Urban crowd distribution analogue simulation method and system

Abstract

The embodiment of the invention provides an urban crowd distribution simulation method and system, and the method is executed by a processor, and comprises the steps: determining a first region and a second region from a to-be-simulated region; obtaining identification crowd distribution and first flow data of the first area; determining a crowd simulation result based on the identified crowd distribution of the first area; and based on the crowd simulation result, determining a planning suggestion, and pushing the crowd simulation result and the planning suggestion to the user terminal.

Description

Technical Field

[0001] The present invention relates to the technical field of crowd simulation, and in particular to a method and system for simulating urban crowd distribution. Background Art

[0002] As the process of urbanization continues to accelerate, the rational management and utilization of urban space has become increasingly important. As an important tool for urban planning, public safety and emergency response, crowd simulation technology is being widely used by major urban management departments. The crowd simulation results that rely on static data and preset model parameters are relatively rigid and often do not conform to reality.

[0003] To solve the above problems, CN107292064B discloses a crowd evacuation simulation method and system based on a multi-bee colony algorithm. Different types of bee colony algorithms can simulate the dynamic behaviors of different types of people at different times and in different environments. At the same time, regular iteration and re-initialization can also capture changes in crowd distribution. This method can simulate crowds based on dynamically changing time to a certain extent, but it still cannot reasonably analyze the reliability of crowd simulation.

[0004] Therefore, a method and system for simulating urban population distribution is provided, which helps to analyze the reliability of simulation while analyzing the changes in population distribution at different times, thereby improving the rationality of urban planning. Summary of the invention

[0005] One of the embodiments of the present specification provides a method for simulating urban crowd distribution, which is executed by a processor, and the method includes: determining a first area and a second area from an area to be simulated, the first area being a sub-area in the area to be simulated whose data completeness meets a preset data condition, and the second area being a sub-area in the area to be simulated whose data completeness does not meet the preset data condition; obtaining the distribution of identified people and first flow data in the first area, the identified population distribution including the number of identified people in the first area within at least one first time period; determining a crowd simulation result based on the identified population distribution in the first area, the crowd simulation result being a visualization result of the simulated crowd distribution in the area to be simulated; determining a planning suggestion based on the crowd simulation result, and pushing the crowd simulation result and the planning suggestion to a user terminal.

[0006] One of the embodiments of the present specification provides a system for simulating urban crowd distribution, the system comprising: a first determination module, configured to determine a first area and a second area from an area to be simulated, the first area being a sub-area in the area to be simulated whose data completeness meets a preset data condition, and the second area being a sub-area in the area to be simulated whose data completeness does not meet the preset data condition; an acquisition module, configured to acquire the distribution of identified people and first flow data in the first area, the distribution of identified people comprising the number of identified people in the first area within at least one first time period; a second determination module, configured to determine a crowd simulation result based on the distribution of identified people in the first area, the crowd simulation result being a visualization result of the simulated crowd distribution in the area to be simulated; a third determination module, configured to determine a planning suggestion based on the crowd simulation result, and push the crowd simulation result and the planning suggestion to a user terminal.

[0007] One of the embodiments of this specification provides a device for simulating the distribution of urban crowds. The device includes at least one storage medium and at least one processor, wherein the at least one storage medium is used to store computer instructions; and the at least one processor is used to execute the computer instructions to implement a method for simulating the distribution of urban crowds.

[0008] One of the embodiments of this specification provides a computer-readable storage medium, wherein the storage medium stores computer instructions, and when the computer instructions are executed by a computer, a method for simulating the distribution of urban crowds is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0010] Figure 1 It is a system module diagram of the urban crowd distribution simulation system shown in some embodiments of this specification;

[0011] Figure 2 is an exemplary flow chart of a method for simulating urban crowd distribution according to some embodiments of this specification;

[0012] Figure 3 is an exemplary schematic diagram of generating crowd simulation results according to some embodiments of this specification;

[0013] Figure 4 is an exemplary schematic diagram of a simulation model according to some embodiments of this specification;

[0014] Figure 5It is an exemplary schematic diagram of determining target simulation results according to some embodiments of this specification. DETAILED DESCRIPTION

[0015] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0016] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0017] As shown in this specification and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "comprise" and "include" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0018] Flowcharts are used in this specification to illustrate the operations performed by the system according to the embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed precisely in order. Instead, the steps may be processed in reverse order or simultaneously. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.

[0019] Crowd simulation technology, as a key tool to support urban planning, public safety strategy formulation and emergency response plan design, needs to reflect the real dynamic crowd distribution. CN107292064B can simulate the dynamic behavior of various types of people at different times and in different environments through different types of bee colony algorithms, but it cannot analyze the reliability of crowd simulation. The simulation of crowd evacuation is also different from the simulation of daily urban crowd distribution.

[0020] In view of this, in some embodiments of the present specification, by obtaining the identified crowd distribution and first flow data of the first area of ​​the area to be simulated, and then determining the crowd simulation result, the changing trend of the crowd size over time and space can be simulated; determining planning suggestions based on the crowd simulation results can provide urban planners with more comprehensive and accurate data support, which helps to improve the rationality of urban planning.

[0021] Figure 1 It is a system module diagram of the urban crowd distribution simulation system shown in some embodiments of this specification.

[0022] like Figure 1 As shown, the urban crowd distribution simulation system 100 may include a first determination module 110 , an acquisition module 120 , a second determination module 130 and a third determination module 140 .

[0023] The first determining module 110 may be configured to determine a first region and a second region from the region to be simulated.

[0024] The acquisition module 120 may be configured to acquire the distribution of identified people in the first area and first flow data.

[0025] The second determination module 130 may be configured to determine a crowd simulation result based on the identified crowd distribution in the first area.

[0026] In some embodiments, the second determination module 130 can be further configured to: construct a first simulation map based on identifying crowd distribution and first flow data; determine the simulated crowd distribution in the first area, the simulated crowd distribution in the second area, and the second flow data through a simulation model based on the first simulation map; determine the simulated crowd distribution in the area to be simulated based on the simulated crowd distribution in the first area, the simulated crowd distribution in the second area, and the second flow data; and generate a crowd simulation result based on the simulated crowd distribution.

[0027] In some embodiments, the second determination module 130 may be further configured to construct a first simulation map based on the distribution of the identified population and the distribution of the confident population in the first area.

[0028] The third determination module 140 may be configured to determine a planning suggestion based on the crowd simulation result, and push the crowd simulation result and the planning suggestion to the user terminal.

[0029] In some embodiments, the urban crowd distribution simulation system 100 may include a storage device, a processor, etc., and the processor may obtain pre-stored data and / or information related to the urban crowd distribution simulation system 100 from the storage device. In some embodiments, the first determination module 110, the acquisition module 120, the second determination module 130, and the third determination module 140 of the urban crowd distribution simulation system 100 may be integrated into the processor. In some embodiments, the urban crowd distribution simulation system 100 may include a network. The processor may obtain data and / or information related to the urban crowd distribution simulation system 100 through the network.

[0030] In some embodiments, the urban crowd distribution simulation system 100 may further include a user terminal. The user terminal may refer to one or more terminal devices or software used by a user. The user may refer to an administrator or operator of the urban crowd distribution simulation system 100. For example, the user terminal may include a mobile phone, a tablet computer, an interactive screen, etc.

[0031] It should be noted that the above description of the urban crowd distribution simulation system and its modules is only for convenience of description and cannot limit this specification to the scope of the embodiments. It is understandable that for those skilled in the art, after understanding the principle of the system, it is possible to arbitrarily combine the modules or form a subsystem to connect with other modules without deviating from this principle.

[0032] Figure 2 It is an exemplary flow chart of the urban population distribution simulation method shown in some embodiments of this specification.

[0033] In some embodiments, the processor can determine a first area and a second area from the area to be simulated; obtain the identified crowd distribution and first flow data of the first area; determine the crowd simulation result based on the identified crowd distribution in the first area; determine planning suggestions based on the crowd simulation result, and push the crowd simulation results and planning suggestions to the user terminal.

[0034] In some embodiments, process 200 may be executed by a processor of a city crowd distribution simulation system. Figure 2 As shown, process 200 includes the following steps:

[0035] Step 210: determine a first area and a second area from the area to be simulated.

[0036] The area to be simulated refers to an area where crowd distribution simulation is required. In some embodiments, the processor may determine the area to be simulated based on user input.

[0037] A sub-region refers to one or more regions divided from the region to be simulated. In some embodiments, the processor may divide the region to be simulated based on the functional structure to determine the sub-region. The functional structure refers to the function of the region to be simulated in urban planning and construction. For example, the functional structure may include roads, lawns, flower beds, etc. The processor may divide the region to be simulated based on the functional structure, and determine a region having the same functional structure and connected as a sub-region. Exemplarily, the processor may determine a road connected in the region to be simulated as a sub-region.

[0038] Data completeness refers to an indicator used to measure whether a sub-area has corresponding data. The data corresponding to the sub-area refers to sensor data captured by sensor devices deployed in the sub-area. Exemplary sensor devices may include cameras, surveillance cameras, etc.

[0039] In some embodiments, the data completeness can be a value of 0 or 1. Wherein, if the data completeness of a sub-region is 1, it means that the sub-region contains corresponding data; if the data completeness of a sub-region is 0, it means that the sub-region does not contain corresponding data, that is, the sub-region has no sensor equipment deployed, or the sensor equipment deployed in the sub-region is damaged and does not capture sensor data. The processor can determine the data completeness of a sub-region that does not have corresponding data as 0; and determine the data completeness of a sub-region that has corresponding data as 1.

[0040] The preset data condition refers to a pre-set judgment condition for determining the first area and the second area. The preset data condition can be preset by the system or manually. An exemplary preset data condition can be: data completeness is 1.

[0041] In some embodiments, the processor may determine the first region and the second region in a variety of ways based on the data completeness of the sub-regions. For example, the processor may determine a region whose data completeness satisfies a preset data condition as the first region, and determine a region whose data completeness does not satisfy the preset data condition as the second region.

[0042] Step 220, obtaining the distribution of identified people in the first area and first flow data.

[0043] The distribution of the identified crowd refers to the number of identified people in the first area. In some embodiments, the distribution of the identified crowd may include the number of identified people in the first area within at least one first time period.

[0044] The first time period refers to a time period in which the quality of the sensor data captured by the sensor device meets a preset quality condition. The preset quality condition can be preset by the system or manually. For example, the preset quality condition can be that the resolution of the sensor data captured by the sensor device is not less than 1080P.

[0045] The first time period can be preset by the user. For example, when the sensor device deployed in the sub-area is a camera without night vision function, the resolution of the sensor data captured by the sensor device during the day is not less than 1080P, and the resolution of the sensor data captured by the sensor device at night is less than 1080P. The user can determine the time period corresponding to the day as at least one first time period. The length of the first time period can also be preset based on the user. For example, the length of the first time period is one hour, and the time period corresponding to the day includes 6 o'clock to 18 o'clock, then 6 o'clock to 7 o'clock can be a first time period.

[0046] The first flow data refers to personnel flow data related to the first area. In some embodiments, the first flow data may include personnel flow direction and personnel flow volume between the first areas and / or between the first area and the second area in at least one first time period. For example, the first flow data may be that 12 people flowed from the first area A to the second area B between 8 o'clock and 9 o'clock.

[0047] In some embodiments, the processor can obtain the distribution of identified people and the first flow data in a variety of ways. For example, the processor can obtain the distribution of identified people in the first area through an image recognition algorithm based on the sensor data captured by the sensor equipment deployed in the first area. For another example, the processor can obtain the flow volume and direction of personnel between the first area and / or between the first area and the second area in the first time period through an image recognition algorithm based on the sensor data captured by the sensor equipment deployed in the first area, combined with the orientation of the first area and the second area, and then determine the first flow data.

[0048] Step 230: Determine a crowd simulation result based on the identified crowd distribution in the first area.

[0049] The crowd simulation result refers to the visualization result of the simulated crowd distribution in the area to be simulated.

[0050] The simulated crowd distribution refers to crowd distribution data predicted by simulation. In some embodiments, the simulated crowd distribution may include the simulated crowd distribution of the first area, the simulated crowd distribution of the second area, and the second flow data. The simulated crowd distribution of the first area may include the number of people in the first area predicted after the first time period, and the simulated crowd distribution of the second area may include the number of people in the second area predicted after the first time period.

[0051] The second flow data refers to personnel flow data related to the first area and the second area. In some embodiments, the second flow data may include personnel flow direction and personnel flow volume between the first areas, or between the first area and the second area, or between the second areas, in at least one first time period. It should be noted that in the first flow data, the personnel flow direction and personnel flow volume between the first areas, or between the first area and the second area in at least one first time period are actually measured through sensor data, etc., while the second flow data is determined based on prediction and inference.

[0052] In some embodiments, the processor may determine the simulated crowd distribution in a variety of ways based on the identified crowd distribution in the first area.

[0053] In some embodiments, the processor can set a corresponding basic number of people for the second area, and determine the basic number of people as the population distribution of the second area before the first time period; based on the first flow data, the identified crowd distribution and the basic number of people corresponding to the second area, the simulated crowd distribution is calculated.

[0054] The basic number of people refers to the predicted distribution of the number of people in the second area before the first time period. In some embodiments, the processor can determine the basic number of people corresponding to the second area through a first preset rule based on the difficulty of data recognition in the first area.

[0055] The difficulty of data recognition refers to the difficulty of identifying the data corresponding to the sub-area. The higher the resolution of the sensor data captured by the sensor device deployed in the sub-area and the shorter the time it takes for the processor to recognize the sensor data captured by the sensor device deployed in the sub-area, the lower the difficulty of data recognition corresponding to the sub-area.

[0056] The first preset rule refers to a pre-set rule for determining the basic number of people corresponding to the second area. An exemplary first preset rule may be: the greater the difficulty of data recognition of the first area adjacent to the second area, the smaller the basic number of people in the corresponding second area; conversely, the smaller the difficulty of data recognition of the first area adjacent to the second area, the closer the basic number of people in the second area is to the mean of the distribution of recognized people in all the first areas in the area to be simulated.

[0057] The processor can calculate the simulated crowd distribution in the first area and the simulated crowd distribution in the second area based on the first flow data, the identified crowd distribution and the basic number of people corresponding to the second area; and calculate the second flow data in combination with the simulated crowd distribution in the first area and the simulated crowd distribution in the second area.

[0058] For example, the first time period is from 8:00 to 9:00, the identified population distribution in the first area A at 8:00 is 20 people, the first flow data is that 12 people flow from the first area A to the second area B in the first time period, and the basic number of people corresponding to the second area B is 10. Then the simulated population distribution in the first area A is 8 people, and the simulated population distribution in the second area B is 22 people.

[0059] For another example, the first time period is from 8:00 to 9:00, the identified population distribution in the first area A is 12 people at 8:00, the basic number of people corresponding to the second area B is 10, the simulated population distribution in the first area A is 10 people, the simulated population distribution in the second area B is 12 people, the simulated population distribution in other first areas is the same as the identified population distribution, and the simulated population distribution in other second areas is the same as the basic number of people. Then the second flow data is: 2 people flow from the first area A to the second area B from 8:00 to 9:00.

[0060] In some embodiments, the processor may determine the simulated population distribution based on the simulation model. For more information, see Figure 3, Figure 4 Related description.

[0061] In some embodiments, the processor may visualize the simulated crowd distribution through simulation visualization techniques such as crowd simulation models to determine crowd simulation results.

[0062] Step 240 , determining planning suggestions based on the crowd simulation results, and pushing the crowd simulation results and planning suggestions to the user terminal.

[0063] Planning suggestions refer to suggestions related to urban planning. For example, planning suggestions may include areas in the city where additional public service projects are needed.

[0064] In some embodiments, the processor may determine planning suggestions based on crowd simulation results. For example, based on the crowd simulation results, the processor may determine areas where the simulated crowd distribution is greater than a threshold number of people or areas where the flow of people output or input in the first flow data is greater than a flow threshold as areas where additional public service projects are required in the planning suggestions. The number threshold and flow threshold may be preset by the system or manually.

[0065] In some embodiments, the processor may push crowd simulation results and planning suggestions to the user terminal based on the network.

[0066] In some embodiments of the present specification, crowd simulation results are determined based on identifying crowd distribution, and then planning recommendations are determined. Incomplete data can be used to predict the complete crowd distribution and determine crowd simulation results, thereby reducing the data acquisition cost in urban planning and construction. The crowd simulation results can evaluate the changes in pedestrian flow in different areas and the aggregation and transfer of crowds, so as to serve as reference data for road, traffic, and municipal facility planning, and can quantify the crowd situation in the area, which is beneficial to urban planning and design and improves the rationality of urban planning.

[0067] In some embodiments, the processor may perform at least one round of simulation model prediction before determining the planning suggestion to obtain at least one candidate simulation result; and determine the target simulation result based on the at least one candidate simulation result.

[0068] The candidate simulation result refers to a crowd simulation result that is a candidate for the target simulation result. In some embodiments, the processor may determine the output of the simulation model as the candidate simulation result based on the simulation model. For relevant descriptions of the simulation model, see Figure 3 and Figure 4 Related description.

[0069] The simulation model prediction refers to predicting crowd simulation results in different first time periods based on the simulation model. The processor can input a first simulation map constructed based on the distribution of identified crowds and first flow data in different first time periods into the simulation model to obtain different candidate simulation results.

[0070] In some embodiments, the number of rounds to perform simulation model predictions may be determined based on environmental data over a preset time period.

[0071] The preset time period refers to a pre-set time period. The preset time period can be preset by the system or manually. For example, the preset time period can be the past 12 hours, the past day, etc.

[0072] Environmental data refers to data related to the environment of the area to be simulated, such as temperature, humidity, etc. The processor can obtain environmental data based on temperature sensors, humidity sensors, etc. deployed in the area to be simulated, or access a meteorological website through the network to obtain environmental data for a preset time period.

[0073] In some embodiments, the processor may determine the number of rounds to perform simulation model prediction based on environmental data in a preset time period in a variety of ways.

[0074] In some embodiments, the processor may determine the environmental change in the preset time period based on the environmental data in the preset time period, and determine the number of rounds of executing the simulation model prediction based on the environmental change. The environmental change in the preset time period refers to the change in the environmental data in the preset time period. The processor may determine the ratio of the standard deviation of the environmental data at multiple time points in the preset time period to the mean of the environmental data as the environmental change in the preset time period.

[0075] Exemplarily, the smaller the environmental change in the preset time period, the fewer rounds of simulation model prediction are executed; conversely, the greater the environmental change in the preset time period, the more rounds of simulation model prediction are executed.

[0076] The greater the environmental change in the preset time period, the more complex the environment of the current area to be simulated, and the greater the difference between the candidate simulation results corresponding to different first time periods. In some embodiments of the present specification, the number of rounds of simulation model prediction is determined based on the environmental data of the preset time period. When the environmental change is large, the number of prediction rounds can be appropriately increased to obtain more candidate simulation results, which is helpful for subsequently determining more accurate target simulation results.

[0077] The target simulation results refer to the crowd simulation results that ultimately serve as the basis for determining planning recommendations.

[0078] In some embodiments, the processor may calculate the similarity of candidate simulation results corresponding to two adjacent rounds of simulation model predictions by using methods such as Euclidean distance and cosine similarity, and use one of the two candidate simulation results with the highest similarity as the target simulation result.

[0079] In some embodiments, the processor can also determine the model based on the data and determine the target simulation result. For more information, see Figure 5 Related description.

[0080] In some embodiments of the present specification, before determining planning recommendations, at least one round of simulation model prediction is performed, and a target simulation result is determined based on at least one candidate simulation result. It is possible to select a crowd simulation result that best meets actual requirements from multiple sets of data through multiple repeated predictions as the basis for determining planning recommendations, thereby helping to improve the accuracy of crowd simulation results and the rationality of urban planning.

[0081] Figure 3 It is an exemplary schematic diagram of generating crowd simulation results according to some embodiments of this specification.

[0082] In some embodiments, the processor can construct a first simulation map 320 based on identifying the crowd distribution 311 and the first flow data 312; based on the first simulation map, determine the simulated crowd distribution 341 of the first area, the simulated crowd distribution 342 of the second area and the second flow data 343 through the simulation model 330; based on the simulated crowd distribution of the first area, the simulated crowd distribution of the second area and the second flow data, determine the simulated crowd distribution 350 of the area to be simulated; based on the simulated crowd distribution, generate a crowd simulation result 360.

[0083] For relevant instructions on identifying crowd distribution, first flow data, simulated crowd distribution in the first area, simulated crowd distribution in the second area, second flow data, simulated crowd distribution, and crowd simulation results, see Figure 2 and its related description.

[0084] The first simulation map is a map for representing the relationship between various information in the first area and the second area. In some embodiments, the processor can construct the first simulation map based on identifying the distribution of the crowd and the first flow data. The first simulation map may include multiple nodes and multiple edges. The nodes of the first simulation map may include a first node and a second node.

[0085] The first node refers to a node corresponding to the first area. In some embodiments, the node attribute of the first node may include the distribution of identified people in the first area corresponding to the first node.

[0086] The second node refers to the node corresponding to the second region. In some embodiments, the node attribute of the second node can be set to a null value.

[0087] In some embodiments, the node attributes of the first node and the second node may further include the region type of the corresponding first region or second region.

[0088] The area type refers to the type of environment in which the sub-area of ​​the area to be simulated is located. For example, the area type may include a park, a school, a business district, a square, a walking path, a sidewalk, a sidewalk, etc. The processor may obtain the area types corresponding to different first areas and second areas based on a map, user input, or a storage device.

[0089] Different area types correspond to different human flows. In some embodiments of this specification, the area type is used as a node attribute of the first node and the second node, which helps to improve the accuracy of the simulated crowd distribution.

[0090] The edge of the first simulation graph is used to connect adjacent nodes, and the edge can reflect the adjacency relationship between a pair of adjacent nodes corresponding to the edge. The edge between the first nodes corresponding to the first flow data, or between the first node and the second node corresponding to the first flow data, has an edge attribute, and its edge attribute is the personnel flow direction and personnel flow amount between the sub-areas corresponding to the adjacent nodes in different first time periods. The edge attributes of the remaining edges, that is, the edges between the nodes corresponding to all areas where there is no personnel flow in the first time period, are empty.

[0091] In some embodiments, the edge attribute of each edge may further include a transition probability interval.

[0092] The transition probability interval of an edge refers to the probability interval of personnel flowing between the sub-areas corresponding to the two nodes connected by the edge.

[0093] In some embodiments, the midpoint of the transition probability interval of each edge is determined based on an average situation of the historical second flow data corresponding to the edge in the area to be simulated.

[0094] The historical second flow data refers to the flow of people and the flow direction of people in the sub-area corresponding to the two nodes connected by the edge in the historical first time period. The processor can determine the average of the flow of people in the sub-area corresponding to the two nodes connected by the edge in multiple historical first time periods as the average of the historical second flow data corresponding to the edge.

[0095] In some embodiments, the processor may determine the average situation of the historical second flow data corresponding to the edge as the midpoint of the transition probability interval of the edge.

[0096] In some embodiments, the interval width of the transition probability interval of each edge may be determined based on the regional difference corresponding to the edge. In some embodiments, the greater the regional difference, the wider the corresponding interval width.

[0097] The regional difference degree corresponding to the edge refers to the difference degree between the sub-regions corresponding to the two nodes connected by the edge. In some embodiments, the processor can determine the regional difference degree by a second preset rule based on the regional types corresponding to the sub-regions corresponding to the two nodes connected by the edge. For relevant descriptions of regional types, see Figure 3 Related description in the previous article.

[0098] The second preset rule refers to a pre-set rule for determining the regional difference, and the second preset rule can be determined by the user based on actual conditions. An exemplary second preset rule can be: the smaller the regional difference is when the regional type is the same, the greater the regional difference is when the regional type is different.

[0099] The processor may determine a transition probability interval based on the interval midpoint and the interval width.

[0100] In some embodiments of the present specification, the transition probability interval is used as an edge attribute, which helps to improve the accuracy of the simulated population distribution.

[0101] In some embodiments, the processor may further construct the first simulation map based on the distribution of identified people and the distribution of confident people in the first area.

[0102] The trusted population distribution refers to the interval of the number of people in a preset first area in multiple different second time periods. In some embodiments, the trusted population distribution includes the interval of the number of people in the first area in at least one second time period.

[0103] The second time period refers to a time period other than the first time period. For example, 8:00-9:00 is the first time period, 10:00-11:00 is the first time period, and 9:00-10:00 is the second time period.

[0104] The confidence number interval refers to a preset number interval that may be included in the first area in the second time period. In some embodiments, the confidence number interval can be preset by the system or manually. For example, the processor can set different number ranges for different area types, and determine the confidence number interval based on the area type and the corresponding number range of the first area. The processor can determine the maximum and minimum values ​​of the distribution of the number of identified people in the first area of ​​the same area type as the upper and lower intervals of the aforementioned number range, and randomly determine the confidence number interval corresponding to the first area of ​​the area type within the number range.

[0105] In some embodiments, the confidence population distribution may be related to the area type of the first area. For example, the processor may determine the highest value of the identification population distribution in all first areas of the same area type as the first area as the confidence population distribution of the first area.

[0106] In some embodiments, the confidence population distribution may also be related to the spatial confidence of the first region.

[0107] Spatial confidence refers to the degree of spatial credibility of the data in the first region.

[0108] In some embodiments, the spatial confidence can be determined based on the amount of data in the corresponding first region and the credibility of the data source. For example, the larger the amount of data in the first region and the higher the credibility of the data source, the higher the corresponding spatial confidence.

[0109] The data volume refers to the amount of data containing crowd information in the first area. The processor can obtain the amount of data containing crowd information corresponding to the first area based on the storage device to determine the data volume of the first area.

[0110] The credibility of the data source refers to the credibility of the data source for obtaining the data containing crowd information in the first area. The data sources of the first area may include multiple types. For example, the data sources of the first area may include sensing equipment, manual monitoring, model prediction, etc. The processor can test each data source in advance to determine the deviation and mean between multiple experimental data obtained from multiple tests of the same data source in the same experimental scenario; the smaller the deviation, the smaller the difference between the mean and the actual value of the data in the experimental scenario, and the higher the credibility of the data source corresponding to the data source. As one of the experimental parameters, the actual value of the data can be determined in advance by the user before the experiment.

[0111] In some embodiments, the higher the spatial confidence of the first region, the smaller the interval width of the confidence population distribution of the first region.

[0112] In some embodiments, the confidence population distribution may also be related to the spatial confidence and the temporal confidence of the first area.

[0113] The time confidence refers to the confidence level of the data of the first region in time. In some embodiments, the processor may determine the time confidence as the ratio of the simulated crowd distribution of the first region to the recognized crowd distribution of the first region at the same time point.

[0114] Exemplarily, 8:00-9:00 is the first time period, and the identified population distributions of the first area A, the first area B, and the first area C at 8:00 are 10 people, 20 people, and 30 people, respectively. During the first time period, 7 people from the first area C flowed to the first area B, and 3 people from the first area A flowed to the first area C. The simulated population distribution of the first area C is 26 people, but at 9:00, the identified population distribution of the first area C is 33 people, and the time confidence of the first area C is 26 / 33.

[0115] In some embodiments, the higher the temporal confidence of the first region, the higher the spatial confidence, and the smaller the interval width of the corresponding confident population distribution.

[0116] Since the identified population distribution and the second flow data of the first area may come from different methods, there may be a conflict between the identified population distribution and the second flow data. Therefore, the time confidence can reflect the degree of fit between the identified population distribution and the second flow data. In some embodiments of the present specification, the confident population distribution is correlated with the time confidence, which helps to improve the accuracy of the subsequent predicted simulated population distribution.

[0117] In some embodiments of the present specification, spatial confidence is correlated with the confident population distribution, and the confidence level of the data in the first region in space can be combined to determine a more realistic confident population distribution, which helps to make a more accurate prediction of the simulated population distribution later.

[0118] In some embodiments, the node attribute of the first node in the first simulation graph may further include the distribution of the trusted population in the first area corresponding to the first node.

[0119] In some embodiments of the present specification, a first simulation map is constructed based on the identified population distribution and the confident population distribution in the first area, and a more accurate prediction of the simulated population distribution can be made in combination with the confident population distribution.

[0120] The simulation model refers to a model for determining the simulated population distribution in the first area, the simulated population distribution in the second area, and the second flow data. In some embodiments, the simulation model can be a machine learning model, such as a graph neural network model (GNN).

[0121] In some embodiments, the input of the simulation model may include a first simulation map, and the output may include a simulated population distribution in a first area, a simulated population distribution in a second area, and second flow data. The simulated population distribution in the first area may be output through a first node, the simulated population distribution in the second area may be output through a second node, and the second flow data may be output through an edge.

[0122] In some embodiments, the processor may train and acquire a simulation model based on a large number of first training samples with first labels. For example, a plurality of first training samples with first labels may be input into an initial simulation model, a loss function may be constructed through the first labels and the prediction results of the initial simulation model, and the initial simulation model may be updated based on the iteration of the loss function. When the loss function of the initial simulation model satisfies a preset condition, the simulation model training is completed, wherein the preset condition may be that the loss function converges, the number of iterations reaches a threshold, etc.

[0123] In some embodiments, the first training sample may include a sample simulation map. The processor may obtain complete crowd distribution data of each sub-area of ​​the area to be simulated based on historical data, and construct a sample simulation map based on the aforementioned crowd distribution data. The complete crowd distribution data includes the actual crowd distribution of each sub-area in the area to be simulated and the flow of people and the direction of people flow between the sub-areas.

[0124] The first label may include the simulated population distribution of the first area corresponding to the first training sample, the simulated population distribution of the second area, and the second flow data. In some embodiments, the processor may randomly select some nodes from the sample simulation map as the second node, and set the node attribute of the second node to empty, and determine the node attribute of the second node before being empty as the simulated population distribution of the second area in the first label; the processor may determine the remaining unselected points from the sample simulation map as the first node, and use the node attribute of the first node as the simulated population distribution of the first area in the first label; the processor may randomly select some edges from the sample simulation map, set the edge attribute of the selected edge to empty, and use the edge attribute of the selected edge before being empty as the second flow data in the first label. It should be noted that the edge randomly selected by the processor may be an edge formed by the first node and the second node, or an edge formed by the second node and the second node, but not an edge formed by the first node and the first node.

[0125] In some embodiments, the training labels in the training data set of the simulation model may be determined based on video data of the area to be simulated.

[0126] Video data refers to video-related data of the area to be simulated. In some embodiments, the processor can determine the video data in a variety of ways. For example, the processor can obtain the video data based on a sensor device or manual shooting.

[0127] In some embodiments, the processor can obtain crowd distribution data in the video data based on manual analysis, machine learning statistics, etc., and then Figure 3 The foregoing method determines a first label in a training data set for a simulation model.

[0128] In some embodiments of the present specification, determining the first label through video data can improve the accuracy of determining the first label and contribute to accurate training of the simulation model.

[0129] In some embodiments, the processor may aggregate the results output by the simulation model to obtain a simulated population distribution.

[0130] The generation of crowd simulation results is similar to the previous article, which can be seen in Figure 2 Related description in the previous article.

[0131] Some embodiments of the present specification improve the input of the first simulation map into the simulation model, which can reflect the topological structure between the sub-areas of the area to be simulated, and use the machine learning model to automatically, quickly and accurately output the simulated population distribution.

[0132] Figure 4 is an exemplary schematic diagram of a simulation model according to some embodiments of the present specification.

[0133] In some embodiments, the simulation model may include a reasoning layer 331 and a correction layer 332 .

[0134] In some embodiments, the reasoning layer can determine the predicted population distribution of the second area based on the first simulation map. The reasoning layer can be a machine learning model, such as a graph neural network model.

[0135] The predicted population distribution refers to the predicted number of people in the second area during the first time period.

[0136] In some embodiments, the input of the reasoning layer may include the first simulation map 320, and the output may include the predicted population distribution 421 of the second area. For more information about the first simulation map, see Figure 3 Related description in the previous article.

[0137] In some embodiments, the input to the reasoning layer may also include environmental data. For more information about environmental data, see Figure 2 Related description in the previous article.

[0138] Environmental data has a certain impact on human outdoor activities. In some embodiments of this specification, by using environmental data as the input of the reasoning layer, the population distribution in the second area can be predicted in combination with the environmental data, which helps to more reasonably determine planning recommendations in combination with the simulated population distribution.

[0139] In some embodiments, the inference layer can be trained based on a large number of second training samples with second labels. Figure 3 The training of the simulation model is similar, see Figure 3 Related description in the previous article.

[0140] In some embodiments, the second training sample may include a sample simulation map. For information on obtaining the sample simulation map, see Figure 3 Related description above. When the input of the reasoning layer includes environmental data, the second training sample may include historical environmental data corresponding to the sample simulation map. The historical environmental data may be obtained based on historical data.

[0141] The second label may be the actual population distribution in the second area corresponding to the second training sample. The second label may be obtained based on historical data.

[0142] In some embodiments, the correction layer can determine the simulated crowd distribution and the second flow data in the second area based on the second simulation map. The correction layer can be a machine learning model, such as a graph neural network model.

[0143] In some embodiments, the input of the correction layer may include the second simulation atlas 420; the output may include the simulated population distribution of the first region, the simulated population distribution of the second region, and the second flow data. The simulated population distribution of the first region may be output through the first node, the simulated population distribution of the second region may be output through the second node, and the second flow data may be output through the edge.

[0144] The second simulation map refers to a map including the predicted population distribution of the second region based on the first simulation map. In some embodiments, the processor can construct the second simulation map based on the first simulation map and the predicted population distribution of the second region.

[0145] The node attributes of the first node of the second simulation graph include the identified population distribution and the trusted population distribution of the first region corresponding to the first node; the node attributes of the second node include the predicted population distribution of the second region corresponding to the second node. The edges of the second simulation graph are similar to those of the first simulation graph, which can be seen in Figure 3 Related description in the previous text.

[0146] In some embodiments, the processor can be trained based on a large number of third training samples with third labels. Figure 3 The training of the simulation model is similar, see Figure 3 Related description in the previous article.

[0147] In some embodiments, the third training sample may include a sample simulation map. For information on obtaining the sample simulation map, see Figure 3 Related description in the previous text.

[0148] The third label may include the simulated crowd distribution in the first area, the simulated crowd distribution in the second area, and the second flow data corresponding to the third training sample. The acquisition of the third label is similar to the acquisition of the first label, which can be seen in Figure 3Related description in the previous article.

[0149] In some embodiments of the present specification, by dividing the simulation model into different layers, the node attributes of the second node missing in the first simulation map can be predicted, thereby improving the data completeness of the input correction layer and facilitating improving the accuracy of the simulated population distribution.

[0150] Figure 5 It is an exemplary schematic diagram of determining target simulation results according to some embodiments of this specification.

[0151] In some embodiments, for the candidate simulation results 510, the processor may determine the demand fit 531 and the data accuracy 532 based on the data determination model 520; determine the simulation effect 540 based on the demand fit and the data accuracy; and determine the target simulation result 550 based on the simulation effect. For more information about the candidate simulation results and the target simulation results, please refer to Figure 2 Related description in the previous article.

[0152] The data determination model refers to a model used to determine demand fit and data accuracy. In some embodiments, the data determination model can be a machine learning model, such as a neural network model.

[0153] Demand fit is an indicator used to measure the degree of fit between candidate simulation results and planning requirements.

[0154] Data accuracy refers to the metric used to measure the accuracy of candidate simulation results.

[0155] In some embodiments, the input of the data determination model may include candidate simulation results, and the output may include requirement fit and data accuracy.

[0156] In some embodiments, the input of the data determination model also includes environmental data 410. For more information about environmental data, see Figure 2 and its related description.

[0157] In some embodiments of the present specification, demand fit and data accuracy are determined based on environmental data, so that the determination of demand fit and data accuracy can be made consistent with actual human activity desires, which helps to improve the accuracy of simulation effect determination.

[0158] In some embodiments, the processor may determine the model based on a large number of fourth training sample training data with fourth labels. The training process of the data determination model is similar to that of the simulation model, which can be seen in Figure 3 Related description in the previous article.

[0159] In some embodiments, the fourth training sample may include a sample population simulation result. The fourth training sample may be obtained based on historical data.

[0160] The fourth label may be the demand fit and data accuracy corresponding to the fourth training sample. The processor may determine the difficulty of determining the planning suggestion based on the sample population simulation result and planning according to the planning suggestion, and determine the inverse of the difficulty of planning as the demand fit in the fourth label; and determine the final planning effect as the data accuracy in the fourth label.

[0161] The processor may determine the difficulty of planning based on the consumption of computing resources and the amount of data required to be collected during planning. The greater the consumption of computing resources and the more data required to be collected, the greater the difficulty. The processor may determine the average of the user's ratings of the final planning effect as the data accuracy.

[0162] In some embodiments, the processor may determine the simulation effect in a variety of ways based on the degree of fit to the requirements and the accuracy of the data. For example, the higher the degree of fit to the requirements and the higher the accuracy of the data, the better the corresponding simulation effect.

[0163] In some embodiments, the processor may determine the simulation effect by weighting based on the demand fit and data accuracy, where the weight is related to the planning demand quantity and the planning demand quality.

[0164] The planning demand refers to the amount of indicators that need to be planned in the simulated area. The planning indicators may include road planning indicators, parking lot planning indicators, rest area planning indicators, etc. The planning demand can be determined based on statistics of citizens, on-site surveys, etc.

[0165] Planning requirement quality refers to the quality of planning for the simulation area. Planning requirement quality can be preset by the system or manually based on actual needs.

[0166] In some embodiments, the weight of demand fit is positively correlated with the planned demand quantity and negatively correlated with the planned demand quality; the weight of data accuracy is negatively correlated with the planned demand quantity and positively correlated with the planned demand quality.

[0167] In some embodiments of the present specification, weights corresponding to demand fit and data accuracy are determined based on planned demand quantity and planned demand quality, and the simulation effect after balancing the planned demand quantity and planned demand quality can be determined in combination with actual demand.

[0168] In some embodiments, the processor may determine the candidate simulation result with the highest simulation effect as the target simulation result.

[0169] In some embodiments of the present specification, a target simulation result is determined based on the simulation effect. Before determining a planning suggestion, a crowd simulation result with the best simulation effect can be selected as a planning basis, thereby improving the rationality of urban planning.

[0170] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements and corrections to this specification. Such modifications, improvements and corrections are suggested in this specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

[0171] At the same time, this specification uses specific words to describe the embodiments of this specification. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of this specification can be appropriately combined.

[0172] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this specification, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and laminar flow hoods in this specification. Although the above disclosure discusses some of the invention embodiments currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0173] Similarly, it should be noted that in order to simplify the description disclosed in this specification and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this specification, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the subject matter of this specification are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.

[0174] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of this specification are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

[0175] Each patent, patent application, patent application publication, and other materials, such as articles, books, specifications, publications, documents, etc., cited in this specification are hereby incorporated by reference in their entirety. Except for application history documents that are inconsistent with or conflicting with the contents of this specification, documents that limit the broadest scope of the claims of this specification (currently or later attached to this specification) are also excluded. It should be noted that if the descriptions, definitions, and / or use of terms in the materials attached to this specification are inconsistent or conflicting with the contents described in this specification, the descriptions, definitions, and / or use of terms in this specification shall prevail.

[0176] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A method for simulating urban crowd distribution, characterized in that: Executed by the processor, including: Determine a first region and a second region from the region to be simulated, wherein the first region is a sub-region of the region to be simulated whose data completeness satisfies a preset data condition, and the second region is a sub-region of the region to be simulated whose data completeness does not satisfy the preset data condition; Acquire the distribution of identified people in the first area and first flow data, wherein the distribution of identified people includes the number of identified people in the first area within at least one first time period; Determine a crowd simulation result based on the identified crowd distribution in the first area, wherein the crowd simulation result is a visualization result of the simulated crowd distribution in the area to be simulated; Based on the crowd simulation result, a planning suggestion is determined, and the crowd simulation result and the planning suggestion are pushed to a user terminal.

2. The method according to claim 1, characterized in that The simulated crowd distribution includes the simulated crowd distribution of the first area, the simulated crowd distribution of the second area, and second flow data, the second flow data includes the flow direction and flow volume of people between the first areas, or between the first area and the second area, or between the second areas in the at least one first time period, and the crowd simulation result determined based on the identified crowd distribution of the first area includes: Based on the identified population distribution and the first flow data, a first simulation map is constructed, wherein the first flow data includes the personnel flow direction and personnel flow volume between the first areas and / or between the first area and the second area within the at least one first time period, the first simulation map includes a first node and a second node and an edge, the first node corresponds to the first area, the second node corresponds to the second area, the edge reflects the adjacency relationship between a pair of adjacent nodes corresponding to the edge, and the node attribute of the first node includes the identified population distribution of the first area corresponding to the first node; Based on the first simulation map, determining the simulated population distribution in the first area, the simulated population distribution in the second area, and the second flow data through a simulation model, wherein the simulation model is a machine learning model; Determining the simulated crowd distribution in the area to be simulated based on the simulated crowd distribution in the first area, the simulated crowd distribution in the second area, and the second flow data; Based on the simulated crowd distribution, the crowd simulation result is generated.

3. The method according to claim 2, characterized in that The method comprises: Based on the identified population distribution and the trusted population distribution in the first area, the first simulation map is constructed, the trusted population distribution includes the trusted population interval of the first area in at least one second time period, and the trusted population distribution in the first area is related to the area type of the first area; the node attributes of the first node in the first simulation map also include the trusted population distribution of the first area corresponding to the first node.

4. The method according to claim 1, characterized in that: The method further comprises: Before determining the planning suggestion, performing at least one round of simulation model prediction to obtain at least one candidate simulation result; Based on the at least one candidate simulation result, a target simulation result is determined.

5. The method according to claim 4, characterized in that The determining the target simulation result based on the at least one candidate simulation result comprises: For the candidate simulation results, determining the degree of demand fit and data accuracy based on a data determination model, wherein the data determination model is a machine learning model; Determining simulation effect based on the demand fit and the data accuracy; Based on the simulation effect, the target simulation result is determined.

6. A city crowd distribution simulation system, characterized in that: include: A first determination module is configured to determine a first area and a second area from the area to be simulated, wherein the first area is a sub-area of ​​the area to be simulated whose data completeness satisfies a preset data condition, and the second area is a sub-area of ​​the area to be simulated whose data completeness does not satisfy the preset data condition; An acquisition module is configured to acquire the distribution of identified people in the first area and first flow data, wherein the distribution of identified people includes the number of identified people in the first area within at least one first time period; A second determination module is configured to determine a crowd simulation result based on the identified crowd distribution in the first area, wherein the crowd simulation result is a visualization result of the simulated crowd distribution in the area to be simulated; The third determination module is configured to determine a planning suggestion based on the crowd simulation result, and push the crowd simulation result and the planning suggestion to the user terminal.

7. The system according to claim 6, characterized in that The simulated crowd distribution includes the simulated crowd distribution of the first area, the simulated crowd distribution of the second area, and second flow data, the second flow data includes the flow direction and flow volume of people between the first areas, or between the first area and the second area, or between the second areas, in the at least one first time period, and the second determination module is further configured as follows: Based on the identified population distribution and the first flow data, a first simulation map is constructed, wherein the first flow data includes the personnel flow direction and personnel flow volume between the first areas and / or between the first area and the second area within the at least one first time period, the first simulation map includes a first node and a second node and an edge, the first node corresponds to the first area, the second node corresponds to the second area, the edge reflects the adjacency relationship between a pair of adjacent nodes corresponding to the edge, and the node attribute of the first node includes the identified population distribution of the first area corresponding to the first node; Based on the first simulation map, determining the simulated population distribution in the first area, the simulated population distribution in the second area, and the second flow data through a simulation model, wherein the simulation model is a machine learning model; Determining the simulated crowd distribution in the area to be simulated based on the simulated crowd distribution in the first area, the simulated crowd distribution in the second area, and the second flow data; Based on the simulated crowd distribution, the crowd simulation result is generated.

8. The system according to claim 7, characterized in that The second determining module is further configured to: Based on the identified population distribution and the trusted population distribution of the first area, the first simulation map is constructed, the trusted population distribution includes the trusted population interval of at least one second time period of the corresponding first area, and the trusted population distribution of the first area is related to the area type of the first area; the node attributes of the first node in the first simulation map also include the trusted population distribution of the first area corresponding to the first node.

9. A device for simulating the distribution of urban crowds, characterized in that: The apparatus comprises at least one processor and at least one memory; The at least one memory is used to store computer instructions; The at least one processor is used to execute at least part of the computer instructions to implement the urban crowd distribution simulation method as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that: The storage medium stores computer instructions. When a computer reads the computer instructions, the computer executes the urban crowd distribution simulation method as described in any one of claims 1 to 5.

Images