CN114612554B - Artificial intelligence-based optimal evacuation path selection method under dust explosion - Google Patents
Artificial intelligence-based optimal evacuation path selection method under dust explosion Download PDFInfo
- Publication number
- CN114612554B CN114612554B CN202210256694.XA CN202210256694A CN114612554B CN 114612554 B CN114612554 B CN 114612554B CN 202210256694 A CN202210256694 A CN 202210256694A CN 114612554 B CN114612554 B CN 114612554B
- Authority
- CN
- China
- Prior art keywords
- dust
- explosion
- area
- dangerous area
- concentration
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of artificial intelligence, in particular to an optimal evacuation path selection method under dust explosion based on artificial intelligence. The method comprises the steps of firstly, collecting monitoring images, determining the position of an explosion source, and dividing a factory into a plurality of areas; collecting the area dust deposition amount and the maximum suspended dust concentration of each area, determining a dangerous area, and calculating the explosion impact force of the dangerous area; obtaining the deposition amount of the pixel point dust corresponding to each pixel point according to the color difference of each pixel point in the dangerous area and the deposition amount of the area dust; calculating to obtain the actual suspended dust concentration according to the explosion impact force and the area dust deposition amount; taking a dangerous area with the actual suspended dust concentration larger than a preset concentration threshold value as a multiple explosion area; and planning an optimal evacuation path according to the explosion impact force and the multiple explosion areas. The invention utilizes the actual suspended dust concentration of each area after the initial explosion to determine the area which can be subsequently exploded, thereby achieving the purpose of avoiding the personnel from being damaged by the subsequent explosion in the evacuation process.
Description
Technical Field
The invention relates to the technical field of machine vision, in particular to an optimal evacuation path selection method under dust explosion based on artificial intelligence.
Background
The dust explosion is a chemical reaction that the suspension dust air mixture is rapidly combusted and the temperature and the pressure are suddenly increased, so that the harm is huge, the casualty of people is very easy to cause, and the dust explosion is easy to generate secondary explosion. Therefore, the rapid and reasonable evacuation of people under the dust explosion is significant.
At present, the evacuation path planning under dust explosion usually only makes personnel evacuation paths according to the situation of single explosion. However, dust explosion is very easy to occur for multiple times, and the intensity of secondary explosion is higher. If only a single explosion is considered to select the evacuation path, the personnel can be damaged by secondary explosion or multiple explosions during the evacuation process.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide an optimal evacuation path selection method under dust explosion based on artificial intelligence, and the adopted technical scheme is as follows:
collecting a plurality of monitoring images in a factory, and determining the position of an explosion source according to the characteristics of the monitoring images when explosion occurs; dividing the plant into a plurality of zones based on the location of the explosive source;
collecting the initial suspended dust concentration, the area dust deposition amount and the maximum suspended dust concentration of each area; the area with the maximum suspended dust concentration larger than the preset concentration threshold is a dangerous area; obtaining the maximum explosive force according to the initial suspended dust concentration; calculating the actual distance between the dangerous area and the position of the explosion source, and calculating the impact time of each dangerous area after explosion; obtaining the explosion impact force of each dangerous area according to the actual distance, the impact time and the maximum explosion force;
acquiring the color difference of each pixel point before and after the dust deposition in the dangerous area based on the monitoring image; normalizing the color difference to obtain the dust deposition degree; obtaining the dust deposition amount of the pixel points according to the area dust deposition amount and the dust deposition degree in the dangerous area;
obtaining the dust lifting ratio according to the explosion impact force of each dangerous area and the area dust deposition amount; obtaining the actual dust raising amount of the dangerous area according to the dust raising proportion and the pixel point dust deposition amount; obtaining the actual dust raising concentration according to the actual dust raising amount and the area volume; adding the actual flying dust concentration and the initial suspended dust concentration to obtain the actual suspended dust concentration of the dangerous area after explosion; the dangerous area with the actual suspended dust concentration larger than the preset concentration threshold is a secondary explosion area;
and planning an optimal evacuation path according to the explosion impact force, the dangerous area and the re-explosion area.
Preferably, the determining the location of the explosion source according to the characteristics of the monitoring image includes:
and determining the position of the explosion source according to the difference value of the pixel value of each pixel point in the monitoring image.
Preferably, the dividing the factory into a plurality of areas based on the location of the explosion source includes:
and dividing the factory into a plurality of areas by different equidistant lines by taking the position of the explosion source as a center.
Preferably, the initial suspended dust concentration is obtained by: and acquiring the initial suspended dust concentration when explosion does not occur according to a dust concentration sensor.
Preferably, the obtaining manner of the maximum suspended dust concentration of each of the regions includes:
obtaining the concentration of the regional suspended dust according to the regional dust deposition amount and the regional volume;
and adding the area suspended dust concentration and the initial suspended dust concentration to obtain the maximum suspended dust concentration of each area.
Preferably, the obtaining the explosive impact force of each dangerous area according to the actual distance, the impacted time and the maximum explosive force includes:
and calculating the explosion impact force of each dangerous area at each moment by using Eschricky equation.
Preferably, the obtaining of the color difference of each pixel point before and after the deposition of the dust in the hazardous area based on the monitoring image includes:
acquiring the original hue of each pixel point on the ground when no dust is deposited in the dangerous area and the deposition hue of each pixel point on the ground when dust is deposited in the dangerous area;
and taking the absolute value of the difference value of the original hue and the deposition hue as the hue difference of each pixel point before and after dust deposition.
Preferably, the obtaining of the dust uplift ratio from the explosion impact force and the area dust deposition amount of each of the dangerous areas includes:
simulating the dust lifting proportion of different areas when the dust deposition amount is subjected to different explosion impact forces in a virtual engine, and constructing a dust lifting model;
and inputting the explosion impact force and the area dust deposition amount of each dangerous area into the dust lifting model to obtain the dust lifting proportion.
The embodiment of the invention at least has the following beneficial effects:
the embodiment of the invention utilizes an artificial intelligence technology, firstly a plurality of monitoring images are collected, the position of an explosion source is determined by the characteristics of the monitoring images, and a factory is divided into a plurality of areas; collecting the initial suspended dust concentration, the area dust deposition amount and the maximum suspended dust concentration of each area; the maximum explosive force is obtained from the initial suspended dust concentration. Taking the area with the maximum suspended dust concentration larger than the preset concentration threshold as a dangerous area, and based on the dangerous area, the subsequent calculation avoids the large increase of the calculated amount caused by the excessive areas; calculating the actual distance between the dangerous area and the position of an explosion source and obtaining the impacted time, and obtaining the explosion impact force of each dangerous area according to the actual distance, the impacted time and the maximum explosion force; and obtaining the dust deposition degree according to the color difference of each pixel point before and after the dust deposition in the dangerous area, and obtaining the dust deposition amount of the pixel point corresponding to each pixel point in the dangerous area. Obtaining the dust lifting proportion of the dangerous area according to the explosion impact force and the area dust deposition amount, and obtaining the actual dust lifting amount of the dangerous area; obtaining actual suspended dust concentration according to the actual dust raising amount, the area volume of each area and the initial suspended dust concentration, and taking the area with the actual suspended dust concentration larger than a preset concentration threshold value as a multiple explosion area, wherein the obtained multiple explosion dangerous area is that the suspended dust concentration in the air is increased due to the dust raising caused by the primary explosion, so that the increased suspended dust concentration reaches the preset concentration threshold value, the secondary explosion can occur, and the secondary explosion area can be accurately obtained; and planning an optimal evacuation path according to the explosion impact force, the dangerous area and the multiple explosion area. Through the dangerous area that takes place many times of explosions such as secondary explosion, cubic explosion that obtains the explosive impact force and the dust volume of raising after the first explosion and lead to, plan subsequent personnel evacuation, avoid only considering single explosion and select the route of fleing to evacuate, and the personnel that lead to receive the injury that the secondary explosion brought when fleing, reached the purpose that plans personnel evacuation route according to the dangerous area of follow-up many times of explosions.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a method for selecting an optimal evacuation path under a dust explosion based on artificial intelligence according to an embodiment of the present invention;
fig. 2 is a flowchart of a method for acquiring a color difference between pixels before and after deposition of dust in a hazardous area according to an embodiment of the present invention;
fig. 3 is a flowchart of a method for obtaining a dust lifting ratio according to an embodiment of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention for achieving the predetermined objects, the following detailed description of the method for selecting an optimal evacuation path under dust explosion based on artificial intelligence according to the present invention with reference to the accompanying drawings and preferred embodiments, the detailed description, the structure, the features and the effects thereof are as follows. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment of the invention provides a specific implementation method of an optimal evacuation path selection method under dust explosion based on artificial intelligence, which is suitable for factories which can generate dust in the equipment production process, such as flour factories, textile factories, feed processing factories, coal mines and the like. The factory building belongs to an unsafe closed space, and some areas have walls and some areas do not have walls. A floor plan of the plant has been acquired and the area and height of each equipment area is known. The amount of the deposited dust produced per unit time of each production facility can be obtained from the operational characteristics of the plant. A plurality of cameras are installed in each equipment area in a factory to guarantee that the position of an explosion source when the first explosion happens can be detected, and a dust concentration sensor is installed in each equipment area and used for detecting the concentration of suspended dust in air when the explosion does not happen in real time. According to the embodiment of the invention, dangerous areas such as secondary explosion, three-four explosion and the like which can generate multiple explosions are obtained by calculating the dust deposition amount of each area, the concentration of suspended dust in air and the explosion impact force at each area position, so that the purpose of planning the evacuation path of personnel according to the dangerous areas of the subsequent multiple explosions is achieved, and the escape personnel are prevented from being damaged by the subsequent multiple explosions.
The following describes a specific scheme of the method for selecting the optimal evacuation path under the dust explosion based on artificial intelligence in detail with reference to the accompanying drawings.
Referring to fig. 1, a flowchart illustrating steps of a method for selecting an optimal evacuation path under a dust explosion based on artificial intelligence according to an embodiment of the present invention is shown, the method includes the following steps:
s100, collecting a plurality of monitoring images in a factory, and determining the position of an explosion source according to the characteristics of the monitoring images when explosion occurs; the factory is divided into a plurality of areas based on the location of the explosive source.
A plurality of monitoring cameras are arranged in a factory, and video image frames collected by the cameras at multiple angles at the moment of explosion are extracted and combined to serve as monitoring images. Due to the fact that the flame color generated during explosion is different from the color of the surrounding environment, the position of an explosion source can be determined according to the difference value of the pixel value of each pixel point in the monitoring image.
Dividing a factory into a plurality of areas, wherein the principle of dividing the areas is as follows: the factory is divided into a plurality of areas by different equidistant lines by taking the position of the explosion source as the center, namely the edge of each area is equal to the position of the explosion source.
S200, collecting the initial suspended dust concentration, the area dust deposition amount and the maximum suspended dust concentration of each area; the area with the maximum suspended dust concentration larger than the preset concentration threshold is a dangerous area; obtaining the maximum explosive force according to the initial suspended dust concentration; calculating the actual distance between the dangerous area and the position of the explosion source, and calculating the impact time of each dangerous area after explosion; and obtaining the explosion impact force of each dangerous area according to the actual distance, the impact time and the maximum explosion force.
Since a factory produces a large amount of dust during production, there is a deposition of dust in the vicinity of production equipment. The deposited dust can be suspended in the air after being impacted by the primary explosion, so that the concentration of the suspended dust in the air is increased, and conditions are provided for secondary explosion. Therefore, the maximum suspended dust concentration when the deposited dust is completely dispersed in the air is calculated, and the area where the maximum suspended dust concentration in the air is greater than the preset concentration threshold value is taken as a dangerous area where subsequent explosion is likely to occur. It should be noted that the preset concentration threshold values of different dust types are different, and the preset concentration threshold values can be specifically set by an implementer according to the types of the dust.
The method comprises the following steps of:
1) And obtaining the concentration of the regional suspended dust according to the regional dust deposition amount and the regional volume.
The dust deposition amount of the area can be known in unit time according to the working property of the factory, and the dust deposition amount of each area at the moment of explosion can be obtained by combining the working time of the production equipment.
The acquisition method of the area volume comprises the following steps: and obtaining the area of each area and the height of the factory building according to the floor layout of the factory building, and obtaining the area volume corresponding to the area.
The zone suspended dust concentration is the ratio of the zone dust deposition amount to the zone volume.
2) And adding the concentration of the regional suspended dust and the initial suspended dust concentration to obtain the maximum suspended dust concentration of each region.
The initial suspended dust concentration is obtained in the following manner: and acquiring the initial suspended dust concentration when explosion does not occur according to the dust concentration sensor.
The maximum suspended dust concentration b m Comprises the following steps:
b m ＝b 0 +a/v
wherein, b 0 Is the initial suspended dust concentration; a is the area dust deposition amount; v is the volume of the region.
The maximum suspended dust concentration b m The area larger than the preset concentration threshold is a dangerous area.
After the initial explosion occurs, explosion impact is generated from the position of an explosion source to the periphery, and because the explosion source is positioned in an incompletely closed space, the impact force applied to each area is related to the distance between the two areas and whether the two areas are shielded by a wall or not due to incomplete blocking of the wall.
Therefore, the obstacle situation and the actual distance between each dangerous area and the position of the explosion source are obtained according to the floor plan, the dangerous areas are divided into two types of shielding without walls and shielding with walls, and the impact force of the initial explosion on the dangerous areas is respectively judged.
Wherein, the Escicke equation is used to calculate the explosion impact force of each dangerous area at each moment.
(1) For hazardous areas that are not shielded by walls, the blast shock wave can be seen as propagating unobstructed in free air.
At maximum explosive force P max On the basis of the total impact force of each dangerous area, the peak value P of the impact force R Inversely proportional to the actual distance d between the hazard zone and the source location of the explosion, i.e. the peak of the impact force
Since the shockwave spreads all around at a supersonic speed from the location of the explosive source, its diffusion speed is set to S =400m/S in the embodiment of the present invention.
Calculating the actual closest distance d between each dangerous area and the position of the explosion source by taking the explosion time as the starting time 1 And a practical maximum distance d 2 The time of impact of the shock wave after the explosion can be calculated according to the actual closest distance
Each dangerous area is influenced by the impact force of an explosion source and the atmospheric pressure P 0 And combining to obtain the impact force corresponding to the dangerous area:
at t < t 0 When the impact wave does not reach the dangerous area, the impact force applied to the dangerous area is 0, and the explosion impact force P = P 0 ；
At t = t 0 When the shock wave just reaches the dangerous area, the impact force applied to the dangerous area is the peak value P of the impact force R Then explosive impact force P = P 0 +P R ；
At t > t 0 When the explosion impact force is reduced to 0 along with the time t, the t can be obtained based on Escherzerg equation 0 The relationship between the explosion impact force P and the time after the moment is as follows:
in addition, c is the explosive amount in the Escherzk equation, but since the relationship between the dust concentration and the peak value of the explosive force is used in the embodiment of the present invention, the dust concentration and the explosive amount are converted, and if the peak value of the explosive impact force generated when the explosive amount is x is y and the peak value of the explosive impact force generated when the dust concentration is z is y, the dust concentration when the peak value of the explosive impact force is y is taken as c.
Thus, a time-impact force model of the dangerous area without wall shielding is constructed.
(2) For hazardous areas that are sheltered by walls, circulating overpressure exists after the blast shock wave bypasses the wall.
The area behind the wall is subjected to a lower explosive impact than the area in front of the wall. Based on the literature, zhang Xiao Wei, zhang Hao and YangmanlinAnd zhangqingming, "law of diffraction and overpressure distribution of explosion shock wave behind explosion-proof wall", university of beijing science and technology, 2021, 41 (4): 372-379. The explosion impact force is reduced by 80% at most after being shielded by the wall, and when the distance behind the wall is less than 1.2 times the height of the wall, the explosion impact force is increased along with the increase of the distance behind the wall, and the peak value of the impact force P is about 1.2 times the height of the wall R Reaches a maximum, after which the explosive impact force begins to decrease as the distance behind the wall increases.
Let d be the actual distance between the location of the explosion source and the hazardous area, and d be the actual distance between the location of the explosion source and the wall 0 Wherein the thickness of the wall is negligible.
The side of the wall facing the source of the explosion is subjected to an impact force peak of
The wall facing away from the source of the explosion is subjected to an impact force with a peak value of ^ 4>
Calculate the explosive impact force received by the rear side of the wall with a height h:
if the actual distance between the hazardous area and the location of the source of the explosion is d 0 ＜d＜d 0 +1.2h, the peak value P of the impact force received by the partial area R Increasing with increasing actual distance, then P R ＝P b [1+(d-d 0 )]；
If the actual distance d = d between the hazard zone and the location of the explosion source 0 +1.2h, the peak value of the impact force received by the position is P R ＝P b (1 + 1.2h); the peak impact force is the maximum value of the peak impact force in the rear wall area and is recorded as
If the actual distance between the danger zone and the position of the explosion source is d > d 0 +1.2h, the peak value P of the impact force received by the partial area R In that
Decreases with increasing distance, then->
If there is more than one wall between the dangerous area and the explosion source, the relationship between the impact force peak value and the distance at each position between the first and second walls can be obtained through the above calculation process (2). Therefore, the impact force peak value of the wall surface of the second face wall facing the explosion source is regarded as a new initial explosion impact force, and the relation between the impact force peak value and the distance of each position behind the second face wall is calculated. By parity of reasoning, the impact force peak value P received by each position of the dangerous area is obtained R 。
According to the step of calculating the time of impact in the step (1), calculating the time of the explosion impact force on the dangerous area shielded by the wall
And t 0 And (4) calculating the explosion impact force before and after the moment. Specifically, the method comprises the following steps:
at t < t 0 When the impact wave does not reach the dangerous area, the impact force applied to the dangerous area is 0, and the explosion impact force P = P 0 ；
At t = t 0 When the shock wave just reaches the dangerous area, the impact force on the dangerous area is the peak value P of the impact force R Then explosive impact force P = P 0 +P R ；
At t > t 0 When the explosion impact force of the dangerous area is gradually reduced along with time after the dangerous area receives the shock wave, the received explosion impact force is gradually reduced to 0 along with time, and then t can be obtained based on Escherzeri equation 0 The relationship between the explosion impact force P and the time after the moment is as follows:
from this, a time-impact force model of the hazard zone when occluded by the wall can be constructed. A time-impact force model for each hazard zone is obtained.
(3) For the position of the explosion source, since it is the explosion starting position of the initial explosion, at time t =0, the explosion impact force received by the explosion source is the largest, and then gradually decreases with the passage of time, and the relation between the explosion impact force P of the position of the explosion source and the time is: p = P 0 +P max e -ct The relation is also based on Escherzkey equation.
The explosion impact force of each dangerous area can be obtained according to the actual distance, the impact time and the maximum explosion force.
Step S300, acquiring the color difference of each pixel point before and after dust deposition in the dangerous area based on the monitoring image; normalizing the color difference to obtain the dust deposition degree; and obtaining the dust deposition amount of the pixel points according to the dust deposition amount and the dust deposition degree of the area in the dangerous area.
Since the dust explosion is very prone to secondary explosion, the possibility of secondary explosion in the dangerous area can be determined according to the deposition amount of the pixel point dust in each dangerous area and the dust raising proportion in the subsequent step S400.
The larger the deposition amount of the dust at the pixel point is, the larger the thickness of the dust is, and the larger the difference between the hue of the dust and the hue of the ground is. Therefore, the deposition amount of the dust of the pixel points at each pixel point can be calculated by obtaining the color difference of each pixel point before and after the dust deposition of the dangerous area in the monitored image.
Referring to fig. 2, the step of obtaining the color difference of each pixel point before and after the deposition of the dust in the hazardous area is as follows:
step S301, acquiring the original hue of each pixel point on the ground when no dust is deposited in the dangerous area and the deposition hue of each pixel point on the ground when dust is deposited.
Firstly, acquiring a monitoring image of a dangerous area:
and obtaining a monitoring image of each dangerous area at the moment before explosion through the monitoring video, and carrying out distortion correction on the monitoring image.
And performing semantic segmentation on the corrected monitoring image by using a DNN (digital noise network) with an Encoder-Decoder structure, marking the pixel value of the ground part in the monitoring image as 1, and marking the pixel values of other parts as 0 to obtain a ground semantic segmentation image.
And multiplying the obtained ground semantic segmentation image with the original monitoring image to obtain a monitoring image M of the ground part.
For each pixel point in the monitoring image M of the ground part, judging the dust deposition degree of the corresponding pixel point according to the color difference before and after dust deposition, specifically:
setting the original hue of each pixel point on the ground as H1= { H1 ] when no dust is deposited 1 ，H1 2 ，...H1 n }。
Obtaining the deposition hue H = { H } of each pixel point on the ground when dust is deposited according to the monitoring image M of the ground part 1 ，H 2 ，...H n }。
Step S302, the absolute value of the difference value between the original hue and the deposition hue is used as the hue difference of each pixel point before and after the dust deposition.
The color difference of each pixel point is as follows: h' = | H-H1|.
The hue difference H 'of each pixel point before and after the dust deposition can be obtained, and for a single pixel point in the monitoring image M, the larger the corresponding hue difference H', the higher the dust deposition degree.
And acquiring the original hue of each pixel point on the ground when no dust is deposited in the dangerous area and the deposition hue of each pixel point on the ground when dust is deposited.
And normalizing the color difference H' corresponding to each pixel point to obtain the dust deposition degree gamma of each pixel point.
And calculating the area dust deposition amount a of the dangerous area and the normalized dust deposition degree gamma of each pixel point to obtain the pixel point dust deposition amount a' = a gamma corresponding to each pixel point.
S400, obtaining the dust lifting ratio according to the explosion impact force of each dangerous area and the dust deposition amount of the area; obtaining the actual dust raising amount of the dangerous area according to the dust raising proportion and the pixel point dust deposition amount; obtaining the actual dust raising concentration according to the actual dust raising amount and the area volume; adding the actual flying concentration of the dust and the initial suspended dust concentration to obtain the actual suspended dust concentration of the dangerous area after explosion; and the dangerous area with the actual suspended dust concentration larger than the preset concentration threshold value is a re-explosion area.
Because the positions and the directions of the dangerous areas relative to the explosion source are different, and the explosion shock waves are diffused to the periphery in a circular manner by taking the position of the explosion source as a center, the magnitude of the explosion impact force received by different positions in the dangerous areas is inconsistent with the impact time.
Since each hazard zone is composed of different equidistant lines: that is, a circle having r as a radius and centered on the position of the explosion source represents the range of action of the explosion shock wave, and the radius r is enlarged as the explosion time progresses.
Setting the actual distance d from different positions of the dangerous area to the position of the explosion source from the floor plan, and setting the actual closest distance d in the dangerous area as d 1 The actual maximum distance is d 2 。
It should be noted that, since the whole danger area can be regarded as being composed of different equidistant lines, since the orientation of each danger area and the position of the explosion source is different, the lengths and the number of the equidistant lines of the different danger areas are different. The actual distance d from each point on each line to the source of the explosion is uniform and has a value of [ d 1 ，d 2 ]And the impact time of the point on each equidistant line subjected to the explosion impact force is consistent, and the value is
In the meantime.
And adding the pixel point dust deposition amount of each pixel point on each equidistant line to obtain the pixel point dust deposition amount on the whole equidistant line.
And simulating the phenomenon that the deposited dust is subjected to explosion shock waves in a simulator to obtain the dust lifting proportion under different explosion shock forces. Specifically, referring to fig. 3, the step of obtaining the dust raising ratio is:
step S401, simulating the dust lifting proportion when the dust deposition amount in different areas is subjected to different explosion impact forces in the virtual engine UE4, and constructing a dust lifting model.
The settings of the simulation environment are: the explosion scene is set to be a cubic area, the center of the bottom surface of the cubic area is used as the position of an explosion source, and deposited dust with the total amount of M consistent with the type of the dust in the embodiment of the invention is randomly distributed on the whole circumference with the distance of D from the explosion source on the ground.
The explosion simulation process comprises the following steps: by explosion source position department produce to expansion impact force f all around, then can obtain the impact force y0 that the deposit dust received, obtain the total amount m of raising the dust simultaneously, can know the impact force that receives at the deposit dust and be f 0's the condition under, the dust raising proportion of deposit dust is:
the impact force f expanding from the position of the explosion source to the periphery is constantly changed, and simultaneously the position of the periphery where the deposited dust is located and the position of the explosion source are changed, so that the deposited dust is subjected to different impact forces f 0 To obtain different impact forces f 0 The dust rising ratio of the lower deposited dust is made to be f 0 The value of (A) includes the impact force P on the deposited dust at each point on different equidistant lines in practical situation R The generalization capability of the simulator is increased, and a more accurate data base is provided for subsequent calculation.
And S402, inputting the explosion impact force and the regional dust deposition amount of each dangerous region into the dust raising model according to the dust raising model to obtain the dust raising proportion.
The dust rising ratio zT of the deposited dust on the equidistant line of each dangerous area at any time T can be obtained, and the amount A of the deposited dust on the equidistant line is combined T The dust raising amount of each equidistant line at the corresponding raising time T is A T z T 。
At time t 0 At the time, the dust-lifting amount of the deposited dust in the dangerous area is
It should be noted that the representation of the summation of the time is that the explosion shock waves corresponding to different times arrive at different equidistant lines, and each dangerous area is formed by a plurality of different equidistant lines, so that the summation of the dust uplifting amounts at a plurality of times is the summation of the dust uplifting amounts on the plurality of equidistant lines.
The initial suspended dust concentration of air in each hazard zone before the initial explosion occurs is b 0 Then at time t 0 When the actual suspended dust concentration of the dangerous area becomes
And taking the dangerous area with the actual suspended dust concentration larger than the preset concentration threshold value as a secondary explosion area.
And when the actual concentration of the suspended dust is greater than the preset concentration threshold value, the dangerous area meets the condition of secondary explosion.
Comparing corresponding time t when each dangerous area reaches a preset concentration threshold value 0 Considering that the dangerous area which reaches the preset concentration threshold value firstly can be subjected to secondary explosion firstly, and determining that the dangerous area is the predicted secondary explosion area.
The explosion time of the secondary explosion area is the corresponding t 0 At the moment of explosion, the suspended dust concentration is t 0 Corresponding in time
The maximum explosive force P which is dissipated outwards during an explosion in the danger zone can likewise be determined from experimental data on the basis of the dust type and the actual suspended dust concentration after the initial explosion max ′。
And under the condition of the predicted secondary explosion, repeating the step S100 to the step S400, updating the explosion impact force and the actual suspended dust concentration of each dangerous area, predicting the subsequent explosion area, the explosion time and the explosion impact force, and updating the time-impact force model of each dangerous area.
In the same step S200, based on the obstacle condition and distance between other areas and the secondary explosion area, and secondary explosionThe maximum explosion impact force Pmax' of the area outward diffusion can be obtained, and the impact force peak value P of each dangerous area exploded by the secondary explosion area can be obtained R ' and an impact time t1 at which the impact force starts to be applied.
An update of the time-impact force model for each danger zone can be made:
1) Setting the actual distance between the position of the secondary explosion area and the dangerous area as d 1 Calculating the time of the explosion impact force to each dangerous area outside the position of the explosion source
In addition to the location of the explosion source, each hazard zone is at time t 1 Before, only the impact force of the primary explosion source is affected, so the corresponding impact force of the explosion is:
at 0 < t 0 When the impact force received by the dangerous area is 0, namely the first shock wave does not reach the dangerous area, the explosion impact force P = P 0 ；
At t = t 0 When the impact wave reaches the dangerous area, the impact force on the dangerous area is the peak value P of the impact force R Then explosive impact force P = P 0 +P R ；
At t > t 0 When the explosion impact force is gradually reduced to 0 along with the time t, the t can be obtained based on Escherzerg equation 0 The relationship between the explosion impact force P and the time after the moment is as follows:
2) Each danger zone except for the primary explosion source and the secondary explosion zone is at t 1 Moment and later moment receive the common influence of the impact force of primary explosion and secondary explosion, then the impact force that corresponds is:
at t = t 1 When subjected to the explosive impact of the primary explosion source is
The explosion impact force of the secondary explosion is P R ', the impact force corresponding to the danger zone is->
At t > t 1 When subjected to the explosive impact of the primary explosion source is
Receives the explosion impact force of the secondary explosion as->
The impact force corresponding to the danger zone is->
3) For the secondary explosion region position, since it is the explosion starting position of the secondary explosion, at t = t 1 During the process, the explosion impact force received by the secondary explosion area position is the largest, and then is gradually reduced along with the time, specifically:
at t = t 1 The corresponding explosion impact force at any moment is as follows:
at t > t 1 The corresponding explosion impact force at any moment is as follows:
4) For the position of the primary explosive source, since it is the initial explosion position of the primary explosion, the explosive force applied to the explosive source is the largest at the time t =0 of the explosive source, and then gradually decreases with the passage of time. However, since it is also affected by the impact force of the secondary explosion shock wave, t = t 1 In time, the explosive force that the primary explosive source received also can change, specific:
at t < t 1 The corresponding explosion impact force at any moment is as follows: p = P 0 +P max e -ct ；
At t = t 1 The corresponding explosion impact force at any moment is as follows:
at t > t 1 The corresponding explosion impact force at the moment is as follows:
an updated time-impact force model for each hazard zone is available.
In the same way, the dangerous area of the third explosion can be obtained in a predictable way, and the final time-impact force model of each dangerous area can be obtained.
And updating the time-impact force model of each dangerous area according to the explosion condition of the third explosion area, and judging the fourth explosion area.
By analogy, the final time-impact force model of each dangerous area can be obtained.
And step S500, planning an optimal evacuation path according to the explosion impact force, the dangerous area and the secondary explosion area.
And obtaining a time-impact force model of each path according to the position relation between each path and each dangerous area, thereby determining the optimal path for people evacuation.
And (3) based on the time-impact force model of the dangerous area passed by each path, obtaining the explosion impact force borne by the path at different moments, and constructing the time-impact force model of the path.
And judging the time period of the personnel on the route when each route is taken as the evacuation route according to the distance relationship between the personnel and each route, so as to obtain the impact force born in the time period.
By comparing the impact force of the personnel passing through each path, the optimal path for evacuating the personnel can be obtained.
And finally, selecting the optimal path when evacuating each person.
In summary, the embodiment of the present invention utilizes an artificial intelligence technique, first collects a plurality of monitoring images, determines the location of an explosion source according to the characteristics of the monitoring images, and divides a factory into a plurality of areas; collecting the initial suspended dust concentration, the area dust deposition amount and the maximum suspended dust concentration of each area; the maximum explosive force is obtained from the initial suspended dust concentration. Taking the area with the maximum suspended dust concentration larger than a preset concentration threshold value as a dangerous area; calculating the actual distance between the dangerous area and the position of the explosion source and obtaining the impacted time, and obtaining the explosion impact force of each dangerous area according to the actual distance, the impacted time and the maximum explosion force; and obtaining the dust deposition degree according to the color difference of each pixel point before and after the dust deposition in the dangerous area, and obtaining the dust deposition amount of the pixel point corresponding to each pixel point in the dangerous area. Obtaining the dust raising proportion of the dangerous area according to the explosion impact force and the area dust deposition amount, and obtaining the actual dust raising amount of the dangerous area; obtaining actual suspended dust concentration according to the actual dust lifting amount, the area volume of each area and the initial suspended dust concentration, and taking the area with the actual suspended dust concentration larger than a preset concentration threshold value as a secondary explosion area; and planning an optimal evacuation path according to the explosion impact force, the dangerous area and the secondary explosion area. Through the dangerous area of multiple explosions such as secondary explosion, tertiary explosion that takes place that obtains the explosion impact force and the dust volume of raising after the first explosion and lead to, plan evacuation of follow-up personnel, avoid only considering single explosion and select the route of fleing to evacuate, and the personnel that lead to receive the injury that secondary explosion brought when fleing.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (7)
1. An optimal evacuation path selection method under dust explosion based on artificial intelligence is characterized by comprising the following steps:
collecting a plurality of monitoring images in a factory, and determining the position of an explosion source according to the characteristics of the monitoring images when explosion occurs; dividing the plant into a plurality of zones based on the location of the explosive source;
collecting the initial suspended dust concentration, the area dust deposition amount and the maximum suspended dust concentration of each area; the area with the maximum suspended dust concentration larger than the preset concentration threshold is a dangerous area; obtaining the maximum explosive force according to the initial suspended dust concentration; calculating the actual distance between the dangerous area and the position of the explosion source, and calculating the impact time of each dangerous area after explosion; obtaining the explosion impact force of each dangerous area according to the actual distance, the impacted time and the maximum explosion force;
acquiring the color difference of each pixel point before and after the dust deposition in the dangerous area based on the monitoring image; normalizing the color difference to obtain the dust deposition degree; obtaining the dust deposition amount of the pixel points according to the area dust deposition amount and the dust deposition degree in the dangerous area;
obtaining the dust lifting ratio according to the explosion impact force of each dangerous area and the area dust deposition amount; obtaining the actual dust raising amount of the dangerous area according to the dust raising proportion and the pixel point dust deposition amount; obtaining the actual dust raising concentration according to the actual dust raising amount and the area volume; adding the actual flying dust concentration and the initial suspended dust concentration to obtain the actual suspended dust concentration of the dangerous area after explosion; the dangerous area with the actual suspended dust concentration larger than the preset concentration threshold is a secondary explosion area;
planning an optimal evacuation path according to the explosion impact force, the dangerous area and the secondary explosion area;
the method for acquiring the color difference of each pixel point before and after the dust deposition in the dangerous area comprises the following steps: acquiring the original hue of each pixel point on the ground when no dust is deposited in the dangerous area and the deposition hue of each pixel point on the ground when dust is deposited in the dangerous area; taking the absolute value of the difference value of the original hue and the deposition hue as the hue difference of each pixel point before and after dust deposition;
the method for obtaining the dust deposition amount of the pixel points according to the area dust deposition amount and the dust deposition degree in the dangerous area comprises the following steps: calculating the area dust deposition amount a of the dangerous area and the normalized dust deposition degree gamma of each pixel point to obtain pixel point dust deposition amount alpha '= alpha gamma corresponding to each pixel point, wherein alpha' is the pixel point dust deposition amount; alpha is the area dust deposition amount; and gamma is the normalized dust deposition degree of each pixel point.
2. The method for selecting the optimal evacuation path under the dust explosion based on the artificial intelligence as claimed in claim 1, wherein the determining the location of the explosion source according to the characteristics of the monitoring image comprises:
and determining the position of the explosion source according to the difference value of the pixel value of each pixel point in the monitoring image.
3. The artificial intelligence based dust explosion optimal evacuation path selection method according to claim 1, wherein the dividing the factory into a plurality of areas based on the location of the explosion source comprises:
and dividing the factory into a plurality of areas by different equidistant lines by taking the position of the explosion source as a center.
4. The artificial intelligence based dust explosion optimal evacuation route selection method according to claim 1, wherein the initial suspended dust concentration is obtained by: and acquiring the initial suspended dust concentration when explosion does not occur according to a dust concentration sensor.
5. The method for selecting the optimal evacuation path under the artificial intelligence based dust explosion according to claim 1, wherein the obtaining manner of the maximum suspended dust concentration of each area comprises:
obtaining the concentration of the regional suspended dust according to the regional dust deposition amount and the regional volume;
adding the regional suspended dust concentration and the initial suspended dust concentration to obtain the maximum suspended dust concentration of each region;
wherein, the concentration of the region suspended dust is the ratio of the deposition amount of the region dust to the volume of the region.
6. The artificial intelligence based dust explosion lower optimal evacuation path selection method according to claim 1, wherein the obtaining of the explosion impact force of each dangerous area according to the actual distance, the impact time and the maximum explosion force comprises:
and calculating the explosion impact force of each dangerous area at each moment by using Eschricky equation.
7. The method of claim 1, wherein the obtaining the dust lifting ratio from the blast impact force and the area dust deposition amount of each dangerous area comprises:
simulating the dust lifting proportion of different areas when the dust deposition amount is subjected to different explosion impact forces in a virtual engine, and constructing a dust lifting model;
and inputting the explosion impact force and the area dust deposition amount of each dangerous area into the dust lifting model to obtain the dust lifting proportion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210256694.XA CN114612554B (en) | 2022-03-16 | 2022-03-16 | Artificial intelligence-based optimal evacuation path selection method under dust explosion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210256694.XA CN114612554B (en) | 2022-03-16 | 2022-03-16 | Artificial intelligence-based optimal evacuation path selection method under dust explosion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114612554A CN114612554A (en) | 2022-06-10 |
| CN114612554B true CN114612554B (en) | 2023-03-24 |
Family
ID=81864069
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210256694.XA Active CN114612554B (en) | 2022-03-16 | 2022-03-16 | Artificial intelligence-based optimal evacuation path selection method under dust explosion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114612554B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2533482C1 (en) * | 2013-06-27 | 2014-11-20 | Федеральное государственное бюджетное учреждение науки ИНСТИТУТ ПРОБЛЕМ КОМПЛЕКСНОГО ОСВОЕНИЯ НЕДР РОССИЙСКОЙ АКАДЕМИИ НАУК (ИПКОН РАН) | Method of forecast of explosion hazard of methane dust mixes during coal destruction |
| CN110503261B (en) * | 2019-08-21 | 2022-12-13 | 东北大学 | Evacuation path optimization method based on dust explosion domino disaster-causing risk |
| CN110751337B (en) * | 2019-10-22 | 2022-09-09 | 中煤科工集团重庆研究院有限公司 | Method for predicting coal and gas outburst and secondary disaster danger area range thereof |
-
2022
- 2022-03-16 CN CN202210256694.XA patent/CN114612554B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN114612554A (en) | 2022-06-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111914819B (en) | Multi-camera fusion crowd density prediction method and device, storage medium and terminal | |
| CN108416968A (en) | Fire alarm method and apparatus | |
| CN109035641B (en) | Dynamic optimization and visualization method for fire evacuation path considering smoke influence | |
| JP7143174B2 (en) | Smoke detection device and smoke identification method | |
| CN112927461B (en) | Early warning decision method and device for charging pile of new energy automobile | |
| TW201435381A (en) | Artificial intelligence earthquake early warning method | |
| CN114612554B (en) | Artificial intelligence-based optimal evacuation path selection method under dust explosion | |
| CN113203409A (en) | Method for constructing navigation map of mobile robot in complex indoor environment | |
| CN109100703B (en) | Method and device for detecting dangerous points of power transmission line | |
| CN112734739A (en) | Visual building crack identification method based on attention mechanism and ResNet fusion | |
| CN109360370B (en) | Robot-based smoke and fire detection method | |
| CN107610159A (en) | Infrared small object tracking based on curvature filtering and space-time context | |
| CN115937508A (en) | Method and device for detecting fireworks | |
| CN106771329A (en) | The detection method of the speed of service in a kind of unmanned plane moderating process | |
| US9607629B2 (en) | Frequency response display | |
| CN107301653A (en) | Video image fire disaster flame detection method based on BP neural network | |
| Bui et al. | Method of defining the parameters for UAV point cloud classification algorithm | |
| CN114037921B (en) | Sag modeling method and system based on intelligent recognition of unmanned aerial vehicle | |
| CN115965773A (en) | Gas leakage detection system based on big data of Internet of things | |
| CN111457895B (en) | Target size calculation and display method for photoelectric load of unmanned aerial vehicle | |
| CN115171006A (en) | Detection method for automatically identifying personnel entering electric power dangerous area based on deep learning | |
| CN113283332A (en) | Fire-fighting robot flame identification method, equipment and storage medium | |
| CN112651902A (en) | Multi-frame filtering implementation method and system in image processing | |
| CN111460646A (en) | Subway station explosion intensity grading early warning method based on deep learning | |
| CN114994634B (en) | Patrol car laser radar probe algorithm |
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 |