CN109242183A - Crowd simulation evacuation method and device based on artificial fish-swarm algorithm and target detection - Google Patents
Crowd simulation evacuation method and device based on artificial fish-swarm algorithm and target detection Download PDFInfo
- Publication number
- CN109242183A CN109242183A CN201811030222.2A CN201811030222A CN109242183A CN 109242183 A CN109242183 A CN 109242183A CN 201811030222 A CN201811030222 A CN 201811030222A CN 109242183 A CN109242183 A CN 109242183A
- Authority
- CN
- China
- Prior art keywords
- evacuation
- individual
- crowd
- speed
- fish
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 40
- 238000001514 detection method Methods 0.000 title claims abstract description 29
- 238000004088 simulation Methods 0.000 title claims abstract description 18
- 241000251468 Actinopterygii Species 0.000 claims abstract description 28
- 230000008569 process Effects 0.000 claims abstract description 12
- 230000006399 behavior Effects 0.000 claims description 33
- 230000006870 function Effects 0.000 claims description 22
- 238000011156 evaluation Methods 0.000 claims description 14
- 239000002245 particle Substances 0.000 claims description 11
- 238000013527 convolutional neural network Methods 0.000 claims description 10
- 238000003860 storage Methods 0.000 claims description 7
- 238000006073 displacement reaction Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000012706 support-vector machine Methods 0.000 claims description 2
- 238000005457 optimization Methods 0.000 description 6
- 230000009471 action Effects 0.000 description 4
- 230000002776 aggregation Effects 0.000 description 4
- 238000004220 aggregation Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 239000003643 water by type Substances 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 239000004459 forage Substances 0.000 description 2
- 230000019637 foraging behavior Effects 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 238000012952 Resampling Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Classifications
-
- 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
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
- G06Q50/265—Personal security, identity or safety
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/277—Analysis of motion involving stochastic approaches, e.g. using Kalman filters
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Multimedia (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Computer Security & Cryptography (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of crowd simulation evacuation method and device based on artificial fish-swarm algorithm and target detection, it the described method comprises the following steps: topological map being established according to evacuation scene plan view, wherein node indicates emergency exit, line indicates two outlets, and marks final outlet；Detection and tracking is carried out according to individual of the video data to evacuation crowd, obtains the individual speed of evacuation crowd；Eating speed in shoal of fish influent pH is indicated using the individual speed, the shoal of fish is initialized, food is indicated with the final outlet position of scene, path planning is carried out to evacuation crowd's individual based on artificial fish-swarm algorithm；When the number of evacuation of final outlet is equal to total number of persons, evacuation process terminates, and obtains evacuation path.The present invention identifies slow individual in evacuation, and crowd is instructed to evacuate, to look after the disadvantaged group in evacuation, while improving evacuation efficiency, avoids excessive congestion.
Description
Technical field
The invention belongs to crowd evacuation emulation field more particularly to a kind of people based on artificial fish-swarm algorithm and target detection
Group's emulation evacuation method and device.
Background technique
In actual evacuation situation, problem is increasingly complicated, and Uncertainty, ignorance factor are numerous, causality
Complexity, and manoeuvre cost is big, the period is long, it is difficult to study rule.Thus domestic and foreign scholars develop for evacuation problem
Many computer models and software.Artificial fish-swarm algorithm is novel colony intelligence optimization algorithm, in a piece of waters, the usual energy of fish
It is enough freely to move about or other individuals is followed to find the place more than food, therefore the most place of fish existence number is generally in waters
It is exactly rich in the place that nutriment is most in this waters, according to this feature, Li Xiaolei proposes artificial fish-swarm algorithm, passes through mould
The behaviors such as look for food, knock into the back, bunching of the quasi- shoal of fish, achieve the purpose that Optimum search.Fish-swarm algorithm is since proposition, in machine
It is widely applied on the problems such as people's path planning, vehicle scheduling, illustrates its good search capability.
When emergency occurs, crowd evacuation efficiency and crowd density have much relations.When one outlet aggregation people very
When more, whole evacuation certainly will be influenced.We from evacuation video in it should be appreciated that individual evacuation speed have otherness,
For example the evacuation speed of old man, child and group of handicapped are slow, if the crowd density of an emergency exit is low, but it is big to evacuate crowd
When part is above-mentioned action slow three classes crowd, evacuation efficiency can also have a greatly reduced quality.Meanwhile if only being made with crowd density
To measure, it will lead to a large amount of crowds toward this outlet aggregation, to keep evacuation slower, or even cause the things such as swarm and jostlement
Therefore.It does not find also to consider individual speed otherness into the correlation technique into crowd evacuation at present.
Summary of the invention
To overcome above-mentioned the deficiencies in the prior art, the present invention provides a kind of based on artificial fish-swarm algorithm and target detection
Crowd simulation evacuation method and device are primarily based on object detection method and obtain pedestrian movement's speed, then original in Artificial Fish
Eating speed one is added on the basis of attribute, and the fast individual of eating speed is analogous to pokesy individual in evacuation, is known
Not Shu San in slow individual, instruct crowd to evacuate, thus look after evacuation in disadvantaged group, while improve dredge
Efficiency is dissipated, excessive congestion is avoided.
To achieve the above object, the present invention adopts the following technical scheme:
A kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection, comprising the following steps:
Topological map is established according to evacuation scene plan view, wherein node indicates that emergency exit, line indicate two outlets
Connection, and mark final outlet；
Detection and tracking is carried out according to individual of the video data to evacuation crowd, obtains the individual speed in real time of evacuation crowd
Degree；
Eating speed in shoal of fish influent pH is indicated using the individual speed, the shoal of fish is initialized, with the final of scene
Outlet port indicates food, carries out path planning to evacuation crowd's individual based on artificial fish-swarm algorithm；
When the number of evacuation of final outlet is equal to total number of persons, evacuation process terminates, and obtains evacuation path.
Further, the individual speed of acquisition evacuation crowd includes:
Acquisition channel exit video data；
The pedestrian target in the video data is detected based on RCNN；
The tracking of pedestrian movement is carried out using particle filter；
For each pedestrian, according to the position difference conversion actual displacement of present frame and the former frame pedestrian；
According to the shooting frame frequency of the actual displacement and video, the speed of the pedestrian is calculated.
Further, described pedestrian target in the video data is detected based on RCNN to include:
Use two-value normalized gradient algorithm picks object candidate area；
Using convolutional neural networks to carrying out feature extraction in candidate region；Wherein, the convolutional neural networks select packet
Network structure containing 5 convolutional layers and 3 full articulamentums；
Feature after extraction is classified using support vector machines, identifies that the candidate region is that pedestrian target is still carried on the back
Scape.
Further, described to include: to evacuation crowd's individual progress path planning based on artificial fish-swarm algorithm
Initialize the shoal of fish, the feed of number, position, field range, moving step length and individual including Artificial Fish
Speed；
For each individual, according to be intended to distance, crowding and the eating speed for exporting to evacuation final outlet of selection compared with
Fast individual amount determines next position of individual；Wherein, the feed is indicated lower than the pedestrian of certain threshold value using speed
The individual of fast speed.
Further, next position of the determining individual includes:
For each individual, next outlet that the individual is selected when different behaviors occur for current location is prejudged,
According to selected next distance for exporting to final emergency exit, crowding and the faster individual amount difference of eating speed
Calculation Estimation function；The evaluation function are as follows:
B(vi)=μ Dis (vi)+λδ+ωNums
Wherein, next distance for exporting to final outletYiIndicate next outlet node viInstitute is in place
The food concentration set；δ indicates the crowding factor；NumsIndicate the number of next faster individual of outlet eating speed；μ,λ
Weight coefficient ,+ω=1 μ+λ are indicated with ω；μ, λ and ω indicate weight coefficient ,+ω=1 μ+λ；
The smallest behavior of evaluation function is selected to execute the movement of individual.
Further, the behavior includes looking for food, bunching, knocking into the back and random behavior.
Further, execute look for food, bunch, the mobile formula for behavior of knocking into the back are as follows:
Execute the mobile formula of random behavior are as follows:
Wherein, rand () is random function, and value is 0~1；Indicate that individual is presently in position, XjIndicate target
Outlet port where body,Indicate the position after individual is mobile；Step indicates moving step length；Visual is the visible model of individual
It encloses.
One or more embodiments provide a kind of computing device, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, the processor are realized when executing described program as claim 1-6 is any
The crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection described in.
One or more embodiments provide a kind of computer readable storage medium, are stored thereon with computer program, should
The crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection is realized when program is executed by processor.
Beneficial effects of the present invention
(1) present invention shows complex environment with topological map, compactly illustrates that the connectivity between region is asked
Topic, can intuitively reflect process and the path of evacuation.
(2) present invention will go in view of the speed difference of child old man and group of handicapped and normal population in practical evacuation
People's speed alternatively evacuates the important indicator of node, using RCNN and particle filter to pedestrian movement's detecting and tracking, can obtain
Pedestrian's real-time speed is obtained, the truth of reflection outlet evacuation provides foundation to evacuate the selection of node, looked after in evacuating
Disadvantaged group, while avoid outlet congestion aggravation, improve evacuation efficiency.
(3) present invention simulates evacuation process using fish-swarm algorithm, there is faster speed of searching optimization and good global optimizing energy
Power.By introducing the judgement to eating speed, Artificial Fish is when selecting next node, in addition to consider that the partner of the position gathers around
Degree is squeezed, it is also contemplated that the problem of partner's eating speed.It corresponds in crowd evacuation, pedestrian examines when selecting next position
Consider influence of pedestrian's speed difference to evacuation efficiency, selection makes objective function obtain the behavior of optimal value, presents a kind of efficient
Optimization method.
Detailed description of the invention
The accompanying drawings constituting a part of this application is used to provide further understanding of the present application, and the application's shows
Meaning property embodiment and its explanation are not constituted an undue limitation on the present application for explaining the application.
Fig. 1 is the flow chart of crowd's evacuation method of the present invention；
Fig. 2 is the 2 d plane picture for evacuating scene；
Fig. 3 is the topological map generated by the connectivity exported；
Fig. 4 is that crowd initializes random distribution schematic diagram in the scene；
Fig. 5 is crowd according to rule evacuation schematic diagram；
Fig. 6 is crowd evacuation finish time schematic diagram.
Specific embodiment
It is noted that described further below be all exemplary, it is intended to provide further instruction to the application.Unless another
It indicates, all technical and scientific terms used herein has usual with the application person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singular
Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
In the absence of conflict, the features in the embodiments and the embodiments of the present application can be combined with each other.
General thought proposed by the present invention: the present invention indicates complicated evacuation environment with topological map, with Points And lines
Relationship represents the connectivity of associated outlet.In view of the action such as old man, children, group of handicapped is relatively slow during evacuation
The truth of individual, the present invention add eating speed one on the basis of Artificial Fish original attribute: due to shoal of fish individual difference
Different, eating speed is different, if certain regional artificial's fish density is small, but individual eating speed is fast, which will be quickly
It reduces, therefore, Artificial Fish should be prejudged before executing behavior of bunching whether there is the too fast individual of eating speed in the shoal of fish, with
Exempt to work hard but to no avail.The fast individual of eating speed be analogous to evacuation in pokesy individual, using individual evacuation speed as
The important indicator for needing to measure in evacuation avoids list and ignores to go out caused by individual evacuation speed difference using density as measurement
Mouth congestion, improves evacuation efficiency.The pedestrian in movement is identified using RCNN, and their speed is further calculated, statistics
The slow individual amount of certain regional action, the selection for pedestrian's emergency exit provide foundation, reflect the truth of evacuation, avoid
Secondary accident occurs.
Topological map is the important representation method of environment in robotics, and indoor environment is expressed as band node and correlation by it
The topology diagram of connecting line, wherein node indicates the critical positions point (turning, door, elevator & stairs etc.) in environment, and side indicates
Connection relationship between node, such as corridor.Topological map is that the simplification of actual environment and regularization indicate, with node and line
Positional relationship is established, there is lower complexity, while can simply and effectively indicate that the selection in crowd evacuation middle outlet path is asked
Topic.
RCNN (Regions with CNN features) is a kind of object detection method, is by convolutional neural networks side
Method is applied to a milestone on target detection problems, it utilizes the feature extraction functions and supporting vector of convolutional neural networks
The classification feature of machine realizes the conversion of target detection problems by Region Proposal method.It is with accuracy and efficiently
Property, it is able to satisfy the demand of real time problems.
Embodiment one
Present embodiment discloses a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection, such as Fig. 1
It is shown, comprising the following steps:
Step 1: the connected relation between region being obtained according to evacuation scene plan view, establishes topological map, wherein node
Emergency exit is represented, line represents the connected relation of two outlets, and marks final outlet.
The 2 d plane picture for obtaining evacuation environment establishes topological map according to the connectivity in region, and emergency exit number is
vi(i=1,2 ... n), and node represents emergency exit, and line represents the connected relation of two outlets.As shown in Fig. 2, wherein marked as
6. 5. node indicate final outlet.
Step 2: detection and tracking being carried out to pedestrian according to video data, obtains pedestrian's real-time speed.
Detecting and tracking is carried out to the pedestrian of certain outlet evacuation using RCNN and particle filter, obtains pedestrian's real-time speed,
The number of the pokesy s individual of certain exit aggregation is calculated, selects outlet to provide foundation for pedestrian.By certain exit s
Body screens, and counts its quantity.
Specifically, evacuation each exit of scene be respectively mounted CCD camera, to pedestrian carry out comprehensive detection with
No matter track, pedestrian walk from which outlet, can measure its real-time speed.After obtaining video data, first with convolutional Neural net
Network extracts pedestrian's feature, is tracked, measures pedestrian's speed.Specific steps include:
Pedestrian target in monitor video is detected using RCNN, is described as follows:
Possible target is chosen using two-value normalized gradient algorithm BING, obtains candidate region；
With convolutional neural networks algorithm (CNN) to carrying out feature extraction in candidate region.Can be selected comprising 5 convolutional layers and
The Alex net network structure of 3 full articulamentums.Input vector x first and filter weights carry out convolution algorithm, calculation formula
For
F=g (∑i∈Mxi×w+b) (1)
Increase deviator b, then use ReLU activation primitive g=max (0, x), finally carries out maximum pond, be input to next
Layer.The last one full articulamentum exports the feature vector of one 4096 dimension.Feature after CNN is extracted finally is divided by SVM
Class, identification candidate region are pedestrian target or background.
Then the predicting tracing that pedestrian movement is carried out using particle filter, is described as follows:
Assuming that Y={ y1,y2,…ytIndicate observational variable sequence, X={ x1,x2…,xtIndicate to be tracked sequence vector,
ytWith xtThe state variable in t moment observational variable and tracked target is respectively represented, weight is wt.So posterior probability p is
Compare current particle and before target similarity degree it can be concluded that next frame target position.Particle is from importance
It is distributed q (xt|x1:1-t,y1:t) in sampling obtain.Particle position with highest similarity determines that weight is more by weight
Newly foundation is
If it is less than a threshold value after right value update, it is necessary to particle resampling, eliminate the too low particle of weight.
Pedestrian is obtained after the coordinate position of next frame, compared with the coordinate of former frame, show that pedestrian is displaced in image
l.It is displaced the corresponding relationship of l and actual displacement according to pedestrian in image, obtains actual displacement l '.Since video camera shooting frame frequency is
Certain, it thus can calculate pedestrian's speed v.
The pedestrian's speed being intended to from some outlet evacuation can be measured by process as above, it, can when pedestrian's speed v is lower than 1m/s
Think that the evacuation speed of the pedestrian is slow, be defined as s, and records the quantity Num of region ss.Setting does not influence certain
The maximum number T of s in the case where outlet evacuation.
Step 3: dividing fast speed and slow pedestrian according to certain threshold value, respectively correspond in shoal of fish influent pH
Feed is relatively slow and the faster individual of feed；
The influent pH that the shoal of fish is defined on the basis of the shoal of fish original attribute is divided into feed quickly individual by eating speed
And general individual, accordingly for evacuating process evacuation speed slow individual and average individual, by the slow individual of evacuation speed at certain
The number statistical of outlet comes out, and pedestrian is instructed to evacuate, and congestion is avoided to aggravate.By the eating speed of Artificial Fish be divided into normal speed and
Quick two kinds, respectively correspond general individual and pokesy individual in evacuation.The too fast individual of eating speed is (slow in action
Individual) be denoted as s, number is denoted as Nums。
Statistical data obtains, in emergency, the speed of travel of ordinary people is 1.25m/s, the walking speed of old man or child
Degree is 0.65m/s, and the speed of the disabled person to want help is 0.57m/s, in the case where not interfering normally to evacuate, takes the 1m/s to be
Threshold value, the speed of travel are denoted as s in threshold value individual below.For evacuation efficiency and security consideration, need to sieve s individual
Choosing, counts its number, to reflect the true evacuation situation and evacuation capacity of outlet, provides foundation for the correct selection of pedestrian, keeps away
Exempt to export congestion aggravation.Certain exit s individual is screened, its quantity is counted.
Step 4: the initialization shoal of fish indicates food with the final outlet position of scene, based on artificial fish-swarm algorithm to evacuation
Crowd's individual carries out path planning.
Crowd evacuation is simulated using the behavior of artificial fish-swarm search food.Artificial Fish judge certain node superiority and inferiority according to
According to including food concentration, the number of the crowding factor and the quick individual of feed.The behavior of Artificial Fish is looked for food, and is bunched, and is knocked into the back
Three kinds, therefrom selection is so that the smallest behavior of target function value.
Initialize the shoal of fish, the feed of number, position, field range, moving step length and individual including Artificial Fish
Speed, wherein individual eating speed be divided into it is very fast and general.Pedestrian is instructed to evacuate according to fish-swarm algorithm.Vector X=(x1,
x2,…,xn) state of Artificial Fish individual is represented, wherein xi(i=1,2 ... n) it is intended to the variable of Optimum search；Y is that food is dense
Degree；di,j=| | Xi-Xj| | for the distance between individual；Visual is the visible range of individual；Step indicates moving step length；δ(0<
δ < 1) it is the crowding factor；Trynum is maximum attempts.
Crowd evacuation is emulated using the behavior of shoal of fish search food, Artificial Fish is got over food distance in fish-swarm algorithm
Small food concentration is bigger, and food position can represent final outlet position in crowd evacuation, certain outlet node viWith finally evacuate out
Distance Dis (the v of mouthi) be represented byThe selection of node is exported by outlet node viApart from final emergency exit
Distance Dis (vi), the number N ums of s individual is codetermined at the crowding factor delta and node of node, and evaluation function is represented by
B(vi)=μ Dis (vi)+λδ+ωNums, wherein+ω=1 μ+λ
Respectively simulation individual look for food in current location, clustering, the next node to knock into the back with selection when random behavior,
Its superiority and inferiority is judged using evaluation function for next node, Artificial Fish selection executes the smallest behavior of evaluation function.
The behavior of Artificial Fish is expressed as follows:
Look for food: people random selection evacuation path during flurried is similar to the foraging behavior of the shoal of fish, evaluation function only by
Next node and final outlet distance Dis (vi) determine.Assuming that pedestrian xiIt is now in vtNode, its within sweep of the eye with
Machine selects node vpIf vpNode food concentration YP>Yt, then can choose vpAs next node, evaluation function is denoted as B
(vi)forage；Otherwise other nodes are selected.After Trynum times repeated as above, Artificial Fish executes random behavior.Mobile formula are as follows:
Rand () is random function, and value is 0~1.
Bunch: people often has group psychology during evacuation, artificial in behavior of bunching similar to the behavior of bunching of the shoal of fish
Fish thinks that the target of selection is the outlet port in the visual field where the fish of center.Assuming that pedestrian xiIt is now in vtNode obtains its visual field
The aggregation central node v of partner in rangecAnd at node partner quantity nfIfProve that center food is sufficient
And partner's density be not it is very high, the relationship of s individual amount and threshold value T in partner is judged at this time, if Nums< T, then can choose vc
As next node, evaluation function is denoted as B (vi)bunching.Mobile formula is expressed as follows:
Otherwise other behaviors are executed.
Knock into the back: the pedestrian during evacuating is easy that other people is followed to do evacuation movement, is similar to fish-swarm algorithm due to blindness
In knock into the back, think that the target of selection is the outlet port in the visual field where the fish of state optimization in behavior of knocking into the back.Assuming that pedestrian xi
It is now in vtNode, obtaining it within sweep of the eye has node v locating for the partner of optimum statepAnd surrounding partner's quantity
nfIfProve that the state has more food and partner's density is not very greatly, further to judge s number of individuals in partner
The relationship of mesh and threshold value T, if Nums< T, then can choose vpAs next node, evaluation function is denoted as B (vi)pursue。
Mobile formula are as follows:
Otherwise other behaviors are selected.
Random: random behavior is a default behavior of foraging behavior, and Artificial Fish is randomly chosen state or partner, evaluation
Function is denoted as B (vi)random.Mobile formula are as follows:
The anticipation that Artificial Fish carries out next step behavior is had no progeny, and selection is so that the behavior that evaluation function is minimized, i.e.,
min(B(vi)forage,B(vi)bunching,B(vi)pursue,B(vi)random)。
Step 5: when the number of evacuation of final outlet is equal to total number of persons, evacuation process terminates, export evacuation path.
That is, terminating when the number come out from final outlet is equal to total number of persons, as shown in Figure 6.
Embodiment two
The purpose of the present embodiment is to provide a kind of computing device.
A kind of computing device including memory, processor and stores the meter that can be run on a memory and on a processor
Calculation machine program, the processor realize following steps when executing described program, comprising:
Topological map is established according to evacuation scene plan view, wherein node indicates that emergency exit, line indicate two outlets
Connection, and mark final outlet；
Detection and tracking is carried out according to individual of the video data to evacuation crowd, obtains the individual speed in real time of evacuation crowd
Degree；
Eating speed in shoal of fish influent pH is indicated using the individual speed, the shoal of fish is initialized, with the final of scene
Outlet port indicates food, carries out path planning to evacuation crowd's individual based on artificial fish-swarm algorithm；
When the number of evacuation of final outlet is equal to total number of persons, evacuation process terminates, and obtains evacuation path.
Embodiment three
The purpose of the present embodiment is to provide a kind of computer readable storage medium.
A kind of computer readable storage medium, is stored thereon with computer program, execution when which is executed by processor
Following steps:
Topological map is established according to evacuation scene plan view, wherein node indicates that emergency exit, line indicate two outlets
Connection, and mark final outlet；
Detection and tracking is carried out according to individual of the video data to evacuation crowd, obtains the individual speed in real time of evacuation crowd
Degree；
Eating speed in shoal of fish influent pH is indicated using the individual speed, the shoal of fish is initialized, with the final of scene
Outlet port indicates food, carries out path planning to evacuation crowd's individual based on artificial fish-swarm algorithm；
When the number of evacuation of final outlet is equal to total number of persons, evacuation process terminates, and obtains evacuation path.
Each step involved in above embodiments two and three is corresponding with embodiment of the method one, and specific embodiment can be found in
The related description part of embodiment one.Term " computer readable storage medium " is construed as including one or more instruction set
Single medium or multiple media；It should also be understood as including any medium, any medium can be stored, encodes or be held
It carries instruction set for being executed by processor and processor is made either to execute in the present invention method.
Beneficial effects of the present invention
(1) present invention shows complex environment with topological map, compactly illustrates that the connectivity between region is asked
Topic, can intuitively reflect process and the path of evacuation.
(2) present invention will go in view of the speed difference of child old man and group of handicapped and normal population in practical evacuation
People's speed alternatively evacuates the important indicator of node, using RCNN and particle filter to pedestrian movement's detecting and tracking, can obtain
Pedestrian's real-time speed is obtained, the truth of reflection outlet evacuation provides foundation to evacuate the selection of node, looked after in evacuating
Disadvantaged group, while avoid outlet congestion aggravation, improve evacuation efficiency.
(3) present invention simulates evacuation process using fish-swarm algorithm, there is faster speed of searching optimization and good global optimizing energy
Power.By introducing the judgement to eating speed, Artificial Fish is when selecting next node, in addition to consider that the partner of the position gathers around
Degree is squeezed, it is also contemplated that the problem of partner's eating speed.It corresponds in crowd evacuation, pedestrian examines when selecting next position
Consider influence of pedestrian's speed difference to evacuation efficiency, selection makes objective function obtain the behavior of optimal value, presents a kind of efficient
Optimization method.
It will be understood by those skilled in the art that each module or each step of aforementioned present invention can be filled with general computer
It sets to realize, optionally, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored
Be performed by computing device in the storage device, perhaps they are fabricated to each integrated circuit modules or by they
In multiple modules or step be fabricated to single integrated circuit module to realize.The present invention is not limited to any specific hardware and
The combination of software.
The foregoing is merely preferred embodiment of the present application, are not intended to limit this application, for the skill of this field
For art personnel, various changes and changes are possible in this application.Within the spirit and principles of this application, made any to repair
Change, equivalent replacement, improvement etc., should be included within the scope of protection of this application.
Above-mentioned, although the foregoing specific embodiments of the present invention is described with reference to the accompanying drawings, not protects model to the present invention
The limitation enclosed, those skilled in the art should understand that, based on the technical solutions of the present invention, those skilled in the art are not
Need to make the creative labor the various modifications or changes that can be made still within protection scope of the present invention.
Claims (9)
1. a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection, which is characterized in that including following step
It is rapid:
Topological map is established according to evacuation scene plan view, wherein node indicates that emergency exit, line indicate that two outlets connect
It is logical, and mark final outlet；
Detection and tracking is carried out according to individual of the video data to evacuation crowd, obtains the individual real-time speed of evacuation crowd；
Eating speed in shoal of fish influent pH is indicated using the individual speed, the shoal of fish is initialized, with the final outlet of scene
Position indicates food, carries out path planning to evacuation crowd's individual based on artificial fish-swarm algorithm；
When the number of evacuation of final outlet is equal to total number of persons, evacuation process terminates, and obtains evacuation path.
2. a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection as described in claim 1, special
Sign is that the individual speed for obtaining evacuation crowd includes:
Acquisition channel exit video data；
The pedestrian target in the video data is detected based on RCNN；
The tracking of pedestrian movement is carried out using particle filter；
For each pedestrian, according to the position difference conversion actual displacement of present frame and the former frame pedestrian；
According to the shooting frame frequency of the actual displacement and video, the speed of the pedestrian is calculated.
3. a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection as claimed in claim 2, special
Sign is, described to detect the pedestrian target in the video data based on RCNN and include:
Use two-value normalized gradient algorithm picks object candidate area；
Using convolutional neural networks to carrying out feature extraction in candidate region；Wherein, it includes 5 that the convolutional neural networks, which are selected,
The network structure of convolutional layer and 3 full articulamentums；
Feature after extraction is classified using support vector machines, identifies that the candidate region is pedestrian target or background.
4. a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection as described in claim 1, special
Sign is, described to include: to evacuation crowd's individual progress path planning based on artificial fish-swarm algorithm
Initialize the shoal of fish, number, position, field range, the eating speed of moving step length and individual including Artificial Fish；
For each individual, according to the selectable distance for exporting to final outlet, the crowding of selectable outlet port and
The faster individual amount of eating speed determines next position of individual；Wherein, pedestrian's table of certain threshold value is lower than using speed
Show that the eating speed is individual faster.
5. a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection as claimed in claim 4, special
Sign is that next position of the determining individual includes:
For each individual, next outlet that the individual is selected when different behaviors occur for current location is prejudged, according to
Selected next distance for exporting to final emergency exit, crowding and the faster individual amount of eating speed calculates separately
Evaluation function；The evaluation function are as follows:
B(vi)=μ Dis (vi)+λδ+ωNums
Wherein, next distance for exporting to final outletYiIndicate next outlet body viThe food of position
Object concentration；δ indicates the crowding factor；NumsIndicate the number of next faster individual of outlet eating speed；μ, λ and ω table
Show weight coefficient ,+ω=1 μ+λ；
The smallest behavior of evaluation function is selected to execute the movement of individual.
6. a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection as claimed in claim 5, special
Sign is, the behavior includes looking for food, bunching, knocking into the back and random behavior.
7. a kind of crowd simulation evacuation method based on artificial fish-swarm algorithm and target detection as claimed in claim 6, special
Sign is, looks for food, bunches, the mobile formula for behavior of knocking into the back are as follows:
Execute the mobile formula of random behavior are as follows:
Wherein, rand () is random function, and value is 0~1；Indicate that individual is presently in position, XjIndicate target individual institute
In outlet port,Indicate the position after individual is mobile；Step indicates moving step length；Visual is the visible range of individual.
8. a kind of computing device including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that the processor is realized when executing described program is based on people as claim 1-7 is described in any item
The crowd simulation evacuation method of work fish-swarm algorithm and target detection.
9. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is held by processor
Such as the described in any item crowd simulation evacuation sides based on artificial fish-swarm algorithm and target detection claim 1-7 are realized when row
Method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811030222.2A CN109242183B (en) | 2018-09-05 | 2018-09-05 | Crowd simulation evacuation method and device based on artificial fish-swarm algorithm and target detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811030222.2A CN109242183B (en) | 2018-09-05 | 2018-09-05 | Crowd simulation evacuation method and device based on artificial fish-swarm algorithm and target detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109242183A true CN109242183A (en) | 2019-01-18 |
| CN109242183B CN109242183B (en) | 2019-06-28 |
Family
ID=65060509
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811030222.2A Active CN109242183B (en) | 2018-09-05 | 2018-09-05 | Crowd simulation evacuation method and device based on artificial fish-swarm algorithm and target detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109242183B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111066698A (en) * | 2020-01-14 | 2020-04-28 | 清华大学 | Square combined multi-layer experimental device for simulating crowd flow of zebra fish |
| CN111709281A (en) * | 2020-05-06 | 2020-09-25 | 北京图创时代科技有限公司武汉分公司 | Intelligent security auxiliary system |
| CN113221379A (en) * | 2021-05-28 | 2021-08-06 | 视伴科技(北京)有限公司 | Method and device for planning number of viewing seats of event activity |
| CN114332170A (en) * | 2021-12-24 | 2022-04-12 | 哈尔滨商业大学 | Novel capital construction site safety monitoring system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103679264A (en) * | 2013-12-23 | 2014-03-26 | 山东师范大学 | Crowd evacuation path planning method based on artificial fish swarm algorithm |
| CN105468801A (en) * | 2014-09-09 | 2016-04-06 | 中国科学院深圳先进技术研究院 | Simulation method and system for crowd evacuation in public place |
| CN107194030A (en) * | 2017-04-20 | 2017-09-22 | 山东师范大学 | A kind of crowd evacuation emulation method driven based on video data and system |
| CN107665269A (en) * | 2017-08-11 | 2018-02-06 | 山东师范大学 | Quick crowd evacuation emulation method and device based on geography information |
| CN107679306A (en) * | 2017-09-26 | 2018-02-09 | 山东师范大学 | The crowd evacuation behavior simulation method and system of video drive |
-
2018
- 2018-09-05 CN CN201811030222.2A patent/CN109242183B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103679264A (en) * | 2013-12-23 | 2014-03-26 | 山东师范大学 | Crowd evacuation path planning method based on artificial fish swarm algorithm |
| CN105468801A (en) * | 2014-09-09 | 2016-04-06 | 中国科学院深圳先进技术研究院 | Simulation method and system for crowd evacuation in public place |
| CN107194030A (en) * | 2017-04-20 | 2017-09-22 | 山东师范大学 | A kind of crowd evacuation emulation method driven based on video data and system |
| CN107665269A (en) * | 2017-08-11 | 2018-02-06 | 山东师范大学 | Quick crowd evacuation emulation method and device based on geography information |
| CN107679306A (en) * | 2017-09-26 | 2018-02-09 | 山东师范大学 | The crowd evacuation behavior simulation method and system of video drive |
Non-Patent Citations (2)
| Title |
|---|
| 蒋颖丽: "基于人工鱼群算法的应急疏散模型及优化研究", 《中国优秀硕士学位论文全文数据库（电子期刊）信息科技辑》 * |
| 连海涛等: "地铁高峰段拥挤人流安全疏散路径规划仿真", 《计算机仿真》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111066698A (en) * | 2020-01-14 | 2020-04-28 | 清华大学 | Square combined multi-layer experimental device for simulating crowd flow of zebra fish |
| CN111709281A (en) * | 2020-05-06 | 2020-09-25 | 北京图创时代科技有限公司武汉分公司 | Intelligent security auxiliary system |
| CN113221379A (en) * | 2021-05-28 | 2021-08-06 | 视伴科技(北京)有限公司 | Method and device for planning number of viewing seats of event activity |
| CN114332170A (en) * | 2021-12-24 | 2022-04-12 | 哈尔滨商业大学 | Novel capital construction site safety monitoring system |
| CN114332170B (en) * | 2021-12-24 | 2022-11-18 | 哈尔滨商业大学 | Novel capital construction site safety monitoring system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109242183B (en) | 2019-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109242183B (en) | Crowd simulation evacuation method and device based on artificial fish-swarm algorithm and target detection | |
| CN111625361B (en) | Joint learning framework based on cooperation of cloud server and IoT (Internet of things) equipment | |
| CN108805083A (en) | The video behavior detection method of single phase | |
| CN107247938A (en) | A kind of method of high-resolution remote sensing image City Building function classification | |
| CN111709285A (en) | Epidemic situation protection monitoring method and device based on unmanned aerial vehicle and storage medium | |
| CN108764085A (en) | Based on the people counting method for generating confrontation network | |
| WO2020135392A1 (en) | Method and device for detecting abnormal behavior | |
| CN105005204A (en) | Intelligent engine system capable of automatically triggering intelligent home and intelligent life scenes and method | |
| CN110472612A (en) | Human bodys' response method and electronic equipment | |
| CN107167144A (en) | A kind of mobile robot indoor environment recognition positioning method of view-based access control model | |
| CN107944386A (en) | Visual scene recognition methods based on convolutional neural networks | |
| CN112469050B (en) | WSN three-dimensional coverage enhancement method based on improved wolf optimizer | |
| CN116258608B (en) | Water conservancy real-time monitoring information management system integrating GIS and BIM three-dimensional technology | |
| CN115775085A (en) | Smart city management method and system based on digital twin | |
| CN106503275A (en) | The tone color collocation method of chat robots and device | |
| CN114599069B (en) | Underwater wireless sensor network routing method based on energy self-collection | |
| CN108985256A (en) | Based on the multiple neural network demographic method of scene Density Distribution, system, medium, terminal | |
| CN112052771A (en) | Object re-identification method and device | |
| CN109620293A (en) | A kind of image-recognizing method, device and storage medium | |
| CN109886146A (en) | Flood information remote-sensing intelligent acquisition method and equipment based on Machine Vision Detection | |
| Gao et al. | 3D skeleton-based video action recognition by graph convolution network | |
| CN108320281A (en) | A kind of image significance detection method and terminal based on multiple features diffusion | |
| CN102779241B (en) | PPI (Point-Point Interaction) network clustering method based on artificial swarm reproduction mechanism | |
| CN113326925B (en) | Method, system, equipment and medium for calculating flower quantity of convolutional neural network based on density map regression | |
| CN104680118B (en) | A kind of face character detection model generation method and system |
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 |