CN112348872A

CN112348872A - Physical process-based infrared cloud accumulation modeling method

Info

Publication number: CN112348872A
Application number: CN202011044260.0A
Authority: CN
Inventors: 黄曦; 雷越; 孙浩; 马向超; 吴鑫; 刘德连; 张建奇; 单仝
Original assignee: Xidian University
Current assignee: Xidian University
Priority date: 2020-09-28
Filing date: 2020-09-28
Publication date: 2021-02-09
Anticipated expiration: 2040-09-28
Also published as: CN112348872B

Abstract

According to the infrared integral cloud modeling method based on the physical process, the distribution of the characteristics of cloud liquid drop density, water vapor density, temperature and the like is obtained through the processes of water vapor transmission, phase change characteristics and the like existing in a calculation grid through a three-dimensional grid model of the integral cloud, then a multiple scattering simplification model is used for classifying and calculating the illumination phenomenon existing in a visible light model of the integral cloud according to the illumination direction and the sight line direction, after a scattering phase function and a basic scattering theory are determined, the radiation energy of multiple scattering is calculated, the spontaneous radiation energy of the integral cloud in an infrared scene is calculated, the total energy is obtained, the radiation energy of the visible light model is corrected, the calculation process of the spontaneous radiation in different wave bands is designed to obtain the integral cloud infrared illumination model, and then the illumination of the cloud is rendered in real time. Therefore, both infrared band cloud accumulation and non-infrared band cloud accumulation are considered, authenticity and real-time performance of cloud accumulation rendering are higher, and accuracy of cloud accumulation rendering is improved.

Description

Physical process-based infrared cloud accumulation modeling method

Technical Field

The invention belongs to the field of infrared waves, and particularly relates to a physical process-based infrared cloud accumulation modeling method.

Background

The infrared scene simulation is an effective means for analyzing and researching the infrared scene, and depends on accurate three-dimensional modeling and strict physical process simulation, so that the infrared virtual scene has strong sense of reality and can meet a plurality of military requirements and civil requirements. Simulating an infrared airborne scene requires analyzing targets and scenes, especially some common airborne scenes, including clouds, sunlight, rain and snow. Clouds, which are typical air scenes, are often present in flight simulation and outdoor scene simulation, but their appearance, such as optical characteristics, shape, etc., is related to altitude, climate, weather, sunlight, and observation angle, etc., so that the simulation of clouds is rather complicated and difficult. In the prior art, a modeling method of accumulating spherical particles in a three-dimensional space is adopted to obtain a cloud accumulation shape with layering, and the result of the cloud accumulation shape is compared with the result of an improved noise modeling method, so that the advantages and the disadvantages of the two methods are analyzed on the basis of ensuring the rendering efficiency. Because the three-dimensional model in the existing cloud simulation system mostly adopts the traditional modeling method, the physical process in the cloud is lack of accurate description, and in the aspect of an illumination model, a simpler forward scattering model is adopted, and various illumination phenomena in the cloud, such as secondary scattering and multiple scattering effects, exist. On the whole, the existing cloud simulation system is difficult to realize both authenticity and real-time performance after integrating three-dimensional modeling and illumination modeling, is mostly suitable for the field of visible light, and is lack of an illumination model adaptive to an infrared band.

Disclosure of Invention

In order to solve the problems in the prior art, the invention provides an infrared cloud accumulation modeling method based on a physical process. The technical problem to be solved by the invention is realized by the following technical scheme:

the embodiment of the invention provides a physical process-based infrared product cloud modeling method, which comprises the following steps:

s1, acquiring a three-dimensional grid model of the cloud;

wherein each mesh in the three-dimensional mesh model stores first variable information for a plurality of particles, the particles representing clouds at respective locations within the mesh, the variable information comprising: droplet density, temperature, and velocity;

s2, determining the cumulus cloud diffused according to the density, the temperature and the speed of the liquid drop based on the density, the temperature and the speed of the liquid drop;

s3, determining a scattering phase function based on an included angle between the incident direction and the scattering direction of the light irradiating the cumulus;

s4, determining an effective phase function of the droplet size distribution based on the width of the droplet size distribution in the cloud and the characteristic radius of the droplet size distribution;

s5, aiming at the clouds in different wave bands, calculating average scattering efficiency factors according to the scattering efficiency factors in different wave bands;

s6, averaging scattering phase functions of the product clouds in different wave bands to determine an average scattering function;

s7, calculating first radiation energy of the light after multiple scattering in the cloud point along the sight line direction and the light direction of the color device after the light enters the cloud point based on the average scattering function and the average scattering efficiency factor;

s8, respectively calculating spontaneous radiation energy corresponding to the cloud accumulations in the far infrared wave band and the intermediate infrared wave band;

s9, summing the spontaneous radiation energy and the first radiation energy to determine the total energy of the cloud in the sight line direction;

s10, multiplying the total energy by the maximum pixel value of each channel in the image three primary color channels respectively to obtain the pixel value of the cloud in the sight line direction;

and S11, rendering the diffused cloud product at the position in the sight line direction according to the pixel values, and repeatedly executing S8-S11 until the diffused cloud product is rendered at each position.

According to the embodiment of the invention, the distribution of the characteristics of cloud, such as droplet density, vapor density, temperature and the like is obtained by calculating the processes of vapor transmission, phase change characteristics and the like existing in a grid through a three-dimensional grid model of an integral cloud, then, a multiple scattering simplified model is used for classifying and calculating the illumination phenomenon existing in a visible light model of the integral cloud according to the illumination direction and the sight line direction, after a scattering phase function and a basic scattering theory are determined, the radiation energy of multiple scattering is calculated, the spontaneous radiation energy of the integral cloud in an infrared scene is calculated, so that the total energy is obtained, the radiation energy of the visible light model is corrected, the calculation process of the spontaneous radiation under different wave bands is designed to obtain the integral cloud infrared illumination model, and then the illumination of the cloud is rendered in real time. Therefore, both infrared band cloud accumulation and non-infrared band cloud accumulation are considered, authenticity and real-time performance of cloud accumulation rendering are higher, and accuracy of cloud accumulation rendering is improved.

The present invention will be described in further detail with reference to the accompanying drawings and examples.

Drawings

Fig. 1 is a schematic flow chart of a modeling method of an infrared product cloud based on a physical process according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a method for performing integration of a line of sight direction and a light direction, and circularly processing spontaneous scattering calculation and scattering calculation of an infrared band to obtain radiation energy according to an embodiment of the present invention;

FIG. 3 is a detailed flowchart for circularly processing spontaneous scattering calculation and scattering calculation of an infrared band to obtain radiation energy by integrating a line of sight direction and a light direction according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a three-dimensional mesh model provided by an embodiment of the invention;

fig. 5 is a flowchart for solving a transport effect caused by a fluid velocity field due to speed, temperature and density by using a semi-lagrangian method according to an embodiment of the present invention;

FIG. 6 is a flow chart of cloud point diffusion provided by an embodiment of the present invention;

FIG. 7 is a schematic diagram of an initial clouding provided by an embodiment of the present invention;

FIG. 8 is a schematic illumination of an integrating cloud provided by embodiments of the present invention;

FIG. 9 is a schematic diagram of calculating total energy by setting sampling points from the direction of a line of sight and the direction of a light according to an embodiment of the present invention;

fig. 10 is a result of infrared illumination simulation of clouds in an infrared scene according to an embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to specific examples, but the embodiments of the present invention are not limited thereto.

Example one

As shown in fig. 1, the method for modeling an infrared product cloud based on a physical process provided by the present invention includes:

s1, acquiring a three-dimensional grid model of the cloud;

wherein, the total energy needs to be quantized, so that the quantized total energy value is in the interval of 0-1.

Referring to fig. 2, after the three-dimensional model is obtained in fig. 2, integration of the sight line direction and the light ray direction needs to be implemented, spontaneous scattering calculation and scattering calculation of the infrared band are processed in a circulating manner, so that radiation energy is obtained, graying of the radiation energy is rendered, and then a formula and a principle of each step are described in more detail.

Example two

As an alternative embodiment of the present invention, the step of S2 includes:

s21, setting an initial cloud accumulation;

s22, initializing the drop density, temperature and speed of the particles in each grid in the initial cloud accumulation;

s23, moving each particle in the grid according to the initialized speed;

s24, calculating the external force of each particle after moving based on the wind force, buoyancy and eddy current restraining force of the particles;

s25, calculating the density and the temperature of the liquid drops of the moved particles based on the current atmospheric pressure of the cloud; 2.26

S26, simplifying the obtained fluid motion equation based on the external force to obtain a simplified fluid motion equation;

s27, solving the simplified fluid motion equation to obtain the speed of the particles in the current unit time;

and S28, replacing the initialized speed with the speed at the current unit time, and repeatedly executing the steps from S23 to S27 until the iteration times are reached to obtain the diffused cloud.

EXAMPLE III

As an alternative embodiment of the present invention, step S26 includes:

acquiring a Navier-Stokes equation as a fluid motion equation;

and setting a minimum unit of a middle speed field and time division, and simplifying the fluid motion equation.

Example four

As an alternative embodiment of the present invention, the step of S4 includes:

acquiring a scattering phase function;

determining different droplet size distributions based on the total particle number in the water droplet, the width of the water size distribution, a gamma distribution function, the characteristic radius of the size distribution and the effective radius corresponding to the size distribution;

and based on the different droplet size distributions, weighting and summing the scattering phase functions of the different droplet sizes to obtain an effective phase function of the droplet size distribution.

EXAMPLE five

As an alternative embodiment of the present invention, the step of S8 includes:

aiming at the cumulus cloud in the far infrared band, calculating the spontaneous radiation energy corresponding to the cumulus cloud by using a Planck formula;

and calculating spontaneous radiation energy aiming at the cloud accumulation of the mid-infrared band based on the set radiation emissivity.

EXAMPLE six

As an optional implementation manner of the present invention, before the step of S8, the modeling method further includes:

and determining the emissivity of the light in the far infrared band and the radiation emissivity of the medium infrared band according to the light absorptivity and the energy conservation law.

EXAMPLE seven

As an alternative embodiment of the present invention, the step of S7 includes:

s71, setting light parameters; the light parameters include: the light projection direction, the illumination direction, the light projection step length and the maximum times of light projection;

s72, determining a sight direction sampling point according to the light projection step length along the sight direction;

s73, determining second variable information at the current sampling point of the sight line direction;

s74, calculating a transmittance in the visual line direction based on the second variable information;

s75, determining second variable information at the current sampling point of the light direction according to the light projection step length along the light direction of the current sampling point;

s76, calculating the transmittance of the current sampling point in the light direction based on the second variable information;

s77, integrating the transmittance;

s78, repeating the steps from S75 to S76 until the light transmittance of all the sampling points in the light direction is calculated;

s79, calculating the first scattering rate and the second scattering rate of the light subjected to scattering action along the light ray direction;

s80, calculating second radiation energy based on the primary scattering rate and the secondary scattering rate;

s81, accumulating the second radiation energies respectively;

s82, taking the next sampling point along the sight direction as the current sampling point, and repeatedly executing the steps from S73 to S82 until the sampling point along the sight direction is polled;

s83, calculating the radiation proportion after multiple scattering along the light direction;

and S84, calculating the first radiation energy based on the radiation proportion and the accumulated second radiation energy.

Referring to fig. 3, the whole process of polling the sight line direction sampling points and the light line direction sampling points is shown in fig. 3, which is a demonstration process of the seventh embodiment, and the specific process is the same as the implementation process of the seventh embodiment.

Example eight

As an alternative embodiment of the present invention, the step of S78 includes:

determining an extinction cross section based on the effective radius of the droplet and the particle number density of the droplet;

based on the matte interface, a light transmittance is determined.

Example nine

As an alternative embodiment of the present invention, the step of S79 includes:

determining a projection function of the light rays passing through the cloud based on the transmittance, the sight line direction and the light ray direction of the light rays passing through the cloud layer of the cloud;

determining a primary scattering function of light passing through the cloud;

determining a secondary scattering function of the light passing through the cloud;

calculating a primary scattering power using the primary scattering function and calculating a secondary scattering power using the secondary scattering function.

Example ten

As an alternative embodiment of the present invention, the step of S83 includes:

when the cloud is in a thermal equilibrium state and in multiple scattering, determining the transmittance of light, a primary scattering proportion, a secondary scattering proportion and an energy conservation equation of the multiple scattering proportion;

when the multiple scattering is uniform, determining the radiation proportion of a single bin based on the energy conservation equation;

when the multiple scattering is non-directional, the proportion of radiation in different directions on a single area source is determined based on the energy conservation equation.

The following is a detailed description of the basic principles in the embodiments of the present invention and the steps in the first to tenth embodiments.

The meaning of the parameters is described first.

_uRepresenting velocity, t time, ρ density, p pressure, f force, u^*Representing the intermediate velocity field, theta the potential temperature, q the particle density, T the Kelvin temperature, II the Exner function, p₀Denotes the standard atmospheric pressure, R_dDenotes the gas constant of dry air, c_pDenotes the constant pressure specific heat capacity of dry air, c_vDenotes the volumetric specific heat capacity of dry air, f_windRepresenting wind action, f_buoShowing buoyancy effect, f_vcRepresents the eddy current suppressing force, θ_vRepresenting the temperature of the virtual potential, theta_v0Denotes a reference virtual temperature, q_cDenotes the droplet density, q_vDenotes the water vapor density, ω denotes the rotation of the fluid velocity field, l denotes the unit length of the grid, q_vsRepresenting saturated vapor density, z representing altitude, Γ representing air temperature vertical decrement rate, L representing latent heat of water vaporization, θ_bRepresenting the intermediate potential temperature, p (theta) representing the scattering phase function, k representing the wavenumber, Csca representing the scattering cross-section, X representing the vector scattering amplitude, N (r) representing the droplet size distribution, gamma representing the width of the size distribution, Γ (gamma) representing the gamma distribution function, N (gamma) representing the scattering phase function₀Denotes the total number of particles in the water droplet, r_nCharacteristic radius, r, representing the size distribution_eDenotes an effective radius, P (theta) denotes an effective phase function, tau (d) denotes transmittance,

the direction of the line of sight is indicated,

denotes the direction of light, Q denotes the scattering efficiency factor, σ denotes the extinction cross-section, κ denotes the extinction coefficient, S₀Denotes transmission, S₁Denotes primary scattering effect, S₂The effect of the secondary scattering is shown,

the total permeation ratio is expressed as a ratio of total permeation,

the ratio of the total primary scattering is expressed,

the ratio of the total secondary scattering is expressed,

the ratio of the total multiple scattering is expressed,

representing the mean effective phase function, M_λbbIndicating the degree of spectral radiation exitance, c₁Denotes a first radiation constant, c₂Representing the second radiation constant, alpha the absorptivity, epsilon the emissivity and G the quantization grey scale.

In the first embodiment, the three-dimensional model includes information in three dimensions, including information in three directions of XYZ axes. In order to facilitate the rapid use and access of such information, a suitable model needs to be designed. The embodiment of the invention adopts a three-dimensional grid to store the information, stores the information such as density, temperature, speed and the like in the cloud, and stores the result in a three-dimensional map so as to facilitate the use and access in the GPU calculation process.

As shown in FIG. 4, the three-dimensional model in this chapter is built under a left-handed coordinate system, in which the XYZ axes are shown in the forward direction. The cube in the figure is a three-dimensional grid for storing cloud model information, the grid is divided into a plurality of uniform grids, 4 × 4 grids are shown in the figure, and the grid fineness is far greater in the actual calculation process. Each of the divided uniform cells is called a voxel (voxel), and each voxel stores cloud information of a current location. Clouds are generated in the grid, and in order to control the shape of the clouds and observe the changes in the clouds, the volume of the clouds is typically smaller than the volume of the grid. When the grid division is fine enough, the voxels in the cloud volume can represent the particles in the cloud, storing the droplet information and the moisture information at the current location.

The coordinate system in fig. 4 is also the coordinate system of the virtual scene simulated by the embodiment of the invention, and the observation camera is generally positioned on the axis in the positive direction of the X axis, and the observation direction is (-1,0, 0). Some azimuthal descriptions of embodiments of the present invention are relative to a cloud grid, which is at the origin of the coordinate axes. Wherein, front refers to the positive direction of the axis relative to the cloud grid, up refers to the positive direction of the Y axis relative to the cloud grid, right refers to the positive direction of the Z axis relative to the cloud grid, and rear, lower and left correspond to the positive direction, the lower and the left respectively.

The implementation of the two steps of the example and the parameters required in the process are detailed below.

(1) Non-compressible condition

True fluids all have different degrees of compressibility, and aerodynamically, the change in density of a gas is negligible, depending on the velocity of the gas flow. Generally, when the gas flow velocity is lower than mach 0.3, a low-velocity aerodynamic method is applied, and the gas can be considered incompressible; high velocity aerodynamic methods are applicable when the gas flow velocity is above mach 0.3, where the gas should be considered compressibly flowing.

(2) Has negligible viscous force

Reynolds number is a dimensionless number that can be used to characterize the flow of a fluid, and when Reynolds number is less than or equal to 1, i.e. the viscosity of a fluid is small and the relative sliding speed is not large, the viscous stress around the boundary layer of a flowing object is small, and the viscous term in the equation can be ignored, i.e. an ideal fluid can be considered. For water, air, and the like under natural conditions, since the moving speed is small and the density is low, this condition is satisfied and the viscous force can be ignored.

The fluid motion equation and the simplified process in the second embodiment of the present invention are described in detail.

(1) Equation of motion of fluid

The Navier-Stokes equation (Navier-Stokes equation, N-S equation for short) is a motion equation for describing momentum conservation of viscous incompressible fluid. Since the fluid expansion speed is much lower than the sonic speed, the viscous force term of the fluid can be ignored, so that the N-S equation can be simplified into the Euler equation, and the form is as follows:

meanwhile, continuity conditions are as follows:

in the formula_uThe expression (2-1) is the expression of conservation of fluid momentum, the right term is the motion update of a fluid velocity field, the second term is the acceleration caused by fluid pressure gradient, and the last term is the acceleration caused by external force action, and generally comprises gravity, wind power and the like. The formula (2-2) is a continuity equation, which shows that the fluid velocity field has no divergence and guarantees conservation of mass.

It is impractical to solve the euler equation completely in real time, and some approximation means is required. The embodiment of the invention divides the formula (2-1) into two parts, and firstly, an intermediate speed field u is calculated^*And assuming that the minimum unit of time division is Δ t, there are:

for equation (2-3), the first term on the right of the equal sign is the transport result of the fluid velocity along with the fluid motion, which can be approximately solved by using the semi-lagrange method, and the same method can be used to update other constants such as temperature and water density (herein, both refer to the water vapor and liquid drops contained in the voxel):

where θ is the potential temperature of the fluid and q represents the density of the particles (gaseous or liquid). The potential temperature is constant under an adiabatic approximate altitude change and can be calculated as:

in the formula (2-6), T is temperature, K is unit, II is an Exner function, and the non-dimensional pressure of the gas can be regarded as. p is a radical of₀Is the standard atmospheric pressure at sea level, and p is the pressure at the height. R_dGas constant of dry air (287 Jkg)^-1K^-1)，c_pAnd c_vRespectively dry air constant pressure specific heat capacity and volumeSpecific heat capacity.

(2) Water vapor transport effect

The process of solving the transportation effect of speed, temperature and density caused by the fluid velocity field by the semi-Lagrange method mainly comprises five steps, as shown in FIG. 5:

taking the temperature as an example, the temperature and the velocity v of the position pos to be updated are obtained first, then the position pos is rewound by half step length backwards to calculate the middle position midpos-v 0.5-dt, and the velocity v of the middle position midpos is obtained_midAnd backwards rewinding one step to calculate the final position finalpos-v_midDt, finally update the temperature of the pos location with the temperature of the finalpos location. Thus, the temperature field transport with the fluid velocity is completed.

(3) Action and correction of forces

After the transmission of the speed-dependent speed field is completed, the acceleration f caused by external force action is considered, the external force action generally comprises the following three items, and the wind force f_windBuoyancy f_buoAnd eddy current suppression force f_vcThus, they have the formula (2-8).

f＝f_vc+f_buo+f_wind (2-8)

For wind force f_windThe simulation is usually independent with respect to the cloud and therefore the setting can be made directly or typical wind speeds under simulated conditions can be taken. Buoyancy force f_buoThe calculation of (a) includes gravity, which is actually the result of the force analysis of the vertical direction of the air mass in one voxel. The higher the water-vapor ratio in the air mass is, the larger the buoyancy is; the higher the droplet fraction, the greater the gravity and the less the buoyancy. In addition, the calculation of the buoyancy is also related to the temperature and the pressure of the environment. The specific calculation is as follows:

wherein g is the acceleration of gravity,

is a unit vector of vertical orientation, q_cIs the drop density, θ_v0Is referred to as a virtual temperature, and generally takes a value between 290K and 300K. Theta_vThe virtual potential temperature represents the potential temperature of dry air under the same density and pressure conditions, and the calculation method is as follows:

θ_v＝θ(1+0.61q_v) (2-10)

wherein q is_vIs the water vapor density.

The equations (2-10) are continuous equations, and only discrete grid calculation can be adopted in the actual calculation process, namely discrete difference equations are used for approximating the solutions of the continuous equations. The deviation of the discrete difference equation from the continuous equation in the result and the relatively coarse gridding of the calculation force requirement determine that the details of the calculation result on a small scale, such as the rotation effect of the fluid, are difficult to embody. To solve this problem, an eddy current suppression force f is introduced_vcThese fluid details are represented using a series of calculations based on the velocity field. The eddy current suppression force can be calculated by first calculating the curl ω of the fluid velocity field as follows:

and then calculating the gradient direction of the rotation field omega and unitizing:

and finally, calculating to obtain the eddy current restraining force:

f_vc＝ηl(N×ω) (2-13)

in the formula (2-13), l is the unit length of the grid, eta is the eddy current suppression coefficient, and the strength of the eddy current suppression force, namely the small-scale effect compensation effect in the motion process of the cumulus is determined, and generally the value is between 0.1 and 10, depending on the specific situation.

(4) Resolution of pressure field

After calculation of the force correction, we have an intermediate velocity field and this velocity field satisfies the equation after neglecting the influence of the density field gradient:

the right side of the above equation is known, so that the equation is a poisson equation, the pressure p can be solved by using an iterative method, and the velocity is corrected by using the pressure, so that a final velocity field can be obtained:

at this point, the motion characteristics of the fluid in the unit time are calculated.

(5) Phase transition characteristics of particles

The phase change of the cloud refers to the change of state of water occurring in the cloud, i.e., the conversion process between liquid droplets, water vapor, and ice crystals. In the embodiment of the present invention, it is assumed that the occurrence of the phase change is in units of voxels in the cloud, that is, a process of changing the state of water molecules in each voxel. Typically, the water vapor density exceeds the saturation point under the current conditions, and the water vapor is converted into droplets, thereby forming a cloud that is seen later. The saturation point of the density of water vapor in the air is related to the temperature T and the atmospheric pressure p of the current position, and can be calculated by the following method:

the pressure in the equation (2-16) needs to be corrected by taking the pressure at the current position into consideration:

z in the formula (2-17) is the altitude of the current position, p₀The standard atmospheric pressure can be 100kPa, T₀For the surface temperature, 280k-310k can be taken, and Γ is 10k/km, which means that the temperature drops by 10k for every 1km rise.

For a cloud, the conservation of water during the phase change process must be considered, and in the cloud, the following continuous conditions are satisfied:

q in the formula (2-18)_vIs the density of water vapor, q_cIs the drop density. C is constant, namely the total amount of water molecules undergoing phase change is constant.

When the density of the water vapor is larger than the saturation point, phase change occurs, and the density of the water vapor and the density of the liquid drops change:

Δq_v＝-C＝min(q_vs-q_vb,q_cb) (2-19)

q_v＝q_vb+Δq_v (2-20)

q_c＝q_cb-Δq_v (2-21)

q in the formula (2-19)_vb、q_cbRefers to the water vapor density and drop density within the voxel at the previous state.

While the adiabatic approximation is a valid approximation of the movement of air with unsaturated water vapor, it cannot be assumed that the potential temperature of saturated air is constant. If the water vapor density in the voxel exceeds the saturation point, the water vapor will condense and release latent heat, causing the voxel air mass to heat up. If the latent heat and cooling due to condensation and evaporation are the only non-adiabatic changes, the first thermodynamic law holds, as follows:

in the formula (2-22), L is latent heat of vaporization of water and is 2.501J/kg at 0 ℃. The above equation shows that the potential temperature change is related to the phase change of the current position in addition to the transport change with the velocity field. So after considering the transport variation, there is the following equation:

theta in the formula (2-23)_bRepresenting the intermediate potential temperature after potential temperature transport with the velocity field. At this point, the change characteristic of the phase change of the particles is solved.

The density field of the cloud under a unit time (represented as a frame in the simulation process) is calculated and rendered, besides the motion, phase change, heat energy conversion and the like of each voxel are required to be calculated, then the partial differential equation is required to be iteratively solved to obtain the pressure, and the velocity field is corrected by using the pressure distribution. Such complex calculations require a considerable number of calculations, but the real-time nature of the simulation requires a limit on the calculation time per frame. In order to realize real-time simulation (the frame frequency reaches at least 30), the embodiment of the invention transfers all simulation calculation into the GPU, and optimizes repeated calculation by utilizing the high parallelism of the GPU, thereby improving the frame frequency to meet the real-time simulation requirement while ensuring high authenticity. A calculation process of each voxel inside the cloud is shown in fig. 3 at one frame time, fig. 6 shows a process of implementing S2 according to the second embodiment, where initially, the initial cloud is shown in fig. 7, a white area represents the initial cloud, and a box represents a grid, and it is necessary to diffuse particles of each grid to obtain a diffused cloud.

The following is a detailed description of the process of solar ray irradiation cloud accumulation emission according to the first to the tenth embodiments.

As shown in fig. 8, which is a schematic view of the illumination of the cloud, the sunlight is not in the same direction as the direction of the line of sight, and the sunlight enters the viewpoint through scattering. For a point P in the cloud, consider the ray that point P enters the viewpoint. Firstly, light enters a point P through a part of cloud layers, and primary scattered light is emitted along the sight line direction after being scattered by the point P; second, the light passes through a portion of the cloud layer to other points, e.g. P₁、P₂The illumination enters the sight line after being scattered for the second time and the third time; finally, diffuse reflection on the surface of the cloud layer, self radiation and the like enter the sight line. In addition, since the sight line ends differently at different points, sky background radiation needs to be considered, for example, sky radiation transmitted in the sight line direction needs to be considered in fig. 8. When the radiation simulates the skyIn the case of background, the sky radiation may be processed with a background radiation metering algorithm. When the direction of the line of sight is the same as the direction of illumination, the solar radiation transmitted in the direction of the line of sight needs to be taken into account.

(a) Scattering phenomena in clouds

(1) Scattering calculation method

The scattering effect in the cloud is calculated by first calculating the scattering phase function. The scattering phase function describes the angular distribution of energy of incident energy after scattering by the particle. And (3) setting an included angle theta between the incident direction and the scattering direction as a scattering angle, and introducing a dimensionless parameter to describe the change of the scattering energy along with the scattering angle, wherein the dimensionless parameter is called as a phase function p (theta). For the phase function, there is a normalization condition:

the Mie scattering phase function (hereinafter referred to as Mie function) can be obtained by solving maxwell's equations, but the process is too complicated and will not be described in detail herein. The scattering phase function p of the final spherical particle at a single wavelength can be expressed as:

p＝2π(|S₁|²+|S₂|²)/(k²Csca) (3-2)

wherein k is wavenumber, Csca ═ -_4πX|/k²d Ω is a scattering cross section, X is a vector scattering amplitude, and S1 and S2 are complex amplitudes, which can be obtained by solving the bezier equation, and will not be described in detail in the embodiments of the present invention.

The low-altitude cloud deposit is formed by mixing water drops with different sizes, and the size of the water drops has great influence on the optical characteristics of the cloud under the assumption that the water drops are spherical: for a certain amount of water, the smaller the water droplet, the more opaque the cloud. Approximately half the size reduction doubles the extinction efficiency, and the Mie scattering phase function also varies with the size of the drop. To analyze this, the droplet size needs to be modeled first. A droplet size distribution in cloud (DSD) model approximates the modified gamma distribution. Is defined as:

wherein N is₀＝300cm^-3γ is the width of the size distribution, and is usually 2. Gamma (gamma) is a gamma distribution function, r_nThe characteristic radius of the size distribution. r is_e＝(γ+2)r_nFor the effective radius corresponding to this distribution, r is taken in the embodiment of the present invention_e7 μm. After the parameters are determined, the droplet size distribution can be obtained.

The effective phase function P (θ) corresponding to the droplet size distribution can be obtained by weighting and summing the phase functions for different droplet sizes, the summation formula being as follows:

finally, the corresponding effective phase function under the designated wavelength can be obtained.

Because the overall calculation process of the effective phase function is quite complex and is difficult to be embedded into real-time calculation, a pre-calculation method is mostly adopted, and the effective phase function is stored in tools such as textures and the like for direct value taking of real-time calculation, or the calculation can be simplified through high-order function simulation, so that the real-time performance of the simulation process is ensured.

After the droplet size model and the phase function are determined, the specific scattering process can be considered. The sight line direction is assumed to be vertical and downward, the sunlight direction is vertical and downward transmitted, and the transmission refers to a process that light rays directly penetrate through the cloud layer without being scattered by particles and finally enter a viewpoint. When the spontaneous radiation of the cloud layer is ignored, only the attenuation effect of the cloud layer on light needs to be considered in the process, the scattering effect of particles does not need to be considered, and the process can be finally expressed as follows:

wherein τ (d) is the transmittance of light passing through the cloud layer in the direction of the line of sight

And the direction of light

The transmission effect does not need to be taken into account at the same time.

Assuming that the individual voxels within the cloud are uniformly distributed, the permeability of a voxel can be determined by the effective radius r of its drop_eAnd the particle number density N of the droplets. With these two parameters, the extinction cross-section can be defined:

where Q is the extinction efficiency factor and is a function of wavelength. Extinction coefficient from extinction cross section:

κ＝Nσ (3-7)

the cloud transmittance can be obtained by this:

where x is the designated ray path and the visible transmittance τ is the integral over that path. Considering all the voxels through which the path passes, a transmission result accumulated by voxels may be obtained, where l represents the corresponding length of the voxel grid. From the frequency perspective, this is the result of attenuation of a beam of light after it has traveled a path; from a probabilistic point of view, this is the probability that a ray will not encounter a particle through a path, i.e., the probability that no scattering will occur through a path. The probability of scattering of a hit particle within this path x is therefore 1- τ, for which the derivative function:

s(x)＝κe^-∫κdx (3-9)

it can be understood as the probability density of hitting a particle at a certain position on path x, so s (x) is the probability density of hitting exactly at the end of a free path of length x.

(2) Primary scattering process

The primary scattering refers to the process that light rays are scattered by only one particle and then penetrate through the cloud layer to finally enter a viewpoint. When the spontaneous radiation of the cloud layer is ignored, the light attenuation effect of the cloud layer and the primary scattering effect of the particles need to be considered in the process, and the effect can be finally expressed as:

S₁＝s(d_l)P(θ_vl)τ(d_v) (3-10)

the first term on the right of the equal sign is the probability that the light directly hits a specified particle (e.g., particle P) through the cloud layer; the second term is a phase function, which represents the probability that a particle scatters in a given direction (e.g., the line-of-sight direction v) at the current angle; the third term is the probability that a ray leaves a particle and enters the viewpoint directly without collision.

(3) Secondary scattering process

The secondary scattering refers to a process that light rays are scattered by two particles and then penetrate through the cloud layer to finally enter a viewpoint. When the spontaneous radiation of the cloud layer is ignored, the light attenuation effect of the cloud layer and the primary and secondary scattering effects of the particles need to be considered in the process, and the effect can be finally expressed as:

for the secondary scattering process, one defined scattering path contains two scattering particles: first scattering particles and second scattering particles. The light rays hit the first scattering particles through the cloud layer, then hit the second scattering particles, finally are transmitted out of the cloud layer and finally enter the viewpoint. As with the primary scattering process, the scattering exit particles, i.e. the secondary scattering particles, are considered. The scattering particles scatter all the primary scattered light hitting the particle, for example, for the point of exit P, P1 and P2 and other particles act as primary scattering particles for P, providing primary scattered light to them.

In fact, all particles in the cloud can be used as primary scattering particles, and the angle and brightness of only primary scattering light rays can be distinguished. However, in the real-time simulation process, because the calculation cost is too high, all possible primary scattering particles cannot be considered for the scattering of each emergent particle, and therefore some approximation algorithms need to be adopted for simplification.

The right term of equation (3-11) is used to integrate the light rays with different angles and different distances, and the integration is too complex, so that the calculation needs to be simplified, and the right two and three equations are approximate calculations. The formula I to formula II is based on the assumption that the radiation weights of different paths under the same scattering angle are the same, and the formula II to formula III is to replace the extinction coefficient of the primary scattering particles with the extinction coefficient of the secondary scattering particles so as to simplify the calculation. The first simplified calculation is mainly established on the basis that the larger the path difference is, the lower the transmittance is, and the smaller the weight is; the second assumption is based primarily on the fact that the population distribution is not very different when the droplets are close together in the cloud.

After approximate integration, the result is independent of the position of the primary scattering point, and the final calculation result is only dependent on the position of the secondary scattering point, similar to the primary scattering calculation. Therefore, the scattering effect of the scattering point as a secondary scattering point can be calculated while the primary scattering is calculated, so that the calculation is avoided for multiple times, and the calculation pressure is reduced.

(4) Multiple scattering calculation

The multiple scattering refers to a process that light rays are scattered by three or more particles and then penetrate through the cloud layer to finally enter a viewpoint. When the spontaneous radiation of the cloud layer is neglected, the attenuation effect of the cloud layer on the light and the three-time scattering effect and above of the particles need to be considered in the process.

Multiple scatterings in clouds can be approximated using methods similar to secondary scatterings, but neither the efficiency of the calculation nor the accuracy of the calculation results are satisfactory. This is mainly due to the following reasons: 1. the multiple scattering effect theoretically includes 3 to infinite scattering process calculations, even if a certain scattering exit point is determined, the scattering effect cannot be calculated one by one, and finally, an approximate algorithm is still adopted to ignore or simplify a part of calculation. Even then, the results obtained are affected by the approximation algorithm. 2. The approximate calculation of the secondary scattering always comprises the characteristics of directionality and symmetry, wherein the directionality refers to that the secondary scattering approximate result is similar to the primary scattering and comprises most of emergent energy at a small angle of an incident direction; symmetry means that the secondary scattering approximation is symmetric, i.e. the scattering results are the same at positive and negative scattering angles. For multiple scattering, the scattering directivity is obviously reduced due to multiple scattering of light, and finally the scattering directivity disappears completely with the increase of the scattering times, and the scattering energy distribution at different angles is the same. Therefore, if the multiple scattering is processed by using the approximation algorithm of the second scattering, the calculated result will also contain strong directivity, which is greatly different from the actual result.

For the multiple scattering calculation results, the approximation is performed with the following assumptions: 1. influence of scattering times on a scattering result is ignored, and scattering effects above a third scattering level are considered in a unified mode. 2. Neglecting the effect of the incident light direction on the multiple scattering effect, the multiple scattering effect is assumed to be uniform. 3. The directionality of the scattering effect is ignored and approximated by a diffuse reflection model.

When the cloud is in thermal equilibrium, based on assumption 1 and the law of conservation of energy, the following formula can be obtained:

in the formula

Which represents the transmittance of light rays, is,

represents the proportion of the primary scattering and the secondary scattering of the whole body,

represents the multiple scattering ratio of

Thus, the total multiple scattering ratio can be obtained, having

Based on the assumption 2 that the radiation ratio of the individual bins A can be obtained, there are

Based on the assumption 3, the radiation ratios in different directions on a single area source can be obtained

And theta is an included angle between the observation direction and the normal vector of the emergent position patch.

When the influence of the shape of the cumulus cloud is neglected, the simplest uniform sphere model is used for approximate calculation, and a faster calculation method can be obtained. The only variables in this method are the direction of the incident light, the size of the cloud, and the average number of particles of the cloud. In a simulation scene, the direction of sunlight is generally fixed, and when the preset shape of the cloud to be simulated is determined, the final stable shape is also fixed, so that the three variables can be preset and pre-calculated, the simulation efficiency is accelerated, and the real-time requirement is met.

(5) Sky radiation

Because the sky radiation effect is much smaller than the sunlight illumination effect, the scattering effect of the sky radiation after entering the cloud is not considered and is treated as background radiation. As shown in fig. 9, a patch at the intersection of the observation direction and the sky emits radiation energy as a main body of sky radiation in the corresponding view direction, and enters the viewpoint through penetration of clouds, which finally affects the observation effect. Therefore, the energy of the sky radiation is related to the position, the temperature and the time period of the patch, the attenuation of the energy under the atmospheric path, the attenuation of the energy through the cloud, and the like, and the sky radiation is only related to the position of the patch and the attenuation of the energy through the cloud under the condition that the atmospheric influence is ignored and the temperature and the position of the sun are not changed in the simulation time period. The position of a small patch, particularly the zenith angle of a viewpoint corresponding to the position of the small patch determines the initial energy of sky radiation, the energy corresponding to the position is stored in texture after the energy corresponding to the position is calculated by adopting a preprocessing method, and the energy is directly obtained by sampling on the texture when the radiation energy is calculated; the attenuation through the cloud can be expressed by using the transmittance of the cloud, which is related to the shape and density distribution of the cloud and can be obtained in the subsequent calculation.

Further details of the self-radiation process in an infrared scene are needed below.

(a) Infrared illumination model

There are many differences in simulation between the cloud in the infrared scene and the visible light scene, the most important of which is the handling of the spontaneous emission of the cloud and secondly the problem of scattering calculations due to the difference in wavelength bands.

(1) Scatter calculation correction

The scattering calculation of the infrared band requires two corrections on the basis of the calculation of the visible band: 1. scattering efficiency factor variation due to different wavebands; 2. the infrared band generally takes into account the width of the band and modifies the scattering phase function accordingly.

For the first point, the average value of the efficiency factors in the wave band is substituted when the formula is calculated, and when the difference of the efficiency factors in the wave band is not large, the efficiency factor corresponding to the intermediate wavelength can be used as the average value. At a wavelength of 4 μm, a refractive index of the droplets of 1.351 and an effective particle radius of 7 μm, an extinction efficiency factor of about 2.6 was obtained.

For the second point, the scattering phase function corresponding to each wavelength in the corresponding band needs to be calculated. After the scattering phase function at a single wavelength is calculated according to the equation (3-4), the average phase function corresponding to the entire band is calculated.

Equation (4-1) can obtain an accurate result of the effective phase function in the corresponding band, but is complex, and in fact, in a shorter band (e.g., 3-5 μm in the mid-infrared band), the variation of the effective phase function is negligible, so that the effective phase function in the entire band can be calculated using the effective phase function corresponding to the mid-wavelength in the band as an average.

(2) Correction of spontaneous emission

In the visible light scene, rendering of the virtual scene is based on RGB three-color operation, that is, the radiation brightness in three bands of red, green and blue of visible light is a calculation object, but according to the planck radiation law, the spontaneous radiation energy of the cloud in the three bands is extremely small, and compared with the light radiation of sunlight and the diffuse reflection effect caused by light, the rendering of the virtual scene can be completely ignored. On the contrary, the spontaneous radiation energy of the cloud under the infrared band, especially under the middle infrared band (3-5 μm) and the far infrared band (8-12 μm) is concentrated, and the scattering effect caused by the illumination cannot be ignored, so the spontaneous radiation effect of the cloud in the infrared scene simulation needs to be considered.

From the two aspects of reality and real-time performance of simulation, the traditional method for processing the spontaneous radiation effect of cloud in infrared scene simulation usually adopts the following two calculation models: a black body approximation model and a gray body approximation model.

The blackbody approximation model is used for processing the cloud infrared simulation problem in a far infrared band (8-12 mu m), the solar radiation intensity is low in the band, the simulation influence of the illumination effect on the cloud is low, and the cloud infrared radiation is mainly spontaneous radiation. Therefore, when only the sun is taken as the only light source and the sky is taken as the simulated background, the cloud can be regarded as a black body, and the spontaneous radiation energy can be directly calculated by the Planck formula (4-2).

Wherein M is_λbbIs the spectral radiant emittance, in W/(m)²μ m), λ is the wavelength, in μm, T is the thermodynamic temperature, in K, h is the Planckian constant, c is the speed of light₁Is a first radiation constant, c₂Is the second radiation constant and is the first radiation constant,

c₂＝hc/K_B≈1.43879×10⁴μm·K。

the gray body approximation model is used for processing the cloud infrared simulation problem under the middle infrared band (3-5 mu m), solar radiation is dominant under the band, the radiation effect of the cloud is mainly scattered solar radiation, and spontaneous radiation is secondary. Thus, when only the sun is considered as the sole light source and the sky is considered as the simulated background, the clouds can be considered as the gray body, and the radiation emissivity is generally between 0.7 and 0.9.

Both models are approximate assumptions based on measured data, and although relatively simple, the simulation results are close to the true infrared image. And referring to the two simple approximate models, performing spontaneous radiation correction on the illumination model provided by the third chapter to ensure the accuracy of the illumination model. First, no matter in the middle infrared band or the far infrared band, the absorption of infrared radiation by cloud needs to be considered, and the absorption rate is assumed to be α, so the equation (3-12) should be changed to:

the absorption rates of black and gray bodies are known from kirchhoff's law,

α＝ε (4-4)

i.e. the absorption is equal to the emissivity, in particular the emissivity of a black body is equal to 1.

Based on a gray body approximation model, the absorption rate in the middle infrared band is 0.7-0.9, the current multiple scattering energy proportion can be calculated by substituting formula (4-3), and the multiple scattering proportion is less than 10% according to the calculation result. In addition, the multiple scattering of the embodiment of the invention uses a processing method similar to diffuse reflection, so that the multiple scattering effect can be completely ignored, and the energy is processed by using a spontaneous radiation method, so that the method has the advantages of

In the far infrared band, because the extinction efficiency factor is smaller, the absorption is stronger, and the multiple scattering effect can be completely ignored. Physically, the strong absorption in the cloud causes the number of scattering times of solar radiation in the cloud to be reduced and the scattering intensity to be weakened, so that the solar radiation is difficult to detect. In addition, for the primary scattering effect and the secondary scattering effect under the far infrared band, as the phase function, the scattering efficiency factor and the solar radiation are smaller under the corresponding wavelength, the phase function, the scattering efficiency factor and the solar radiation can be ignored in comparison with the spontaneous radiation within the range allowed by the error, and therefore, in the far infrared band, the cloud exhibits the black body radiation characteristic.

Given the calculation method of the spontaneous emission, it is necessary to determine the relevant simulation scheme, which is included in the previous simulation method. Similar to the transmission effect of light scattered on particles, spontaneous radiation propagates from the particles along an observation path, is absorbed and scattered by the path, and finally enters a viewpoint to generate an observation result. As shown in FIG. 9, P₂The particles generate spontaneous radiation, and the radiation effect is along P₂A enters the viewpoint, likewise P₁The particles at the position P generate spontaneous radiation, and the radiation effect is along the position P₁A. The PA comes into view, with the difference that the transmission is different due to the different path lengths. The simulation scheme can be determined by the characteristic of spontaneous radiation, similar to the illumination model, the spontaneous radiation characteristic is placed in a rendering shader determined by the three-dimensional model, and spontaneous radiation is accumulated when integration is performed by the observation pathAnd (4) integrating the radiation, and finally outputting the total spontaneous radiation amount under the path.

After calculating the transmitted, scattered and spontaneous radiant energy, the total energy in the path is obtained, as shown in fig. 2. Unlike the direct calculation of RGB three colors for visible light simulation, infrared simulation requires a certain quantization scale. And (3) quantizing the radiation output result within the determined scale, outputting the final quantization result, and analyzing and simulating the experiment quantitatively. Assume that the radiance output of an object is L_outputUpper limit of quantization of simulation scene L_maxLower limit of L_minThe final quantization result G can be obtained, which is

The resulting quantization result G of equation (4-6) is generally represented by gray scale, and in a typical 0-255 gray scale scene, the available radiance output is L_outputCorresponding gray scale G · 255.

Before the display is carried out, firstly, the quantization scale and the temperature setting of a simulation scene are determined, the scene quantization scale is a radiance scale, the simulation wave band is a middle infrared wave band (3-5 mu m), and the lower limit is set to be black body radiance corresponding to 270K, namely 0.6064W/(sr m)²) The upper limit is the blackbody radiation brightness corresponding to 330K, i.e. 5.3518W/(sr m)²). The ground scene temperature is 300K, the cloud accumulation bottom height is 1000m, therefore, the cloud accumulation grid bottom temperature is 290K, and a water vapor generation area is arranged in the center of the grid. The sunlight illumination direction is set to be (1, -1,0), the observation direction is set to be (-1,0,0), and the surface temperature of the sun is 5500K and is taken as blackbody approximation treatment.

Three illumination effects, namely primary scattering, secondary scattering and spontaneous radiation effects, should exist in the infrared illumination simulation. Unlike the visible light illumination simulation, since the spontaneous emission of the cloud in the infrared band is not negligible, the radiation energy concentrated in the primary scattering, the secondary scattering, and the multiple scattering becomes weak due to the existence of the spontaneous emission. The scattered attenuated energy is absorbed by the particles and output as spontaneous radiation. Scene rendering is performed by combining the simulation scheme mentioned in the first section, and the result is displayed in two aspects of strong directional illumination (primary scattering and secondary scattering) and spontaneous radiation.

Fig. 10 shows the simulation result of the infrared illumination of the cloud in the infrared scene, which is the comprehensive effect of the simulation illumination, the highly directional illumination (primary scattering, secondary scattering), and the spontaneous emission effect from top to bottom. As is apparent from the figure, compared with the dominant position of sunlight in visible light simulation, in an infrared simulation scene, the proportion of strong directional illumination (primary scattering and secondary scattering) is obviously reduced, but the strong directional illumination is still the main body of illumination energy, and due to the illumination direction of the sunlight, the illumination effect generated in the current simulation visual angle is extremely strong in directivity; part of the energy in the simulation results comes from the spontaneous radiation, and the spontaneous radiation changes due to the change of the cloud three-dimensional model, but the energy of the spontaneous radiation is scattered and absorbed, which causes the spontaneous radiation results to change with time, namely, the spontaneous radiation results obviously change along with the divergence of the cloud.

Analyzing a simulation result of a stable state, wherein the highest gray level 255 appears at the upper edge of the cloud, mainly because strong forward scattering occurs, scattered light contains a large amount of solar energy within a small angle, and therefore the radiation brightness is large; average gray level 35 in spontaneous emission effect corresponds to radiation brightness 1.3508W/(sr m)²) Corresponding to the black body temperature 287K, the temperature is slightly lower than the atmospheric temperature of the height of the cloud body, and the actual measurement result of the cloud emissivity is met.

As can be seen from fig. 10, the spontaneous emission effect has no uniformity of the multiple scattering effect, but rather, the cloud three-dimensional model changes to produce a significant change. This is mainly because in the simulation scheme, the calculation object of the spontaneous radiation is cloud particles in the cloud, so the spontaneous radiation starts from the particles, and enters the viewpoint after propagating a certain path, and the effects of absorption and scattering and the like are still generated on the path, which results in the obvious dynamic property of the simulation result.

In addition, the simulation result in the graph has an obvious characteristic that the spontaneous radiation effect of a partial area of the cloud center is stronger than that of other areas in the simulation process. This is mainly because the central region is the location of the water vapour supply, which leads to two consequences: 1. the phase change process of the central area is more severe than that of other areas, and the temperature is increased due to heat generated by water vapor condensation; 2. the concentration in the central region is higher than in the other regions, and thus the transmittance is relatively low. This result is not evident in the visible simulation, but in the infrared simulation, since it only affects the spontaneous emission. During the simulation process, the phenomenon can be effectively weakened by reducing the initial concentration of the water vapor and reducing the water vapor supply area, but the final simulation result of the three-dimensional model is influenced.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.

In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples described in this specification can be combined and combined by those skilled in the art.

The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.

Claims

1. A modeling method of infrared cloud accumulation based on a physical process is characterized by comprising the following steps:

s1, obtaining a three-dimensional grid model of the cloud product, each grid in the three-dimensional grid model storing first variable information of a plurality of particles, the particles representing clouds at respective positions in the grid, the variable information including: droplet density, temperature, and velocity;

2. The modeling method of claim 1, wherein the step of S2 includes:

s21, setting an initial cloud accumulation;

s23, moving each particle in the grid according to the initialized speed;

s25, calculating the density and the temperature of the liquid drops of the moved particles based on the current atmospheric pressure of the cloud;

3. The modeling method of claim 2, wherein the step of S26 includes:

acquiring a Navier-Stokes equation as a fluid motion equation;

4. The modeling method of claim 1, wherein the step of S4 includes:

acquiring a scattering phase function;

5. The modeling method of claim 4, wherein the step of S8 includes:

6. The modeling method of claim 5, wherein prior to the step of S8, the modeling method further comprises:

7. The modeling method of claim 1, wherein the step of S7 includes:

s77, integrating the transmittance;

s81, accumulating the second radiation energies respectively;

8. The modeling method of claim 7, wherein the step of S78 includes:

based on the matte interface, a light transmittance is determined.

9. The modeling method of claim 7, wherein the step of S79 includes:

determining a primary scattering function of light passing through the cloud;

10. The modeling method of claim 7, wherein the step of S83 includes: