CN104866651A - Modeling and simulation method for network artificial social model based on agent - Google Patents
Modeling and simulation method for network artificial social model based on agent Download PDFInfo
- Publication number
- CN104866651A CN104866651A CN201510212556.1A CN201510212556A CN104866651A CN 104866651 A CN104866651 A CN 104866651A CN 201510212556 A CN201510212556 A CN 201510212556A CN 104866651 A CN104866651 A CN 104866651A
- Authority
- CN
- China
- Prior art keywords
- wealth
- network
- agent
- circle
- social
- 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
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000004088 simulation Methods 0.000 title claims abstract description 11
- 230000034994 death Effects 0.000 claims abstract description 26
- 238000009825 accumulation Methods 0.000 claims abstract description 14
- 230000033001 locomotion Effects 0.000 claims abstract description 7
- 230000008859 change Effects 0.000 claims description 34
- 239000004567 concrete Substances 0.000 claims description 22
- 230000010429 evolutionary process Effects 0.000 claims description 16
- 230000010287 polarization Effects 0.000 abstract description 14
- 230000008569 process Effects 0.000 abstract description 7
- 238000005094 computer simulation Methods 0.000 abstract description 2
- 230000007246 mechanism Effects 0.000 description 6
- 238000011160 research Methods 0.000 description 5
- 230000033228 biological regulation Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000015572 biosynthetic process Effects 0.000 description 3
- 230000007423 decrease Effects 0.000 description 3
- 230000004083 survival effect Effects 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000013467 fragmentation Methods 0.000 description 2
- 238000006062 fragmentation reaction Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000003997 social interaction Effects 0.000 description 2
- 230000006641 stabilisation Effects 0.000 description 2
- 238000011105 stabilization Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 235000003642 hunger Nutrition 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000004060 metabolic process Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 239000011371 regular concrete Substances 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000009154 spontaneous behavior Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a modeling and simulation method for a network artificial social model based on agent, and belongs to the technical field of computer simulation. The modeling and simulation method comprises the following steps: firstly, initializing a network, and establishing an acquaintance circle sub-network and a friend circle sub-network on a map of a sugarscape model based on a social circle theory; then, establishing populations of different levels based on the wealth amount, and dividing individuals in the space into three different levels according to the wealth amount; enabling the individuals to move and look for resources; enabling the individuals to carry out motion and wealth accumulation through motion and wealth acquiring manners; adding the birth rate, death rate and social mobile population; judging whether the individual wealth changes or not, and simulating the network evolution process through computation of structure parameters of the network; finally, modeling and simulating wealth variation of the social populations. The modeling and simulation method provides a reference model for relieving the serious problem of social polarization between the rich and the poor, and provides a preliminary theoretical basis for governments to take measures for relieving the polarization between the rich and the poor.
Description
Technical field
The present invention relates to a kind of modeling and simulating method based on the artificial social model of agent network, belong to computer simulation technique field.
Background technology
A community network is exactly the system that group or group link together by certain relation and form.The wealth of society acquisition occurred in community network and the community network mechanism of influencing each other can utilize complex network model to study.Research for complex network model creates the network model of many classics.As the modularity etc. of Small-world Characters, uncalibrated visual servo characteristic, high cluster feature and network.The social economic system of these research one side mankind is complex adaptive systems, and the spontaneous behavior of people in society and action modes being carried out artificial setting has larger gap with actual community network.On the other hand mainly discusses the wealth of society regularity of distribution from dynamic social relationships configuration aspects, but the interact relation of the wealth of society to networking does not but relate to, in society, wealth and network influence each other dynamic change.
The social economic system of the mankind is complex adaptive systems, utilizes the Lattice strain method based on agent to study the mechanism of Evolution of wealth in social economic system and obtains many concerns.On the basis of Sugascape model, introduce and to generate in a structure closer to actual community network based on social sphere's sub-network of agent and there is acquaintance's network of dynamic perfromance, then sharing rule by wealth selects friend to generate friend networks at acquaintance's network, individual agent obtains wealth distributed intelligence from friend networks, agent utilizes this information searching wealth, the change of agent wealth distribution can affect again acquaintance's network and the friend networks of agent, the final formation wealth of society and community network coevolution mechanism.So the change of single research network change or the wealth of society, all do not conform to the actual conditions.
Summary of the invention
The invention provides a kind of modeling and simulating method based on the artificial social model of agent network, for model of offering reference for the serious social polarization between the rich and the poor problem of alleviation, alleviating polarization between the rich and the poor for government formulate measure provides preliminary theoretical foundation.
The modeling and simulating method that the present invention is based on the artificial social model of agent network is achieved in that and first carries out initialization to network, then on the map of sugared domain model, sets up acquaintance's circle network and circle of friends sub-network based on social circle theory; Then set up the population of different levels based on the number of wealth amount, how many individualities in space is divided into the individuality of three class different estates by wealth amount; Then agent moves and finds resource; Then agent is undertaken moving by mobile and wealth obtain manner and accumulates with wealth; Add birth rate, mortality ratio and social mobile population again; Judge whether agent wealth changes, then by the calculating of the structural parameters of network, analog network evolutionary process; Finally utilize acquaintance's circle network of foundation and social sphere's sub-network to the wealth change modeling and simulating of social population.
The concrete steps of the described modeling and simulating method based on the artificial social model of agent network are as follows:
Step1, first initialization is carried out to network, then on the map of sugared domain model, set up acquaintance's circle network and circle of friends sub-network based on social circle theory, built the social circle of Different Individual by the size arranging the social radius of Different Individual; Netinit comprise setting initial agent sum, setting population society radius, setting society mobile, agent birth death is set;
Step2, set up the population of different levels based on the number of wealth amount: by the individuality in space, how much be divided into poverty by wealth amount, medium, be rich in three classes, concrete dividing mode is as follows:
If W
i>M
1, then for being rich in population; If W
i<M
2, be then poor population; If M
1<W
i<M
2for medium population, get e=0.3, f=0.5;
Wherein, M
1=C+D
1, M
2=C-D
2, C=M/N, D
1=Ce, D
2=Cf;
In formula, C represents whole society member revenue-center value, and M represents whole society member gross income in a certain period, and N represents whole society member total population in a certain period, D
1represent the starting point extreme difference be rich in, e represents the starting point extreme difference coefficient be rich in, D
2represent the starting point extreme difference that extreme poverty is poor, f represents the starting point extreme difference coefficient that extreme poverty is poor;
Step3, the mobile mode finding resource of setting;
Give individual agents move mode to find resource, each individual agent, by judging that the information in sugared territory around finds optimum position, collects sugar amount and the sugar amount of agent is constantly increased;
Described move mode the concrete steps of M} are:
The four direction of each individuality in oneself radius is found out the position that sugar amount is maximum, and judge that this position does not have other agent to occupy, if do not have other agent to occupy, then chosen distance oneself nearest and the maximum position of sugar amount, move to this position and on the position that this is new, sugar eaten up; If there are other agent to occupy, continue to find from oneself is nearest and maximum position measured by the sugar be not occupied, move to the sugar that this position is eaten up in this position;
Step4, setting movement and wealth obtain manner carry out moving and wealth accumulation:
In the acquaintance's circle network generated and circle of friends sub-network community network, agent obtains rule according to mobile and wealth to carry out moving and wealth accumulation, and wealth consumption;
Mobile and wealth obtain manner Z}, represent all individualities on sugared domain model map, find resource and Gains resources to survive, the concrete steps of its obtain manner are as follows:
First in a model, each agent in circle of friends sub-network
iutilize wealth judgment formula W
i>SPW
jor W
j>SPW
ifind the agent that the resource that is connected with oneself in circle of friends sub-network is maximum
j, agent after finding
ishift to agent
j, agent
iat agent
jposition be in the circle of radius with DIS=1, according to rule, { doubly, wherein, Mom is between 1-10, and now the wealth amount of each individuality is W for the tile that M} finds resource maximum the Mom eating up the resource on tile
i=W
i+ Mom*Psugar;
If the agent be connected with oneself in circle of friends sub-network
jwhen resource is not maximum, then agent
idirectly according to rule, { M} obtains wealth;
In formula, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, Psugar is the sugar amount in the map tile of sugared territory, agent
jbe i-th agent individuality, agent
ia jth agent is individual;
Step5, according to birth death mode B} adds birth rate, mortality ratio:
Birth rate, mortality ratio is introduced, for reducing society more really in the acquaintance's circle network generated and circle of friends sub-network community network; Wherein, birth rate represents new born population, in order to replace dead adult, maintains social total number of people; Mortality ratio representative is along with the adult of passage of time natural death or the individuality of emergency situations death;
Birth death mode B} is:
If agent wealth is less than or equal to 0, then agent is dead, and be born a new agent simultaneously, and the age of new agent is set to 0, and wealth is set to 4, and other attributes are identical with former generation;
If the agent age is greater than max age, then agent is dead, and be born a new agent simultaneously, and the new agent age is set to 0, and other attributes are identical with former generation, wherein, and any number of max age between 60 to 100 of each agent;
Step6, judge whether agent wealth changes; If changed, then repeat Step1, Step2, Step3, Step4, Step5 step, otherwise perform Step7;
The structural parameters analog network evolutionary process of Step7, computational grid:
By the cluster coefficients C of computational grid
i, node degree K, Ji Du ?degree correlation coefficient r, simulate the evolutionary process of network;
Step8, the acquaintance's circle network utilizing foundation and circle of friends sub-network are to the wealth change modeling and simulating of social population;
Adjust social radius by adjustment model, friends coefficient, simulation social population wealth move towards trend:
Individual wealth value constantly changes, and agent sets up new friends network according to wealth judgment formula, and acquaintance network is also ceaselessly upgrading;
According to mobile and wealth obtain manner, { Z} simulates social population's wealth change, thus sets up wealth change modeling and simulating.
In described step Step1, the concrete steps of described generation acquaintance circle network are:
Step1.1, the social radius length that first setting two is individual, judge whether the distance of two agent central point lines is less than or equal to social radius length less in two individualities;
If be less than or equal to, so two individualities are acquaintance, and two individual solid lines that adopt connect, and then form acquaintance's circle network;
If be greater than, so it doesn't matter for two individualities, and two individualities do not connect;
Step1.2, on the basis of acquaintance, the few node of setting wealth becomes the ratio SP that many friends agent of wealth need meet, then according to wealth judgment formula W
i>SPW
jor W
j>SPW
ijudge whether two individualities are friends;
If W
i>SPW
jor W
j>SPW
i, then two individualities are friends, adopt dotted line to connect between two individualities, and then form circle of friends sub-network;
If W
i≤ SPW
jor W
j≤ SPW
i, then two individualities are only acquaintance, and two individual solid lines that adopt connect;
Wherein, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, SP is the ratio that node that wealth is few becomes strong friend agent of wealth and need meet.
In described step Step7, described C
icomputing formula be: Ci=2M
i/ [k
i(k
i-1)]; Wherein, M
ifor in esse limit number, k
ifor the nodes with this node link;
The computing formula of described K is:
wherein, N is nodes all in network, k
ifor the nodes with this node link;
The computing formula of described r is:
wherein K
i, K
jrepresent limit e respectively
ijtwo nodes, M represents total limit number of network.
Carry out matching network change trend by substituting into three parameter calculation formulas: wherein, K distribution presents exponential distribution form, and the people that namely wealth is less distributes more, along with the increase number of wealth is fewer and feweri; R>0, illustrates the positive correlation of network moderate degree, and the node of namely spending large node large with degree is connected, and degree is the limit of node; C
idistribution presents the distribution of continuous long-tail, 0≤C
i≤ 1, work as C
i=0, the people in society may completely independently not connect with other individualities, works as C
i=1, individuality is connected with all individualities in regional area.
Principle of the present invention is:
Step1, first initialization is carried out to network, then on the map of sugared domain model, set up acquaintance's circle network and circle of friends sub-network based on social circle theory, built the social circle of Different Individual by the size arranging the social radius of Different Individual; What sugar domain model well explained people and outside resources needs the relation of depending on each other for existence mutually.The mankind want survival and development necessary constantly from extraneous Gains resources.And the maintenance of resource needs the intervention of people.After improving, sugared domain model explains the relation of people and resource to a great extent.If the people in society is familiar with each other, and all in the social radius of oneself, then connects, otherwise do not connect.According to everyone wealth amount size, all individualities in acquaintance's circle are divided three classes, are respectively poor, medium and rich.The social circle size of each individuality represents with social radius t-r.In society, rich somebody is often wide than the social scope of the people of poverty, is therefore divided three classes by t-r: microassociation's radius (small-reach), medium society radius (middle-reach) and great society radius (high-reach).Circle of friends sub-network is set up on the basis of acquaintance's network, just can realize wealth resource information between the individuality of regulation only on friend networks and share.There are people in actual society and tend to the network connection system that the individuality close with oneself social resources be connected, the basis of the interpersonal friend each other of regulation be respective social status whether on same level or the level that is more or less the same, thus study the wealth of society and the coevolution relation between distribution and community network thereof.And the judgement of level is defined often by respective wealth amount size.Therefore, friends is defined by the degree of closeness of wealth stock number.
Step2, set up the population of different levels based on the number of wealth amount: by the individuality in space, how much poverty is divided into by wealth amount, medium, are rich in three classes, for the ease of the evolution trend of the number of different wealth stratum in artificial social product's model that direct vision builds, in the artificial social network built, agents is simply divided into three classes according to the wealth value of agent by us, and concrete dividing mode is as follows:
If W
i>M
1, then for being rich in population; If W
i<M
2, be then poor population; If M
1<W
i<M
2for medium population, get e=0.3, f=0.5;
Wherein, M
1=C+D
1, M
2=C-D
2, C=M/N, D
1=Ce, D
2=Cf;
In formula, C represents whole society member revenue-center value, and M represents whole society member gross income in a certain period, and N represents whole society member total population in a certain period, D
1represent the starting point extreme difference be rich in, e represents the starting point extreme difference coefficient be rich in, D
2represent the starting point extreme difference that extreme poverty is poor, f represents the starting point extreme difference coefficient that extreme poverty is poor;
Step3, the mobile mode finding resource of setting;
Give individual agents move mode to find resource, each individual agent is by judging that the information in sugared territory around finds optimum position, best position refers to, the position of the maximum sugar amount within the scope of the vision of agent current present position, agent range of vision refers to the scope of r unit on the four direction of the current location four corners of the world, r is the random value of 1 to 4), collect sugar amount and the sugar amount of agent is constantly increased;
Described move mode the concrete steps of M} are:
The four direction of each individuality in oneself radius is found out the position that sugar amount is maximum, and judge that this position does not have other agent to occupy, if do not have other agent to occupy, then chosen distance oneself nearest and the maximum position of sugar amount, move to this position and on the position that this is new, sugar eaten up; If there are other agent to occupy, continue to find from oneself is nearest and maximum position measured by the sugar be not occupied, move to the sugar that this position is eaten up in this position;
Simple, { M} can define rule like this: find from the position in all range of vision from oneself is nearest and maximum position measured by the sugar be not occupied, move to the sugar that this position is eaten up in this position.
Step4, setting movement and wealth obtain manner carry out moving and wealth accumulation:
In the acquaintance's circle network generated and circle of friends sub-network community network, agent obtains rule according to mobile and wealth to carry out moving and wealth accumulation, and wealth consumption;
Draw in the community network generated, { Z} carries out moving and wealth is accumulated, and wealth consumption according to mobile and wealth obtain manner for agent; The process that individual agent finds resource is by indirectly obtaining wealth compared with the introduction of powerhouse in circle of friends network.In circle of friends, all agent are friends, must have one or certain several comparatively powerhouse in the middle of all friends, if these friends are ready to obtain chance and the information of wealth and oneself to share, oneself also just can obtain the chance of earning wealth very soon, and then realizes the accumulation of wealth and experience;
Mobile and wealth obtain manner Z}, represent all individualities on sugared domain model map, find resource and Gains resources to survive, its obtain manner concrete steps are as follows:
First in a model, each agent in circle of friends sub-network
iutilize wealth judgment formula W
i>SPW
jor W
j>SPW
ifind the agent that the resource that is connected with oneself in circle of friends sub-network is maximum
j, agent after finding
ishift to agent
j, agent
iat agent
jposition be in the circle of radius with DIS=1, according to rule, { doubly, wherein, Mom is between 1-10, and now the wealth amount of each individuality is W for the tile that M} finds resource maximum the Mom eating up the resource on tile
i=W
i+ Mom*Psugar;
If the agent be connected with oneself in circle of friends sub-network
jwhen resource is not maximum, then agent
idirectly according to rule, { M} obtains wealth;
In formula, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, Psugar is the sugar amount in the map tile of sugared territory, agent
jbe i-th agent individuality, agent
ia jth agent is individual;
Step5, according to birth death mode B} adds birth rate, mortality ratio:
Birth rate, mortality ratio is introduced, for reducing society more really in the acquaintance's circle network generated and circle of friends sub-network community network; Wherein, birth rate represents new born population, in order to replace dead adult, maintains social total number of people; Mortality ratio representative is along with the adult of passage of time natural death or the individuality of emergency situations death;
Agent sugar amount constantly increases, and simultaneously owing to there is metabolism, the agent sugar amount that often makes a move also can reduce accordingly.If in some moment, the sugared quantitative change of agent is 0 or is less than 0, be then considered as this agent and died of hunger, will be removed network.If agent will not survived by hungry to death always.Each agent also has oneself age limit, in society, individual can not unlimited survival, if the agent age has arrived the max age of oneself, be then considered as death and also will be removed network.There is death also must have birth;
Birth death rule B} is:
If agent wealth is less than or equal to 0, then agent is dead, and be born a new agent simultaneously, and the age of new agent is set to 0, and wealth is set to 4, and other attributes are identical with former generation;
If the agent age is greater than max age, then agent is dead, and be born a new agent simultaneously, and the new agent age is set to 0, and other attributes are identical with former generation, wherein, and any number of max age between 60 to 100 of each agent;
Step6, judge whether agent wealth changes; If changed, then repeat Step1, Step2, Step3, Step4, Step5 step, otherwise perform Step7;
The structural parameters analog network evolutionary process of Step7, computational grid:
By the cluster coefficients C of computational grid
i, node degree K, Ji Du ?degree correlation coefficient r, simulate the evolutionary process of network; The evolutionary process deriving network can adopt Matlab, and cluster coefficients refers to the probability also interlinked with two nodes of same node link in a network, and this coefficient is used for portraying the Local Structure character of network; Degree is the simple and important concept describing separate nodes attribute, node face limit number degree of being; Network Du ?degree related coefficient for explaining that network is positive correlation or negative correlation or uncorrelated;
Step8, the acquaintance's circle network utilizing foundation and circle of friends sub-network are to the wealth change modeling and simulating of social population; Society's radius can affect the size of acquaintance's circle network and circle of friends sub-network, and then affects friend's quantity of each individuality, and the channel of the individual obtaining information that friend is more can increase, and the chance of accumulating wealth is larger; Friends coefficient determines the condition that two individualities become friend, and the individuality that resource is many might not be ready that the individuality few with resource becomes friend;
Adjust social radius by adjustment model, friends coefficient, simulation social population wealth move towards trend:
Individual wealth value constantly changes, and agent sets up new friends network according to wealth judgment formula, and acquaintance network is also ceaselessly upgrading;
According to mobile and wealth obtain manner, { Z} simulates social population's wealth change, thus sets up wealth change modeling and simulating.
The invention has the beneficial effects as follows:
1, it is a kind of method can adding up multiple actual community network feature, and it is built by agent and management social relationships itself, more close to the operating mechanism of actual society, can describe the mechanism of large scale community network coupling evolution flexibly.
What 2, sugared domain model well explained people and outside resources needs the relation of depending on each other for existence mutually.The mankind want survival and development necessary constantly from extraneous Gains resources.And the maintenance of resource needs the intervention of people.After improving, sugared domain model explains the relation of people and resource to a great extent.
3, individual networks scale produces the higher and scale dynamic change in time of difference, cluster coefficients due to the difference of social radius; The global density of network is relatively low, degree of having degree positive correlation, social structure and the feature such as average path is short, so individual networks parameter in social sphere's sub-network and overall network parameter are closer to reality.
4, the change of population wealth can change the strong ties relation of acquaintance's circle network and circle of friends sub-network inside, also can change the Weak link between circle simultaneously.Population wealth Baidu changing factor all can cause the disconnection of old relation and the foundation of new relation, thus causes the dynamic evolution of social interactions network, have mapped the dynamic property of network.
5, the difference of friends coefficient in network, the circle of friends sub-network formed varies in size, the Weak link between circle can be formed, and tend to be connected to each other between the individuality that wealth amount is suitable, the community of people's social interaction circle and community network complexity and the trend of the weak associating of combination among the strong ones can be expressed.
Accompanying drawing explanation
Fig. 1 is contact situation schematic diagram one individual in the present invention society circle;
Fig. 2 is contact situation schematic diagram two individual in the present invention society circle;
Fig. 3 is the agent visual field of the present invention schematic diagram three;
Fig. 4 is simulation contact surface of the present invention;
Internet pricing Distribution evolution when Fig. 5 is SP=20% of the present invention;
Internet pricing Distribution evolution when Fig. 6 is SP=50% of the present invention;
Internet pricing Distribution evolution when Fig. 7 is SP=90% of the present invention;
Network wealth when Fig. 8 is SP=20% of the present invention develops;
Lorentz curve when Fig. 9 is SP=20% of the present invention;
Network wealth when Figure 10 is SP=50% of the present invention develops;
Lorentz curve when Figure 11 is SP=50% of the present invention;
Network wealth when Figure 12 is SP=90% of the present invention develops;
Lorentz curve when Figure 13 is SP=90% of the present invention;
Figure 14 is the Gini index contrast under the different SP value of the present invention.
Embodiment
Embodiment 1: as represented in figures 1 through 14, a kind of modeling and simulating method based on the artificial social model of agent network, first carries out initialization to network, then on the map of sugared domain model, sets up acquaintance's circle network and circle of friends sub-network based on social circle theory; Then set up the population of different levels based on the number of wealth amount, how many individualities in space is divided into the individuality of three class different estates by wealth amount; Then agent moves and finds resource; Then agent is undertaken moving by mobile and wealth obtain manner and accumulates with wealth; Add birth rate, mortality ratio and social mobile population again; Judge whether agent wealth changes, then by the calculating of the structural parameters of network, analog network evolutionary process; Finally utilize acquaintance's circle network of foundation and social sphere's sub-network to the wealth change modeling and simulating of social population.
The concrete steps of the described modeling and simulating method based on the artificial social model of agent network are as follows:
Step1, first initialization is carried out to network, then on the map of sugared domain model, set up acquaintance's circle network and circle of friends sub-network based on social circle theory, built the social circle of Different Individual by the size arranging the social radius of Different Individual; Netinit comprise setting initial agent sum, setting population society radius, setting society mobile, agent birth death is set;
Step2, set up the population of different levels based on the number of wealth amount: by the individuality in space, how much be divided into poverty by wealth amount, medium, be rich in three classes, concrete dividing mode is as follows:
If W
i>M
1, then for being rich in population; If W
i<M
2, be then poor population; If M
1<W
i<M
2for medium population, get e=0.3, f=0.5;
Wherein, M
1=C+D
1, M
2=C-D
2, C=M/N, D
1=Ce, D
2=Cf;
In formula, C represents whole society member revenue-center value, and M represents whole society member gross income in a certain period, and N represents whole society member total population in a certain period, D
1represent the starting point extreme difference be rich in, e represents the starting point extreme difference coefficient be rich in, D
2represent the starting point extreme difference that extreme poverty is poor, f represents the starting point extreme difference coefficient that extreme poverty is poor;
Step3, the mobile mode finding resource of setting;
Give individual agents move mode to find resource, each individual agent, by judging that the information in sugared territory around finds optimum position, collects sugar amount and the sugar amount of agent is constantly increased;
Described move mode the concrete steps of M} are:
The four direction of each individuality in oneself radius is found out the position that sugar amount is maximum, and judge that this position does not have other agent to occupy, if do not have other agent to occupy, then chosen distance oneself nearest and the maximum position of sugar amount, move to this position and on the position that this is new, sugar eaten up; If there are other agent to occupy, continue to find from oneself is nearest and maximum position measured by the sugar be not occupied, move to the sugar that this position is eaten up in this position;
Step4, setting movement and wealth obtain manner carry out moving and wealth accumulation:
In the acquaintance's circle network generated and circle of friends sub-network community network, agent obtains rule according to mobile and wealth to carry out moving and wealth accumulation, and wealth consumption;
Mobile and wealth obtain manner Z}, represent all individualities on sugared domain model map, find resource and Gains resources to survive, the concrete steps of its obtain manner are as follows:
First in a model, each agent in circle of friends sub-network
iutilize wealth judgment formula W
i>SPW
jor W
j>SPW
ifind the agent that the resource that is connected with oneself in circle of friends sub-network is maximum
j, agent after finding
ishift to agent
j, agent
iat agent
jposition be in the circle of radius with DIS=1, according to rule, { doubly, wherein, Mom is between 1-10, and now the wealth amount of each individuality is W for the tile that M} finds resource maximum the Mom eating up the resource on tile
i=W
i+ Mom*Psugar;
If the agent be connected with oneself in circle of friends sub-network
jwhen resource is not maximum, then agent
idirectly according to rule, { M} obtains wealth;
In formula, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, Psugar is the sugar amount in the map tile of sugared territory, agent
jbe i-th agent individuality, agent
ia jth agent is individual;
Step5, according to birth death mode B} adds birth rate, mortality ratio:
Birth rate, mortality ratio is introduced, for reducing society more really in the acquaintance's circle network generated and circle of friends sub-network community network; Wherein, birth rate represents new born population, in order to replace dead adult, maintains social total number of people; Mortality ratio representative is along with the adult of passage of time natural death or the individuality of emergency situations death;
Birth death mode B} is:
If agent wealth is less than or equal to 0, then agent is dead, and be born a new agent simultaneously, and the age of new agent is set to 0, and wealth is set to 4, and other attributes are identical with former generation;
If the agent age is greater than max age, then agent is dead, and be born a new agent simultaneously, and the new agent age is set to 0, and other attributes are identical with former generation, wherein, and any number of max age between 60 to 100 of each agent;
Step6, judge whether agent wealth changes; If changed, then repeat Step1, Step2, Step3, Step4, Step5 step, otherwise perform Step7;
The structural parameters analog network evolutionary process of Step7, computational grid:
By the cluster coefficients C of computational grid
i, node degree K, Ji Du ?degree correlation coefficient r, simulate the evolutionary process of network;
Step8, the acquaintance's circle network utilizing foundation and circle of friends sub-network are to the wealth change modeling and simulating of social population;
Adjust social radius by adjustment model, friends coefficient, simulation social population wealth move towards trend:
Individual wealth value constantly changes, and agent sets up new friends network according to wealth judgment formula, and acquaintance network is also ceaselessly upgrading;
According to mobile and wealth obtain manner, { Z} simulates social population's wealth change, thus sets up wealth change modeling and simulating.
In described step Step1, the concrete steps of described generation acquaintance circle network are:
Step1.1, the social radius length that first setting two is individual, judge whether the distance of two agent central point lines is less than or equal to social radius length less in two individualities;
If be less than or equal to, so two individualities are acquaintance, and two individual solid lines that adopt connect, and then form acquaintance's circle network;
If be greater than, so it doesn't matter for two individualities, and two individualities do not connect;
Step1.2, on the basis of acquaintance, the few node of setting wealth becomes the ratio SP that many friends agent of wealth need meet, then according to wealth judgment formula W
i>SPW
jor W
j>SPW
ijudge whether two individualities are friends;
If W
i>SPW
jor W
j>SPW
i, then two individualities are friends, adopt dotted line to connect between two individualities, and then form circle of friends sub-network;
If W
i≤ SPW
jor W
j≤ SPW
i, then two individualities are only acquaintance, and two individual solid lines that adopt connect;
Wherein, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, SP is the ratio that node that wealth is few becomes strong friend agent of wealth and need meet.
In described step Step7, described C
icomputing formula be: Ci=2M
i/ [k
i(k
i-1)]; Wherein, M
ifor in esse limit number, k
ifor the nodes with this node link;
The computing formula of described K is:
wherein, N is nodes all in network, k
ifor the nodes with this node link;
The computing formula of described r is:
wherein K
i, K
jrepresent limit e respectively
ijtwo nodes, M represents total limit number of network.
Embodiment 2: as represented in figures 1 through 14, a kind of modeling and simulating method based on the artificial social model of agent network, first carries out initialization to network, then on the map of sugared domain model, sets up acquaintance's circle network and circle of friends sub-network based on social circle theory; Then set up the population of different levels based on the number of wealth amount, how many individualities in space is divided into the individuality of three class different estates by wealth amount; Then agent moves and finds resource; Then agent is undertaken moving by mobile and wealth obtain manner and accumulates with wealth; Add birth rate, mortality ratio and social mobile population again; Judge whether agent wealth changes, then by the calculating of the structural parameters of network, analog network evolutionary process; Finally utilize acquaintance's circle network of foundation and social sphere's sub-network to the wealth change modeling and simulating of social population.
The concrete steps of the described modeling and simulating method based on the artificial social model of agent network are as follows:
Step1, first initialization is carried out to network, then on the map of sugared domain model, set up acquaintance's circle network and circle of friends sub-network based on social circle theory, built the social circle of Different Individual by the size arranging the social radius of Different Individual;
Initial parameter design is as shown in table 1:
The initial preset parameter of table 1
Society's circle theory provision: create contact between individuality time, require that both sides are familiar with thinking that both are understanding just now, as shown in Figure 1, if the social radius ratio B of A is large, so B is included in the circle of A, can regard as: A is familiar with B, but B is not familiar with A; As shown in Figure 2, two individual social radiuses are identical, and so A and B is familiar with mutually.
Two kinds of circle specific explanations are as follows:
(1) acquaintance's circle network is generated
Utilize the community network acquaintance circle internetwork connection mode based on social sphere's sub-network.If the people in society is familiar with each other, and all in the social radius of oneself, then sets up solid line and connect, otherwise do not connect.In addition, according to everyone wealth amount size, all individualities in acquaintance's circle are divided three classes, are respectively poor, medium and rich.Can be red by color respectively in two-dimensional space network, green and blueness represents.
(2) circle of friends sub-network is generated
Circle of friends sub-network is set up on the basis of acquaintance's network, just can realize wealth resource information between the individuality of regulation only on friend networks and share.If the wealth difference between two individualities is in the scope required, then both are friend's care and then connect, otherwise can not form friend.In two-dimensional space network, the line between friend is set to dotted line by us.
Step2, set up the population of different levels based on the number of wealth amount: by the individuality in space, how much be divided into poverty by wealth amount, medium, be rich in three classes, concrete dividing mode is as follows:
If W
i>M
1, then for being rich in population; If W
i<M
2, be then poor population; If M
1<W
i<M
2for medium population, get e=0.3, f=0.5;
Wherein, M
1=C+D
1, M
2=C-D
2, C=M/N, D
1=Ce, D
2=Cf;
In formula, C represents whole society member revenue-center value, and M represents whole society member gross income in a certain period, and N represents whole society member total population in a certain period, D
1represent the starting point extreme difference be rich in, e represents the starting point extreme difference coefficient be rich in, D
2represent the starting point extreme difference that extreme poverty is poor, f represents the starting point extreme difference coefficient that extreme poverty is poor;
Step3, the mobile mode finding resource of setting;
Give individual agents move mode to find resource, each individual agent, by judging that the information in sugared territory around finds optimum position, collects sugar amount and the sugar amount of agent is constantly increased;
Described move mode the concrete steps of M} are:
Be illustrated in figure 3 the range of vision of agent, the four direction of each individuality in oneself radius is found out the position that sugar amount is maximum, and judge that this position does not have other agent to occupy, if do not have other agent to occupy, then chosen distance oneself nearest and the maximum position of sugar amount, move to this position and on the position that this is new, sugar eaten up; If there are other agent to occupy, continue to find from oneself is nearest and maximum position measured by the sugar be not occupied, move to the sugar that this position is eaten up in this position;
Step4, setting movement and wealth obtain manner carry out moving and wealth accumulation:
In the acquaintance's circle network generated and circle of friends sub-network community network, agent obtains rule according to mobile and wealth to carry out moving and wealth accumulation, and wealth consumption;
Mobile and wealth obtain manner Z}, represent all individualities on sugared domain model map, find resource and Gains resources to survive, its regular concrete steps are as follows:
First in a model, each agent in circle of friends sub-network
iutilize wealth judgment formula W
i>SPW
jor W
j>SPW
ifind the agent that the resource that is connected with oneself in circle of friends sub-network is maximum
j, agent after finding
ishift to agent
j, agent
iat agent
jposition be in the circle of radius with DIS=1, according to rule, { doubly, wherein, Mom is between 1-10, and now the wealth amount of each individuality is W for the tile that M} finds resource maximum the Mom eating up the resource on tile
i=W
i+ Mom*Psugar;
If the agent be connected with oneself in circle of friends sub-network
jwhen resource is not maximum, then agent
idirectly according to rule, { M} obtains wealth;
In formula, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, Psugar is the sugar amount in the map tile of sugared territory, agent
jbe i-th agent individuality, agent
ia jth agent is individual;
Step5, according to birth death mode B} adds birth rate, mortality ratio:
Birth rate, mortality ratio is introduced, for reducing society more really in the acquaintance's circle network generated and circle of friends sub-network community network; Wherein, birth rate represents new born population, in order to replace dead adult, maintains social total number of people; Mortality ratio representative is along with the adult of passage of time natural death or the individuality of emergency situations death;
Birth death mode B} is:
If agent wealth is less than or equal to 0, then agent is dead, and be born a new agent simultaneously, and the age of new agent is set to 0, and wealth is set to 4, and other attributes are identical with former generation;
If the agent age is greater than max age, then agent is dead, and be born a new agent simultaneously, and the new agent age is set to 0, and other attributes are identical with former generation, wherein, and any number of max age between 60 to 100 of each agent;
Step6, judge whether agent wealth changes; If changed, then repeat Step1, Step2, Step3, Step4, Step5 step, otherwise perform Step7;
The structural parameters analog network evolutionary process of Step7, computational grid:
By the cluster coefficients C of computational grid
i, node degree K, Ji Du ?degree correlation coefficient r, simulate the evolutionary process of network;
Step8, the acquaintance's circle network utilizing foundation and circle of friends sub-network are to the wealth change modeling and simulating of social population;
Adjust social radius by adjustment model, friends coefficient, simulation social population wealth move towards trend:
Individual wealth value constantly changes, and agent sets up new friends network according to wealth judgment formula, and acquaintance network is also ceaselessly upgrading;
According to mobile and wealth obtain manner, { Z} simulates social population's wealth change, thus sets up wealth change modeling and simulating.
In described step Step1, the concrete steps of described generation acquaintance circle network are:
Step1.1, the social radius length that first setting two is individual, judge whether the distance of two agent central point lines is less than or equal to social radius length less in two individualities;
If be less than or equal to, so two individualities are acquaintance, and two individual solid lines that adopt connect, and then form acquaintance's circle network;
If be greater than, so it doesn't matter for two individualities, and two individualities do not connect;
Step1.2, on the basis of acquaintance, the few node of setting wealth becomes the ratio SP that many friends agent of wealth need meet, then according to wealth judgment formula W
i>SPW
jor W
j>SPW
ijudge whether two individualities are friends;
If W
i>SPW
jor W
j>SPW
i, then two individualities are friends, adopt dotted line to connect between two individualities, and then form circle of friends sub-network;
If W
i≤ SPW
jor W
j≤ SPW
i, then two individualities are only acquaintance, and two individual solid lines that adopt connect;
Wherein, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, SP is the ratio that node that wealth is few becomes strong friend agent of wealth and need meet.
In described step Step7, described C
icomputing formula be: Ci=2M
i/ [k
i(k
i-1)]; Wherein, M
ifor in esse limit number, k
ifor the nodes with this node link;
The computing formula of described K is:
wherein, N is nodes all in network, k
ifor the nodes with this node link;
The computing formula of described r is:
wherein K
i, K
jrepresent limit e respectively
ijtwo nodes, M represents total limit number of network.
Process flow diagram as shown in Figure 4, first sets agent sum, the social radius of each agent individuality, agent attribute, agent behavior, initial wealth value; In order to the situation that people in simulating reality life may occur, we also set birth rate, mortality ratio; By setting friends coefficient S P, in order to characterize the degree of congestion of underclass and upper class passage, thus the extremes of poverty and wealth of reflection society.Wherein SP represents, comparatively poor individual wealth value must account for rich individual more than SP, just can become rich individual friend.
By the value of adjustment model SP, and fix the value of other parameters, under obtaining different SP value, the influence curve of social polarization between the rich and the poor:
The first step: during SP=20%, the wealth matter of the less agent of individual wealth be connected that the dynamic evolution of network: SP=20% represents in acquaintance's network accounts for more than 20% of the larger agent of wealth, and both are friends.Network presents community phenomenon gradually, and the individual amount be free on outside community constantly reduces, the community that final formation two is large, and occupies two larger places of resource in space respectively.Community inside is interconnected, and is all friends between all agent.Although be dispersed with poverty in network, medium and to be rich in three classes individual, this social structure does not have appearance three class community.But rich or poor two community embraced together.And although isolated individuality is distributed in the less area of resource more, they are not but the poor individual distribution yet having a small amount of medium and rich individuality entirely.This illustrates when network reaches stable, and the cognition that a part isolates joins in the middle of community, and individuality also can isolate out from community, and network is a dynamic steady state (SS).This and social real-life match.
Be illustrated in figure 5 the degree Distribution evolution process of network.Initial time, most of individual angle value is between 1 to 3, and a few peoples' angle value is larger.Along with the change of time and condition, engender little community between individuality, as shown in Fig. 5 (b).There is multiple peak value in degree distribution curve, can find out that little peak value is approximate obey Poisson distribution from the peak value figure amplified, and the conclusion of Poisson distribution and the degree of the community network in the past research conclusion that distributes is consistent.When network reach stable after, forms two large community groups, although the peak value existence now in curve, increasing number, peak value diminishes.Maximum angle value is about 55.This shows, each large community is converged by much little community to form.There is strong ties in little community inside, is connected between little community by Weak link.The quantity that each agent links friend is more or less the same.
Second step: the dynamic evolution of network during SP=50%: equally also occurred community phenomenon during network stabilization, but compared with network during SP=20%, community is now no longer mainly concentrate on two community, but multiple community.Du-Du related coefficient when network is tending towards dynamic stability is 0.038 is same distribution network, and cluster coefficients is 0.795, and sight when comparing SP=20% declines to some extent.This shows that the weak link tightness degree between the strong link of social structure inside and community declines all to some extent.
As shown in Figure 6, be degree Distribution evolution process.The degree distribution of network initial time Fig. 6 (a) presents disparity, and the individuality that angle value is less is in the great majority, and breaks up comparatively serious.Along with passage of time, also there is multiple peak value in the degree of Fig. 6 (b) network, each peak value follows Poisson distribution equally.When network reaches dynamic steady state, compare with during SP=20%, be not difficult to find out that number of peaks now increases further, and occur the place that many places peak value is larger; Maximum angle value is also reduced to about 50 by 55, although this illustrates that friends network scale reduces quantity and is being on the increase, tightness degree declines.{ individuality that F} makes wealth differ within 50% can become friends and then set up friends network rule, and the probability that the rich are connected with the rich strengthens, combination among the strong ones intensifying trend.
3rd step: the dynamic evolution of network during SP=90%: the evolutionary process of network is similar with twice evolution above.Network reaches the dynamic stability moment, and community quantity increases further, and scale reduces further, is difficult to form larger community again.Now Du-Du the related coefficient of network is 0.032, is equally also same distribution network.The cluster coefficients of network is 0.746, and with 0.818,0.786 compares, and numerical value reduces again, and the tightness degree that network connects declines, and community quantity increases further.
As shown in Figure 7, be degree Distribution evolution process.Initial time is as shown in Fig. 7 (a), and the unbalanced phenomena of internet pricing distribution is strengthened further.When network evolution is to intermediate time, because the tightness degree of network declines, although there is the appearance of community, be not clearly.When network reaches dynamic stability, as shown in Fig. 7 (c), maximum angle value is about 40, and decline obviously, peak value mainly appears at beginning and the position, end of curve, this shows, the individuality that angle value is less and the larger individuality of angle value show little community phenomenon, the further fragmentation of community respectively, only have wealth quantity minority individuality closely just can form friend's community network, combination among the strong ones, weak associating trend is obvious.
4th step: the wealth distribution of network during SP=20%: as shown in Figure 8, due to initial time, by geographical environment and class origin to affect polarization between the rich and the poor comparatively serious.As time goes on, poor individual amount reduces, and nearly all individuality is obtained for the accumulation of wealth, and wealth value constantly increases.When network is close to equilibrium state, the number of each stratum is suitable, polarization between the rich and the poor obtains to be alleviated significantly (although it is to be noted that network reaches stable here, but the activity that individuality is accumulated wealth remains to be carried out, this is the process of a mobile equilibrium), the wealth distribution of whole network presents equalization phenomenon.This also can be confirmed by the Lorentz curve of Fig. 9.
Fig. 9 is the change of the Lorentz curve of network evolution different times, Lorentz curve is in order to compare and to analyze a country at different times or country variant in the unequal degree of the wealth in same epoch, and this curve is a graphical method easily of summing up income and Wealthy distribution information.By network human relations hereby curve, can intuitively to a national distribution of earnings equality or unequal situation.As shown in Figure 9 (solid line and skew lines enclose the Lorentz curve that area is initial time; Area between some septal line and skew lines is the Lorentz curve of network evolution intermediate time; Dashed curve and skew lines to enclose area be the Lorentz curve that network reaches the dynamic stability moment), the area between curve and skew lines is more and more less.This shows that individual social increasing extent is extensive, and wealth flow range becomes large.The more wealth of increasing poor individual acquisition, the unbalanced phenomenon of the wealth of society is eased.
5th step: the wealth distribution of network during SP=50%, as shown in Figure 10, the wealth substep of initial time still presents great lack of uniformity, and the poor's quantity is a lot, and polarization between the rich and the poor is comparatively serious.As time goes on, although each individuality is also ceaselessly being accumulated wealth, network is reaching the dynamic stability moment, and as shown in Figure 10 (c), although polarization between the rich and the poor obtains certain alleviation, its degree decreases compared with SP=20%.SP=50% represents: according to rule { F
p, weak person's resource at least just will have an opportunity become with it friends and then connect at powerhouse's more than 50%, and SP value is larger, the individuality showing to be in the bottom of society and upper strata individuality to exchange degree lower.
As can be seen from the Lorentz curve of Figure 11, the unbalanced phenomenon of the wealth of society obtains alleviation equally.When network reaches dynamic steady state, when the area between dashed curve and skew lines is greater than parameter SP=20% the network stabilization moment area (wherein in Figure 11 solid line and skew lines enclose the Lorentz curve that area is initial time; Area between some septal line and skew lines is the Lorentz curve of network evolution intermediate time; Dashed curve and skew lines to enclose area be the Lorentz curve that network reaches the dynamic stability moment).
6th step: the wealth distribution of network during SP=90%, as shown in figure 12, initial time, still also exists polarization.As time goes on wealth differentiation weakens, but same SP=20%, and 50% compares, though individual wealth obtains accumulation journey equally, as shown in Figure 12 (b), the wealth value of each stratum increases.When network reaches dynamic stability, the degree of the alleviation of polarization much smaller than SP=20%, the network of 50%.Combination among the strong ones intensifying trend, the as easy as rolling off a log agglomerate of the rich, connects underclass relative with the passage of the upper class inaccessible.
As can be seen from Lorentz curve, as long as friends network exists, polarization between the rich and the poor must present reduction trend.Although individual wealth is tending towards balanced, the individual wealth otherness that network reaches the dynamic stability moment is not quite similar along with the change of parameter SP.As shown in figure 13 (solid line and skew lines enclose the Lorentz curve that area is initial time; Area between some septal line and skew lines is the Lorentz curve of network evolution intermediate time; Dashed curve and skew lines to enclose area be the Lorentz curve that network reaches the dynamic stability moment), dashed curve and area that skew lines encloses are greater than parameter SP=20%, the Lorentz curve in 50% moment.Under different parameters, the variation tendency of wealth equalization well can be portrayed with Gini coefficient.
7th step: after the analysis and regulation first step to the parameter of the 6th step, draw network G ini index contrast figure, as shown in figure 14, from figure, we can find out no matter SP value is how many, and Gini coefficient is all on a declining curve.This from be consistent for the conclusion of the Lorentz curve reflected under different SP value before, no matter what value SP gets, and social polarization between the rich and the poor is all constantly reducing.But the difference of SP but result in the difference of the fall off rate of Gini coefficient.SP is larger, and the wealth value ratio that two individualities become friends is larger, and require that wealth value differs less, and then circle of friends sub-network scale is less, the chance of accumulating wealth reduces.Therefore want to allow Gini coefficient decline rapidly at short notice, SP value must be reduced as far as possible, strengthen the interchange between individuality, allow more people can enter into the shared community network of wealth.
In order to study influencing each other between community network and the wealth of society, propose an artificial social model CASugarscape of complicated agent network to realize community network and the simulation of wealth of society coevolution.Introduce and to generate in a structure closer to actual community network based on social sphere's sub-network of agent and there is acquaintance's network of dynamic perfromance, then sharing rule by wealth selects friend to generate friend networks at acquaintance's network, individual agent obtains wealth distributed intelligence from friend networks, agent utilizes this information searching wealth, the change of agent wealth distribution can affect again acquaintance's network and the friend networks of agent, the final formation wealth of society and community network coevolution mechanism.Research shows that wealth shares rule or wealth matched rule is distributed with material impact to community network and wealth: the wealth matched rule that (1) forms friend is stronger, namely become friend and require that wealth amount is more close, the cluster coefficients of network reduces gradually, degree distribution presents multimodality, network maximal degree reduces gradually, but network degree of being degree is correlated with, so the tightness degree that network connects declines, community quantity increases further, the further fragmentation of community.(2) the wealth matched rule forming friend is stronger, and the possibility of the individuality composition friend community network that wealth quantity is more close is stronger.Namely become friend and require that wealth amount is more close, then more may become friend between the rich grasping resource in social network network, the individuality not grasping social resources can only become friend with the people of this stratum, combination among the strong ones, and weak associating trend is obvious.Wealth distribution polarization increases, and Gini coefficient increases.
By reference to the accompanying drawings the specific embodiment of the present invention is explained in detail above, but the present invention is not limited to above-mentioned embodiment, in the ken that those of ordinary skill in the art possess, various change can also be made under the prerequisite not departing from present inventive concept.
Claims (4)
1. based on a modeling and simulating method for the artificial social model of agent network, it is characterized in that: first initialization is carried out to network, then on the map of sugared domain model, set up acquaintance's circle network and circle of friends sub-network based on social circle theory; Then set up the population of different levels based on the number of wealth amount, how many individualities in space is divided into the individuality of three class different estates by wealth amount; Then agent moves and finds resource; Then agent is undertaken moving by mobile and wealth obtain manner and accumulates with wealth; Add birth rate, mortality ratio and social mobile population again; Judge whether agent wealth changes, then by the calculating of the structural parameters of network, analog network evolutionary process; Finally utilize acquaintance's circle network of foundation and social sphere's sub-network to the wealth change modeling and simulating of social population.
2. the modeling and simulating method based on the artificial social model of agent network according to claim 1, is characterized in that: the concrete steps of the described modeling and simulating method based on the artificial social model of agent network are as follows:
Step1, first initialization is carried out to network, then on the map of sugared domain model, set up acquaintance's circle network and circle of friends sub-network based on social circle theory, built the social circle of Different Individual by the size arranging the social radius of Different Individual;
Step2, set up the population of different levels based on the number of wealth amount: by the individuality in space, how much be divided into poverty by wealth amount, medium, be rich in three classes, concrete dividing mode is as follows:
If W
i>M
1, then for being rich in population; If W
i<M
2, be then poor population; If M
1<W
i<M
2for medium population, get e=0.3, f=0.5;
Wherein, M
1=C+D
1, M
2=C-D
2, C=M/N, D
1=Ce, D
2=Cf;
In formula, C represents whole society member revenue-center value, and M represents whole society member gross income in a certain period, and N represents whole society member total population in a certain period, D
1represent the starting point extreme difference be rich in, e represents the starting point extreme difference coefficient be rich in, D
2represent the starting point extreme difference that extreme poverty is poor, f represents the starting point extreme difference coefficient that extreme poverty is poor;
Step3, the mobile mode finding resource of setting;
Give individual agents move mode to find resource, each individual agent, by judging that the information in sugared territory around finds optimum position, collects sugar amount and the sugar amount of agent is constantly increased;
Described move mode the concrete steps of M} are:
The four direction of each individuality in oneself radius is found out the position that sugar amount is maximum, and judge that this position does not have other agent to occupy, if do not have other agent to occupy, then chosen distance oneself nearest and the maximum position of sugar amount, move to this position and on the position that this is new, sugar eaten up; If there are other agent to occupy, continue to find from oneself is nearest and maximum position measured by the sugar be not occupied, move to the sugar that this position is eaten up in this position;
Step4, setting movement and wealth obtain manner carry out moving and wealth accumulation:
In the acquaintance's circle network generated and circle of friends sub-network community network, agent obtains rule according to mobile and wealth to carry out moving and wealth accumulation, and wealth consumption;
Mobile and wealth obtain manner Z}, represent all individualities on sugared domain model map, find resource and Gains resources to survive, the concrete steps of its obtain manner are as follows:
First in a model, each agent in circle of friends sub-network
iutilize wealth judgment formula W
i>SPW
jor W
j>SPW
ifind the agent that the resource that is connected with oneself in circle of friends sub-network is maximum
j, agent after finding
ishift to agent
j, agent
iat agent
jposition be in the circle of radius with DIS=1, according to rule, { doubly, wherein, Mom is between 1-10, and now the wealth amount of each individuality is W for the tile that M} finds resource maximum the Mom eating up the resource on tile
i=W
i+ Mom*Psugar;
If the agent be connected with oneself in circle of friends sub-network
jwhen resource is not maximum, then agent
idirectly according to rule, { M} obtains wealth;
In formula, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, Psugar is the sugar amount in the map tile of sugared territory, agent
jbe i-th agent individuality, agent
ia jth agent is individual;
Step5, according to birth death mode B} adds birth rate, mortality ratio:
Birth rate, mortality ratio is introduced, for reducing society more really in the acquaintance's circle network generated and circle of friends sub-network community network; Wherein, birth rate represents new born population, in order to replace dead adult, maintains social total number of people; Mortality ratio representative is along with the adult of passage of time natural death or the individuality of emergency situations death;
Birth death mode B} is:
If agent wealth is less than or equal to 0, then agent is dead, and be born a new agent simultaneously, and the age of new agent is set to 0, and wealth is set to 4, and other attributes are identical with former generation;
If the agent age is greater than max age, then agent is dead, and be born a new agent simultaneously, and the new agent age is set to 0, and other attributes are identical with former generation, wherein, and any number of max age between 60 to 100 of each agent;
Step6, judge whether agent wealth changes; If changed, then repeat Step1, Step2, Step3, Step4, Step5 step, otherwise perform Step7;
The structural parameters analog network evolutionary process of Step7, computational grid:
By the cluster coefficients C of computational grid
i, node degree K, Ji Du ?degree correlation coefficient r, simulate the evolutionary process of network;
Step8, the acquaintance's circle network utilizing foundation and circle of friends sub-network are to the wealth change modeling and simulating of social population;
Adjust social radius by adjustment model, friends coefficient, simulation social population wealth move towards trend:
Individual wealth value constantly changes, and agent sets up new friends network according to wealth judgment formula, and acquaintance network is also ceaselessly upgrading;
According to mobile and wealth obtain manner, { Z} simulates social population's wealth change, thus sets up wealth change modeling and simulating.
3. the modeling and simulating method based on the artificial social model of agent network according to claim 1, is characterized in that: in described step Step1, and the concrete steps of described generation acquaintance circle network are:
Step1.1, the social radius length that first setting two is individual, judge whether the distance of two agent central point lines is less than or equal to social radius length less in two individualities;
If be less than or equal to, so two individualities are acquaintance, and two individual solid lines that adopt connect, and then form acquaintance's circle network;
If be greater than, so it doesn't matter for two individualities, and two individualities do not connect;
Step1.2, on the basis of acquaintance, the few node of setting wealth becomes the ratio SP that many friends agent of wealth need meet, then according to wealth judgment formula W
i>SPW
jor W
j>SPW
ijudge whether two individualities are friends;
If W
i>SPW
jor W
j>SPW
i, then two individualities are friends, adopt dotted line to connect between two individualities, and then form circle of friends sub-network;
If W
i≤ SPW
jor W
j≤ SPW
i, then two individualities are only acquaintance, and two individual solid lines that adopt connect;
Wherein, W
ibe the wealth value of i-th agent, W
jfor the wealth value of a jth agent, SP is the ratio that node that wealth is few becomes strong friend agent of wealth and need meet.
4. the modeling and simulating method based on the artificial social model of agent network according to claim 1, is characterized in that: in described step Step7, described C
icomputing formula be: Ci=2M
i/ [k
i(k
i-1)]; Wherein, M
ifor in esse limit number, k
ifor the nodes with this node link;
The computing formula of described K is:
wherein, N is nodes all in network, k
ifor the nodes with this node link;
The computing formula of described r is:
wherein K
i, K
jrepresent limit e respectively
ijtwo nodes, M represents total limit number of network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510212556.1A CN104866651A (en) | 2015-04-29 | 2015-04-29 | Modeling and simulation method for network artificial social model based on agent |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510212556.1A CN104866651A (en) | 2015-04-29 | 2015-04-29 | Modeling and simulation method for network artificial social model based on agent |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104866651A true CN104866651A (en) | 2015-08-26 |
Family
ID=53912476
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510212556.1A Pending CN104866651A (en) | 2015-04-29 | 2015-04-29 | Modeling and simulation method for network artificial social model based on agent |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104866651A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110136041A (en) * | 2019-05-09 | 2019-08-16 | 中国科学院自动化研究所 | Artificial population synthetic method, system, device based on multiple social relationships constraint |
| CN110543671A (en) * | 2019-07-25 | 2019-12-06 | 太原科技大学 | city safety livability modeling method based on artificial society |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110072052A1 (en) * | 2008-05-28 | 2011-03-24 | Aptima Inc. | Systems and methods for analyzing entity profiles |
| CN104199845A (en) * | 2014-08-08 | 2014-12-10 | 杭州电子科技大学 | On-line comment sentiment classification method based on agent model |
| CN104217100A (en) * | 2014-07-24 | 2014-12-17 | 昆明理工大学 | Modeling simulation method of language competition model based on agent social circle network |
-
2015
- 2015-04-29 CN CN201510212556.1A patent/CN104866651A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110072052A1 (en) * | 2008-05-28 | 2011-03-24 | Aptima Inc. | Systems and methods for analyzing entity profiles |
| CN104217100A (en) * | 2014-07-24 | 2014-12-17 | 昆明理工大学 | Modeling simulation method of language competition model based on agent social circle network |
| CN104199845A (en) * | 2014-08-08 | 2014-12-10 | 杭州电子科技大学 | On-line comment sentiment classification method based on agent model |
Non-Patent Citations (1)
| Title |
|---|
| 路兰等: "基于异质性的产业群交易网络", 《复杂系统与复杂性科学》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110136041A (en) * | 2019-05-09 | 2019-08-16 | 中国科学院自动化研究所 | Artificial population synthetic method, system, device based on multiple social relationships constraint |
| CN110136041B (en) * | 2019-05-09 | 2021-05-04 | 中国科学院自动化研究所 | Artificial population synthesis method, system and device based on multiple social relationship constraints |
| CN110543671A (en) * | 2019-07-25 | 2019-12-06 | 太原科技大学 | city safety livability modeling method based on artificial society |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107968962B (en) | A kind of video generation method of the non-conterminous image of two frames based on deep learning | |
| CN109299436A (en) | A kind of ordering of optimization preference method of data capture meeting local difference privacy | |
| CN114091667A (en) | Federal mutual learning model training method oriented to non-independent same distribution data | |
| Stadtfeld | The micro-macro link in social networks | |
| Lorenz | Fostering consensus in multidimensional continuous opinion dynamics under bounded confidence | |
| Moudoud et al. | Towards a secure and reliable federated learning using blockchain | |
| CN107295453A (en) | A kind of wireless sensor network data fusion method | |
| CN106373015A (en) | Intimacy determination method and system in social network | |
| CN116362327A (en) | Model training method and system and electronic equipment | |
| CN104866651A (en) | Modeling and simulation method for network artificial social model based on agent | |
| Zhang et al. | R $^{2} $ fed: resilient reinforcement federated learning for industrial applications | |
| Zou et al. | Fast comprehensive flood risk assessment based on game theory and cloud model under parallel computation (P-GT-CM) | |
| Xu et al. | Competitive search algorithm: a new method for stochastic optimization | |
| CN115879542A (en) | Federal learning method oriented to non-independent same-distribution heterogeneous data | |
| Peng et al. | Research on the random multi-attribute decision-making methods with trapezoidal fuzzy probability based on prospect theory | |
| CN112699402B (en) | Wearable device activity prediction method based on federal personalized random forest | |
| CN109194509A (en) | A method of the prediction network based on path strong or weak relation and community information connects side | |
| CN107465571A (en) | Tactical network simulated training background service flow generation method based on statistical property | |
| CN110263643A (en) | A kind of fast video people counting method based on sequential relationship | |
| Wu et al. | A novel emergence model of public opinion based on small-world network | |
| Wang et al. | Research on social resilient control structure based on SFZ | |
| CN109255376A (en) | A method of the prediction network based on the punishment of adaptive response and community information connects side | |
| Gao et al. | The Dynamics of Two-State Public Opinion Propagation On Signed Networks | |
| CN109489665B (en) | Method and system for predicting position of people refuge track based on resource allocation | |
| Xia et al. | Structural evolution in knowledge transfer network: an agent-based model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150826 |
|
| RJ01 | Rejection of invention patent application after publication |