CN108491972A - A kind of crowd evacuation emulation method and device based on Sarsa algorithms - Google Patents
A kind of crowd evacuation emulation method and device based on Sarsa algorithms Download PDFInfo
- Publication number
- CN108491972A CN108491972A CN201810233963.4A CN201810233963A CN108491972A CN 108491972 A CN108491972 A CN 108491972A CN 201810233963 A CN201810233963 A CN 201810233963A CN 108491972 A CN108491972 A CN 108491972A
- Authority
- CN
- China
- Prior art keywords
- crowd
- evacuation
- model
- leader
- sarsa
- 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.)
- Pending
Links
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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
- G06Q50/265—Personal security, identity or safety
Abstract
The invention discloses a kind of crowd evacuation emulation method and device based on Sarsa algorithms, this method include：Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and person model is imported in evacuation model of place, crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；Crowd after initialization is grouped and filters out the leader of each group, microcosmic crowd movement's guidance is carried out using social force model is improved, path is most preferably evacuated using the Sarsa algorithms selections for presetting modifiable decay value, leader leads normal pedestrian in group to be moved according to best evacuation path.Social force model and the Sarsa algorithms in intensified learning are combined by the present invention, and the guidance of evacuation path can be provided to crowd, improves evacuation efficiency.
Description
Technical field
The invention belongs to the technical fields of crowd evacuation emulation, are dredged more particularly, to a kind of crowd based on Sarsa algorithms
Dissipate emulation mode and device.
Background technology
With the continuous quickening of urbanization process, city incity building and people El density are also rapidly increasing, therewith
And that comes is that public place personnel largely assemble, and in densely populated place region, once accident occurs, easily cause such as crowd
Congestion such as tramples at the malignant events, often causes the dead group of group if it effectively cannot evacuate crowd and hinders serious accident.How prominent
Hair event effectively carries out the condition of a disaster control and crowd evacuation when occurring, to reduce or remit casualties and property loss is both at home and abroad all
The problem of highest attention.China is a populous nation, and social structure complexity is unbalanced, and especially China is in society now
Transitional period, urbanization process are accelerated, population increase in city.These realities make public place crowd massing problem more
Prominent, casualties caused by after accident generation is also more serious.In these public places, caused by densely populated place
Injures and deaths case also happens occasionally.And scene modeling, optimum path search and crowd movement's row are carried out by computer simulation technique
, can be while reaching best evacuation rehearsal effect by cost minimization, therefore for modeling, computer simulation, which becomes, to be ground
Study carefully the main method of crowd evacuation under accident.
There are mainly two types of crowd evacuation emulation models, macromodel and micromodel.Macromodel is not examined from entirety
Consider the local detail information of individual behavior.The interaction of micromodel each individual and environment from the point of view of individual, can be more
Mend deficiency of the macromodel to pedestrian's datail description.Social force model is exactly a kind of typical micromodel, and social force model is
New pedestrian's flow model that Helbing et al. is proposed in nineteen ninety-five, is described as power in social force model by pedestrian movement
Effect as a result, pedestrian movement by reciprocal force drives jointly between reciprocal force, individual and environment between own drive power, individual.Wherein,
The expectation that own drive power description individual is moved to target；Reciprocal force reflection individual is to other people psychological repellence and physics between individual
Repel, makes to keep certain distance between individual, realize that the collision of pedestrian movement avoids；Reciprocal force ensures individual between individual and environment
Safe distance between barrier makes smooth avoiding barrier during pedestrian movement, and improved social force model is in original
Outlet attraction and friend's attraction are introduced on the basis of beginning social force model, respectively describe sucking action of the outlet to pedestrian
And the phenomenon of gathering in groups band together of pedestrian.
Sarsa algorithms are a kind of important intensified learning methods, it was proposed in 1994 by Rummer and Niranjan
Based on model algorithm, be initially referred to as improved Q- learning algorithms., still using Q value iteration, Sarsa is that one kind exists for it
Line strategy TD study (on-policy TD).Agent is walked in each study, is acted a according to ε-Greedy strategy determination first, is obtained
Heuristics and training example (st, at, st+1, rt+1, at+1)；And then state s is determined according to ε-Greedy strategyt+1When action
at+1, then carry out value function modification；By determining at+1The next action taken as agent.Obviously, Sarsa algorithms
With Q- learning algorithms the difference is that Q- learns to be iterated using the maximum value of value function, and Sarsa then using
Actual Q values are iterated, and in addition to this, Sarsa study is acted in each study step agent according to current Q values determination, therefore
It is a kind of strategy of on-line TD study to claim Sarsa.
In conclusion how will be in social force model and intensified learning in the crowd evacuation emulation of the prior art
The problem of Sarsa algorithms are conjointly employed in crowd evacuation emulation method, and how to improve the effect of crowd evacuation in public place
The problem of safety of evacuation crowd under rate and crisis situations, still lack effective solution.
Invention content
For the deficiencies in the prior art, how solve in the prior art will be in social force model and intensified learning
The problem of Sarsa algorithms are conjointly employed in crowd evacuation emulation method, and how to improve the effect of crowd evacuation in public place
The problem of safety of evacuation crowd under rate and crisis situations, it is thin that the present invention proposes a kind of crowd based on Sarsa algorithms
Emulation mode and device are dissipated, social force model and the Sarsa algorithms in intensified learning are combined, can be provided to crowd
Path guidance is evacuated, evacuation efficiency is improved.
The first object of the present invention is to provide a kind of crowd evacuation emulation method based on Sarsa algorithms.
To achieve the goals above, the present invention is using a kind of following technical solution：
A kind of crowd evacuation emulation method based on Sarsa algorithms, this method include：
Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；
Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and person model is imported and is dredged
It dissipates in model of place, crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；
Crowd after initialization is grouped to and is filtered out the leader of each group, social force model progress is micro- using improving
Crowd movement's guidance is seen, path, leader's led cluster are most preferably evacuated using the Sarsa algorithms selections for presetting modifiable decay value
Normal pedestrian is according to the movement of best evacuation path in group.
Scheme as a further preference, the real scene database include several real scene video informations, are utilized
KLT tracing algorithms extract crowd's parameter information, crowd's parameter information from the video information of the real scene database
Including crowd evacuation motion path, initialization coordinate, initial velocity.
Scheme as a further preference screens leading for each group according to crowd and outlet port in the method
Person, the leader are the pedestrian near outlet port in each group.
Scheme as a further preference, in the method, the outlet if leader arrives safe and sound is drawn in selection group again
Neck person, until optional without pedestrian's individual in each group.
Scheme as a further preference, the improvement social force model have modified the phase on the basis of traditional society's power model
It hopes speed, structure is carried out according to self related desired speed function of distance between remaining time, pedestrian and surrounding pedestrian's speed
It builds.
Scheme as a further preference, in the method, in each emergency exit of the evacuation model of place of establishment
Place's setting counter, crowd's number of individuals for counting each outlet evacuation.
Scheme as a further preference according to crowd's number of individuals of each outlet evacuation, calculates separately in the method
The crowding in exit；The crowding is under normal circumstances by the crowd of the predetermined number of outlet and corresponding outlet evacuation
The ratio of body number total value, when exporting crowding more than the crowding threshold value, it is believed that congestion has occurred in outlet；
Leader selects most preferably to export as thin according to the crowding of distance and each outlet apart from each outlet port
Dissipate target.
Scheme as a further preference, according to the leader current location, direction of action, using Sarsa algorithm meters
The income of all and current location unicom next target point is calculated, and calculates average Q value and maximum Q values, is weighted and asks
With, the point of Income Maximum is chosen as next evacuation target point, while updating income storage table, target is evacuated until reaching,
Complete income storage table is obtained, determines best evacuation path.
Scheme as a further preference is preset and is changed in the utilization Sarsa algorithms selections most preferably evacuate path
Decay value, the decay value value range be 0-1, when need Sarsa algorithms single step update when, the decay value, which is arranged, is
0, when wanting the newer dynamics of the everything of Sarsa algorithms consistent, it is 1 that the decay value, which is arranged, and the decay value value is
When between 0-1, the size for presetting decay value is directly proportional to the action update intensity closer from target point.
The second object of the present invention is to provide a kind of computer readable storage medium.
To achieve the goals above, the present invention is using a kind of following technical solution：
A kind of computer readable storage medium, wherein being stored with a plurality of instruction, described instruction is suitable for by terminal device equipment
Processor load and execute following processing：
Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；
Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and person model is imported and is dredged
It dissipates in model of place, crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；
Crowd after initialization is grouped to and is filtered out the leader of each group, social force model progress is micro- using improving
Crowd movement's guidance is seen, path, leader's led cluster are most preferably evacuated using the Sarsa algorithms selections for presetting modifiable decay value
Normal pedestrian is according to the movement of best evacuation path in group.
The third object of the present invention is to provide a kind of terminal device.
To achieve the goals above, the present invention is using a kind of following technical solution：
A kind of terminal device, including processor and computer readable storage medium, processor is for realizing each instruction；It calculates
Machine readable storage medium storing program for executing is suitable for being loaded by processor and executing following processing for storing a plurality of instruction, described instruction：
Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；
Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and person model is imported and is dredged
It dissipates in model of place, crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；
Crowd after initialization is grouped to and is filtered out the leader of each group, social force model progress is micro- using improving
Crowd movement's guidance is seen, path, leader's led cluster are most preferably evacuated using the Sarsa algorithms selections for presetting modifiable decay value
Normal pedestrian is according to the movement of best evacuation path in group.
Beneficial effects of the present invention：
1, a kind of crowd evacuation emulation method and device based on Sarsa algorithms of the present invention, by Sarsa algorithms
It is combined with social force model is improved, macroscopical path planning is carried out using Sarsa algorithms, carried out using improvement social force model microcosmic
Individual movement instructs, the common crowd evacuation emulation completed under complex scene.
2, a kind of crowd evacuation emulation method and device based on Sarsa algorithms of the present invention, sharp Sarsa (λ) are calculated
Method, be arranged a decay value λ, when λ between zero and one, value is bigger, and the action update intensity bigger closer from target point is in this way
Just do not have to be limited to that single step is newer can only to update nearest step action every time, all related steps of update that can be more efficiently
Action.
3, a kind of crowd evacuation emulation method and device based on Sarsa algorithms of the present invention, are being familiar with scene
It is grouped evacuation under the guiding of navigation, the utilization rate in channel and personnel's peace under crisis situations in public place can be effectively improved
Quan Xing is conducive to design evacuation prediction scheme, and help is provided for true evacuation rehearsal.
Description of the drawings
The accompanying drawings which form a part of this application are used for providing further understanding of the present application, and the application's shows
Meaning property embodiment and its explanation do not constitute the improper restriction to the application for explaining the application.
Fig. 1 is flow chart of the method for the present invention；
Fig. 2 is crowd evacuation schematic diagram in certain classroom of the embodiment of the present invention 2；
Fig. 3 is crowd evacuation schematic diagram in certain school corridor of the embodiment of the present invention 2；
Fig. 4 be the embodiment of the present invention 2 emulation experiment in crowd evacuation initialization schematic diagram；
Fig. 5 be the embodiment of the present invention 2 crowd's grouping after lead individual choice to evacuate target, and to the mobile signal in outlet
Figure；
Fig. 6 is that the individual in leading individual to execute Sarsa algorithms, organizing of the embodiment of the present invention 2 is followed and led close to outlet
Schematic diagram；
Fig. 7 is the embodiment of the present invention 2 in evacuation finish time schematic diagram.
Specific implementation mode：
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the application.Unless another
It indicates, all technical and scientific terms that the present embodiment uses have and the application person of an ordinary skill in the technical field
Normally understood identical meanings.
It should be noted that term used herein above is merely to describe specific implementation mode, 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 singulative
It is also 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 combination thereof.
It should be noted that flowcharts and block diagrams in the drawings show according to various embodiments of the present disclosure method and
The architecture, function and operation in the cards of system.It should be noted that each box in flowchart or block diagram can represent
A part for a part for one module, program segment, or code, the module, program segment, or code may include one or more
A executable instruction for realizing the logic function of defined in each embodiment.It should also be noted that some alternately
Realization in, the function that is marked in box can also occur according to the sequence different from being marked in attached drawing.For example, two connect
The box even indicated can essentially be basically executed in parallel or they can also be executed in a reverse order sometimes,
This depends on involved function.It should also be noted that each box in flowchart and or block diagram and flow chart
And/or the combination of the box in block diagram, it can be come using the dedicated hardware based system for executing defined functions or operations
It realizes, or can make to combine using a combination of dedicated hardware and computer instructions to realize.
In the absence of conflict, the features in the embodiments and the embodiments of the present application can be combined with each other with reference to
The invention will be further described with embodiment for attached drawing.
Embodiment 1：
The purpose of the present embodiment 1 is to provide a kind of crowd evacuation emulation method based on Sarsa algorithms.
To achieve the goals above, the present invention is using a kind of following technical solution：
As shown in Figure 1,
A kind of crowd evacuation emulation method based on Sarsa algorithms, this method include：
Step (1)：Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；
Step (2)：Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and by personage's mould
Type imports in evacuation model of place, and crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；
Step (3)：Crowd after initialization is grouped to and is filtered out the leader of each group, using improvement social force mould
Type carries out microcosmic crowd movement's guidance, most preferably evacuates path using the Sarsa algorithms selections for presetting modifiable decay value, leads
Person leads normal pedestrian in group to be moved according to best evacuation path.
The present embodiment the step of in (1), obtains and the video record in region is specified to form real scene database, it is described true
Real scene database includes several real scene video informations, utilizes KLT tracing algorithms regarding from the real scene database
Crowd's parameter information is extracted in frequency information, crowd's parameter information includes crowd evacuation motion path, initializes coordinate, is initial
Speed.
The present embodiment the step of in (2), setting evacuation scenario parameters information creates evacuation model of place and personage's mould
Type, and person model is imported in evacuation model of place, environment space of the evacuation model of place as crowd evacuation is described
Person model is as evacuation crowd.The semantic information of extraction evacuation model of place, by crowd's parameter information under the evacuation scene
Crowd's initialization is carried out as default evacuation crowd parameter information, and according to default evacuation crowd parameter information.
The present embodiment the step of in (2), it is arranged at each emergency exit of the evacuation model of place of establishment and counts
Device, crowd's number of individuals for counting each outlet evacuation.According to crowd's number of individuals of each outlet evacuation, outlet is calculated separately
The crowding at place；The crowding is under normal circumstances by crowd's number of individuals of the predetermined number of outlet and corresponding outlet evacuation
The ratio of total value, when exporting crowding more than the crowding threshold value, it is believed that congestion has occurred in outlet；Leader is according to distance
The crowding of the distance of each outlet port and each outlet selects most preferably to export as evacuation target.
The present embodiment the step of in (3), evacuation crowd is grouped, one is filtered out in each group and is led
Person, the leader of each group is screened according to crowd and outlet port, and the leader is in each group near outlet position
The pedestrian set.Microcosmic crowd movement's guidance is carried out using social force model is improved, using Sarsa algorithms in the crowd evacuation
Best evacuation path, leader is selected to lead the interior normal pedestrian of group according to the movement of best evacuation path in path；Preserve each group most
Good evacuation path carries out crowd evacuation emulation as the recommendation paths of evacuation rehearsal.
The present embodiment the step of in (3), normal pedestrian follows movement according to improved social force model in group.This
The improvement social force model used in invention is that outlet attraction and friend's attraction are introduced on the basis of primitive society's power model
Power respectively describes outlet to the sucking action of pedestrian and the phenomenon of gathering in groups band together of pedestrian.Considering relative velocity to the social heart
On the basis of managing power influence, desired speed is had modified, it is proposed that distance and surrounding pedestrian's speed between remaining time, pedestrian
Self related desired speed function is spent, improvement social force model is resulted in, which can be used for more truly mould
Large-scale crowd behavior in quasi- accident.
Improving social force model formula is：
The final desired speed of pedestrian is combined by self desired speed and surrounding pedestrian's average speed, pedestrian self
Desired speed is exactly that pedestrian oneself wants with great speed, and direction here is not necessarily the direction for being directed toward outlet, is merely representative of
Self desired route of pedestrian, improved desired speed formula：
Improved self drive is made of the quality of pedestrian, final desired speed, current velocity Huo existing speed and reaction time, from
Driving force illustrates that the pedestrian it is expected to move to export direction with what kind of speed, and final desired speed here is that have with direction
The speed of size, improved self drive model formation：
Improved interpersonal active force represents pedestrian j to the active force of pedestrian i, consists of two parts:The heart
Manage power and physical force, when i and j not in contact with when, indicate to only exist psychological forces between two pedestrians at this time；When i and j are contacted, this is indicated
When two pedestrians between there is only psychological forces, there is also physical force, improved interpersonal active force formula is：
The interaction force of pedestrian and barrier represents active force of the barrier to pedestrian i, also by psychological forces and physics
Power forms, when i and barrier not in contact with when, indicate that barrier only exists psychological force effect to pedestrian at this time；When i and barrier connect
When touching, public affairs indicate that there is only psychological forces effects to pedestrian for barrier at this time, there is also physics force effect, pedestrian and barrier
Interaction force formula is：
fiw=[1+cg (- vi·niw)]Ai exp[(ri-diw)/Bi]niw+kg(rij-diw)niw-kg(ri-diw)(vi·
tiw)tiw
In initial social force model, the interpersonal active force description of all pedestrians is the same, and in reality
This is false, and friend and stranger can be distinguished by introducing friend's attraction, can be simulated realistically and be tied in groups
Team's phenomenon, the formula that outlet attraction is sent out with friend's attraction are：
fis=Ci exp[(ri-dis)/Di]nis
fiq=Eexp [(riq-diq)/Fi]niq
The present embodiment the step of in (3), according to the leader current location, direction of action, using Sarsa algorithms
The income of all and current location unicom next target point is calculated, and calculates average Q value and maximum Q values, is weighted and asks
With, the point of Income Maximum is chosen as next evacuation target point, while updating income storage table, target is evacuated until reaching,
Complete income storage table is obtained, determines best evacuation path.
Sarsa algorithms are a kind of important intensified learning methods, it equally uses trial-and-error method, need not establish environment and appoint
The precise information of business describes, by learning agent can from system mode, action, reward useful information in grasp it is a set of
Optimisation strategy and knowledge, the estimation of Sarsa algorithms is action value function, that is to say, that estimation is free position s under tactful π
The action value function Q of all executable action aπ(s, a) for the state S of each nonterminalt, reach next state St+1
Afterwards, formula update Q (S be may be byt, at), and if StIt is final state, then enables Q (St+1=0, at+1), algorithm is final
Obtain stateful-action pair Q functions, and optimal policy π is exported according to Q functions.Sarsa is introduced into crowd evacuation emulation
System is advantageous in that:In face of specific environment, after training study, pedestrian can directly obtain from Sarsa algorithms
Experience in find a best evacuation path and accelerate crowd evacuation so as to avoid computing repeatedly.
The interaction of row human and environment can regard the process of a sampling four-tuple as, i.e.,：
<st, at, r, st+1,at+1>
Wherein, stExpression state, atExpression acts, and r indicates return, st+1Indicate next state, at+1It indicates next
The action taken when state.
The value function of Sarsa algorithms more new formula：
Q(st,at)←Q(st,at)+α[rt+1+λQ(st+1,at+1)-Q(st,at)]
In the present embodiment, using Sarsa algorithms come update evacuation during each step information, state stIt is exactly leader
Current position, i.e. the distance between leader and target point；Then atBe exactly leader from current location to next position
Transition, i.e., leader action direction；From starting point stState starts, and updates each state-action with Sarsa algorithms
Value function Q, strategy use ε-greedy methods, until reaching aiming spot.
Using Sarsa algorithms, the value in Q table is constantly updated, then according to new value come judge will be at some
What kind of action state takes, from being currently at stStart, has just calculated current at, and the s of next stept+1With
at+1Also it is calculated, such iteration, until reaching target point, finally obtains complete value function Q tables, the path passed by
As optimal path.
The present embodiment the step of in (3), most preferably evacuate path using Sarsa algorithms selections described, preset and can be changed
Decay value more, the decay value value range are 0-1, and when needing the single step of Sarsa algorithms to update, the decay value is arranged
It is 0, when wanting the newer dynamics of the everything of Sarsa algorithms consistent, it is 1 that the decay value, which is arranged, the decay value value
When between 0-1, the size for presetting decay value is directly proportional to the action update intensity closer from target point.
Using Sarsa (λ) algorithm, a decay value λ is set, chooses a number between 0 and 1, when λ is equal to 0,
The single step update for having reformed into Sarsa algorithms can only update nearest step action；When λ be equal to 1, reformed into bout update,
Is just as to the newer dynamics of all steps action and works as λ between zero and one, value is bigger, the action closer from target point update
The bigger .Sarsa of dynamics (λ) can update the preceding λ steps for getting reward, this sample embodiment does not just have to be limited to single step update
Can only update nearest step action every time, all related steps of the update that the present embodiment can be more efficiently act.
The present embodiment the step of in (3), the outlet if leader arrives safe and sound, leader in selection group again, until
It is optional without pedestrian's individual in each group.
Embodiment 2：
The purpose of the present embodiment 2 is to carry out experimental verification to method using the present embodiment.
In the present embodiment, the running environment of system be using Visual Studio 2012+OSG as developing instrument,
It is carried out under Windows7 operating system environments, realizes the crowd evacuation emulation under complex scene.Wait for that evacuation individual exists by 350
Crowd evacuation emulation simulation is carried out on the teaching region of 300*150, as shown in Figure 4-Figure 7.
Fig. 2 is crowd evacuation video recording sectional drawing in certain classroom；Fig. 3 is crowd evacuation video recording sectional drawing in certain school corridor；
Fig. 4 is the initialization schematic diagram of crowd evacuation in emulation experiment；Fig. 5 be crowd grouping after lead individual choice evacuate target, and
The schematic diagram mobile to outlet；Fig. 6 is that individual is being led to execute Sarsa algorithms, and the interior individual of group, which follows, leads showing close to outlet
It is intended to；Fig. 7 is in evacuation finish time schematic diagram.
Embodiment 3：
The purpose of the present embodiment 3 is to provide a kind of computer readable storage medium.
To achieve the goals above, the present invention is using a kind of following technical solution：
A kind of computer readable storage medium, wherein being stored with a plurality of instruction, described instruction is suitable for by terminal device equipment
Processor load and execute following processing：
Step (1)：Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；
Step (2)：Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and by personage's mould
Type imports in evacuation model of place, and crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；
Step (3)：Crowd after initialization is grouped to and is filtered out the leader of each group, using improvement social force mould
Type carries out microcosmic crowd movement's guidance, most preferably evacuates path using the Sarsa algorithms selections for presetting modifiable decay value, leads
Person leads normal pedestrian in group to be moved according to best evacuation path.
Embodiment 4：
The purpose of the present embodiment 4 is to provide a kind of terminal device.
To achieve the goals above, the present invention is using a kind of following technical solution：
A kind of terminal device, including processor and computer readable storage medium, processor is for realizing each instruction；It calculates
Machine readable storage medium storing program for executing is suitable for being loaded by processor and executing following processing for storing a plurality of instruction, described instruction：
Step (1)：Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；
Step (2)：Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and by personage's mould
Type imports in evacuation model of place, and crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；
Step (3)：Crowd after initialization is grouped to and is filtered out the leader of each group, using improvement social force mould
Type carries out microcosmic crowd movement's guidance, most preferably evacuates path using the Sarsa algorithms selections for presetting modifiable decay value, leads
Person leads normal pedestrian in group to be moved according to best evacuation path.
These computer executable instructions make the equipment execute according to each reality in the disclosure when running in a device
Apply method or process described in example.
In the present embodiment, computer program product may include computer readable storage medium, containing for holding
The computer-readable program instructions of row various aspects of the disclosure.Computer readable storage medium can be kept and store
By the tangible device for the instruction that instruction execution equipment uses.Computer readable storage medium for example can be-- but it is unlimited
In-- storage device electric, magnetic storage apparatus, light storage device, electromagnetism storage device, semiconductor memory apparatus or above-mentioned
Any appropriate combination.The more specific example (non exhaustive list) of computer readable storage medium includes：Portable computing
Machine disk, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM or
Flash memory), static RAM (SRAM), Portable compressed disk read-only memory (CD-ROM), digital versatile disc
(DVD), memory stick, floppy disk, mechanical coding equipment, the punch card for being for example stored thereon with instruction or groove internal projection structure, with
And above-mentioned any appropriate combination.Computer readable storage medium used herein above is not interpreted instantaneous signal itself,
The electromagnetic wave of such as radio wave or other Free propagations, the electromagnetic wave propagated by waveguide or other transmission mediums (for example,
Pass through the light pulse of fiber optic cables) or pass through electric wire transmit electric signal.
Computer-readable program instructions described herein can be downloaded to from computer readable storage medium it is each calculate/
Processing equipment, or outer computer or outer is downloaded to by network, such as internet, LAN, wide area network and/or wireless network
Portion's storage device.Network may include copper transmission cable, optical fiber transmission, wireless transmission, router, fire wall, interchanger, gateway
Computer and/or Edge Server.Adapter or network interface in each calculating/processing equipment are received from network to be counted
Calculation machine readable program instructions, and the computer-readable program instructions are forwarded, for the meter being stored in each calculating/processing equipment
In calculation machine readable storage medium storing program for executing.
Computer program instructions for executing present disclosure operation can be assembly instruction, instruction set architecture (ISA)
Instruction, machine instruction, machine-dependent instructions, microcode, firmware instructions, condition setup data or with one or more programmings
Language arbitrarily combines the source code or object code write, the programming language include the programming language-of object-oriented such as
C++ etc., and conventional procedural programming languages-such as " C " language or similar programming language.Computer-readable program refers to
Order can be executed fully, partly be executed on the user computer, as an independent software package on the user computer
Execute, part on the user computer part on the remote computer execute or completely on a remote computer or server
It executes.In situations involving remote computers, remote computer can include LAN by the network-of any kind
(LAN) or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as utilize internet
Service provider is connected by internet).In some embodiments, believe by using the state of computer-readable program instructions
Breath comes personalized customization electronic circuit, such as programmable logic circuit, field programmable gate array (FPGA) or programmable logic
Array (PLA), the electronic circuit can execute computer-readable program instructions, to realize the various aspects of present disclosure.
It should be noted that although being referred to several modules or submodule of equipment in the detailed description above, it is this
Division is merely exemplary rather than enforceable.In fact, in accordance with an embodiment of the present disclosure, two or more above-described moulds
The feature and function of block can embody in a module.Conversely, the feature and function of an above-described module can be with
It is further divided into and is embodied by multiple modules.
Beneficial effects of the present invention：
1, a kind of crowd evacuation emulation method and device based on Sarsa algorithms of the present invention, by Sarsa algorithms
It is combined with social force model is improved, macroscopical path planning is carried out using Sarsa algorithms, carried out using improvement social force model microcosmic
Individual movement instructs, the common crowd evacuation emulation completed under complex scene.
2, a kind of crowd evacuation emulation method and device based on Sarsa algorithms of the present invention, sharp Sarsa (λ) are calculated
Method, be arranged a decay value λ, when λ between zero and one, value is bigger, and the action update intensity bigger closer from target point is in this way
Just do not have to be limited to that single step is newer can only to update nearest step action every time, all related steps of update that can be more efficiently
Action.
3, a kind of crowd evacuation emulation method and device based on Sarsa algorithms of the present invention, are being familiar with scene
It is grouped evacuation under the guiding of navigation, the utilization rate in channel and personnel's peace under crisis situations in public place can be effectively improved
Quan Xing is conducive to design evacuation prediction scheme, and help is provided for true evacuation rehearsal.
The foregoing is merely the preferred embodiments of the application, are not intended to limit this application, for the skill of this field
For art personnel, the application can have various modifications and variations.Within the spirit and principles of this application, any made by repair
Change, equivalent replacement, improvement etc., should be included within the protection domain of the application.Therefore, the present invention is not intended to be limited to this
These embodiments shown in text, and it is to fit to widest range consistent with the principles and novel features disclosed in this article.
Claims (10)
1. a kind of crowd evacuation emulation method based on Sarsa algorithms, which is characterized in that this method includes：
Real scene database is received, the video crowd in database that tracks obtains crowd's parameter information；
Model of place and person model are evacuated according to default evacuation scenario parameters information creating, and person model is imported into evacuation field
In scape model, crowd's initialization is carried out using crowd's parameter information as default evacuation crowd's parameter information；
Crowd after initialization is grouped to and is filtered out the leader of each group, microcosmic people is carried out using social force model is improved
Group's exercise guidance most preferably evacuates path using the Sarsa algorithms selections for presetting modifiable decay value, and leader leads in group
Normal pedestrian is according to the movement of best evacuation path.
2. the method as described in claim 1, which is characterized in that the real scene database includes several real scene videos
Information extracts crowd's parameter information, the crowd using KLT tracing algorithms from the video information of the real scene database
Parameter information includes crowd evacuation motion path, initialization coordinate, initial velocity.
3. the method as described in claim 1, which is characterized in that in the method, screened according to crowd and outlet port each
The leader of group, the leader are the pedestrian near outlet port in each group.
If the outlet or, leader arrives safe and sound, leader in selection group again, until optional without pedestrian's individual in each group.
4. the method as described in claim 1, which is characterized in that the improvement social force model is on traditional society's power model basis
On have modified desired speed, according to distance between remaining time, pedestrian and self related desired speed of surrounding pedestrian's speed
Function is built.
5. the method as described in claim 1, which is characterized in that in the method, each of model of place is evacuated in establishment
Counter is set at a emergency exit, crowd's number of individuals for counting each outlet evacuation.
6. method as claimed in claim 5, which is characterized in that in the method, according to crowd's individual of each outlet evacuation
Number, calculates separately the crowding in exit；The crowding is under normal circumstances by the predetermined number of outlet and corresponding outlet
The ratio of crowd's number of individuals total value of evacuation, when exporting crowding more than the crowding threshold value, it is believed that gathered around outlet
It is stifled；
Leader selects most preferably to export as evacuation mesh according to the crowding of distance and each outlet apart from each outlet port
Mark.
7. the method as described in claim 1, which is characterized in that according to the leader current location, direction of action, use
Sarsa algorithms calculate the income of all and current location unicom next target point, and calculate average Q value and maximum Q values, into
Row weighted sum chooses the point of Income Maximum as next evacuation target point, while updating income storage table, is dredged until reaching
Target is dissipated, complete income storage table is obtained, determines best evacuation path.
8. the method as described in claim 1, which is characterized in that in the utilization Sarsa algorithms selections most preferably evacuate path,
Modifiable decay value is preset, the decay value value range is 0-1, and when needing the single step of Sarsa algorithms to update, institute is arranged
It is 0 to state decay value, and when wanting the newer dynamics of the everything of Sarsa algorithms consistent, it is 1 that the decay value, which is arranged, described to decline
When variate value is between 0-1, the size for presetting decay value is directly proportional to the action update intensity closer from target point.
9. a kind of computer readable storage medium, wherein being stored with a plurality of instruction, which is characterized in that described instruction is suitable for by terminal
The processor of equipment equipment loads and executes the method according to any one of claim 1-8.
10. a kind of terminal device, including processor and computer readable storage medium, processor is for realizing each instruction；It calculates
Machine readable storage medium storing program for executing is for storing a plurality of instruction, which is characterized in that described instruction is appointed for executing according in claim 1-8
Method described in one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810233963.4A CN108491972A (en) | 2018-03-21 | 2018-03-21 | A kind of crowd evacuation emulation method and device based on Sarsa algorithms |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810233963.4A CN108491972A (en) | 2018-03-21 | 2018-03-21 | A kind of crowd evacuation emulation method and device based on Sarsa algorithms |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108491972A true CN108491972A (en) | 2018-09-04 |
Family
ID=63318893
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810233963.4A Pending CN108491972A (en) | 2018-03-21 | 2018-03-21 | A kind of crowd evacuation emulation method and device based on Sarsa algorithms |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108491972A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543285A (en) * | 2018-11-20 | 2019-03-29 | 山东师范大学 | A kind of crowd evacuation emulation method and system of fused data driving and intensified learning |
| CN109670270A (en) * | 2019-01-11 | 2019-04-23 | 山东师范大学 | Crowd evacuation emulation method and system based on the study of multiple agent deeply |
| CN113536613A (en) * | 2021-09-17 | 2021-10-22 | 深圳市城市交通规划设计研究中心股份有限公司 | Crowd evacuation simulation method and device, terminal equipment and storage medium |
| CN113642978A (en) * | 2021-06-30 | 2021-11-12 | 山东师范大学 | Crowd evacuation method and system based on crowd sensing trust management mechanism |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104658297A (en) * | 2015-02-04 | 2015-05-27 | 沈阳理工大学 | Central type dynamic path inducing method based on Sarsa learning |
| CN105740514A (en) * | 2016-01-22 | 2016-07-06 | 山东师范大学 | Computer simulation system for large-size crowd evacuation and method therefor |
| CN107292064A (en) * | 2017-08-09 | 2017-10-24 | 山东师范大学 | A kind of crowd evacuation emulation method and system based on many ant colony algorithms |
| CN107464021A (en) * | 2017-08-07 | 2017-12-12 | 山东师范大学 | A kind of crowd evacuation emulation method based on intensified learning, device |
-
2018
- 2018-03-21 CN CN201810233963.4A patent/CN108491972A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104658297A (en) * | 2015-02-04 | 2015-05-27 | 沈阳理工大学 | Central type dynamic path inducing method based on Sarsa learning |
| CN105740514A (en) * | 2016-01-22 | 2016-07-06 | 山东师范大学 | Computer simulation system for large-size crowd evacuation and method therefor |
| CN107464021A (en) * | 2017-08-07 | 2017-12-12 | 山东师范大学 | A kind of crowd evacuation emulation method based on intensified learning, device |
| CN107292064A (en) * | 2017-08-09 | 2017-10-24 | 山东师范大学 | A kind of crowd evacuation emulation method and system based on many ant colony algorithms |
Non-Patent Citations (3)
| Title |
|---|
| JACK GUEST,ETC.: "Visual Analysis of Situationally Aware Building Evacuations", 《VISUALIZATION AND DATA ANALYSIS 2013》 * |
| TRAN XUAN SANG,ETC.: "Path Finding Algorithms for Autonomous Robots Based on Reinforcement Learning", 《INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN COMPUTER ENGINEERING & TECHNOLOGY》 * |
| 汪蕾: "社会力模型的改进研究", 《南京理工大学学报(自然科学版)》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543285A (en) * | 2018-11-20 | 2019-03-29 | 山东师范大学 | A kind of crowd evacuation emulation method and system of fused data driving and intensified learning |
| CN109670270A (en) * | 2019-01-11 | 2019-04-23 | 山东师范大学 | Crowd evacuation emulation method and system based on the study of multiple agent deeply |
| CN113642978A (en) * | 2021-06-30 | 2021-11-12 | 山东师范大学 | Crowd evacuation method and system based on crowd sensing trust management mechanism |
| CN113536613A (en) * | 2021-09-17 | 2021-10-22 | 深圳市城市交通规划设计研究中心股份有限公司 | Crowd evacuation simulation method and device, terminal equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108491972A (en) | A kind of crowd evacuation emulation method and device based on Sarsa algorithms | |
| WO2021073046A1 (en) | Parallel smart emergency collaboration method and system, and electronic device | |
| CN109670270A (en) | Crowd evacuation emulation method and system based on the study of multiple agent deeply | |
| CN108491598B (en) | Crowd evacuation simulation method and system based on path planning | |
| CN110399983A (en) | Shape similarity analysis | |
| CN109101694B (en) | A kind of the crowd behaviour emulation mode and system of the guidance of safe escape mark | |
| CN107480320B (en) | Crowd evacuation simulation method and system based on topological map and visual influence | |
| Wong et al. | Optimized evacuation route based on crowd simulation | |
| CN107403049B (en) | A kind of Q-Learning pedestrian's evacuation emulation method and system based on artificial neural network | |
| CN110334245A (en) | A kind of short video recommendation method and device of the figure neural network based on Temporal Order | |
| CN110415521A (en) | Prediction technique, device and the computer readable storage medium of traffic data | |
| CN108446469B (en) | Video-driven group behavior evacuation simulation method and device | |
| CN101216951A (en) | Intelligent group motion simulation method in virtual scenes | |
| Liu et al. | A perception‐based emotion contagion model in crowd emergent evacuation simulation | |
| CN107480821A (en) | The multi-Agent cooperation crowd evacuation emulation method and device of instance-based learning | |
| Sun et al. | Crowd evacuation simulation method combining the density field and social force model | |
| KR102284862B1 (en) | Method for providing video content for programming education | |
| WO2021138761A1 (en) | Task execution method and apparatus for virtual avatar, and terminal device | |
| Barnett et al. | Coordinated crowd simulation with topological scene analysis | |
| CN109584667A (en) | A kind of subway large passenger flow rehearsal simulation training system and method | |
| Karbovskii et al. | Multimodel agent-based simulation environment for mass-gatherings and pedestrian dynamics | |
| WO2021102615A1 (en) | Virtual reality scene and interaction method therefor, and terminal device | |
| KR20140137068A (en) | Evacuation simulation system and providing method thereof | |
| US20110161060A1 (en) | Optimization-Based exact formulation and solution of crowd simulation in virtual worlds | |
| Zhang et al. | Knowledge-based crowd motion for the unfamiliar environment |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180904 |
|
| RJ01 | Rejection of invention patent application after publication |