CN110704789B - Population dynamic measurement and calculation method and system based on 'urban superconcephalon' computing platform - Google Patents
Population dynamic measurement and calculation method and system based on 'urban superconcephalon' computing platform Download PDFInfo
- Publication number
- CN110704789B CN110704789B CN201910760691.8A CN201910760691A CN110704789B CN 110704789 B CN110704789 B CN 110704789B CN 201910760691 A CN201910760691 A CN 201910760691A CN 110704789 B CN110704789 B CN 110704789B
- Authority
- CN
- China
- Prior art keywords
- space
- population
- urban
- influence degree
- spatial
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
Abstract
The invention provides a population dynamic measuring and calculating method and system automatically depending on an urban superconcephalon computing platform, wherein the method comprises the following steps: s1, presetting population types; s2, establishing an influence degree model of the space environment and the space relation of the urban space area relative to the dynamic distribution of different population types; s3, collecting a certain number of samples of real city space areas; s4, respectively training the influence degree model established in S2 by using the samples collected in S3; and S5, substituting the space environment and space relation parameters of any real urban space region into the trained influence degree model in the S4, and acquiring various types of population dynamic distribution in the urban space region. The method is beneficial to scientifically analyzing and judging the dynamic distribution conditions of different types of population with pertinence, predicting the population quantity of urban spatial areas and mining population dense points.
Description
Technical Field
The invention relates to the field of intelligent population management, in particular to a population dynamic measuring and calculating method and system based on an urban superconcephalon computing platform.
Background
The population presents high-mobility dynamic distribution in urban space, the population quantity in a future period of time is measured and calculated aiming at a space area of a city, and then dense population points are excavated, so that the allocation of resources such as transportation tools, police force, communication signal vehicles and the like is facilitated, and the accidental conditions such as congestion, trampling, communication interruption and the like of the dense population points are avoided. However, currently, a scientific quantitative analysis method is lacked for predicting the dynamic distribution of high mobility of regional population, and only the evaluation prediction can be performed by depending on experience.
At present, the city superconcephalon is taken as an intelligent big data computing system, city reality, history, time and space data can be gathered on a unified space-time coordinate system based on the infrastructure of the Internet and the Internet of things, different industry knowledge is learned by artificial intelligence, data association relation is mined, understanding of city development and operation systems is realized, global instant analysis and simulation can be carried out, and further, the optimal configuration of public resources of a physical reality city, the fine and orderly social management, the improvement of resident time quality, the efficient city operation and the sustainable development are promoted.
Therefore, how to determine the influence degree of dynamic distribution brought by the spatial environment of the urban spatial area and the spatial relationship parameters thereof to different types of population by using the urban superconcephalon system, and further measure and calculate the predicted inflow and outflow amounts of different types of population in the urban spatial area according to the influence degree, thereby realizing dynamic measurement and calculation of population is a problem to be solved by those skilled in the art.
Disclosure of Invention
In view of the above, the invention provides a population dynamic measurement and calculation method and system based on an urban superconcephalon computing platform, which aggregates spatial environment and spatial relationship parameters of a real urban spatial region on the basis of real-time big data acquisition of an internet of things architecture, runs a population distribution influence degree model facing various population types in real time through the superior performance of the urban superconcephalon computing system, measures and calculates the expected inflow and outflow amounts of different types of populations in any urban spatial region, realizes population dynamic measurement and calculation, obtains the population number of the urban spatial region in a period of time in the future, is used for discovering dense population points, and is beneficial to allocation of resources such as transportation tools, police force, communication signal vehicles and the like.
In order to achieve the purpose, the invention adopts the following technical scheme:
a population dynamic measuring and calculating method based on an urban superconcephalon computing platform comprises the following steps:
s1, presetting population types;
s2, establishing an influence degree model of the space environment and the space relation of the urban space area relative to the dynamic distribution of different population types;
s3, collecting a certain number of samples of real city space areas;
s4, respectively training the influence degree model established in S2 by using the samples collected in S3;
and S5, substituting the space environment and space relation parameters of any real urban space region into the trained influence degree model in the S4, and acquiring various types of population dynamic distribution in the urban space region.
Preferably, in S1, n population types, which are n types, are preset in the 1 st and 2 nd … th population types according to the individual endogenous differences, such as income level, occupation, age, travel mode, occupation, and the like.
Preferably, the influence model of dynamic distribution in S2 is a BP multi-layer neural network model applicable to various urban spatial regions, and for different population types, influence models of dynamic distribution of environmental relationships and spatial relationships of the urban spatial regions with respect to each of the population types 1 and 2 … n are respectively established, each influence model corresponding to one population type.
The degree of influence of the dynamic distribution generated for each type of population is quite different because of the spatial environment of the urban spatial region itself and the spatial relationship between the urban spatial region and the primary spatial targets within a certain distance of the perimeter. For example, the following steps are carried out: different types of population travel modes are divided into bus subway travel, driving travel, bicycle travel, walking travel and the like, when one subway or bus station is arranged in one urban space area, the population whose travel mode is bus or subway travel can be greatly influenced in dynamic distribution, and the population traveling through other modes is not obviously influenced; different types of people have different income levels, and when a certain urban spatial area is close to a high-grade commercial area, the dynamic distribution of people with high-grade commodity consumption capacity is greatly influenced, and the dynamic distribution of other types of people is not obviously influenced. Therefore, n population types are preset, n influence degree models are established for the n types of population, each influence degree model corresponds to one population type, and accurate prediction on the dynamic distribution of the personnel of the specific population types is facilitated.
In addition, the influence degree model adopts a BP multilayer neural network model, the BP neural network aiming at each population type consists of two processes of information forward propagation and error backward propagation, each neuron of an input layer is responsible for receiving multi-dimensional characteristic quantities of space environment and space relation of an input real city space region from the outside and transmitting the input information to each neuron of a middle layer, the middle layer is an internal information processing layer and is responsible for information transformation according to the requirement of information change capability, and the middle layer can be designed into a single hidden layer or a multi-hidden layer structure; and the information transmitted to each neuron of the output layer by the last hidden layer is further processed to complete a forward propagation processing process of learning, and the influence degree of dynamic distribution aiming at each population type is output to the outside by the output layer. And entering a back propagation stage of the error when the influence degree of the output dynamic distribution for each population type is not consistent with the dynamic distribution data of the corresponding population type in the sample in the urban space area. And the error passes through the output layer, the weight of each layer is corrected in a mode of error gradient reduction, and the error is reversely transmitted to the hidden layer and the input layer by layer. The repeated information forward propagation and error backward propagation processes are processes of continuously adjusting weights of all layers and neural network learning training, and the processes are carried out until the errors output by the network are reduced to an acceptable degree. Therefore, the influence degree model adopts a BP multi-layer neural network model, and the influence degrees of the space environment and the space relation of the urban space area on different population types can be accurately output after training.
Preferably, the urban spatial area sample collected in S3 includes the spatial environment characteristic quantity, the spatial relationship characteristic quantity, and the dynamic distribution data of various types of population in the urban spatial area; the characteristic quantity of the urban area space environment comprises but is not limited to the natural area, the greening area, the total road mileage, the number of public transportation stations, the number of business circles, the number of public service institutions such as schools or medical treatment and the like of the urban area space; the characteristic quantity of the spatial relationship of the urban area reflects the radiation influence degree of main spatial targets existing in a certain distance range around the urban area, including but not limited to important public transportation and subway hubs, large commercial centers and industrial parks, on the urban spatial area, and can be expressed as follows:
α1(d1)*β1，α2(d2)*β2，...，αi(di)*βi
wherein, "1, 2, …, i" represents the 1 st, 2 nd, … th to the i th type of primary space target existing within a certain range around the urban space region;
βithe value of the degree of influence that the ith type of primary spatial target has itself is expressed, for example: the ith main space target is a bus and subway junction, the more bus stops of the bus and subway junction, the corresponding betaiThe larger the value is, the larger the radiation influence degree of the main space target on the urban space area is; the ith main space target is a shopping center, the larger the scale of building the shopping center is, the corresponding betaiThe larger the value is, the larger the radiation influence degree of the main space target on the urban space area is;
αi(di) Representing the influence radiation coefficient of the ith main space target on the urban space region, the coefficient being related to diOf (d) an inverse proportional function of (b), wherein diRepresents the distance between the primary space target and the city space region, i.e. the distance d between the ith primary space target and the city space regioniThe larger the radiation impact of the ith primary spatial target on the urban spatial region.
Preferably, the specific steps of S4 are as follows:
s41, taking the space environment characteristic quantity and the space relation characteristic quantity contained in the urban space area sample collected in S3 as the multidimensional characteristic quantity of the space environment and the space relation of the urban space area, bringing the multidimensional characteristic quantity into the influence degree model established in S2, and outputting the influence degree of dynamic distribution aiming at each population type by the influence degree model;
s42, judging whether the influence degree of the output dynamic distribution of each population type is matched with the actual dynamic distribution data of the corresponding population type in the sample;
and S43, if the output influence degree of the dynamic distribution of each population type is not matched with the actual dynamic distribution data of the corresponding population type in the sample, adjusting the neuron parameters of the influence degree model according to the judgment result, then inputting the multidimensional characteristic quantity of the spatial environment and the spatial relationship of the urban spatial region contained in the sample again, outputting the influence degree of the dynamic distribution of each population type by the influence degree model and comparing the influence degree with the actual dynamic distribution data corresponding to the sample until the two are matched, and finishing the training of the influence degree model.
For sample collection, taking the example of collecting real samples of 1000 population of 100 cities as an example, 1000 population needs to cover population types of different travel modes, professions, ages, income levels, physical conditions and the like, the samples correspond to n models of different population types respectively, and a large number of real samples are utilized to train a neural network model, so that the matching degree of the spatial environment and spatial relationship of the urban spatial region represented by the model to the dynamic distribution of each population type and the real samples is maximized. Preferably, the spatial environment and spatial relationship parameters of any real urban spatial region are substituted into the trained influence degree model in S4, the influence degree for each type of population is obtained, and the dynamic distribution of each type of population in the urban spatial region in the future is analyzed and predicted.
Based on the method, the invention designs the following system:
a population dynamic measuring and calculating system based on an urban superconcephalon computing platform comprises: the system comprises a population type preset model, an influence degree model establishing module, a sample collecting module, a training module and a population dynamic distribution acquiring module; wherein the content of the first and second substances,
the population type preset model is used for presetting population types;
the influence degree model establishing module is used for establishing an influence degree model of the space environment and the space relation of the urban space region relative to the dynamic distribution of different population types;
the sample acquisition module is used for acquiring samples of a certain number of real urban space areas;
the training module is used for respectively training the influence degree model established by the influence degree model establishing module by using the samples acquired by the sample acquiring module;
the population dynamic distribution acquisition module is used for substituting the space environment and space relation parameters of any real urban space region into the influence degree model trained by the training module to acquire various types of population dynamic distributions in the urban space region.
Preferably, the population type presetting model presets the population types in the 1 st population, the 2 nd population … n in total according to the internal cause difference of individuals, such as income level, occupation, age, travel mode, occupation and other factors.
Preferably, the influence degree model of the dynamic distribution in the influence degree model building module adopts a BP multilayer neural network model suitable for various urban spatial regions, and for different population types, the influence degree model building module respectively builds influence degree models of the environmental relationship and the spatial relationship of the urban spatial regions relative to n dynamic distributions of the 1 st population type and the 2 nd population type … n, and each influence degree model corresponds to one population type.
Preferably, the urban space area samples collected by the sample collection module include spatial environment coefficients and spatial relationship coefficients of the urban space area, and dynamic distribution data of various types of population in the urban space area.
Preferably, the training module comprises: the device comprises an initial unit, a judgment unit and a parameter correction unit; wherein the content of the first and second substances,
the initial unit is used for substituting the space environment characteristic quantity and the space relation characteristic quantity contained in the urban space area sample collected by the sample collection module into the influence degree model established by the influence degree model establishing module as the multidimensional characteristic quantity of the space environment and the space relation of the urban space area, and outputting the influence degree aiming at the dynamic distribution of each population type by the influence degree model; the judging unit is used for judging whether the influence degree of the output dynamic distribution of each population type is matched with the actual dynamic distribution data of the corresponding population type in the sample;
and when the output influence degree of the dynamic distribution of each population type is not matched with the actual dynamic distribution data of the corresponding population type in the sample, the parameter correction unit is used for adjusting the neuron parameters of the influence degree model according to the judgment result, then inputting the multidimensional characteristic quantity of the space environment and the space relation of the urban space region contained in the sample again, outputting the influence degree of the dynamic distribution of each population type by the influence degree model and comparing the influence degree with the actual dynamic distribution data corresponding to the sample until the influence degree is matched with the actual dynamic distribution data, and finishing the training of the influence degree model.
The invention has the following beneficial effects:
according to the technical scheme, based on the prior art, the invention provides the population dynamic measuring and calculating method based on the urban superconcephalon computing platform, and the population dynamic measuring and calculating system based on the urban superconcephalon computing platform is designed according to the method.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flow chart of a population dynamic measuring and calculating method based on an urban superconcephalon computing platform;
FIG. 2 is a structural block diagram of a population dynamic measuring and calculating system based on an urban superconcephalon computing platform.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, the present invention provides the following method:
a population dynamic measuring and calculating method based on an urban superconcephalon computing platform comprises the following steps:
s1, presetting population types;
in order to further optimize the technical characteristics, n population types of 1 st, 2 nd, … nd are preset in S1 according to the internal cause difference of the individual, such as income level, age, travel mode, occupation and the like. The population type division is a preorder step for realizing sample collection, establishment and training of an influence degree model, and on the basis of population type division, the steps of sample collection classification and establishment and training of an influence degree model can be respectively carried out according to the population types in subsequent steps.
S2, establishing an influence degree model of the space environment and the space relation of the urban space area relative to the dynamic distribution of different population types;
in order to further optimize the above technical features, the influence degree model of dynamic distribution in S2 adopts a BP multilayer neural network model suitable for various urban spatial regions, that is, the influence degree model established in step S2 does not define a model for the actual situation of the spatial environment and spatial relationship of a specific spatial region, but reflects the influence of the spatial environment and spatial relationship on the general regularity of dynamic distribution of human mouth for a universal model established for any spatial region; after the spatial environment and the spatial relationship of a specific spatial region are substituted into the model, the influence degree of the spatial region on the dynamic distribution of the population facing each population type can be output; of course, the model is also an objective model which represents the spatial environment of the urban spatial region and the corresponding statistical probability relationship between the spatial relationship and the population dynamic distribution on the basis of big data.
And aiming at different population types, respectively establishing n dynamically distributed influence degree models of the environmental relationship and the spatial relationship of the urban spatial region relative to the 1 st population type and the 2 nd population type … nth population type, wherein each influence degree model corresponds to one population type. The degree of influence of the dynamic distribution generated for each type of population is quite different because of the spatial environment of the urban spatial region itself and the spatial relationship between the urban spatial region and the primary spatial targets within a certain distance of the perimeter. For example, the following steps are carried out: different types of population travel modes are divided into bus subway travel, driving travel, bicycle travel, walking travel and the like, when one subway or bus station is arranged in one urban space area, the population whose travel mode is bus or subway travel can be greatly influenced in dynamic distribution, and the population traveling through other modes is not obviously influenced; different types of people have different income levels, and when a certain urban spatial area is close to a high-grade commercial area, the dynamic distribution of people with high-grade commodity consumption capacity is greatly influenced, and the dynamic distribution of other types of people is not obviously influenced. Therefore, n population types are preset, n influence degree models are established for the n types of population, each influence degree model corresponds to one population type, and accurate prediction on the dynamic distribution of the personnel of the specific population types is facilitated.
The BP multi-layer neural network model is a network formed by connecting a plurality of neurons together according to a certain rule, and a neural network comprises an input layer, a hidden layer (middle layer) and an output layer. The number of neurons in the input layer is the same as the dimension of input data, the number of neurons in the output layer is the same as the number of data to be fitted, and the number of neurons and the number of layers in the hidden layer need to be set according to some rules and targets expected to be reached by the model.
In the invention, aiming at the influence degree model of each population type, the spatial environment characteristic of any urban spatial region is described as a spatial environment characteristic quantity E ═<e1，e2，...ek>The space environment characteristic quantity E is a multi-dimensional (k dimensions) characteristic vector, wherein a value of each dimension represents a characteristic quantity obtained by evaluating the urban space region in a quantitative manner on one aspect factor of the space environment; for example, the characteristic quantity of the spatial environment of the urban spatial region includes, but is not limited to, the natural area, the green area, the total road mileage, the number of public transportation sites, the number of business circles, the number of public service institutions such as schools or medical treatment, etc. of the urban spatial region, and the numerical value of the quantitative evaluation of the natural area is represented as e1The numerical value for quantitative evaluation of greening area is represented as e2And so on, finally forming the spatial environment characteristic quantity E of the urban spatial region<e1，e2，...ek>。
The invention describes the radiation influence degree of any main space target existing in a certain distance range around the urban area, including but not limited to important public transportation and subway hubs, large commercial centers, industrial parks and the like, on the urban area as a spatial relation characteristic quantity S<s1，s2，...si>The spatial relation characteristic quantity S is a multi-dimensional (i-dimensional) characteristic vector, wherein the value of each dimension represents the characteristic quantity obtained by quantitatively evaluating the radiation influence degree of the urban spatial area by each type of main spatial target existing in a certain distance range (for example, in a range of 5KM around the urban area) around the urban area; for example, the radiation influence degree of any public transportation and subway hub, large commercial center and industrial park in the 5KM range around the urban spatial region on the urban spatial region is considered, and the numerical value of quantitative evaluation of the radiation influence degree of the industrial park on the urban spatial region is represented as s1The numerical value of quantitative evaluation of the radiation influence degree of the large-scale business center on the urban spatial region is expressed as s2Etc. ofAnd finally forming a spatial relation characteristic quantity S of the urban spatial region<s1，s2，...si>. More specifically, s can beiIs denoted by si＝αi(di)*βi，βiThe value of the degree of influence that the ith type of primary spatial target has itself is expressed, for example: the ith main space target is a bus and subway junction, the more bus stops of the bus and subway junction, the corresponding betaiThe larger the value is, the larger the radiation influence degree of the main space target on the urban space area is; the ith main space target is a shopping center, the larger the scale of building the shopping center is, the corresponding betaiThe larger the value is, the larger the radiation influence degree of the main space target on the urban space area is; alpha is alphai(di) Representing the influence radiation coefficient of the ith main space target on the urban space region, the coefficient being related to diOf (d) an inverse proportional function of (b), wherein diRepresents the distance between the primary space target and the city space region, i.e. the distance d between the ith primary space target and the city space regioniThe larger the radiation impact of the ith primary spatial target on the urban spatial region. Of course, if there are a plurality of primary spatial objects of the same type within the 5KM range, for example, there are a plurality of shopping centers, then α for each spatial object can be calculated separatelyi(di)*βiThen, the average value is calculated as the value s of the corresponding dimension of the type space targeti。
And then, for the defined influence model based on the BP multilayer neural network architecture, weighting and combining the spatial environment characteristic quantity E and the spatial relation characteristic quantity S of any urban spatial region, taking the obtained multidimensional characteristic vector as the input of the influence model, and inputting the input into an input layer of the BP multilayer neural network. In particular, the merged input feature vector may be represented as
Namely multiplying each dimension of the space environment characteristic quantity E and the space relation characteristic quantity SAnd combining the input feature vectors into a multi-dimensional input feature vector after the corresponding weight values are obtained. The input layer of the BP neural network model is provided with N input neurons, the hidden layer is provided with K hidden layer neurons, the output layer is provided with M output neurons, and M-dimensional feature vectors are output, so that the N-dimensional input feature vectors X are converted into the influence degree model based on the BP multi-layer neural network architectureNConverting into an M-dimensional feature vector representing the degree of influence on the type population dynamics, denoted XM。
And the influence degree model connects the BP multilayer neural network in series with a softmax multi-class classifier, and an M-dimension characteristic vector X representing the influence degree on the dynamic distribution of the population of the classMInputting a softmax multi-class classifier, and calculating the distribution classes of the population of the type in the urban space area and the probability of each distribution class, wherein the softmax multi-class classifier defines three distribution classes of moving-in, moving-out and keeping, so that the three distribution classes can be obtained according to the M-dimension feature vector XMAnd determining the probability of the population of the type moving in and out of the urban spatial region and keeping the three distribution states due to the influence of the spatial environment characteristic quantity E and the spatial relationship characteristic quantity S. Wherein, the softmax multi-class classifier is based on the M-dimension feature vector XMThe resulting probability is:
wherein P isjRepresenting the probability of this type of population falling into the jth distribution category, wjAnd wiWeight matrices representing the j-th and i-th distribution categories, respectively, bjAnd biBias terms representing the j-th and i-th distribution categories, respectively.
S3, collecting a certain number of samples of real city space areas;
in order to optimize the technical characteristics, for sample collection, taking a real sample of 100000 population in 100 urban spatial areas as an example, 100000 individual population types need to be covered by the sample, which corresponds to n models of different population types, respectively, and a large number of real samples are used to train a neural network model, which is beneficial to maximizing the degree of influence of the spatial environment and spatial relationship of the urban spatial areas represented by the model on the dynamic distribution of each population type and the matching degree of the real sample.
Specifically, the urban spatial area sample collected in S3 includes the spatial environment characteristic quantity, the spatial relationship characteristic quantity, and the dynamic distribution data of various types of population in the urban spatial area; the characteristic quantity of the urban area space environment comprises but is not limited to the natural area, the greening area, the total road mileage, the number of public transportation stations, the number of business circles, the number of public service institutions such as schools or medical treatment and the like of the urban area space; the characteristic quantity of the spatial relationship of the urban area reflects the radiation influence degree of main spatial targets existing in a certain distance range around the urban area, including but not limited to important public transportation and subway hubs, large commercial centers and industrial parks, on the urban spatial area, and can be expressed as follows:
α1(d1)*β1，α2(d2)*β2，...，αi(di)*βi
wherein the content of the first and second substances,
"1, 2, …, i" represents the 1 st, 2 nd, … th through the i th type of primary space target existing within a certain range around the urban space region;
βithe value of the degree of influence that the ith type of primary spatial target has itself is expressed, for example: the ith main space target is a bus and subway junction, the more bus stops of the bus and subway junction, the corresponding betaiThe larger the value is, the larger the radiation influence degree of the main space target on the urban space area is; the ith main space target is a shopping center, the larger the scale of building the shopping center is, the corresponding betaiThe larger the value is, the larger the radiation influence degree of the main space target on the urban space area is;
αi(di) Representing the influence radiation coefficient of the ith main space target on the urban space region, the coefficient being related to diOf (d) an inverse proportional function of (b), wherein diRepresents the distance between the primary space target and the city space region, i.e. the distance d between the ith primary space target and the city space regioniThe larger the radiation impact of the ith primary spatial target on the urban spatial region.
S4, respectively training the influence degree model established in S2 by using the samples collected in S3;
in order to further optimize the above technical features, the specific steps of S4 are as follows:
s41, taking the space environment characteristic quantity and the space relation characteristic quantity contained in the urban space area sample collected in S3 as the multidimensional characteristic quantity of the space environment and the space relation of the urban space area, substituting the multidimensional characteristic quantity into the influence degree model established in S2, and outputting the influence degree of dynamic distribution aiming at each population type by the influence degree model; as described above, the influence of the dynamic distribution for each population type output by the influence model is represented by the probability of each of the three distribution states of the population of the type moving in, moving out and remaining in the urban spatial region
S42, judging whether the influence degree of the output dynamic distribution of each population type is matched with the actual dynamic distribution data of the corresponding population type in the sample; specifically, counting the probabilities of actual moving in, moving out and remaining in the urban space region of the corresponding population types in the sample, and comparing whether the probabilities are matched with the probabilities output by the influence degree model;
and S43, if the output influence degree of the dynamic distribution of each population type is not matched with the actual dynamic distribution data of the corresponding population type in the sample, adjusting the neuron parameters of the BP neural network of the influence degree model according to the judgment result, then inputting the multidimensional characteristic quantity of the space environment and the space relation of the urban space region contained in the sample again, outputting the influence degree of the dynamic distribution of each population type by the influence degree model and comparing the influence degree with the actual dynamic distribution data corresponding to the sample until the influence degree is matched with the actual dynamic distribution data of the population type, and finishing the training of the influence degree model.
More specifically, the input feature vector is represented as XN＝{xp1...xpNIn which xp1，xp2，......xpNAs the value of the input feature vector in each dimension; substituting the model into the BP neural network model, and sequentially calculating the numerical values of the hidden layer and the output layer as follows:
wherein w1nkIs the weight between the nth neuron of the input layer and the kth neuron of the hidden layer, O1pkIs the output of the k neuron of the hidden layer; w2kmIs the weight between the k neuron of the hidden layer and the m neuron of the output layer, O2pmIs the output, activation function of the mth output layer neuron
i represents the ith round of training; judging whether the deviation between the influence degree of the dynamic distribution of each population type output by the influence degree model after the training of the current round (the ith round) and the corresponding actual dynamic distribution in the sample is less than or equal to a preset allowable deviation epsilon, if so, stopping iteration, and if not, continuing the following process; and performing reverse calculation through a training feedback unit corresponding to the neural network unit:
wherein the learning rate is mu, the learning rate is,
-δpm(i)＝(tpm-O2pm(i))O2pm(i)(1-O2pm(i))，
the weight is changed as follows:
w1nk(i+1)＝w1nk(i)+Δw1nk(i+1)
w2km(i+1)＝w2km(i)+Δw2km(i+1)
and repeatedly learning, and continuously adjusting the weight values among the neurons until the deviation reaches less than or equal to the preset allowable deviation epsilon, and finishing the training of the BP neural network model.
And S5, substituting the space environment and space relation parameters of any real urban space region into the trained influence degree model in the S4, and acquiring various types of population dynamic distribution in the urban space region.
In order to further optimize the technical characteristics, the spatial environment and spatial relationship parameters of any real urban spatial region are substituted into the influence degree model trained in the step S4, the influence degree for each type of population is obtained, and the dynamic distribution of each type of population in the urban spatial region in the future is analyzed and predicted.
As shown in fig. 2, based on the above method, the present invention designs the following system:
a population dynamic measuring and calculating system based on an urban superconcephalon computing platform comprises: the system comprises a population type preset model, an influence degree model establishing module, a sample collecting module, a training module and a population dynamic distribution acquiring module; wherein the content of the first and second substances,
the population type presetting model 1 is used for presetting population types;
the influence degree model establishing module 2 is used for establishing an influence degree model of the space environment and the space relation of the urban space region relative to the dynamic distribution of different population types;
the sample acquisition module 3 is used for acquiring samples of a certain number of real urban space areas;
the training module 4 is used for training the influence degree model established by the influence degree model establishing module 2 by using the samples acquired by the sample acquiring module 3;
the population dynamic distribution acquisition module is used for substituting the space environment and the space relation parameters of any real urban space region into the influence degree model trained by the training module to acquire various types of population dynamic distributions in the urban space region.
In order to further optimize the technical characteristics, the population type preset model presets the population types in the No. 1 and No. 2 … n according to the internal cause difference of individuals, such as income level, occupation, age, travel mode, occupation and the like.
In order to further optimize the technical characteristics, the influence degree model of the dynamic distribution in the influence degree model establishing module 2 adopts a BP multilayer neural network model suitable for various urban spatial regions, and for different population types, the influence degree model establishing module 2 respectively establishes influence degree models of the environmental relationship and the spatial relationship of the urban spatial regions relative to n dynamic distributions of the 1 st population type and the 2 nd population type … nth population type, and each influence degree model corresponds to one population type.
In order to further optimize the above technical solution, the urban spatial area samples collected by the sample collection module 3 include spatial environment coefficients and spatial relationship coefficients of the urban spatial area, and dynamic distribution data of various types of population in the urban spatial area.
In order to further optimize the above technical solution, the training module 4 includes: the device comprises an initial unit, a judgment unit and a parameter correction unit; wherein the content of the first and second substances,
the initial unit is used for substituting the space environment characteristic quantity and the space relation characteristic quantity contained in the urban space area sample collected by the sample collection module into the influence degree model established by the influence degree model establishing module as the multidimensional characteristic quantity of the space environment and the space relation of the urban space area, and outputting the influence degree aiming at the dynamic distribution of each population type by the influence degree model;
the judging unit is used for judging whether the influence degree of the output dynamic distribution of each population type is matched with the actual dynamic distribution data of the corresponding population type in the sample;
and when the output influence degree of the dynamic distribution of each population type is not matched with the actual dynamic distribution data of the corresponding population type in the sample, the parameter correction unit is used for adjusting the neuron parameters of the influence degree model according to the judgment result, then inputting the multidimensional characteristic quantity of the space environment and the space relation of the urban space region contained in the sample again, outputting the influence degree of the dynamic distribution of each population type by the influence degree model and comparing the influence degree with the actual dynamic distribution data corresponding to the sample until the influence degree is matched with the actual dynamic distribution data in the sample, and finishing the training of the influence degree model.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. A population dynamic measurement and calculation method based on an urban superconcephalon computing platform is characterized by comprising the following steps:
s1, presetting population types;
s2, establishing the influence of the spatial environment and spatial relationship of the urban spatial region relative to the dynamic distribution of different population typesA degree model; the influence degree model of the dynamic distribution adopts a BP multilayer neural network model which is suitable for various urban space regions and reflects the influence of space environment and space relation on the general regularity of the dynamic distribution of the human mouth; after substituting the space environment and the space relation of a specific space region into the model, outputting the influence degree of the space region on the dynamic distribution of population facing each population type; and, for each population type's influence model, the spatial environment feature of any urban spatial region is described as a spatial environment feature quantity E ═<e1，e2，...ek>The spatial environment characteristic quantity E is a multi-dimensional characteristic vector, wherein the value of each dimension represents the characteristic quantity obtained by quantitatively evaluating one aspect factor of the spatial environment of the urban spatial region, and the characteristic quantity obtained by quantitatively evaluating one aspect factor of the spatial environment of the urban spatial region comprises the natural area, the greening area, the total road mileage, the number of public transport stations, the number of business circles and the number of public service institutions of the urban spatial region; the radiation influence degree of a space target existing in a certain distance range around any urban area on the urban space area is described as a space relation characteristic quantity S ═<s1，s2，...si>The space relation characteristic quantity S is a multi-dimensional characteristic vector, wherein the value of each dimension represents the characteristic quantity obtained by quantitatively evaluating the radiation influence degree of each type of space target existing in a certain distance range around the urban area on the urban space area, and S isi＝αi(di)*βi，βiIndicating the value of the degree of influence, α, that the ith type of spatial target has on its owni(di) Representing the influence radiation coefficient of the ith space target on the urban space region, the coefficient being related to diOf (d) an inverse proportional function of (b), wherein diRepresents the distance between the space target and the city space region, i.e. the distance d between the ith space target and the city space regioniThe larger the target is, the smaller the radiation influence degree of the ith space target on the urban space area is; for defined BP-based messagesThe influence degree model of the layer neural network architecture is characterized in that space environment characteristic quantity E and space relation characteristic quantity S of any urban space area are weighted and combined, and the obtained multidimensional characteristic vector is used as input X of the influence degree modelN(ii) a Influence degree model based on BP multi-layer neural network architecture is used for converting N-dimensional input feature vector XNConverting into an M-dimensional feature vector representing the degree of influence on the dynamic distribution of population types, denoted XM(ii) a The influence degree model is characterized in that the BP multi-layer neural network is connected with a softmax multi-class classifier in series, and M-dimension characteristic vector X representing influence degree on human mouth type dynamic distributionMInputting the softmax multi-class classifier, and calculating the distribution classes of the population types in the urban space area and the probability of each distribution class, wherein the softmax multi-class classifier defines three distribution classes of moving in, moving out and keeping, so that the three distribution classes are obtained according to the M-dimensional feature vector XMDetermining the probability of the population type moving in and out of the urban spatial region and keeping the three distribution categories due to the influence of the spatial environment characteristic quantity E and the spatial relationship characteristic quantity S;
s3, collecting a certain number of samples of real city space areas; the urban space area sample comprises the space environment characteristic quantity and the space relation characteristic quantity of the urban space area and dynamic distribution data of various types of population in the urban space area;
s4, respectively training the influence degree model established in S2 by using the samples collected in S3;
and S5, substituting the space environment and space relation parameters of any real urban space region into the trained influence degree model in S4, acquiring the influence degree aiming at each population type, and analyzing and predicting the dynamic distribution of various types of population in the urban space region in the future.
2. The method for dynamically calculating population on the basis of the urban superconcephalon computing platform according to the claim 1, wherein n population types, namely the 1 st population type and the 2 nd population type … n population type, are preset according to the internal cause difference of an individual in S1.
3. The method for dynamically measuring and calculating the population based on the urban superconcephalon computing platform according to the claim 1, which is characterized in that influence degree models of the environmental relationship and the spatial relationship of urban spatial regions relative to n dynamic distributions of the 1 st population type and the 2 nd population type … nth population type are respectively established for different population types, and each influence degree model corresponds to one population type.
4. The method for measuring and calculating the population dynamics based on the urban superconcephalon computing platform according to claim 1, wherein the specific steps of S4 are as follows:
s41, taking the space environment characteristic quantity and the space relation characteristic quantity contained in the urban space area sample collected in S3 as the multidimensional characteristic quantity of the space environment and the space relation of the urban space area, bringing the multidimensional characteristic quantity into the influence degree model established in S2, and outputting the influence degree of dynamic distribution aiming at each population type by the influence degree model;
s42, judging whether the influence degree of the output dynamic distribution of each population type is matched with the actual dynamic distribution data of the corresponding population type in the sample;
and S43, if the output influence degree of the dynamic distribution of each population type is not matched with the actual dynamic distribution data of the corresponding population type in the sample, adjusting the neuron parameters of the influence degree model according to the judgment result, then inputting the multidimensional characteristic quantity of the spatial environment and the spatial relationship of the urban spatial region contained in the sample again, outputting the influence degree of the dynamic distribution of each population type by the influence degree model and comparing the influence degree with the actual dynamic distribution data corresponding to the sample until the two are matched, and finishing the training of the influence degree model.
5. A dynamic population measuring and calculating system based on an urban superconcephalon computing platform is characterized by comprising: the system comprises a population type preset model (1), an influence degree model establishing module (2), a sample collecting module (3), a training module (4) and a population dynamic distribution obtaining module (5); wherein the content of the first and second substances,
the population type preset model (1) is used for presetting population types;
the influence degree model establishing module (2) is used for establishing an influence degree model of the space environment and the space relation of the urban space region relative to the dynamic distribution of different population types; the influence degree model of the dynamic distribution adopts a BP multilayer neural network model which is suitable for various urban space regions and reflects the influence of space environment and space relation on the general regularity of the dynamic distribution of the human mouth; after substituting the space environment and the space relation of a specific space region into the model, outputting the influence degree of the space region on the dynamic distribution of population facing each population type; and, for each population type's influence model, the spatial environment feature of any urban spatial region is described as a spatial environment feature quantity E ═<e1，e2，...ek>The spatial environment characteristic quantity E is a multi-dimensional characteristic vector, wherein the sampling value of each dimension represents the characteristic quantity obtained by quantitatively evaluating one aspect factor of the spatial environment of the urban spatial region, and the characteristic quantity obtained by quantitatively evaluating one aspect factor of the spatial environment of the urban spatial region comprises the natural area, the greening area, the total road mileage, the number of public transportation stations, the number of business circles and the number of public service institutions of the urban spatial region; the radiation influence degree of a space target existing in a certain distance range around any urban area on the urban space area is described as a space relation characteristic quantity S ═<s1，s2，...si>The space relation characteristic quantity S is a multi-dimensional characteristic vector, wherein the value of each dimension represents the characteristic quantity obtained by quantitatively evaluating the radiation influence degree of each type of space target existing in a certain distance range around the urban area on the urban space area, and S isi＝αi(di)*βi，βiIndicating the value of the degree of influence, α, that the ith type of spatial target has on its owni(di) Representing the influence radiation coefficient of the ith space target on the urban space region, the coefficient being related to diOf (d) an inverse proportional function of (b), wherein diRepresenting spatial objects and said city skyDistance between regions, i.e. distance d between ith space object and the city space regioniThe larger the target is, the smaller the radiation influence degree of the ith space target on the urban space area is; for the defined influence model based on the BP multilayer neural network architecture, weighting and combining the space environment characteristic quantity E and the space relation characteristic quantity S of any city space region, and taking the obtained multidimensional characteristic vector as the input X of the influence modelN(ii) a Influence degree model based on BP multi-layer neural network architecture is used for converting N-dimensional input feature vector XNConverting into an M-dimensional feature vector representing the degree of influence on the dynamic distribution of population types, denoted XM(ii) a The influence degree model is characterized in that the BP multi-layer neural network is connected with a softmax multi-class classifier in series, and M-dimension characteristic vector X representing influence degree on human mouth type dynamic distributionMInputting the softmax multi-class classifier, and calculating the distribution classes of the population types in the urban space area and the probability of each distribution class, wherein the softmax multi-class classifier defines three distribution classes of moving in, moving out and keeping, so that the three distribution classes are obtained according to the M-dimensional feature vector XMDetermining the probability of the population type moving in and out of the urban spatial region and keeping the three distribution categories due to the influence of the spatial environment characteristic quantity E and the spatial relationship characteristic quantity S;
the sample acquisition module (3) is used for acquiring samples of a certain number of real urban space areas; the urban space area sample comprises the space environment characteristic quantity and the space relation characteristic quantity of the urban space area and dynamic distribution data of various types of population in the urban space area;
the training module (4) is used for respectively training the influence degree model established by the influence degree model establishing module (2) by using the samples acquired by the sample acquiring module (3);
the population dynamic distribution acquisition module (5) is used for substituting the space environment and space relation parameters of any real urban space region into the influence degree model trained by the training module (4), acquiring the influence degree aiming at each population type, and analyzing and predicting various types of population dynamic distribution in the urban space region in the future.
6. The system for dynamically measuring and calculating the population based on the urban superconcephalon computing platform is characterized in that the population type presetting model (1) presets … th populations of 1 st population type and 2 nd population types of n population types according to the factors such as income level, occupation, age, travel mode, occupation and the like of individuals.
7. The system for dynamically measuring and calculating the population based on the urban superconcephalon computing platform according to the claim 5, wherein the influence degree model establishing module (2) respectively establishes influence degree models of the environmental relationship and the spatial relationship of the urban spatial region relative to n dynamic distributions of the 1 st population type and the 2 nd population type … nth population type aiming at different population types, and each influence degree model corresponds to one population type.
8. The system for demographic measurement and calculation on an urban superconcephalon computing platform according to claim 5, characterized in that the training module (4) comprises: the device comprises an initial unit, a judgment unit and a parameter correction unit; wherein the content of the first and second substances,
the initial unit is used for substituting the space environment characteristic quantity and the space relation characteristic quantity contained in the urban space region sample collected by the sample collection module (3) into the influence degree model built by the influence degree model building module (2) as the multidimensional characteristic quantity of the space environment and the space relation of the urban space region, and outputting the influence degree of dynamic distribution aiming at each population type by the influence degree model; the judging unit is used for judging whether the influence degree of the output dynamic distribution of each population type is matched with the actual dynamic distribution data of the corresponding population type in the sample;
and when the output influence degree of the dynamic distribution of each population type is not matched with the actual dynamic distribution data of the corresponding population type in the sample, the parameter correction unit is used for adjusting the neuron parameters of the influence degree model according to the judgment result, then inputting the multidimensional characteristic quantity of the space environment and the space relation of the urban space region contained in the sample again, outputting the influence degree of the dynamic distribution of each population type by the influence degree model and comparing the influence degree with the actual dynamic distribution data corresponding to the sample until the influence degree is matched with the actual dynamic distribution data, and finishing the training of the influence degree model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910760691.8A CN110704789B (en) | 2019-08-16 | 2019-08-16 | Population dynamic measurement and calculation method and system based on 'urban superconcephalon' computing platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910760691.8A CN110704789B (en) | 2019-08-16 | 2019-08-16 | Population dynamic measurement and calculation method and system based on 'urban superconcephalon' computing platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110704789A CN110704789A (en) | 2020-01-17 |
| CN110704789B true CN110704789B (en) | 2021-04-13 |
Family
ID=69193589
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910760691.8A Active CN110704789B (en) | 2019-08-16 | 2019-08-16 | Population dynamic measurement and calculation method and system based on 'urban superconcephalon' computing platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110704789B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113642807B (en) * | 2021-09-01 | 2022-04-12 | 智慧足迹数据科技有限公司 | Population mobility prediction method and related device |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101089884A (en) * | 2007-07-12 | 2007-12-19 | 中国科学院地理科学与资源研究所 | Human body data space dynamic modeling method |
| CN101694706A (en) * | 2009-09-28 | 2010-04-14 | 深圳先进技术研究院 | Modeling method of characteristics of population space-time dynamic moving based on multisource data fusion |
| CN103218517A (en) * | 2013-03-22 | 2013-07-24 | 南京信息工程大学 | GIS (Geographic Information System)-based region-meshed spatial population density computing method |
| KR101488069B1 (en) * | 2012-11-28 | 2015-01-29 | 한국수력원자력 주식회사 | System and method for analyzing population data in order to for off-site consequence analysis |
| WO2016142681A1 (en) * | 2015-03-06 | 2016-09-15 | Micromass Uk Limited | Spectrometric analysis of microbes |
| CN106708962A (en) * | 2016-11-30 | 2017-05-24 | 中山大学 | Urban population distribution method based on building properties |
| CN109376969A (en) * | 2018-12-14 | 2019-02-22 | 中南大学 | City fining population distribution dynamic prediction method and device based on deep learning |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105678407A (en) * | 2015-12-31 | 2016-06-15 | 国网上海市电力公司 | Daily electricity consumption prediction method based on artificial neural network |
| CN107798077A (en) * | 2017-10-09 | 2018-03-13 | 中国电子科技集团公司第二十八研究所 | A kind of Population surveillance method and system |
-
2019
- 2019-08-16 CN CN201910760691.8A patent/CN110704789B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101089884A (en) * | 2007-07-12 | 2007-12-19 | 中国科学院地理科学与资源研究所 | Human body data space dynamic modeling method |
| CN101694706A (en) * | 2009-09-28 | 2010-04-14 | 深圳先进技术研究院 | Modeling method of characteristics of population space-time dynamic moving based on multisource data fusion |
| KR101488069B1 (en) * | 2012-11-28 | 2015-01-29 | 한국수력원자력 주식회사 | System and method for analyzing population data in order to for off-site consequence analysis |
| CN103218517A (en) * | 2013-03-22 | 2013-07-24 | 南京信息工程大学 | GIS (Geographic Information System)-based region-meshed spatial population density computing method |
| WO2016142681A1 (en) * | 2015-03-06 | 2016-09-15 | Micromass Uk Limited | Spectrometric analysis of microbes |
| CN106708962A (en) * | 2016-11-30 | 2017-05-24 | 中山大学 | Urban population distribution method based on building properties |
| CN109376969A (en) * | 2018-12-14 | 2019-02-22 | 中南大学 | City fining population distribution dynamic prediction method and device based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 丽江古城旅游商业人口和空间分布的关系研究;黄珏等;《中国园林》;20091231;第25卷(第5期);第23-26页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110704789A (en) | 2020-01-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109544932B (en) | Urban road network flow estimation method based on fusion of taxi GPS data and gate data | |
| CN110570651B (en) | Road network traffic situation prediction method and system based on deep learning | |
| Shepelev et al. | A Study of the Travel Time of Intersections by Vehicles using Computer Vision. | |
| CN113362598B (en) | Traffic flow prediction method for expressway service area | |
| CN102081846A (en) | Expressway charge data track matching based traffic state recognition method | |
| CN109784752A (en) | A kind of reservoir area of Three Gorges water eutrophication risk assessment early warning system and its analysis method | |
| CN111709549A (en) | Navigation reminding method for short-time traffic flow prediction based on SVD-PSO-LSTM | |
| CN114187766B (en) | Road service level evaluation method based on saturation rate | |
| Chen et al. | A multiscale-grid-based stacked bidirectional GRU neural network model for predicting traffic speeds of urban expressways | |
| CN113205698A (en) | Navigation reminding method based on IGWO-LSTM short-time traffic flow prediction | |
| CN112884014A (en) | Traffic speed short-time prediction method based on road section topological structure classification | |
| CN110704789B (en) | Population dynamic measurement and calculation method and system based on 'urban superconcephalon' computing platform | |
| CN109190948B (en) | Correlation analysis method for operation of large-scale aviation hub and urban traffic jam | |
| CN113920729A (en) | Method for evaluating perception capability of traffic participants based on roadside perception system | |
| CN109147322B (en) | Multi-source data self-adaptive fusion method in urban traffic big data processing | |
| CN113536665B (en) | Road surface temperature short-term prediction method and system based on characteristic engineering and LSTM | |
| CN109389240A (en) | Trip mode discrimination method based on big data machine learning | |
| CN117131991A (en) | Urban rainfall prediction method and platform based on hybrid neural network | |
| CN110309953B (en) | Urban security monitoring layout system and method adopting target mobility distribution prediction | |
| CN112149922A (en) | Method for predicting severity of accident in exit and entrance area of down-link of highway tunnel | |
| CN112101132A (en) | Traffic condition prediction method based on graph embedding model and metric learning | |
| Rodríguez-Rueda et al. | Origin–Destination matrix estimation and prediction from socioeconomic variables using automatic feature selection procedure-based machine learning model | |
| CN111640294A (en) | Method for predicting passenger flow change of urban bus line under influence of newly-built subway line | |
| CN113688200A (en) | Decision tree-based special population action track collection method and system | |
| CN114444795A (en) | Single-line bus passenger travel data generation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |