CN113822000A - Flow field real-time prediction system and method based on edge calculation - Google Patents
Flow field real-time prediction system and method based on edge calculation Download PDFInfo
- Publication number
- CN113822000A CN113822000A CN202111187792.4A CN202111187792A CN113822000A CN 113822000 A CN113822000 A CN 113822000A CN 202111187792 A CN202111187792 A CN 202111187792A CN 113822000 A CN113822000 A CN 113822000A
- Authority
- CN
- China
- Prior art keywords
- flow field
- neural network
- calculation
- particles
- fluid
- 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
- 238000004364 calculation method Methods 0.000 title claims abstract description 74
- 238000000034 method Methods 0.000 title claims abstract description 45
- 239000002245 particle Substances 0.000 claims abstract description 98
- 239000012530 fluid Substances 0.000 claims abstract description 58
- 238000013528 artificial neural network Methods 0.000 claims abstract description 56
- 238000006073 displacement reaction Methods 0.000 claims abstract description 41
- 230000001133 acceleration Effects 0.000 claims abstract description 11
- 239000007787 solid Substances 0.000 claims description 18
- 230000006870 function Effects 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 8
- 230000004913 activation Effects 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 5
- 238000004806 packaging method and process Methods 0.000 claims description 4
- 238000012856 packing Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 230000001502 supplementing effect Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 238000004088 simulation Methods 0.000 abstract description 7
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 238000004891 communication Methods 0.000 description 6
- 238000002360 preparation method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 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
- G06F30/28—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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/25—Design optimisation, verification or simulation using particle-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- 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
- G06F2119/14—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)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Fluid Mechanics (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Algebra (AREA)
- Molecular Biology (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A flow field real-time prediction system based on edge calculation and a method thereof are provided, wherein the system comprises a Windows server, Atlas200DK and a digital sensor. The method comprises the following steps: acquiring flow field calculation parameters; establishing a discretized particle model according to the flow field parameters; calculating the intermediate velocity and the intermediate displacement of the particles according to the viscosity force and the physical force in the control equation; constructing input data required by solving a neural network of the pressure Poisson equation according to the intermediate displacement; transmitting the obtained input data to an Atlas200DK AI acceleration module for calculation and pressure data solving; transmitting the pressure data back to the Windows server; correcting the particle velocity and position according to the pressure gradient; and entering the next time step, and re-executing the steps. Compared with the prior art, the invention has the advantages that: the system and the method of the invention combine the neural network technology and the leading edge artificial intelligence equipment to realize the real-time simulation of the complex flow field and solve the problems of long calculation time of the complex flow field and difficult realization of real-time flow field analysis in the field of fluid calculation.
Description
Technical Field
The invention relates to the technical field of edge calculation and computational fluid mechanics, in particular to a flow field real-time prediction system and a flow field real-time prediction method aiming at edge calculation.
Background
At present, mainstream fluid calculation methods are mainly divided into two types: a mesh-based euler method and a particle-based lagrange method. The Euler method divides the calculation area into grids and disperses the differential equation to obtain the physical quantity change of the grid points at each moment; the Lagrange method is used for researching the change of the physical quantity of the particles along with time, and is suitable for more complex boundary conditions or large-deformation fluid simulation scenes.
At present, the grid method is the most mature fluid simulation scheme, but the grid method is difficult to deal with complex boundaries and large deformation fluid simulation processing. The common flow field conditions in actual life or production are quite complex, and the problems can be easily solved by adopting a gridding-free Lagrange method.
Fluid computation with the meshless method requires greater computational expense than with the meshy method. At present, the rapid flow field simulation is generally realized by using a computing hardware cluster with strong performance or a professional cloud computing server. However, in practical engineering applications, expensive computing hardware or cloud computing servers make fast fluid computing difficult to implement.
Disclosure of Invention
In order to solve the problems of high fluid calculation consumption and high calculation cost in the background technology, the invention provides a flow field real-time prediction system based on edge calculation and a method thereof. The system and the method solve the problems of long calculation time of the complex flow field and difficulty in realizing the real-time flow field analysis in the field of calculating the fluid, and have the advantages of low cost, high speed, reliability and suitability for complex scenes.
In order to achieve the above purpose, the flow field real-time prediction system based on edge calculation of the present invention has the technical scheme that:
a real-time flow field prediction system based on edge calculation comprises
The server is used for acquiring and storing flow field calculation parameters, establishing a particle model, calculating the intermediate speed and the intermediate displacement of the particle model and input data of a neural network, transmitting the input data of the neural network to a developer suite to calculate pressure data, and correcting the intermediate speed and the intermediate displacement of the particles by using the pressure data;
the developer kit comprises an acceleration module, a server and a data processing module, wherein the acceleration module is used for calculating input data of a neural network, solving pressure data and transmitting the pressure data to the server;
and the digital sensors comprise a digital temperature sensor and a digital flow rate sensor and are used for acquiring the temperature and the flow rate of the flow field.
The invention also provides a flow field real-time prediction method based on edge calculation, which adopts the technical scheme that:
a flow field real-time prediction method based on edge calculation comprises the following steps:
s1, acquiring flow field calculation parameters and storing the flow field calculation parameters in a server;
s2, establishing a discretized particle model in a server according to the flow field parameters;
s3, solving a particle model in the server according to the viscosity and the physical force in the control equation, and calculating the intermediate speed and the intermediate displacement of the particles;
s4, constructing input data of a neural network according to the intermediate displacement, wherein the neural network is used for solving a pressure Poisson equation;
s5, transmitting the input data to a neural network of a developer suite for calculation, and solving pressure data; then transmitting the pressure data back to the server;
s6, correcting the intermediate speed and the intermediate displacement of the particles according to the pressure gradient;
s7, entering the next time step, and re-executing the steps S3-S7.
Further, the flow field parameters include constants in an empirical formula of viscosity coefficients, flow field boundary positions, solid wall boundary positions, density information, and temperature and flow rate of the fluid.
Further, step S2 specifically includes the following steps:
s21, comparing the memory of the developer suite according to the flow field calculation parameters, and ensuring that the total memory is less than 70% to determine a proper particle spacing;
s22, constructing four layers of common particles on the fixed wall boundary and the outer side according to the particle spacing and the fixed wall boundary position, and uniformly dispersing the rest into fluid particles; wherein, two layers of the common particles close to the fluid are first solid particles, and two layers far away from the fluid are second solid particles; the first type solid particles and the second type solid particles do not update speed displacement, and the second type solid particles do not participate in pressure solving;
s23, marking fluid particles positioned on the boundary of the flow field according to the position of the boundary of the flow field and the flow velocity of the fluid, and marking the fluid particles escaping from the boundary of the flow field as common particles in the calculation process; then judging whether fluid particles exist in the boundary position of the flow field, and if so, marking the fluid particles; and if not, supplementing the fluid particles with the flow rate into the boundary position of the flow field according to the fluid flow rate.
Further, step S3 is specifically: decomposing a control equation into a pre-estimation step and a correction step by using a non-grid method-a moving particle semi-implicit method; in the pre-estimating step, a particle model is solved according to the viscous force and the physical force in the control equation, and the pre-estimated intermediate speed v is calculated*And a middle displacement r*。
Further, step S4 specifically includes the following steps:
s41, calculating the particle number density n according to the intermediate displacement*The calculation formula is as follows:
wherein ,
the median displacement of i particles is shown, W the kernel function, h the support set radius, is 2.1l0；
S42, converting the solving process of the pressure Poisson equation into a regression problem, wherein the calculation formula is as follows:
wherein k represents the operation to the k-th step,
representing the divergence of the i particle at the intermediate velocity of step k.
Further, the neural network construction method comprises the following steps: to be provided with
Constructing a neural network structure taking a multilayer perceptron as a main body as input data of the neural network; the neural network with each layer of perceptron as the main body is expressed as:
f(x)＝H(ω*x+b)
wherein ,x(x1,x2,…,xn) An input vector representing the layer, ω represents a weight, b represents a bias, H represents an activation function, f () represents the layer output, and the neural network employs ReLU as the activation function.
Further, step S5 specifically includes the following steps:
s51, packing the neural network input data of all the fluid particles;
s52, transmitting the neural network input data to a developer suite;
s53, receiving and unpacking data by the developer suite, and calling a neural network to solve pressure data;
s54, packaging the solution result, then returning to the server, and entering the step S6.
Further, step S6 is more specifically: according to pressure data transmitted back by a developer suite, solving a pressure gradient item of a control equation, and correcting speed and displacement, wherein a calculation formula is as follows:
wherein ,v*Represents an intermediate speed; r is*Represents the intermediate displacement; v represents a velocity; r represents a displacement; Δ t represents a time step; p represents pressure, ρ represents density;
and then outputting the calculation result of the current time step.
Compared with the prior art, the invention has the advantages and beneficial effects that: the real-time flow field prediction system and method based on edge calculation realize the real-time simulation of the complex flow field by combining the neural network technology and the leading edge artificial intelligence equipment, and solve the problems of long calculation time of the complex flow field and difficulty in realizing the real-time flow field analysis in the field of fluid calculation. In addition, the method has the advantages of low cost, high speed, reliability and applicability to the real-time calculation of the fluid in a complex scene.
Drawings
FIG. 1 is a schematic view of a flow field modeling process;
FIG. 2 is a neural network architecture for use with the present invention;
FIG. 3 is a program structure of the Ubuntu/Windows server of the present invention;
FIG. 4 shows the program structure of Atlas200DK of the invention;
fig. 5 is a schematic diagram of physical data preparation.
Detailed Description
The drawings are for illustrative purposes only and are not to be construed as limiting the patent. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In this embodiment, as a preferable mode, the server is an Ubuntu/Windows server, and the developer suite is an Atlas200DK developer suite of hua corporation, which is in a developer board form with an Atlas200AI acceleration module as a core. Where Atlas200AI acceleration module integrates the Ascend 310AI processor. The neural network reasoning for solving the pressure Poisson equation is transplanted to an Atlas200DK development board to run, the rest part is completed by a Ubuntu/Windows server, and the two realize rapid data exchange by utilizing a Socket communication protocol through a gigabit Ethernet.
The technical solution of the present invention is further described below with reference to fig. 1 to 5 and the embodiment.
Example 1
A flow field real-time prediction system based on edge calculation comprises a Ubuntu/Windows server, Atlas200DK and a digital sensor. The Ubuntu/Windows server is used for acquiring and storing flow field calculation parameters, establishing a particle model, calculating the intermediate speed and the intermediate displacement of the particle model and input data of a neural network, transmitting the input data of the neural network to a developer suite to calculate pressure data, and correcting the intermediate speed and the intermediate displacement of the particles by using the pressure data. The Atlas200DK includes an acceleration module therein for calculating input data of the neural network, solving pressure data, and transmitting the pressure data to the server. The digital sensor comprises a digital temperature sensor and a digital flow rate sensor and is used for acquiring the temperature and the flow rate of the flow field.
As a preferred mode, the program structure is (see fig. 3-4 for a corresponding flowchart):
Ubuntu/Windows server:
1. acquiring flow field calculation parameters;
2. calculating the intermediate velocity and the intermediate displacement of the particles according to the viscosity force and the physical force in the control equation;
3. according to the intermediate displacement, constructing input data of a neural network for solving a pressure Poisson equation;
4. transmitting the input data to an AI acceleration module of Atlas200DK by using a Socket communication protocol, and calculating pressure data;
5. receiving pressure data sent from Atlas200DK by using a Socket communication protocol;
6. correcting the particle velocity and position according to the pressure gradient;
7. judging whether the target calculation time is reached, if the target calculation time is reached, outputting a calculation result, and analyzing the calculation result; and if the target calculation time is not reached, returning to the step 2 and starting the next time step.
Atlas200DK end:
1. generating a model file suitable for solving the neural network by the AIcore;
2. initializing a computing environment;
3. receiving input data by using a Socket protocol;
4. transmitting the input data into an AIcore solving neural network to solve pressure data;
5. packaging the pressure data and transmitting the pressure data back to the Ubuntu/windows server;
6. and returning to the step 2, and waiting for reasoning the next time step.
Example 2
A flow field real-time prediction method based on edge calculation comprises the following steps:
s1, flow field calculation parameters including constants, flow field boundary positions, solid wall boundary positions, density information, fluid temperature and fluid flow rate in a viscosity coefficient empirical formula are obtained, and then the data are stored in a server.
The constants in the empirical formula of the viscosity coefficient can be obtained by selecting a table look-up mode or an experimental mode according to different empirical formulas and fluid types and stored in a database.
Acquiring boundary position information by adopting an engineering measurement mode at a flow field boundary position and a fixed wall boundary position, storing the boundary position information in a database of Atlas200DK equipment, and directly reading boundary data from the database during calculation;
fluid viscosity, calculated using an empirical formula based on temperature; the method comprises the steps of firstly collecting a fluid sample, obtaining relevant parameters except temperature of an empirical formula, connecting a serial port type digital temperature sensor with an Atlas200DK serial port, and obtaining temperature data in real time during calculation and substituting the temperature data into the empirical formula to obtain the viscosity of the fluid.
The temperature and the flow rate of the fluid are respectively measured by a digital temperature sensor and a digital flow rate sensor which are externally connected with Atlas200DK, data processing is carried out through the Atlas200DK, the viscosity coefficient of the fluid is calculated, the flow rate boundary of the flow field is determined, relevant physical data are transmitted back to a computer for storage through a socket communication protocol, and the physical data are prepared as shown in figure 5.
When a flow velocity inlet exists, the digital flow velocity sensor is connected to an Atlas200DK expansion GPIO interface in an instant measurement mode, and the Atlas200DK is used for reading a flow velocity signal during calculation to obtain instant flow velocity boundary information.
S2, establishing a discretized particle model in the server according to the flow field parameters, namely modeling the flow field;
s21, according to flow field calculation parameters, comparing the internal memory of Atlas200DK, and ensuring that the total internal memory is less than 70% to determine the proper particle spacing.
S22, constructing four layers of common particles on the fixed wall boundary and the outer side according to the particle spacing and the fixed wall boundary position, and uniformly dispersing the rest into fluid particles. Wherein, two layers of the common particles close to the fluid are first type solid particles, and two layers far away from the fluid are second type solid particles. The first type solid particles and the second type solid particles do not update the speed displacement, and the second type solid particles do not participate in pressure solving.
And S23, marking the fluid particles positioned on the boundary of the flow field according to the position of the boundary of the flow field and the flow velocity of the fluid, and marking the fluid particles escaping from the boundary of the flow field as common particles in the calculation process. Judging whether fluid particles exist in the boundary position of the flow field, and if so, marking the fluid particles; and if not, supplementing the fluid particles with the flow rate into the boundary position of the flow field according to the fluid flow rate. The positional relationship between the boundary particles and the fluid particles is shown in fig. 1.
S3, in the server, calculating the intermediate speed and the intermediate displacement of the particles according to the viscosity and the physical strength in the control equation;
the control equation is decomposed into estimation steps and correction step distribution for solving by a non-grid Method-Moving Particle Semi-implicit Method (MPS). In the pre-estimating step, a particle model is solved according to the viscous force and the physical force in the control equation, and the pre-estimated intermediate speed v is calculated*And a middle displacement r*The calculation formula is as follows:
r*＝rk+Δtv*
wherein ,vkRepresents the speed of the kth step (current step); r iskRepresents the displacement of the k-th step (current step), and v represents the kinematic viscosity of water; f represents the physical force to which the particles are subjected; typically only the gravitational acceleration g.
The control equation is:
wherein ,
representing the satellite derivative to some physical quantity.
The traditional steps for solving the control equation are specifically as follows:
I. calculating intermediate speed and intermediate displacement according to the viscous force and the physical force in the control equation;
II, calculating the particle number density by using the intermediate displacement obtained in the estimation step;
constructing a pressure Poisson equation through an incompressible principle, and solving to obtain pressure data;
wherein ,
the laplacian operator representing pressure, specifically expanded as follows; the right side is the source item; gamma represents the coefficient of the mixed source, and is 0-0.2; n is0Representing the initial population density.
Using the obtained pressure to solve the pressure gradient term correction intermediate speed and intermediate displacement of the control equation;
v, entering the next time step, and solving a control equation at the next moment according to the steps III-IV.
S4, constructing input data of a neural network for solving the pressure Poisson equation according to the intermediate displacement, wherein the input data comprises
S41, calculating the particle number density n according to the intermediate displacement*The calculation formula is as follows:
wherein ,
the median displacement of i particles is shown, W the kernel function, h the support set radius, is 2.1l0；
S42, converting the solving process of the pressure Poisson equation into a regression problem, wherein the calculation formula is as follows:
wherein k represents the operation to the k-th step,
representing the divergence of the i particle at the intermediate velocity of step k.
The formula of the conventional pressure poisson equation is as follows:
wherein D represents a system dimension;
w represents a weight function in MPS; p represents pressure; n is0Representing the initial population density.
S5, transmitting the input data to an Atlas200DK AI acceleration module for calculation, and solving pressure data; the pressure data is then transmitted back to the Ubuntu/Windows server.
S51, packing the neural network input data of all the fluid particles;
s52, sending the input data of the neural network to an inference program on Atlas200DK by adopting a Socket communication protocol;
and S53, receiving and unpacking the input data of the neural network by the inference program, and then calling the built-in AI core to solve the neural network to solve the pressure data. The neural network is provided with
Constructing a neural network structure taking a multilayer perceptron as a main body as input data of the neural network; the neural network with each layer of perceptron as the main body is expressed as:
f(x)＝H(ω*x+b)
wherein ,x(x1,x2,…,xn) An input vector representing the layer, ω represents a weight, b represents a bias, H represents an activation function, f () represents the layer output, and the neural network employs ReLU as the activation function.
Before the program is operated in an inference mode, a model file suitable for the AIcore to solve the neural network is generated, namely the neural network (shown in figure 2) is converted into a form which can be used by an AI chip, and the process comprises two steps of model preparation and model conversion. The model preparation specifically comprises: the training data set was acquired using the traditional MPS algorithm in S3, and the neural network was constructed and trained with tenserflow. The model conversion is specifically as follows: the pb file of the tensierflow save network is converted using the Huawei OMG tool, and the om model supported by the AI chip is obtained.
After the model file is generated, a program is written to realize reasoning, namely the built-in AI core is used for solving the neural network to solve pressure data, and the method specifically comprises the following steps:
I. initializing a model, wherein the model comprises a Huawei Matrix framework such as a Graph object, an Engine object and a model manager which are needed for running an om file;
reconstructing input data sent by the Ubuntu/Windows server to enable the input data to become NNTensor objects suitable for model stewards;
pushing the NNTensor object into a model housekeeper, calling AIcore to solve a neural network for reasoning, and solving pressure data;
obtaining a reasoning result;
v, packaging results and sending the results back to the Ubuntu/Windows server;
and VI, returning to the step II, and waiting for receiving the data of the next time step.
And S54. after the solution of the AIcore solving neural network is finished, packing the inference result by an inference program, and sending the inference result back to the Ubuntu/Windows server in a Socket communication mode.
S6, correcting the speed and the position of the particles according to the pressure gradient;
1. solving the pressure gradient term correction speed and displacement of the control equation according to the pressure data transmitted back to the Ubuntu/Windows server; the specific calculation formula is as follows:
and 2, the Ubuntu/Windows server outputs the calculation result of the current time step, namely the result of the real-time prediction of the flow field, including the particle position, the speed, the pressure and the like.
S7, entering the next time step, and re-executing the steps S3-S7.
And when the calculation termination condition is reached, outputting a calculation result and analyzing the calculation result. If the calculation termination condition is not reached, returning to the step 2 and starting the next time step.
Compared with the prior art, the beneficial effect of this embodiment is: the real-time flow field prediction system and method based on edge calculation in the embodiment realize real-time simulation of a complex flow field by combining a neural network technology and leading edge artificial intelligence equipment, and solve the problems of long calculation time and difficulty in realizing real-time flow field analysis of the complex flow field in the field of fluid calculation. When sudden flood disasters come, disaster development is rapidly analyzed for some flood disaster-prone scenes using the professional computing server unconditionally, a coping strategy is made more rapidly, and losses are reduced. In addition, the system and the method thereof have the advantages of low cost, high speed, reliability and applicability to the real-time calculation of the fluid in complex scenes.
It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.
Claims (9)
1. A flow field real-time prediction system based on edge calculation is characterized in that: comprises that
The server is used for acquiring and storing flow field calculation parameters, establishing a particle model, calculating the intermediate speed and the intermediate displacement of the particle model and input data of a neural network, transmitting the input data of the neural network to a developer suite to calculate pressure data, and correcting the intermediate speed and the intermediate displacement of the particles by using the pressure data;
the developer kit comprises an acceleration module, a server and a data processing module, wherein the acceleration module is used for calculating input data of a neural network, solving pressure data and transmitting the pressure data to the server;
and the digital sensors comprise a digital temperature sensor and a digital flow rate sensor and are used for acquiring the temperature and the flow rate of the flow field.
2. A flow field real-time prediction method based on edge calculation is characterized by comprising the following steps:
s1, acquiring flow field calculation parameters and storing the flow field calculation parameters in a server;
s2, establishing a discretized particle model in a server according to the flow field parameters;
s3, solving a particle model in the server according to the viscosity and the physical force in the control equation, and calculating the intermediate speed and the intermediate displacement of the particles;
s4, constructing input data of a neural network according to the intermediate displacement, wherein the neural network is used for solving a pressure Poisson equation;
s5, transmitting the input data to a neural network of a developer suite for calculation, and solving pressure data; then transmitting the pressure data back to the server;
s6, correcting the intermediate speed and the intermediate displacement of the particles according to the pressure gradient;
s7, entering the next time step, and re-executing the steps S3-S7.
3. The method of claim 2, wherein the flow field parameters include constants in an empirical formula for viscosity coefficient, flow field boundary position, solid wall boundary position, density information, and temperature and flow rate of the fluid.
4. The method according to claim 3, wherein step S2 specifically comprises the steps of:
s21, comparing the memory of the developer suite according to the flow field calculation parameters to ensure that the total memory is less than 70 percent to determine the proper inter-particle distance l0；
S22, constructing four layers of common particles on the fixed wall boundary and the outer side according to the particle spacing and the fixed wall boundary position, and uniformly dispersing the rest into fluid particles; wherein, two layers of the common particles close to the fluid are first solid particles, and two layers far away from the fluid are second solid particles; the first type solid particles and the second type solid particles do not update speed displacement, and the second type solid particles do not participate in pressure solving;
s23, marking fluid particles positioned on the boundary of the flow field according to the position of the boundary of the flow field and the flow velocity of the fluid, and marking the fluid particles escaping from the boundary of the flow field as common particles in the calculation process; then judging whether fluid particles exist in the boundary position of the flow field, and if so, marking the fluid particles; and if not, supplementing the fluid particles with the flow rate into the boundary position of the flow field according to the fluid flow rate.
5. The method according to claim 2, wherein step S3 is specifically: decomposing a control equation into a pre-estimation step and a correction step by using a non-grid method-a moving particle semi-implicit method; in the pre-estimating step, a particle model is solved according to the viscous force and the physical force in the control equation, and the pre-estimated intermediate speed v is calculated*And a middle displacement r*。
6. The method according to claim 3, wherein step S4 specifically comprises the steps of:
s41, calculating the particle number density n according to the intermediate displacement*The calculation formula is as follows:
wherein ,
represents the intermediate displacement of the i particle; w represents a kernel function; h represents the support radius of 2.1l0；
S42, converting the solving process of the pressure Poisson equation into a regression problem, wherein the calculation formula is as follows:
wherein k represents the operation to the k step;
representing the divergence of the i particle at the intermediate velocity of step k.
7. The method of claim 6, wherein: the neural network construction method comprises the following steps: to be provided with
Constructing a neural network structure taking a multilayer perceptron as a main body as input data of the neural network; the neural network with each layer of perceptron as the main body is expressed as:
f(x)＝H(ω*x+b)
wherein ,x(x1,x2,…,xn) An input vector representing the layer, ω represents a weight, b represents a bias, H represents an activation function, f () represents the layer output, and the neural network employs ReLU as the activation function.
8. The method according to claim 2, wherein step S5 specifically comprises the steps of:
s51, packing the neural network input data of all the fluid particles;
s52, transmitting the neural network input data to a developer suite;
s53, receiving and unpacking data by the developer suite, and calling a neural network to solve pressure data;
s54, packaging the solution result, then returning to the server, and entering the step S6.
9. The method according to claim 2, wherein step S6 is more specifically: according to pressure data transmitted back by a developer suite, solving a pressure gradient item of a control equation, and correcting speed and displacement, wherein a calculation formula is as follows:
wherein ,v*Represents an intermediate speed; r is*Represents the intermediate displacement; v represents a velocity; r represents a displacement; Δ t represents a time step; p represents pressure, ρ represents density;
and then outputting the calculation result of the current time step.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111187792.4A CN113822000B (en) | 2021-10-12 | 2021-10-12 | Flow field real-time prediction system and method based on edge calculation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111187792.4A CN113822000B (en) | 2021-10-12 | 2021-10-12 | Flow field real-time prediction system and method based on edge calculation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113822000A true CN113822000A (en) | 2021-12-21 |
| CN113822000B CN113822000B (en) | 2023-10-13 |
Family
ID=78920210
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111187792.4A Active CN113822000B (en) | 2021-10-12 | 2021-10-12 | Flow field real-time prediction system and method based on edge calculation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113822000B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103970989A (en) * | 2014-04-15 | 2014-08-06 | 昆明理工大学 | Immersing boundary flow field calculation method based on fluid/solid interface consistency |
| CN110222828A (en) * | 2019-06-12 | 2019-09-10 | 西安交通大学 | A kind of Unsteady Flow method for quick predicting based on interacting depth neural network |
| WO2021163309A1 (en) * | 2020-02-11 | 2021-08-19 | Siemens Aktiengesellschaft | Frequency-compensted neural networks for fluid dynamics design problems |
| US20210279386A1 (en) * | 2020-03-05 | 2021-09-09 | International Business Machines Corporation | Multi-modal deep learning based surrogate model for high-fidelity simulation |
| CN113420513A (en) * | 2021-07-01 | 2021-09-21 | 西北工业大学 | Underwater cylinder turbulent flow partition flow field prediction method based on deep learning |
-
2021
- 2021-10-12 CN CN202111187792.4A patent/CN113822000B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103970989A (en) * | 2014-04-15 | 2014-08-06 | 昆明理工大学 | Immersing boundary flow field calculation method based on fluid/solid interface consistency |
| CN110222828A (en) * | 2019-06-12 | 2019-09-10 | 西安交通大学 | A kind of Unsteady Flow method for quick predicting based on interacting depth neural network |
| WO2021163309A1 (en) * | 2020-02-11 | 2021-08-19 | Siemens Aktiengesellschaft | Frequency-compensted neural networks for fluid dynamics design problems |
| US20210279386A1 (en) * | 2020-03-05 | 2021-09-09 | International Business Machines Corporation | Multi-modal deep learning based surrogate model for high-fidelity simulation |
| CN113420513A (en) * | 2021-07-01 | 2021-09-21 | 西北工业大学 | Underwater cylinder turbulent flow partition flow field prediction method based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 陈粤 等: "基于SolidWorks Flow Simulation的换热器流场仿真分析及优化", 软件导刊, vol. 16, no. 3, pages 11 - 14 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113822000B (en) | 2023-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111311685B (en) | Motion scene reconstruction unsupervised method based on IMU and monocular image | |
| US11853894B2 (en) | Meta-learning for multi-task learning for neural networks | |
| US20200082041A1 (en) | Estimating physical parameters of a physical system based on a spatial-temporal emulator | |
| CN110427654B (en) | Landslide prediction model construction method and system based on sensitive state | |
| WO2019034129A1 (en) | Neural network structure generation method and device, electronic equipment and storage medium | |
| CN115456160A (en) | Data processing method and data processing equipment | |
| CN114493050B (en) | Multi-dimensional fusion new energy power parallel prediction method and device | |
| CN116451584B (en) | Thermal stress prediction method and system based on neural network | |
| CN106023286A (en) | Fluid animation accelerated generation method based on data driving | |
| Janny et al. | Eagle: Large-scale learning of turbulent fluid dynamics with mesh transformers | |
| CN116310219A (en) | Three-dimensional foot shape generation method based on conditional diffusion model | |
| CN115983148A (en) | CFD simulation cloud picture prediction method, system, electronic device and medium | |
| WO2021070204A1 (en) | Machine learning, deep learning and artificial intelligence for physical transport phenomenon in thermal management | |
| CN116758219A (en) | Region-aware multi-view stereo matching three-dimensional reconstruction method based on neural network | |
| CN113822000A (en) | Flow field real-time prediction system and method based on edge calculation | |
| CN116051699A (en) | Dynamic capture data processing method, device, equipment and storage medium | |
| CN113779116B (en) | Object ordering method, related equipment and medium | |
| CN114781207A (en) | Heat source layout temperature field prediction method based on uncertainty and semi-supervised learning | |
| CN114298961A (en) | Image processing method, device, equipment and storage medium | |
| Cui et al. | The atmospheric model of neural networks based on the improved Levenberg-Marquardt algorithm | |
| WO2024071377A1 (en) | Information processing device, information processing method, and program | |
| US20240086598A1 (en) | Transforming a model in a first language to a surrogate in a second language for simulation | |
| CN112561977B (en) | Point cloud sharp feature normal vector estimation method based on depth feature classification and neighborhood optimization | |
| CN115345387B (en) | Wind field wind speed prediction method and device and storage medium | |
| CN114154679B (en) | Spark-based PCFOA-KELM wind power prediction method and device |
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 |