CN116341397A - Fire extinguishing bomb temperature field simulation method based on deep neural network - Google Patents
Fire extinguishing bomb temperature field simulation method based on deep neural network Download PDFInfo
- Publication number
- CN116341397A CN116341397A CN202310620346.0A CN202310620346A CN116341397A CN 116341397 A CN116341397 A CN 116341397A CN 202310620346 A CN202310620346 A CN 202310620346A CN 116341397 A CN116341397 A CN 116341397A
- Authority
- CN
- China
- Prior art keywords
- fire extinguishing
- extinguishing bomb
- neural network
- temperature field
- deep neural
- 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 35
- 238000004088 simulation Methods 0.000 title claims abstract description 29
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 28
- 238000012549 training Methods 0.000 claims abstract description 39
- 238000012546 transfer Methods 0.000 claims abstract description 28
- 239000000463 material Substances 0.000 claims abstract description 23
- 238000003062 neural network model Methods 0.000 claims abstract description 23
- 238000004364 calculation method Methods 0.000 claims abstract description 10
- 238000013507 mapping Methods 0.000 claims abstract description 10
- 238000012544 monitoring process Methods 0.000 claims abstract description 7
- 238000013178 mathematical model Methods 0.000 claims abstract description 6
- 230000008878 coupling Effects 0.000 claims abstract description 4
- 238000010168 coupling process Methods 0.000 claims abstract description 4
- 238000005859 coupling reaction Methods 0.000 claims abstract description 4
- 230000010006 flight Effects 0.000 claims abstract description 4
- 239000012530 fluid Substances 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 230000004907 flux Effects 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 9
- 238000012795 verification Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 6
- 230000006403 short-term memory Effects 0.000 claims description 5
- 230000003213 activating effect Effects 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 238000013213 extrapolation Methods 0.000 claims description 3
- 238000011084 recovery Methods 0.000 claims description 3
- 210000004027 cell Anatomy 0.000 description 4
- 108010066057 cabin-1 Proteins 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 208000000044 Amnesia Diseases 0.000 description 1
- 208000031091 Amnestic disease Diseases 0.000 description 1
- 230000006986 amnesia Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 108010066114 cabin-2 Proteins 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- A—HUMAN NECESSITIES
- A62—LIFE-SAVING; FIRE-FIGHTING
- A62C—FIRE-FIGHTING
- A62C19/00—Hand fire-extinguishers in which the extinguishing substance is expelled by an explosion; Exploding containers thrown into the fire
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Geometry (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Computer Hardware Design (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Algebra (AREA)
- Mathematical Analysis (AREA)
- Fluid Mechanics (AREA)
- Public Health (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a fire extinguishing bomb temperature field simulation method based on a deep neural network, which comprises the following steps: s1: acquiring shape characteristics and material structures of the fire extinguishing bomb; s2: according to the shape, material and structure of fire extinguishing bullet, dividing the grid; s3: establishing a heat transfer mathematical model of multi-field coupling based on an N-S equation and a Newton heat flow formula; s4: acquiring relevant parameters of the fire extinguishing bomb; s5: collecting monitoring data after a plurality of bullet carrying flights, and establishing a sample data set of a fire extinguishing bullet temperature field; s6: constructing a deep neural network model, training according to the data in the training set, and establishing a mapping relation between parameters and an external temperature field; s7: and (3) carrying out simulation calculation on the temperature field inside the fire extinguishing bomb according to the heat transfer model in the step (S3), and deducing the temperature distribution conditions of different areas inside the fire extinguishing bomb. The scheme is based on the deep neural network model and the historical monitoring data, a temperature field can be obtained, and the distribution conditions of the internal temperature field and the external temperature field of the fire extinguishing bomb in different scenes can be accurately and rapidly predicted.
Description
Technical Field
The invention relates to the technical field of fire extinguishing bomb temperature field simulation, in particular to a fire extinguishing bomb temperature field simulation method based on a deep neural network.
Background
The airborne accurate guidance fire extinguishing bullet has flexible high-speed, accurate and efficient fire extinguishing capability, and has important significance for preventing and extinguishing fires in forest, mountain areas and other areas. In order to ensure the fire extinguishing precision of the fire extinguishing bomb, fragile equipment such as a control chip is often assembled in the fire extinguishing bomb, and in the flying process of the fire extinguishing bomb, the temperature suddenly drops to a certain degree along with the increase of the height, the requirements on the control chip and each hardware are higher, and the fire extinguishing effect of the fire extinguishing bomb is greatly influenced, so that the temperature field of the fire extinguishing bomb is required to be simulated, the distribution condition of the temperature field of the fire extinguishing bomb under different parameter conditions is researched, and the performance and the precision of the fire extinguishing bomb are improved. However, the prior art has few researches on the temperature field simulation of the fire extinguishing bomb, and the temperature field change condition of the fire extinguishing bomb under different parameters and different pneumatic environments cannot be fully analyzed.
Disclosure of Invention
In view of the above, the invention provides a fire extinguishing bomb temperature field simulation method based on a deep neural network, which aims to fully utilize available monitoring data, realize the description of mapping relations between different parameters and fire extinguishing bomb temperature fields by using a deep learning technology, and simulate the distribution of the temperature fields by combining aerodynamic and thermodynamic rules.
The invention provides a fire extinguishing bomb temperature field simulation method based on a deep neural network, which comprises the following steps:
s1: acquiring shape characteristics, material structures and working condition variables of the fire extinguishing bomb;
s2: according to the shape, material and structure of fire extinguishing bullet, dividing the grid;
s3: establishing a heat transfer mathematical model of multi-field coupling based on an N-S equation and a Newton heat flow formula;
s4: acquiring relevant parameters of the fire extinguishing bomb, including: fly height, fly speed, and fly angle of attack, etc.;
s5: collecting monitoring data after a plurality of bullet carrying flights, and establishing a sample data set of a fire extinguishing bullet temperature field;
s6: constructing a deep neural network model, training according to the data in the training set, and establishing a mapping relation between parameters and an external temperature field;
s7: and (3) carrying out simulation calculation on the temperature field inside the fire extinguishing bomb according to the heat transfer model in the step (S3), and deducing the temperature distribution conditions of different areas inside the fire extinguishing bomb.
Based on the above technical solution, preferably, the specific process of grid division according to the shape, material and structure of the fire extinguishing bomb in step S2 is as follows:
s21: establishing a three-dimensional model of the fire extinguishing bomb and a three-dimensional rectangular coordinate system respectively having x, y and z coordinate axes according to the shape characteristics of the fire extinguishing bomb obtained in the step S1, and carrying out layered modeling according to the material structure and the heat transfer characteristics;
s22: dividing the whole area into a plurality of sub-areas according to the overall outer diameter of the three-dimensional model, and respectively carrying out grid division on each sub-area;
s23: setting inlet boundary, outlet boundary and wall boundary conditions of the grid; the boundary condition of the entrance is set as a free infinity boundary condition; outlet boundary changeThe magnitude value is determined by the calculation result of the internal flow, and zero gradient extrapolation is adopted; the wall boundary adopts isothermal non-slip wall physical condition, and the initial wall condition is set
Temperature (temperature)T= T wall ，（/>
Wall temperature), normal pressure gradient is +.>
, wherein />
Respectively is the air outside the wall surface>
A velocity component in the axial direction.
Preferably, step S3 establishes a multi-field coupled heat transfer mathematical model based on the N-S equation and the Newton' S heat flow equation, comprising:
s31: construction of three-dimensional N-S equation differential form
, wherein ,
is the conservation variable vector of the fluid, E, F, G are respectively +.>
Flux vector in direction, +.>
Respectively->
A viscous flux vector in the direction; let->
Is the total energy of flow per unit volume,/->
Is specific heat ratio->
Is the pressure of the fluid>
Is air density->
Each of the traffic equations is then of the form: />
；
Let Planet number be
The constant pressure heat capacity is ∈>
，/>
Is a viscosity coefficient according to the Sutherland formula
Calculation of->
Is the air viscosity coefficient at 0 ℃ under one atmosphere pressure, < >>
，
Is air density, coefficient of thermal conductivity->
Then the respective viscous flux equation is of the form:
wherein ,
representing vector transpose->
The method comprises the following steps:
wherein ,
is a partial differential symbol.
S32: constructing a boundary layer Newton heat flow formula, and calculating heat generated by convection heat transfer of a fire extinguishing bomb boundary layer in the flight process
. Let the convection heat transfer coefficient be->
Recovery temperature is +.>
The wall temperature is->
The incoming flow temperature is->
Restoring factor->
Then: />
；
wherein ,
mach number>
Is the speed of sound of the fluid.
S33: according to the law of the heat transfer process of the fire extinguishing bomb, a numerical simulation model of three-dimensional heat transfer of the fire extinguishing bomb is established, and according to different heat transfer materials of the fire extinguishing bomb, the density of the different materials is as follows
Specific heat of different materials is +.>
Radial direction is +.>
The radius of the fire extinguishing bomb is +.>
Then respectively establishing different heat conduction equations under a cylindrical coordinate system: />
。
Preferably, step S6 builds a deep neural network model, performs training according to data in the training set, and builds a mapping relationship between parameters and an external temperature field, which specifically includes:
s61: construction of deep neural network model
； wherein />
To be in coordinates +.>
Position on the shaft, ">
For extinguishing bullet flying speed>
For flying angle of attack>
For model parameters +.>
Is a model structure of the deep neural network;
s62: dividing training data and test data according to the constructed data sets in the steps S4 and S5, training the constructed deep neural network model, and giving a training data set and a test set;
s63: dividing all data in a training dataset into
The shares, one of which is taken as the verification set in each round
Others as training set->
Training and verification, and traversing each of the shares. Inputting training parameters and verification parameters into deep neural network model +.>
Wherein, training predicted temperature is obtained>
And verifying the predicted temperature->
。
S64: calculating an optimal model with root mean square error MSE as a loss functionParameters (parameters)
：
wherein ,
，/>
is the number of samples.
S65: obtaining optimal model parameters according to S64
And then, verifying an optimal neural network model by using a test data set to fit the mapping relation between parameters such as space coordinates, flight height, flight speed and the like of the fire extinguishing bomb and the temperature field, and obtaining a final temperature field prediction model based on the depth neural network.
Preferably, the deep neural network model constructed in step S61
Is a long and short term memory neural network, and controls the circulation and loss of the characteristics by introducing a gating mechanism. The network structure is made up of a series of units, each structure including an input gate, an output gate, a hidden layer and a forget gate. For arbitrary->
At the moment, wherein the door is input
The value of +.>
, in the formula />
Is->
Input data of time of day->
For the hidden layer data of the last moment, +.>
and />
For learning parameters in the input gate, +.>
Is->
The short-term memory value of the moment in time,
, in the formula />
and />
Learning parameters for cell states; forgetting door
The values of (2) are: />
，/>
and />
Learning parameters in the forget gate; output door->
The value of +.>
，/>
and />
Outputting learning parameters in the door; hidden layer->
The value of +.>
, wherein ,/>
In the case of a cell state value,
activating a function for sigmoid->
Is a hyperbolic tangent activation function.
Compared with the prior art, the fire extinguishing bomb temperature field simulation method based on the deep neural network has the following beneficial effects:
(1) The scheme is based on the deep neural network model and the historical monitoring data, a temperature field can be obtained, and the distribution conditions of the internal temperature field and the external temperature field of the fire extinguishing bomb in different scenes can be accurately and rapidly predicted;
(2) According to the scheme, the fluid model and the thermodynamic model are built, so that the heat conduction simulation among different materials in the fire extinguishing bomb can be realized;
(3) The scheme has the advantages of high prediction efficiency and prediction accuracy, can shorten the research period of layering characteristics of the temperature field and distribution of the temperature field of the fire extinguishing bomb, and paves the way for temperature adjustment of the follow-up fire extinguishing bomb and accurate temperature compensation of a control system.
Drawings
FIG. 1 is a schematic flow chart of a fire extinguishing bomb temperature field simulation method based on a deep neural network;
FIG. 2 is a schematic diagram of an onboard precision guided fire extinguishing bomb structure according to the fire extinguishing bomb temperature field simulation method based on a deep neural network of the present invention;
fig. 3 is a schematic structural diagram of an electronic device of a fire extinguishing bomb temperature field simulation method based on a deep neural network.
Reference numerals: 1. a guide head cabin; 2. fire extinguishing cabin; 3. and a guidance cabin.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment of the application provides a fire extinguishing bomb temperature field simulation method based on a deep neural network. The execution main body of the fire extinguishing bomb temperature field simulation method based on the deep neural network comprises, but is not limited to, at least one of a service end, a terminal and the like which can be configured to execute the method provided by the embodiment of the application. In other words, the fire extinguishing bomb temperature field simulation method based on the deep neural network can be implemented by software or hardware installed on a terminal device or a server device. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like.
As shown in fig. 1 to 3, the invention provides a fire extinguishing bomb temperature field simulation method based on a deep neural network, which comprises the following steps:
s1: the shape characteristics, the material structure and the working condition variables of the fire extinguishing bomb are obtained, as shown in fig. 2, and the airborne accurate guided fire extinguishing bomb adopts three-section split type structural layout: mainly comprises three modules of a seeker cabin 1, a fire extinguishing cabin 2 and a guidance cabin 3. The guiding cabin 1 is designed into a hemispherical shape, the fire extinguishing cabin is designed into a cylindrical section, the guiding cabin 3 is arranged at the tail end, the fins are fixed on the outer surface of the fire extinguishing cabin, the fins are distributed in an X shape, the installation space of the steering engine and the avionics system is provided, and the guiding cabin is designed into a cylindrical shape.
S2: according to the shape, material and structure of the fire extinguishing bomb, grid division is carried out, and in the case, high-quality multi-block structured grids are generated by utilizing ANSYS self-contained grid division Mesh software;
the specific flow is as follows:
s21: establishing a three-dimensional model of the fire extinguishing bomb and a three-dimensional rectangular coordinate system respectively having x, y and z coordinate axes according to the shape characteristics of the fire extinguishing bomb obtained in the step S1, and carrying out layered modeling according to the material structure and the heat transfer characteristics;
s22: dividing the whole area into a plurality of sub-areas according to the overall outer diameter of the three-dimensional model, and respectively carrying out grid division on each sub-area;
s23: the inlet boundary, outlet boundary and wall boundary conditions of the mesh are set. The boundary condition of the entrance is set as a free infinity boundary condition; the value of the outlet boundary variable is determined by the calculation result of the internal flow, and zero gradient extrapolation is adopted; the wall boundary adopts isothermal non-slip wall physical condition, and the initial wall condition is set
Temperature (temperature)T= T wall ，（T wall Wall temperature), normal pressure gradient is +.>
, wherein />
Respectively is the air outside the wall surface>
A velocity component in the axial direction.
S3: establishing a heat transfer mathematical model of multi-field coupling based on an N-S equation and a Newton heat flow formula; the method specifically comprises the following steps:
s31: construction of three-dimensional N-S equation differential form
, wherein ,
is the conservation variable vector of the fluid, E, F, G are respectively +.>
Flux vector in direction, +.>
Respectively->
A viscous flux vector in the direction; let->
Is the total energy of flow per unit volume,/->
Is specific heat ratio->
Is the pressure of the fluid>
Is air density->
Each of the traffic equations is then of the form: />
；
Let Planet number be
The constant pressure heat capacity is ∈>
，/>
Is a viscosity coefficient according to the Sutherland formula
Calculation of->
Is the air viscosity coefficient at 0 ℃ under one atmosphere pressure, < >>
，
Is air density, coefficient of thermal conductivity->
Then the respective viscous flux equation is of the form:
wherein ,
representing vector transpose->
The method comprises the following steps:
wherein ,
is a partial differential symbol;
s32: constructing a boundary layer Newton heat flow formula, and calculating heat generated by convection heat transfer of a fire extinguishing bomb boundary layer in the flight process
. Let the convection heat transfer coefficient be->
Recovery temperature is +.>
The wall temperature is->
The incoming flow temperature is->
Restoring factor->
Then: />
；
wherein ,
mach number>
Is the speed of sound of the fluid.
S33: according to the law of the heat transfer process of the fire extinguishing bomb, a numerical simulation model of three-dimensional heat transfer of the fire extinguishing bomb is established, and according to different heat transfer materials of the fire extinguishing bomb, the density of the different materials is as follows
Specific heat of different materials is +.>
Radial direction is +.>
The radius of the fire extinguishing bomb is +.>
Then respectively establishing different heat conduction equations under a cylindrical coordinate system: />
；
S4: acquiring relevant parameters of the fire extinguishing bomb, including: fly height, fly speed, and fly angle of attack, etc.;
s5: as shown in fig. 3, the database collects monitoring data after a plurality of bullet-carrying flights and establishes a sample data set of a fire extinguishing bullet temperature field;
s6: as shown in fig. 3, the server calculates the control center to construct a deep neural network model, trains according to the data in the training set, and establishes a mapping relation between parameters and an external temperature field; the specific process is as follows:
s61: construction of deep neural network model
； wherein />
To be in coordinates +.>
Position on the shaft, ">
For extinguishing bullet flying speed>
For flying angle of attack>
For model parameters +.>
Is a model structure of the deep neural network; the Python language and Pytorch are used as deep learning frames, wherein the maximum training frequency max epoch is set to 100 during model training, the learning rate is 0.001, the optimizer selects adam, the number of neurons in a hidden layer is 16, and the batch size is 128. Meanwhile, when the variation value of the verification error is smaller than +.>
When the training is performed, the loss function is regarded as convergence, training is stopped, and the phenomenon of overfitting of training is avoided;
s62: dividing training data and test data according to the constructed data sets in the steps S4 and S5, training the constructed deep neural network model, and giving a training data set and a test set;
s63: dividing all data in a training dataset into
Parts, in each round, one of them is taken as validation set +.>
Others as training set->
Training and verification, and traversing each of the shares. Inputting training parameters and verification parameters into deep neural network model +.>
Wherein, training predicted temperature is obtained>
And verifying the predicted temperature->
。
S64: calculating optimal model parameters by taking root mean square error MSE as loss function
：
wherein ,
，/>
is the number of samples.
S65: obtaining optimal model parameters according to S64
Then, verifying the optimal neural network model by using the test data setAnd obtaining a final temperature field prediction model based on the depth neural network by fitting the mapping relation between parameters such as space coordinates, flight height, flight speed and the like of the fire extinguishing bomb and the temperature field.
Preferably, the deep neural network model constructed in step S61
Is a long and short term memory neural network, and controls the circulation and loss of the characteristics by introducing a gating mechanism. The network structure is made up of a series of units, each structure including an input gate, an output gate, a hidden layer and a forget gate. For arbitrary->
At the moment, wherein the door is input
The value of +.>
, in the formula />
Is->
Input data of time of day->
For the hidden layer data of the last moment, +.>
and />
For learning parameters in the input gate, < ->
Is->
Short-term memory value of time of day->
, in the formula />
and />
A learnable parameter that is a cell state; amnesia door->
The values of (2) are: />
，/>
and />
Is a learnable parameter in the forget gate; output door->
The value of +.>
，/>
and />
Is a learnable parameter in the output gate; hidden layer->
The value of +.>
, wherein ,/>
In the case of a cell state value,
activating a function for sigmoid->
Is a hyperbolic tangent activation function.
S7: and (3) carrying out simulation calculation on the temperature field inside the fire extinguishing bomb according to the heat transfer model in the step (S3), and deducing the temperature distribution conditions of different areas inside the fire extinguishing bomb.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (5)
1. A fire extinguishing bomb temperature field simulation method based on a deep neural network, which is characterized by comprising the following steps:
s1: acquiring shape characteristics, materials and structures of the fire extinguishing bomb;
s2: grid division is carried out according to the shape, material and structure of the fire extinguishing bomb;
s3: establishing a heat transfer mathematical model of multi-field coupling based on an N-S equation and a Newton heat flow formula;
s4: acquiring relevant parameters of the fire extinguishing bomb, including: fly height, fly speed, and fly angle of attack;
s5: collecting monitoring data after a plurality of bullet carrying flights, and establishing a sample data set of a fire extinguishing bullet temperature field;
s6: constructing a deep neural network model, training according to the data in the training set, and establishing a mapping relation between parameters and an external temperature field;
s7: and (3) carrying out simulation calculation on the temperature field inside the fire extinguishing bomb according to the heat transfer model in the step (S3), and deducing the temperature distribution conditions of different areas inside.
2. The fire extinguishing bomb temperature field simulation method based on the deep neural network as claimed in claim 1, wherein the specific flow of the grid division according to the shape, the material and the structure of the fire extinguishing bomb in the step S2 is as follows:
s21: establishing a three-dimensional model of the fire extinguishing bomb and a three-dimensional rectangular coordinate system respectively having x, y and z coordinate axes according to the shape characteristics of the fire extinguishing bomb obtained in the step S1, and carrying out layered modeling according to the material structure and the heat transfer characteristics;
s22: dividing the whole area into a plurality of sub-areas according to the overall outer diameter of the three-dimensional model, and respectively carrying out grid division on each sub-area;
s23: setting inlet boundary, outlet boundary and wall boundary conditions of the grid; the boundary condition of the entrance is set as a free infinity boundary condition; the value of the outlet boundary variable is determined by the calculation result of the internal flow, and zero gradient extrapolation is adopted; the wall boundary adopts isothermal non-slip wall physical condition, and the initial wall condition is set
Temperature (temperature)T=T wall ，T wall For the wall temperature, the normal pressure gradient is +.>
, wherein />
Respectively is the air outside the wall surface>
A velocity component in the axial direction.
3. The fire extinguishing bomb temperature field simulation method based on the deep neural network according to claim 2, wherein the step S3 establishes a multi-field coupled heat transfer mathematical model based on an N-S equation and a newton heat flow formula, and includes:
s31: construction of three-dimensional N-S equation differential form
, wherein ,
is the conservation variable vector of the fluid, E, F and G are respectively/>
Flux vector in direction, +.>
Respectively->
A viscous flux vector in the direction; let->
Is the total energy of flow per unit volume,/->
Is specific heat ratio->
Is the pressure of the fluid>
Is air density->
Each of the traffic equations is then of the form: />
；
Let Planet number be
The constant pressure heat capacity is ∈>
，/>
Is a viscosity coefficient according to the Sutherland formula
Calculation of->
Is the air viscosity coefficient at 0 ℃ under one atmosphere pressure, < >>
Thermal conductivity->
Then the respective viscous flux equation is of the form:
wherein ,
representing vector transpose->
The method comprises the following steps:
wherein ,
is a partial differential symbol;
s32: constructing a boundary layer Newton heat flow formula, and calculating heat generated by convection heat transfer of a fire extinguishing bomb boundary layer in the flight process
Let the convection heat transfer coefficient be ∈>
Recovery temperature is +.>
The incoming flow temperature is->
Restoring factor->
Then:
；
wherein ,
mach number>
Is the speed of sound of the fluid;
s33: according to the law of the heat transfer process of the fire extinguishing bomb, a numerical simulation model of three-dimensional heat transfer of the fire extinguishing bomb is established, and according to different heat transfer materials of the fire extinguishing bomb, the density of the different materials is as follows
Specific heat of different materials is +.>
Radial direction is +.>
The radius of the fire extinguishing bomb is
Then respectively establishing different heat conduction equations under a cylindrical coordinate system: />
。
4. The fire extinguishing bomb temperature field simulation method based on the deep neural network as claimed in claim 3, wherein the step S6 is to construct a deep neural network model, train according to the data in the training set, and build a mapping relation of parameter-appearance temperature field, and the specific process is as follows:
s61: construction of deep neural network model
； wherein />
To be in coordinates +.>
Position on the shaft, ">
For extinguishing bullet flying speed>
For flying angle of attack>
For model parameters +.>
Is a model structure of the deep neural network;
s62: dividing training data and test data according to the constructed data sets in the steps S4 and S5, training the constructed deep neural network model, and giving a training data set and a test set;
s63: dividing all data in a training dataset into
The shares, one of which is taken as the verification set in each round
Others as training set->
Training and verifying, and traversing each part, and inputting training parameters and verification parameters into the deep neural network model +.>
Wherein, training predicted temperature is obtained>
And verifying the predicted temperature->
；
S64: calculating optimal model parameters by taking root mean square error MSE as loss function
：
wherein ,
，/>
is the number of samples;
s65: obtaining optimal model parameters according to S64
And then, verifying an optimal neural network model by using a test data set to fit the mapping relation between parameters such as space coordinates, flight height, flight speed and the like of the fire extinguishing bomb and the temperature field, and obtaining a final temperature field prediction model based on the depth neural network.
5. The fire extinguishing bomb temperature field simulation method based on the deep neural network as claimed in claim 4, wherein the deep neural network model constructed in step S61
The deep neural network structure consists of a series of units, and each structure comprises an input gate, an output gate, a hidden layer and a forgetting gate; for arbitrary->
At the moment, wherein the door is entered->
The value of (2) is
, in the formula />
Is->
Input data of time of day->
For the hidden layer data of the last moment, +.>
and />
For learning parameters in the input gate, +.>
Is->
The short-term memory value of the moment in time,
, in the formula />
and />
Learning parameters for cell states; forgetting door
The values of (2) are: />
，/>
and />
Learning parameters in the forget gate; output door->
The value of +.>
，/>
and />
Outputting learning parameters in the door; hidden layer->
The value of +.>
, wherein ,/>
In the case of a cell state value,
activating a function for sigmoid->
Is a hyperbolic tangent activation function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310620346.0A CN116341397B (en) | 2023-05-30 | 2023-05-30 | Fire extinguishing bomb temperature field simulation method based on deep neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310620346.0A CN116341397B (en) | 2023-05-30 | 2023-05-30 | Fire extinguishing bomb temperature field simulation method based on deep neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116341397A true CN116341397A (en) | 2023-06-27 |
| CN116341397B CN116341397B (en) | 2023-09-05 |
Family
ID=86879101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310620346.0A Active CN116341397B (en) | 2023-05-30 | 2023-05-30 | Fire extinguishing bomb temperature field simulation method based on deep neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116341397B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107808065A (en) * | 2017-11-23 | 2018-03-16 | 南京航空航天大学 | The solid hot quick calculation method of 3 D complex profile high-speed aircraft stream |
| US20200148395A1 (en) * | 2018-01-26 | 2020-05-14 | Dalian University Of Technology | Method for prediction of key performance parameters of aero-engine in transition condition |
| US20200242202A1 (en) * | 2019-01-29 | 2020-07-30 | Shenzhen Fugui Precision Ind. Co., Ltd. | Fire development situation prediction device and method |
| CN113901733A (en) * | 2021-08-26 | 2022-01-07 | 北京航天长征飞行器研究所 | Method for determining thermal environment numerical value of gas flow field |
| CN114021501A (en) * | 2021-11-09 | 2022-02-08 | 华东理工大学 | Fire temperature field reconstruction method, system, computer equipment, medium and terminal |
| CN114722732A (en) * | 2022-06-09 | 2022-07-08 | 华中科技大学 | Method for predicting temperature field of fuel tank of hypersonic aircraft based on point cloud network |
| US20230102815A1 (en) * | 2020-02-21 | 2023-03-30 | Suzhou Inspur Intelligent Technology Co., Ltd | Turbulence field update method and apparatus, and related device thereof |
| CN115994491A (en) * | 2023-03-24 | 2023-04-21 | 湖南光华防务科技集团有限公司 | Method and system for compensating hanging flight temperature of fire extinguishing bomb |
-
2023
- 2023-05-30 CN CN202310620346.0A patent/CN116341397B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107808065A (en) * | 2017-11-23 | 2018-03-16 | 南京航空航天大学 | The solid hot quick calculation method of 3 D complex profile high-speed aircraft stream |
| US20200148395A1 (en) * | 2018-01-26 | 2020-05-14 | Dalian University Of Technology | Method for prediction of key performance parameters of aero-engine in transition condition |
| US20200242202A1 (en) * | 2019-01-29 | 2020-07-30 | Shenzhen Fugui Precision Ind. Co., Ltd. | Fire development situation prediction device and method |
| US20230102815A1 (en) * | 2020-02-21 | 2023-03-30 | Suzhou Inspur Intelligent Technology Co., Ltd | Turbulence field update method and apparatus, and related device thereof |
| CN113901733A (en) * | 2021-08-26 | 2022-01-07 | 北京航天长征飞行器研究所 | Method for determining thermal environment numerical value of gas flow field |
| CN114021501A (en) * | 2021-11-09 | 2022-02-08 | 华东理工大学 | Fire temperature field reconstruction method, system, computer equipment, medium and terminal |
| CN114722732A (en) * | 2022-06-09 | 2022-07-08 | 华中科技大学 | Method for predicting temperature field of fuel tank of hypersonic aircraft based on point cloud network |
| CN115994491A (en) * | 2023-03-24 | 2023-04-21 | 湖南光华防务科技集团有限公司 | Method and system for compensating hanging flight temperature of fire extinguishing bomb |
Non-Patent Citations (3)
| Title |
|---|
| 何运康等: "基于注意力机制和双向长短时记忆网络的横波速度预测方法及应用", 《石油物探》, pages 225 - 235 * |
| 王志超;张龙;姚琳;: "高速飞行器结构气动热计算与温度场模拟", 四川兵工学报, no. 11, pages 49 - 52 * |
| 郭曼丽等: "N-S方程耦合雷诺平均湍流模型无网格算法研究", 《南京理工大学学报》, pages 548 - 555 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116341397B (en) | 2023-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zheng et al. | The computational fluid dynamic modeling of downwash flow field for a six-rotor UAV | |
| CN105182743A (en) | Robust H-infinity-based variable-gain decoupling control method | |
| Egorova et al. | Estimation of gaseous plume concentration with an unmanned aerial vehicle | |
| CN116341397B (en) | Fire extinguishing bomb temperature field simulation method based on deep neural network | |
| Shen et al. | Review on wind resistance for quadrotor UAVs: Modeling and controller design | |
| Lei et al. | Moving time UCAV maneuver decision based on the dynamic relational weight algorithm and trajectory prediction | |
| Xu et al. | A study of Nash-EGO algorithm for aerodynamic shape design optimizations | |
| Dai et al. | Integrated morphing strategy and trajectory optimization of a morphing waverider and its online implementation based on the neural network | |
| Bramantya et al. | Study of the effect of 4-digit NACA variation on airfoil performance using computation fluid dynamics | |
| CN116414153A (en) | Unmanned aerial vehicle take-off and landing correction method based on artificial intelligence | |
| Sun et al. | Multifidelity multidisciplinary design optimization of integral solid propellant ramjet supersonic cruise vehicles | |
| Tabassum et al. | Variance reduction of quadcopter trajectory tracking in turbulent wind | |
| Lu et al. | Quadrotor inverted Pendulum control based on improved Particle Swarm Optimization | |
| Carlson et al. | A flight simulator for agile fighter aircraft and nonlinear aerodynamics | |
| Zhang et al. | Reentry trajectory optimization of hypersonic vehicle with enhancing parametrization method | |
| CN114355779A (en) | Global sliding mode control method for structural thermal test nonlinear extended state observer | |
| Vural et al. | A comparison of longitudinal controllers for autonomous UAV | |
| Carreno Ruiz et al. | Assessment of Quadrotor Near-Wall behaviour using six-Degrees of Freedom CFD simulations | |
| Dutta et al. | An ensemble kalman filter based explicit nonlinear model predictive control design for two degree freedom of helicopter model | |
| Shern et al. | CFD Study of A Micro Air Vehicle (MAV) | |
| Haitao et al. | Research on Attitude Control for UAV Based on Improved PSO Algorithm | |
| Liu et al. | Research on UAV Air Combat Maneuver Decision Based on Decision Tree CART Algorithm | |
| CN117519257B (en) | Supersonic speed cruising altitude control method based on back-stepping method | |
| Akbal et al. | In Flight Ice Shape Prediction with Data Fit Surrogate Models | |
| PATEL | CFD Analysis of Propeller Blade Design for Quadrotor |
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 |