obtain an attention weight coefficient, and the formula is as follows:
and carrying out weighted summation on all the input Attention values and the corresponding weight coefficients to obtain an output vector Attention (h), wherein the formula is as follows:
s33: an inverse parametric solution model of fluid instability and turbulence effects was constructed using fully connected neural networks that draw attention mechanisms and automatic differential variational reasoning.
Example 2
A machine learning solving method of a flame propagation model of a flame explosion under a sparse barrier specifically comprises the following steps:
step 1, constructing a gas explosion experiment system under non-uniform distribution of obstacles, and acquiring experimental data of explosion pressure monitoring points:
the first step: the gas explosion experiment system under the non-uniform distribution of the barriers is formed by the experiment model, the dangerous gas charging system, the ignition system and the data collection system.
And a second step of: and carrying out an explosion experiment on pressure monitoring points which are arranged in advance, and acquiring pressure data of different positions of the model, namely the experimental data.
Step 2, constructing an explosion model which is the same as the gas explosion experimental system by using OpenFOAM, sampling Latin for parameters of fluid instability and turbulence effect, and performing simulation to form a simulation training data set;
the first step: model construction and grid division are carried out by using the snappyhexmesh functions in CAD and OpenFOAM, grid sensitivity analysis is carried out, and the number of the final grids is 723 ten thousand.
And a second step of: ANOVAA analysis is performed on the empirical parameters in the turbulence model, and finally the important empirical parameters are determined as。
And a third step of: at the position ofLatin sampling was performed in the range of 20% up and down, forming 50 sets of simulated training data sets.
Step 3, performing sensitivity analysis on the models, and performing simulation on the models with different grid numbers respectively to finally determine that the number of the most convergent grids is 723 ten thousand;
step 4, utilizing machine learning to carry out inverse solution on parameters of fluid instability and turbulence effect, thereby solving a turbulent flame propagation model, and embedding the turbulent flame propagation model into a control equation and the turbulent model;
and 5, under the sparse distribution of the obstacles, the representation of a multi-space-time evolution mechanism and overpressure characteristics of oil gas explosion is realized, the turbulence effect jointly induced by fluid instability and the obstacles is analyzed, and the establishment of preventive measures is guided. According to the evolution and distribution of flame and pressure, the following conclusion is drawn:
(1) Under sparsely distributed obstacles, flame instability combined with obstacle-induced turbulence acts on the propagation of flame and overpressure.
(2) The restriction of the obstacle plays an important role in the transition of the flame model.
The technical features of the present invention that are not described in the present invention may be implemented by or using the prior art, and are not described in detail herein, but the above description is not intended to limit the present invention, and the present invention is not limited to the above examples, but is also intended to be within the scope of the present invention by those skilled in the art.
Claims (4)
1. The machine learning solving method of the flame propagation model of the explosion under the sparse barrier is characterized by comprising the following steps of:
s1, constructing a gas explosion experiment system under non-uniform distribution of obstacles, and acquiring experiment data of explosion pressure monitoring points;
s2, constructing an explosion model identical to the gas explosion experimental system by using OpenFOAM, sampling Latin for parameters of fluid instability and turbulence effect, and performing simulation to form a simulation training data set;
s3, utilizing machine learning to carry out inverse solution on parameters of fluid instability and turbulence effect, thereby solving a turbulent flame propagation model, and embedding the turbulent flame propagation model into a control equation and the turbulent model;
and S4, under the sparse distribution of the obstacles, the representation of a multi-space-time evolution mechanism and overpressure characteristics of oil gas explosion is realized, the turbulence effect jointly induced by fluid instability and the obstacles is analyzed, and the establishment of preventive measures is guided.
2. The machine learning solution method of a sparse under-obstacle flame propagation model according to claim 1, wherein the step of obtaining experimental data of the explosion pressure monitoring point in S1 is as follows:
s11: the gas explosion experiment system under the non-uniform distribution of the barriers is formed by an experiment model, a dangerous gas charging system, an ignition system and a data collection system, and a plurality of pressure detection points are arranged at different positions in the gas explosion experiment system;
s12: and carrying out an explosion experiment on the pressure monitoring points which are arranged in advance, and acquiring pressure data of different positions of the model, namely the experimental data.
3. The machine learning solution method of a sparse barrier under-flame propagation model according to claim 1, wherein in S2, the same flame propagation model as the gas flame experiment system is constructed by using OpenFOAM, and the specific steps of latin sampling parameters of fluid instability and turbulence effect are as follows:
s21: carrying out model construction and grid division by using a snappyhexmesh function in CAD and OpenFOAM, and carrying out grid sensitivity analysis;
s22: performing ANOVA analysis on parameters of fluid instability and turbulence effect in the turbulence model to obtain parameters with high importance;
s23: and defining upper and lower range boundaries for the parameters of the fluid instability and the turbulence effect, and sampling Latin from the boundary ranges to form a simulation combination.
4. The machine learning solution method of a sparse under-obstacle flame propagation model according to claim 1, wherein the specific steps of S3 are as follows:
s31: using python to implement automatic differential variation reasoning;
s32: introducing an attention mechanism into the fully connected neural network;
s33: an inverse parametric solution model of fluid instability and turbulence effects was constructed using fully connected neural networks that draw attention mechanisms and automatic differential variational reasoning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311507623.3A CN117236198B (en) | 2023-11-14 | 2023-11-14 | Machine learning solving method of flame propagation model of blasting under sparse barrier |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311507623.3A CN117236198B (en) | 2023-11-14 | 2023-11-14 | Machine learning solving method of flame propagation model of blasting under sparse barrier |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117236198A true CN117236198A (en) | 2023-12-15 |
| CN117236198B CN117236198B (en) | 2024-02-27 |
Family
ID=89082927
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311507623.3A Active CN117236198B (en) | 2023-11-14 | 2023-11-14 | Machine learning solving method of flame propagation model of blasting under sparse barrier |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117236198B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1457771A2 (en) * | 2003-03-12 | 2004-09-15 | Nissan Motor Co., Ltd. | Flame propagation modeling method |
| CN102608285A (en) * | 2012-02-21 | 2012-07-25 | 南京工业大学 | Prediction method for explosion characteristics of organic mixture based on support vector machine |
| WO2014123440A1 (en) * | 2013-02-06 | 2014-08-14 | Некоммерческое Партнерство По Научной, Образовательной И Инновационной Деятельности "Центр Импульсного Детонационного Горения" | Method and device for initiating detonation in a tube containing a combustible mixture |
| WO2017031356A1 (en) * | 2015-08-19 | 2017-02-23 | D-Wave Systems Inc. | Discrete variational auto-encoder systems and methods for machine learning using adiabatic quantum computers |
| CN109359833A (en) * | 2018-09-27 | 2019-02-19 | 中国石油大学(华东) | A kind of ocean platform based on ABC-BRANN model fires risk analysis method |
| EP3467718A1 (en) * | 2017-10-04 | 2019-04-10 | Prowler.io Limited | Machine learning system |
| CN111506814A (en) * | 2020-04-09 | 2020-08-07 | 苏州大学 | Sequence recommendation method based on variational self-attention network |
| CN113821908A (en) * | 2021-08-19 | 2021-12-21 | 常州大学 | Numerical simulation method in explosion venting process of gas pipeline |
| CN113962164A (en) * | 2021-09-10 | 2022-01-21 | 中国石油大学(华东) | Organic gas leakage diffusion real-time intelligent early warning method considering uncertainty reasoning |
| CN114021404A (en) * | 2021-10-29 | 2022-02-08 | 大连海事大学 | Hydrogen turbulent flow deflagration simulation dynamic modeling method |
| CN115935841A (en) * | 2022-12-02 | 2023-04-07 | 山东科技大学 | Numerical simulation method for explosion propagation characteristic of multi-element gas in semi-closed space |
-
2023
- 2023-11-14 CN CN202311507623.3A patent/CN117236198B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1457771A2 (en) * | 2003-03-12 | 2004-09-15 | Nissan Motor Co., Ltd. | Flame propagation modeling method |
| CN102608285A (en) * | 2012-02-21 | 2012-07-25 | 南京工业大学 | Prediction method for explosion characteristics of organic mixture based on support vector machine |
| WO2014123440A1 (en) * | 2013-02-06 | 2014-08-14 | Некоммерческое Партнерство По Научной, Образовательной И Инновационной Деятельности "Центр Импульсного Детонационного Горения" | Method and device for initiating detonation in a tube containing a combustible mixture |
| WO2017031356A1 (en) * | 2015-08-19 | 2017-02-23 | D-Wave Systems Inc. | Discrete variational auto-encoder systems and methods for machine learning using adiabatic quantum computers |
| EP3467718A1 (en) * | 2017-10-04 | 2019-04-10 | Prowler.io Limited | Machine learning system |
| CN109359833A (en) * | 2018-09-27 | 2019-02-19 | 中国石油大学(华东) | A kind of ocean platform based on ABC-BRANN model fires risk analysis method |
| CN111506814A (en) * | 2020-04-09 | 2020-08-07 | 苏州大学 | Sequence recommendation method based on variational self-attention network |
| CN113821908A (en) * | 2021-08-19 | 2021-12-21 | 常州大学 | Numerical simulation method in explosion venting process of gas pipeline |
| CN113962164A (en) * | 2021-09-10 | 2022-01-21 | 中国石油大学(华东) | Organic gas leakage diffusion real-time intelligent early warning method considering uncertainty reasoning |
| CN114021404A (en) * | 2021-10-29 | 2022-02-08 | 大连海事大学 | Hydrogen turbulent flow deflagration simulation dynamic modeling method |
| CN115935841A (en) * | 2022-12-02 | 2023-04-07 | 山东科技大学 | Numerical simulation method for explosion propagation characteristic of multi-element gas in semi-closed space |
Non-Patent Citations (3)
| Title |
|---|
| 于梅春;张登春;: "障碍物场中预混燃烧火焰的数值模拟", 中国安全科学学报, no. 03 * |
| 侯炳旭;俞冀阳;徐沾杰;江光明;邹志强;: "利用高斯过程回归对燃爆单元宽度的预测方法研究", 核动力工程, no. 02 * |
| 王公忠;张建华;李登科;陈先锋;: "障碍物对预混火焰特性影响的大涡数值模拟", 爆炸与冲击, no. 01 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117236198B (en) | 2024-02-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Qiu et al. | Atmospheric dispersion prediction and source estimation of hazardous gas using artificial neural network, particle swarm optimization and expectation maximization | |
| CN106056210B (en) | A kind of PM2.5 concentration value Forecasting Methodologies based on hybrid neural networks | |
| Su et al. | Smart performance-based design for building fire safety: Prediction of smoke motion via AI | |
| Zheng et al. | Inverse calculation approaches for source determination in hazardous chemical releases | |
| Bastani et al. | Contaminant source identification within a building: toward design of immune buildings | |
| CN107341576A (en) | A kind of visual air pollution of big data is traced to the source and trend estimate method | |
| CN109118409B (en) | Natural gas pipeline fire accident hazard analysis method based on risk degree | |
| Guichard | Assessment of an improved Random Flow Generation method to predict unsteady wind pressures on an isolated building using Large-Eddy Simulation | |
| Mao et al. | Comparison of the impacts of empirical power-law dispersion schemes on simulations of pollutant dispersion during different atmospheric conditions | |
| Cheng et al. | Source term estimation with deficient sensors: Traceability and an equivalent source approach | |
| CN113962164B (en) | Organic gas leakage diffusion real-time intelligent early warning method considering uncertainty reasoning | |
| CN104750953A (en) | Method for simulating middle-and-small-scale airborne substance atmosphere transportation ensemble diffusion | |
| Zhong et al. | Modelling segregation effects of heterogeneous emissions on ozone levels in idealised urban street canyons: Using photochemical box models | |
| Mao et al. | Improving source inversion performance of airborne pollutant emissions by modifying atmospheric dispersion scheme through sensitivity analysis combined with optimization model | |
| CN117236198B (en) | Machine learning solving method of flame propagation model of blasting under sparse barrier | |
| Li et al. | Emergency evacuation risk assessment method for educational buildings based on improved extreme learning machine | |
| Alattar et al. | Evaluating Particulate Matter (PM2. 5 and PM10) Impact on Human Health in Oman Based on a Hybrid Artificial Neural Network and Mathematical Models | |
| Alattar et al. | Neural and Mathematical Predicting Models for Particulate Matter Impact on Human Health in Oman | |
| Lauret et al. | Atmospheric turbulent dispersion modeling methods using machine learning tools | |
| Jeong et al. | Application of data assimilation to improve the forecasting capability of an atmospheric dispersion model for a radioactive plume | |
| Shen et al. | Spatial-temporal evolution prediction of gas distribution based on PSO-Elman neural network | |
| Lauret et al. | 2D modeling of turbulent flow around a cylindrical storage tank by artificial neural networks | |
| Ghayoumi Zadeh et al. | A Bidirectional Long Short-Term Neural Network Model to Predict Air Pollutant Concentrations: A Case Study of Tehran, Iran | |
| Berendt-Marchel et al. | Does the zero carry essential information for artificial neural network learning to simulate the contaminant transport in urban areas? | |
| Konstantinov et al. | Approaches to creating environment safety automation control system of the industrial complex |
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 |