CN115510732B - Shelter infrared characteristic simulation rapid algorithm based on deep learning - Google Patents
Shelter infrared characteristic simulation rapid algorithm based on deep learning Download PDFInfo
- Publication number
- CN115510732B CN115510732B CN202210632586.8A CN202210632586A CN115510732B CN 115510732 B CN115510732 B CN 115510732B CN 202210632586 A CN202210632586 A CN 202210632586A CN 115510732 B CN115510732 B CN 115510732B
- Authority
- CN
- China
- Prior art keywords
- shelter
- model
- radiation
- simulation
- infrared
- 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.)
- Active
Links
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
-
- 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
- G06N3/08—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
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Fluid Mechanics (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pure & Applied Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Algebra (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Radiation Pyrometers (AREA)
Abstract
A shelter infrared characteristic simulation rapid algorithm based on deep learning comprises the following steps: building a cubic cabin temperature distribution calculation model; reading in the temperature field distribution of the model, and calculating the self infrared radiation intensity of the surface of the model; reading the intensity of solar radiation received by the surface of the shelter, ground background radiation and sky background radiation, and calculating the reflected radiation of the surface of the model; obtaining a correlation database of total radiation quantity and environmental parameters of the shelter surface; and taking the environmental parameters as input and the shelter radiation distribution field as output to train the convolutional neural network model, thereby obtaining the shelter infrared characteristic simulation rapid algorithm based on deep learning. According to the invention, a model is trained through deep learning, a plurality of case libraries are obtained, different parameters are input, namely, database results can be adjusted, and after new parameters are input, the neural network can train new results, so that the neural network can perform cyclic reciprocation, has the capability of rapid calculation, less influence of the parameters, and obtains more accurate calculation results.
Description
Technical Field
The invention relates to the technical field of infrared simulation, in particular to a shelter infrared characteristic simulation rapid algorithm based on deep learning.
Background
With the development of computer technology, infrared simulation technology is widely applied to different fields. Because of the relation between actual conditions and funds, the infrared simulation is impossible to test in a real environment, and the computer technology is used for carrying out the infrared simulation, so that physical characteristics and mathematical models can be provided for the infrared system simulation. Because the shelter can generate a series of heat and mass transfer processes with different mechanisms in different environments, different shelter temperature distribution and the influence factors of the infrared radiation characteristics of the shelter are many under the conditions of different time points, different weather and the like.
The infrared radiation characteristic calculation analysis of the shelter relates to a plurality of application situations, and the infrared radiation characteristic research work in China mainly aims at airplanes, tanks, ships, ground objects, ocean backgrounds and the like. There are two main approaches to infrared radiation property studies: firstly, a radiation experience model is established according to experimental measurement results, which is time-consuming and labor-consuming; secondly, theoretical modeling research is carried out by analyzing physical models of targets and environments, but accurate results are difficult to obtain.
The infrared simulation of current shelter has a lot of defects, includes: the influence parameters are more, such as: the influences of meteorological conditions, the size, structural characteristics, material characteristics, internal factors, external environmental factors and the like of the shelter; the computing resources are large, and enough large workstations are needed to support the computing; the calculation takes a long time, and the calculation result time is as long as several days or one month; and is heavily dependent on the experience of the engineer.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a shelter infrared characteristic simulation rapid algorithm based on deep learning, which solves the technical problems of more influencing parameters, large calculation occupied resources, long calculation time consumption, inaccurate calculation result and the like in the prior art, and specifically adopts the following scheme:
the invention relates to a shelter infrared characteristic simulation rapid algorithm based on deep learning, which is characterized in that: the method comprises the following steps:
step 1, establishing a shelter temperature distribution calculation model;
step 2, reading in the temperature field distribution of the model, and calculating the self infrared radiation intensity of the surface of the model;
step 3, reading in solar radiation intensity, ground background radiation and sky background radiation received by the surface of the shelter, and calculating reflected radiation of the surface of the model to obtain a correlation database of total radiation quantity and environmental parameters of the surface of the shelter;
and 4, establishing an infrared radiation model, which comprises the following steps of:
step A, establishing a shelter geometric model;
step B, extracting simulation data of a temperature field, a radiation field and monitoring points;
step C, establishing a radiation distribution field of the cubic cabin and an external environment model through deep learning;
and D, taking the environmental parameters as input, taking the shelter radiation distribution field as output, training a convolutional neural network model, and obtaining a shelter infrared characteristic simulation rapid algorithm based on deep learning.
The invention relates to a shelter infrared characteristic simulation rapid algorithm technical scheme based on deep learning, which is further preferably characterized in that:
1. the construction of the shelter geometric model is carried out according to the following steps:
establishing a three-dimensional geometric model of the shelter according to the selected shelter data;
and setting certain external conditions and environment parameters for the shelter three-dimensional geometric model to form an infrared simulation calculation domain.
2. The simulation technology is adopted to simulate the model in full time period and various environmental backgrounds, and a simulation database is established, and the method specifically comprises the following steps:
dividing the wind speed, the temperature and the time of the external environment of the shelter into N groups, and dividing the wind speed, the temperature and the time of the external environment of the shelter into M weather conditions according to solar radiance and atmospheric irradiance to obtain different working conditions composed of N.M parameters;
and calculating the temperature field display and the infrared radiation display of the shelter under different working conditions by adopting a simulation technology method, and building an infrared simulation database of the shelter.
3. The simulation techniques include flow solid heat transfer, radiant heat transfer, and computational fluid dynamics simulation techniques.
4. According to the simulation data of the extracted temperature field, the radiation field and the monitoring points, the infrared radiation field distribution and the external environment model of the cubic cabin are built through deep learning, and the method specifically comprises the following steps:
a. the environmental parameters comprise environmental temperature, solar irradiance and environmental wind speed; the material parameters include: ground material density, ground material thermal conductivity, ground material constant pressure heat capacity; defining relative Cartesian coordinates of the monitoring points and the shelter model as input data of simulation data, and defining shelter surface temperature distribution, infrared radiation distribution and monitoring point temperature as an objective function of the simulation data;
b. dividing the training set and the testing set in the same proportion for a plurality of times;
c. defining convolutional neural network parameters, a loss function, an activation function and a pooling layer;
d. introducing the training set into a convolutional neural network, and optimizing the calculation accuracy of the convolutional neural network by adopting a linear correction unit function;
e. the test set is brought into the optimized convolutional neural network, and a loss function is calculated;
f. modifying the convolution layer number and the neuron number of each layer of the convolution neural network, modifying the pooling method and parameters of the pooling layer, repeating d and e, and enabling the loss function value to be minimum.
Compared with the prior art, the invention has the beneficial effects that:
1. establishing a three-dimensional calculation model aiming at an infrared radiation field of the ground shelter, and carrying out numerical simulation on a shelter target model and a surrounding environment thereof by adopting multi-physical field coupling simulation to obtain a temperature field and an infrared radiation field database related to the calculation domain; according to the simulation data matrix in the simulation database, a convolutional neural network is adopted to construct a mapping relation between external environment parameters and boundary conditions of the ground shelter target and a plurality of physical parameters (sets) by a machine learning method. And obtaining a data prediction model of transient infrared radiation field distribution. The accurate computing capability of computational fluid mechanics and thermodynamic technology on three-dimensional complex multi-physical-field environmental parameters and the reconstruction capability of a machine learning algorithm on a high-dimensional data matrix are fully utilized.
2. Compared with the traditional calculation method, the method for calculating the infrared radiation field distribution of the shelter by using the model has the advantages that the model is not required to be generalized by data driving, and the method has finer parameter calculation capability, more visual data visualization method and more rapid calculation time.
3. When the calculation model generates temperature field and infrared radiation field data to be calculated according to actual measurement point data, existing environment parameters and infrared radiation field simulation data in a database are directly called, an infrared characteristic simulation rapid calculation model is constructed according to the existing simulation data, the simulation data of the infrared radiation field to be predicted is used as prediction data, real-time calculation of the simulation data is not needed, historical data can be fully mined, calculation hardware and time cost are greatly reduced, and the whole simulation rapid algorithm has high efficiency.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 shows different monitoring points of the present invention: and comparing temperature change comparison graphs of two monitoring points of the left cabin and the right cabin of the shelter.
Detailed Description
The technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiment 1, a shelter infrared characteristic simulation rapid algorithm based on deep learning, comprising the following steps:
step 1, establishing a shelter temperature distribution calculation model;
step 2, reading in the temperature field distribution of the model, and calculating the self infrared radiation intensity of the surface of the model;
step 3, reading in solar radiation intensity, ground background radiation and sky background radiation received by the surface of the shelter, and calculating reflected radiation of the surface of the model to obtain a correlation database of total radiation quantity and environmental parameters of the surface of the shelter;
and 4, establishing an infrared radiation model, which comprises the following steps of:
step A, establishing a shelter geometric model;
step B, extracting simulation data of a temperature field, a radiation field and monitoring points;
step C, establishing a radiation distribution field of the cubic cabin and an external environment model through deep learning;
and D, taking the environmental parameters as input, taking the shelter radiation distribution field as output, training a convolutional neural network model, and obtaining a shelter infrared characteristic simulation rapid algorithm based on deep learning.
According to the invention, environmental parameters are used as input, the shelter radiation distribution field is used as output to train a convolutional neural network model, and a shelter infrared characteristic simulation rapid algorithm based on deep learning is obtained. The infrared radiation model is built by the following steps: building a geometric model of the cubic cabin; simulating the model by adopting simulation technologies such as fluid-solid heat transfer, radiation heat transfer, computational fluid dynamics and the like in a full period of time and various environmental backgrounds, and establishing a simulation database; extracting simulation data of a temperature field, a radiation field and monitoring points; and building a cube cabin infrared radiation field distribution and external environment model through deep learning. According to the technical scheme, the model results are trained through deep learning, countless case libraries are obtained, different parameters are input, the database results can be called out, and after new parameters are input, the neural network can train new results again, so that the neural network can perform cyclic reciprocation, has the capability of faster calculation and less influence of the parameters, and obtains more accurate calculation results. The infrared simulation technology mainly utilizes computer simulation software to calculate the infrared radiation characteristic distribution of a target, and because the infrared radiation temperature of the shelter is received by the infrared imaging detector instead of the real temperature of the shelter, a shelter infrared radiation characteristic theoretical model is built on the basis of a shelter temperature distribution model, the infrared radiation characteristic of the shelter is analyzed and calculated, a training set is brought into a convolutional neural network through deep learning, the calculation accuracy of the convolutional neural network is optimized, and an accurate calculation result can be obtained.
Embodiment 2, in the shelter infrared characteristic simulation fast algorithm based on deep learning described in embodiment 1: the construction of the shelter geometric model is carried out according to the following steps:
establishing a three-dimensional geometric model of the shelter according to the selected shelter data;
and setting certain external conditions and environment parameters for the shelter three-dimensional geometric model to form an infrared simulation calculation domain.
Embodiment 3, in the fast algorithm for simulation of infrared characteristics of a shelter based on deep learning according to embodiment 1 or 2: the simulation technology is adopted to simulate the model in full time period and various environmental backgrounds, and a simulation database is established, and the method specifically comprises the following steps:
dividing the wind speed, the temperature and the time of the external environment of the shelter into N groups, and dividing the wind speed, the temperature and the time of the external environment of the shelter into M weather conditions according to solar radiance and atmospheric irradiance to obtain different working conditions composed of N.M parameters;
and calculating the temperature field display and the infrared radiation display of the shelter under different working conditions by adopting a simulation technology method, and building an infrared simulation database of the shelter.
Embodiment 4, the deep learning-based shelter infrared characteristic simulation fast algorithm of any one of embodiments 1-3: the simulation techniques include flow solid heat transfer, radiant heat transfer, and computational fluid dynamics simulation techniques.
a. the environmental parameters comprise environmental temperature, solar irradiance and environmental wind speed; the material parameters include: ground material density, ground material thermal conductivity, ground material constant pressure heat capacity; defining relative Cartesian coordinates of the monitoring points and the shelter model as input data of simulation data, and defining shelter surface temperature distribution, infrared radiation distribution and monitoring point temperature as an objective function of the simulation data;
b. dividing the training set and the testing set in the same proportion for a plurality of times;
c. defining convolutional neural network parameters, a loss function, an activation function and a pooling layer;
d. introducing the training set into a convolutional neural network, and optimizing the calculation accuracy of the convolutional neural network by adopting a linear correction unit function;
e. the test set is brought into the optimized convolutional neural network, and a loss function is calculated;
f. modifying the convolution layer number and the neuron number of each layer of the convolution neural network, modifying the pooling method and parameters of the pooling layer, repeating d and e, and enabling the loss function value to be minimum.
Embodiment 6, a shelter infrared characteristic simulation rapid algorithm based on deep learning, comprises the following steps:
building a cubic cabin temperature distribution calculation model;
reading in the temperature field distribution of the model, and calculating the self infrared radiation intensity of the surface of the model;
reading the intensity of solar radiation received by the surface of the shelter, ground background radiation and sky background radiation, and calculating the reflected radiation of the surface of the model;
obtaining a correlation database of total radiation quantity and environmental parameters of the shelter surface;
building a geometric model of the cubic cabin;
extracting simulation data of a temperature field, a radiation field and monitoring points;
establishing a cube cabin infrared radiation field distribution and an external environment model through deep learning;
taking environmental parameters as input and taking shelter radiation distribution field as output to train a convolutional neural network model to obtain shelter infrared characteristic simulation rapid algorithm based on deep learning
Embodiment 7, a shelter infrared characteristic simulation rapid algorithm based on deep learning, comprises the following steps:
and 1, establishing a geometric model of a shelter, wherein the shelter is made of a universal commercial automobile model, the material is low carbon steel, and the vehicle window is glass. And (3) making a ground geometric model, wherein the ground geometric model is made of concrete, and the length, the width and the thickness of the ground model are respectively 12 multiplied by 10 multiplied by 1 (meters).
Step 2, carrying out unstructured tetrahedron grid division on the whole shelter numerical simulation calculation domain, wherein the total grid amount is 57493, the ground part is thicker, and the grid number is 12312; the shelter portion is thin, and the grid number of the shelter portion is 45181, so that the simulation time cost and the accurate simulation of the vehicle surface temperature distribution are guaranteed.
And 3, reading in the distribution of the temperature field of the model, receiving the solar radiation intensity, the ground background radiation and the sky background radiation on the surface of the shelter, and calculating the self infrared radiation intensity and the reflected radiation on the surface of the model.
Dividing the wind speed, temperature and time of the external environment of the shelter into N groups, and dividing the solar radiance into M weather conditions to obtain different working conditions composed of N.M parameters;
the wind speed of the external environment of the shelter is generally in the range of 0-5m/s, and the temperature is generally in the range of 0-30 0 ) The time is 0-24 hours. In the example, the wind speed and the temperature range are divided into nine types, wherein two adjacent wind speed intervals are increased by 1m/s, and two adjacent temperature intervals are increased by 10 0 . Solar radiation is divided into three weather conditions, namely, the radiation is respectively 150 on cloudy days, 550 on cloudy days and 1000 on sunny days. Accordingly, the external environment of the shelter is represented as a database of 9*3 =27 operating conditions.
And 4, calculating the temperature field display and the infrared radiation display of the shelter under different working conditions by using methods such as flowing solid heat transfer, radiation heat transfer, computational fluid dynamics and the like, and building an infrared simulation database of the shelter.
Simulation data of monitoring points are extracted from a shelter infrared simulation database, through deep learning, training of a neural network is conducted from 27 working conditions, a shelter radiation distribution field is used as output to train a convolutional neural network model, and the neural network is optimized to obtain a plurality of case libraries.
Only external environment parameters are input and selected, the time is set to be 10 points, the solar radiation degree is 1000 on sunny days, the wind speed is 2m/s, and the temperature is 20 0 The upper limit and the lower limit of the temperature scale are set so as to better compare and observe the change, and the temperature distribution is clicked and drawn, so that the result can be displayed directly to the shelter temperature field through a model trained by the neural network. And simultaneously, setting external environment parameters of the radiation field, setting the upper limit and the lower limit of a radiation scale as the same as the setting of the temperature field, clicking to draw radiation distribution, and directly displaying a result on the shelter radiation field.
Meanwhile, different monitoring points can be quickly checked through monitoring point measurement: the shelter left cabin and the shelter right cabin are arranged in the same external environment: solar radiation sunny day 1000 and wind speed 2m/s, temperature 20 0 The temperature change comparison curve of the two monitoring points is shown in fig. 1, the left cabin of the shelter refers to a left line with the peak at the left, and the right cabin of the shelter refers to a right line with the peak at the right.
Embodiment 8, an electronic device using the simulation of the fast algorithm of embodiments 1-7, comprising:
one or more processors;
a storage means for storing one or more programs;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the deep learning based shelter infrared characteristic simulation express algorithm provided by embodiment 1.
Embodiment 9, a computer readable storage medium storing a computer program implementing the deep learning based shelter infrared characteristic simulation express algorithm of embodiments 1-7 when executed by a processor using the simulation express algorithm of embodiments 1-7.
The above description is only of the preferred embodiments of the present invention, but the protection scope of the present invention is not limited thereto, and any person skilled in the art, within the scope of the present invention, can substitute or change the technical solution and the inventive conception of the present invention equally within the scope of the present invention.
Claims (5)
1. The shelter infrared characteristic simulation rapid algorithm based on deep learning is characterized by comprising the following steps of:
step 1, establishing a shelter temperature distribution calculation model;
step 2, reading in the temperature field distribution of the model, and calculating the self infrared radiation intensity of the surface of the model;
step 3, reading in solar radiation intensity, ground background radiation and sky background radiation received by the surface of the shelter, and calculating reflected radiation of the surface of the model to obtain a correlation database of total radiation quantity and environmental parameters of the surface of the shelter;
and 4, establishing an infrared radiation model, which comprises the following steps of:
step A, establishing a shelter geometric model;
step B, extracting simulation data of a temperature field, a radiation field and monitoring points;
step C, building a radiation distribution field of the cubic cabin and an external environment model through a convolutional neural network;
and D, taking the environmental parameters as input, taking the shelter radiation distribution field as output, training a convolutional neural network model, and obtaining a shelter infrared characteristic simulation rapid algorithm based on deep learning.
2. The deep learning-based shelter infrared characteristic simulation rapid algorithm of claim 1, wherein: the construction of the shelter geometric model is carried out according to the following steps:
establishing a three-dimensional geometric model of the shelter according to the selected shelter data;
and setting certain external conditions and environment parameters for the shelter three-dimensional geometric model to form an infrared simulation calculation domain.
3. The deep learning-based shelter infrared characteristic simulation rapid algorithm of claim 1, wherein: the simulation technology is adopted to simulate the model in full time period and various environmental backgrounds, and a simulation database is established, and the method specifically comprises the following steps:
dividing the wind speed, the temperature and the time of the external environment of the shelter into N groups, and dividing the wind speed, the temperature and the time of the external environment of the shelter into M weather conditions according to solar radiance and atmospheric irradiance to obtain different working conditions composed of N.M parameters;
and calculating the temperature field display and the infrared radiation display of the shelter under different working conditions by adopting a simulation technology method, and building an infrared simulation database of the shelter.
4. A deep learning based shelter infrared characteristic simulation rapid algorithm as claimed in claim 3, wherein: the simulation techniques include flow solid heat transfer, radiant heat transfer, and computational fluid dynamics simulation techniques.
5. The deep learning-based shelter infrared characteristic simulation rapid algorithm of claim 1, wherein: according to the simulation data of the extracted temperature field, the radiation field and the monitoring points, a cube cabin infrared radiation field distribution and external environment model is built through a convolutional neural network, and the method specifically comprises the following steps:
a. the environmental parameters comprise environmental temperature, solar irradiance and environmental wind speed; the material parameters include: ground material density, ground material thermal conductivity, ground material constant pressure heat capacity; defining relative Cartesian coordinates of the monitoring points and the shelter model as input data of simulation data, and defining shelter surface temperature distribution, infrared radiation distribution and monitoring point temperature as an objective function of the simulation data;
b. dividing the training set and the testing set in the same proportion for a plurality of times;
c. defining convolutional neural network parameters, a loss function, an activation function and a pooling layer;
d. introducing the training set into a convolutional neural network, and optimizing the calculation accuracy of the convolutional neural network by adopting a linear correction unit function;
e. the test set is brought into the optimized convolutional neural network, and a loss function is calculated;
f. modifying the convolution layer number and the neuron number of each layer of the convolution neural network, modifying the pooling method and parameters of the pooling layer, repeating d and e, and enabling the loss function value to be minimum.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210632586.8A CN115510732B (en) | 2022-06-07 | 2022-06-07 | Shelter infrared characteristic simulation rapid algorithm based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210632586.8A CN115510732B (en) | 2022-06-07 | 2022-06-07 | Shelter infrared characteristic simulation rapid algorithm based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115510732A CN115510732A (en) | 2022-12-23 |
| CN115510732B true CN115510732B (en) | 2023-07-11 |
Family
ID=84500827
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210632586.8A Active CN115510732B (en) | 2022-06-07 | 2022-06-07 | Shelter infrared characteristic simulation rapid algorithm based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115510732B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116718277A (en) * | 2023-06-06 | 2023-09-08 | 中国人民解放军92493部队试验训练总体研究所 | Island shore environment infrared target characteristic simulation method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113420639A (en) * | 2021-06-21 | 2021-09-21 | 南京航空航天大学 | Method and device for establishing near-ground infrared target data set based on generation countermeasure network |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101976275B (en) * | 2010-09-21 | 2012-11-07 | 北京航空航天大学 | Airplane infrared radiation and atmospheric transmittance modeling method |
| CN103186906A (en) * | 2011-12-28 | 2013-07-03 | 中国科学院沈阳自动化研究所 | Real-time infrared dynamic scene simulation method for multiple objects in sea and sky background |
| CN102968521B (en) * | 2012-10-25 | 2015-04-22 | 西安电子科技大学 | Infrared reflection characteristic simulation method based on GPU (graphics processing unit) programming |
| CN112215957A (en) * | 2020-09-16 | 2021-01-12 | 哈尔滨新光光电科技股份有限公司 | Aerial target infrared characteristic simulation method |
-
2022
- 2022-06-07 CN CN202210632586.8A patent/CN115510732B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113420639A (en) * | 2021-06-21 | 2021-09-21 | 南京航空航天大学 | Method and device for establishing near-ground infrared target data set based on generation countermeasure network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115510732A (en) | 2022-12-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yang et al. | Urban microclimate and its impact on built environment–a review | |
| CN112099110B (en) | Ocean internal wave forecasting method based on machine learning and remote sensing data | |
| CN110222828A (en) | A kind of Unsteady Flow method for quick predicting based on interacting depth neural network | |
| CN114462718A (en) | CNN-GRU wind power prediction method based on time sliding window | |
| CN111881851A (en) | Garden seedling intelligent detection and counting method based on UAV and convolutional neural network | |
| CN115510732B (en) | Shelter infrared characteristic simulation rapid algorithm based on deep learning | |
| CN108764527B (en) | Screening method for soil organic carbon library time-space dynamic prediction optimal environment variables | |
| CN103353295B (en) | A kind of method of accurately predicting dam dam body vertical deformation amount | |
| CN112100911A (en) | Solar radiation prediction method based on deep BISLTM | |
| CN115471679A (en) | Method and intelligent system for synchronously assimilating water level and flow of natural river | |
| CN115081677B (en) | Wind speed prediction method and device, electronic equipment and readable storage medium | |
| CN112100922A (en) | Wind resource prediction method based on WRF and CNN convolutional neural network | |
| CN115293424A (en) | New energy maximum power generation capacity calculation method, terminal and storage medium | |
| CN102509096B (en) | Extracting and processing method for inclination angles of corn plant leaves | |
| CN201716727U (en) | Geographical simulation system based on remote sensing and GIS | |
| CN110414734B (en) | Method for forecasting and evaluating wind resource utilization rate | |
| TWI717796B (en) | System, method and storage medium for estimating the amount of sunshine in geographic location by artificial intelligence | |
| CN115455838B (en) | High-spatial-resolution flow field reconstruction method for time-course data | |
| CN114638048A (en) | Three-dimensional spray pipe flow field rapid prediction and sensitivity parameter analysis method and device | |
| CN112749470B (en) | Layout optimization fitting method for structural deformation sensor | |
| Bai et al. | On the efficiency of a cfd-based full convolution neural network for the postprocessing of field data | |
| Jiang et al. | Optimization of sensor location for improving wind power prediction accuracy | |
| CN116108761B (en) | Regional climate simulation method and system for coupling deep learning and HASM | |
| CN110991685A (en) | Meteorological temperature prediction method | |
| CN115796076B (en) | Nuclear power temperature drainage reference temperature determining method and device fused with hydrodynamic model |
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 |