CN-121980997-A - Rapid simulation method and system for aerosol diffusion

Abstract

The invention belongs to the technical field of air pollutant diffusion numerical simulation, and particularly relates to a rapid simulation method and system for aerosol diffusion, wherein aerosol concentration time sequence data of each point in a target space are obtained through computational fluid dynamics simulation; the method comprises the steps of discretizing a target space into cube grids, calculating average concentration of each grid, selecting reference grids from all grids to form a space proportion database, taking concentration time sequence data of the reference grids as a training set to obtain a time sequence function, inquiring corresponding space proportion coefficients from the space proportion database, and multiplying the space proportion coefficients with the obtained time sequence function. The invention reduces the computational complexity by a plurality of orders of magnitude through the one-time early-stage high-precision computational fluid mechanics simulation solidification space diffusion mode and the subsequent computation is only simple function evaluation and multiplication operation, and simultaneously maintains the space generalization precision which is highly consistent with the computational fluid mechanics data, thereby fundamentally solving the technical contradiction of quick and quasi-incoordination.

Inventors

• GUO MING
• Cao Shangen
• HUANG HUANG
• XIE XIN
• HUANG GUI
• LI YAN
• LI ZIWEI
• ZHONG GUOBO
• FENG XIAOFEI
• HU FENGYING
• ZHU MIN
• ZHANG JIJUN
• WU FEI
• XU ZIJIAN
• WANG SHENGAO
• HUANG YIBO
• LI BIAO

Assignees

• 中国人民解放军海军工程大学

Dates

Publication Date

20260505

Application Date

20260115

Claims (6)

1. 1. A rapid simulation method for aerosol diffusion, comprising the following steps: S1, acquiring aerosol concentration time sequence data of each point of a target space in a preset time period through computational fluid dynamics simulation; s2, discretizing a target space into cube grids with preset scales, and calculating the average concentration of each grid at each moment; S3, selecting a grid meeting preset conditions from all grids as a reference grid, calculating the average ratio of the concentration time sequence data of each grid to the concentration time sequence data of the reference grid in time, taking the average ratio as the space ratio coefficient of the grid, and forming a space ratio database which does not change with time by the space ratio coefficients of all grids; s4, using the concentration time sequence data of the reference grid in the step S3 as a training set, and fitting the training set by adopting a chi-square mixed model, wherein the time sequence function expression is as follows: Wherein, the For the number of model components, Respectively the first Weight parameters, degree of freedom parameters, position parameters and scale parameters of the individual components; S5, for any target position in the space, inquiring the corresponding space proportion coefficient from the space proportion database according to the grid where the target position is located And the spatial proportionality coefficient obtained by inquiry And the time sequence function obtained in the step S4 Multiplying to obtain the predicted concentration time sequence of the target position 。
2. 2. The rapid simulation method of aerosol diffusion according to claim 1, wherein in step S2, the simulation area is focused on a target height range, and the space within the height range is discretized into a cubic grid of a preset scale.
3. 3. The aerosol diffusion rapid simulation method according to claim 1, wherein the preset condition in step S3 is that the concentration sequence of the grid is higher than a set threshold value in the whole simulation period, and the concentration accumulated value is the largest in a preset search area.
4. 4. The aerosol diffusion rapid simulation method according to claim 1, further comprising the step of performing global optimization on parameters of the chi-square hybrid model in step S4 by using an optimization algorithm to obtain a fitted time sequence function.
5. 5. The method of claim 4, wherein the optimization algorithm is a multi-agent genetic algorithm.
6. 6. An aerosol diffusion rapid simulation system comprising a memory and a processor, the memory having instructions stored therein for causing the processor to perform the method of any of claims 1-5.

Description

Rapid simulation method and system for aerosol diffusion Technical Field The invention belongs to the technical field of numerical simulation of air pollutant diffusion, and particularly relates to a rapid simulation method and system for aerosol diffusion. Background In various scenes such as nuclear facility operation, industrial production and the like, once radioactive substances leak or harmful aerosol is released in an indoor environment, the accuracy of diffusion behavior is directly related to the scientificity of emergency decision making, the safety of personnel evacuation and the effectiveness of training exercise. The rapid and accurate simulation of the diffusion track and distribution state of aerosol in indoor space has become a key technical support for guaranteeing public safety and environmental safety. Currently, indoor aerosol diffusion simulation relies mainly on two types of technical paths. The method is based on a hydrodynamic basic equation, can carry out high-fidelity numerical solution on a complex flow field and a particulate matter transportation process, and becomes a standard tool in scientific research of related fields by virtue of excellent simulation precision. However, the calculation process of the technology involves complex meshing and iterative solving steps, so that the calculation efficiency is low, and a large amount of time is required to be consumed for single three-dimensional transient simulation, so that the calculation cost is high. The defect makes computational fluid mechanics simulation difficult to apply in a scene requiring real-time interaction or quick response, and cannot meet real-time requirements of practical applications such as emergency training. The other technical path is an experience or analysis model such as Gaussian smoke plume, and the model is constructed based on simplified assumption conditions and analysis solutions, has extremely high calculation speed, and can rapidly output simulation results. The derivation of such models is typically premised on an unbounded or semi-infinite space, with the core applicable scenario being an open atmosphere environment. In indoor environments, the presence of walls, various obstacles and ventilation systems makes the flow field exhibit highly non-uniform characteristics, and complex backflow and vortex phenomena are formed in the space, which results in the basic assumption of the gaussian model no longer being true in indoor environments. If the model is directly applied to indoor aerosol diffusion simulation, errors of simulation results are obviously increased, real diffusion rules of pollutants in the indoor cannot be accurately reflected, and the requirement of practical application on simulation precision is difficult to meet. In summary, the path in the prior art has obvious contradiction that computational fluid dynamics simulation with high precision cannot meet the real-time requirement, but experience capable of meeting the real-time requirement or the simulation precision of an analytical model in an indoor environment is insufficient. Therefore, developing a fast simulation method that can achieve both simulation accuracy and calculation efficiency in an indoor environment is a key technical problem to be solved in the current field. Disclosure of Invention The aerosol diffusion rapid simulation method and system provided by the invention can effectively solve the problems in the background technology. The invention provides a rapid simulation method for aerosol diffusion, which comprises the following steps: S1, acquiring aerosol concentration time sequence data of each point of a target space in a preset time period through computational fluid dynamics simulation; s2, discretizing a target space into cube grids with preset scales, and calculating the average concentration of each grid at each moment; S3, selecting a grid meeting preset conditions from all grids as a reference grid, calculating the average ratio of the concentration time sequence data of each grid to the concentration time sequence data of the reference grid in time, taking the average ratio as the space ratio coefficient of the grid, and forming a space ratio database which does not change with time by the space ratio coefficients of all grids; s4, using the concentration time sequence data of the reference grid in the step S3 as a training set, and fitting the training set by adopting a chi-square mixed model, wherein the time sequence function expression is as follows: Wherein, the For the number of model components, Respectively the firstWeight parameters, degree of freedom parameters, position parameters and scale parameters of the individual components; S5, for any target position in the space, inquiring the corresponding space proportion coefficient from the space proportion database according to the grid where the target position is located And the spatial proportionality coefficient obtained by inquiryAnd t