A method, system, device and medium for inferring population density of a city

Abstract

The application discloses a kind of urban population density deduction method, system, equipment and medium, it is related to urban management field, the method is: static population distribution benchmark model is constructed, the traffic congestion data of each road section in city is acquired, the congestion index of each grid unit at current time is calculated based on the traffic congestion data;The standardized deviation of the congestion index and historical congestion benchmark is calculated, based on the standardized deviation and preset proportion coefficient, the dynamic change amount indicating the effective integration degree of each grid unit is calculated;The dynamic change amount is added to the weighted space integration degree in the preset static population distribution benchmark model, to deduce the real-time population density of each grid unit is obtained.Therefore, by implementing the present application, the population density of city can be accurately deduced, and high-time-efficiency, high-precision population distribution data support can be provided for urban planning, traffic management, emergency response and other urban management decisions.

Description

Technical Field

[0001] This invention relates to the field of urban management, and in particular to a method, system, device and medium for predicting urban population density. Background Technology

[0002] Projecting urban population density distribution is a crucial foundation for advancing smart city construction and modernizing urban governance. In various urban governance and operation scenarios, including urban planning, traffic management, public safety early warning, business decision support, and emergency response dispatch, high-precision, near-real-time population distribution information can provide a scientific basis for resource allocation, facility layout, service optimization, and risk prevention, serving as key data support for achieving refined urban management and dynamic decision-making.

[0003] Currently, existing technologies mainly rely on traditional population statistics methods based on administrative units or fixed grids, spatial syntax models based on spatial topology analysis, and various static population distribution models. While traditional grid-based statistical methods are simple in structure, they are limited by the large scale of statistical units and long update cycles, failing to reflect micro-scale population differences and real-time dynamics. Spatial syntax models only consider road network topology, neglecting the substantial impact of urban functions such as commerce, employment, and residence on population distribution, leading to significant discrepancies between predicted results and actual population flow. Static models, unable to integrate high-frequency dynamic data such as real-time traffic flow, struggle to respond promptly to the instantaneous impact of sudden events like traffic accidents and large-scale activities on population distribution, making accuracy difficult to guarantee. Summary of the Invention

[0004] This invention provides a method, system, device, and medium for predicting urban population density, which can accurately predict urban population density.

[0005] This invention provides a method for extrapolating urban population density, comprising: Spatial syntactic analysis is performed on each grid unit based on historical urban road network data to obtain the spatial syntactic integration degree of each grid unit. Kernel density analysis is performed on the interest point data of each functional category to obtain the urban functional density of each grid unit. The spatial syntactic integration degree and the urban functional density are fused to obtain the weighted spatial integration degree of each grid unit. A static population distribution benchmark model is constructed based on the weighted spatial integration degree. Obtain traffic congestion data for each road segment in the city, and calculate the congestion index of each grid unit at the current time based on the traffic congestion data; The standardized deviation between the congestion index and the historical congestion benchmark is calculated. Based on the standardized deviation and a preset proportional coefficient, the dynamic change representing the effective integration degree of each grid unit is calculated. The historical congestion benchmark is calculated based on historical congestion data, and the proportional coefficient is obtained by training with historical traffic anomaly data and population change data. The dynamic changes are superimposed on the weighted spatial integration degree in the static population distribution benchmark model to deduce the real-time population density of each grid cell.

[0006] This invention provides a high-frequency, high-spatial-resolution dynamic data source for simulation by acquiring and calculating the traffic congestion index of each grid unit in real time. This directly reflects the current operating status of the urban road network, which is a prerequisite for capturing the real-time flow of population caused by changes in traffic conditions. By comparing the current congestion index with a benchmark value obtained from historical data statistics, a standardized deviation is calculated. Combined with a proportional coefficient trained using historical data, the abstract traffic flow anomaly is precisely quantified into a dynamic change in the spatial effective integration degree, thereby achieving an objective and quantitative characterization of the impact of external events on population distribution. Finally, by superimposing the above dynamic changes onto a pre-trained static population distribution benchmark model based on multi-source data, a "static benchmark + dynamic correction" simulation framework is creatively constructed. The static model ensures an accurate depiction of the stable relationship between urban spatial structure and functional attractiveness under normal conditions, while the dynamic correction introduces spatial accessibility disturbances caused by real-time traffic anomalies. The combination of the two allows the simulation results to reflect both the long-term laws of urban population distribution and the sensitive response to short-term dynamic changes, thus significantly improving the spatiotemporal accuracy and real-time performance of population density simulation.

[0007] Furthermore, the spatial syntactic analysis of each grid unit based on historical urban road network data is performed to obtain the spatial syntactic integration degree of each grid unit. Kernel density analysis is performed based on interest point data for each functional category to obtain the urban functional density of each grid unit. The spatial syntactic integration degree and the urban functional density are fused to obtain the weighted spatial integration degree of each grid unit. A static population distribution benchmark model is constructed based on the weighted spatial integration degree, including: Acquire the historical urban road network data, population data of each grid unit, the point of interest data, and regional interest surface data; Based on the historical urban road network data and the regional interest surface data, the spatial syntactic integration degree of each grid cell is calculated; Kernel density analysis is performed on the point of interest data to obtain the urban function density of each function category in each grid cell, wherein the function categories include at least residential, employment and commercial. Calculate the regional functional strength of each grid cell based on the regional interest surface data; By integrating the spatial syntactic integration degree, the urban functional density, and the regional functional intensity, the weighted spatial integration degree of each grid unit is calculated. Using the population data as the dependent variable and the weighted spatial integration degree as the independent variable, a geographically weighted regression model is trained to obtain a static population distribution baseline model.

[0008] By integrating road network topology (spatial syntactic integration), urban functional distribution (POI kernel density), and regional characteristics (AOI functional intensity), a weighted spatial integration is constructed, and a static benchmark model is trained based on geographic weighted regression. This model accurately quantifies the impact of spatial structure and functional attractiveness on population distribution, providing a high-precision foundation for dynamic projection and significantly improving the accuracy of population density projection.

[0009] Furthermore, the spatial syntactic analysis of each grid unit based on historical urban road network data is used to obtain the spatial syntactic integration degree of each grid unit, including: Based on the aggregation processing of the historical urban road network data and the regional interest surface data, an aggregated road network is obtained; Spatial syntactic analysis is performed on the aggregated road network to obtain the global integration degree of each line segment unit in the urban road network; The global integration degree of each line segment unit is distributed to the corresponding grid unit through spatial connection or weighted average to obtain the spatial syntactic integration degree of each grid unit.

[0010] Furthermore, the urban functional density includes residential density, employment density, and commercial density. The kernel density analysis of the point-of-interest data to obtain the urban functional density of each grid cell includes: Based on the functional type of the interest point data, a subset of interest points for each functional category is obtained; Kernel density analysis is performed on the subsets of interest points for each functional category to calculate the residential density, employment density, and business density within each grid cell.

[0011] Furthermore, the weighted spatial integration degree of each grid cell is calculated by integrating the spatial syntactic integration degree, the urban functional density, and the regional functional intensity, including: Using the entropy weight method, the weight coefficients for residential, employment, and commercial categories are determined based on the degree of variation in urban functional density for each functional category. Based on the regional functional strength of each grid cell, calculate the interest enhancement factor for the corresponding functional category; The weighted spatial integration degree of each grid unit is calculated by weighting and fusing the spatial syntactic integration degree, the residential category weight coefficient, the employment category weight coefficient, the commercial category weight coefficient, the urban functional density corresponding to each functional category, and the interest enhancement factor.

[0012] Furthermore, the historical congestion benchmark includes a mean and a standard deviation. The calculation of the standardized deviation between the congestion index and the historical congestion benchmark, based on the standardized deviation and a preset scaling factor, yields a dynamic change representing the effective integration degree of each grid cell, including: Obtain historical traffic congestion data for the same period in each grid cell, and calculate the average value and standard deviation of the historical congestion index based on the traffic congestion data; Calculate the difference between the congestion index of each grid cell and the average value, and divide the difference by the standard deviation to obtain the standardized deviation; The standardized deviation is compared with a preset anomaly detection threshold. If the absolute value of the standardized deviation is greater than the anomaly detection threshold, the corresponding grid cell is determined to be in a traffic anomaly state. The standardized deviation corresponding to the grid cell in an abnormal state is multiplied by the scaling factor to obtain the dynamic change in the effective integration degree of the corresponding grid cell.

[0013] By establishing a dynamic anomaly monitoring mechanism, the degree to which real-time traffic deviates from historical norms (standardized deviation) is quantified as a change in spatial accessibility. This enables the model to perceive and respond to the actual impact of dynamic events such as traffic accidents and large-scale events on population distribution, thereby achieving dynamic correction based on the static baseline model and significantly improving the accuracy and timeliness of population density projection in spatiotemporal changing scenarios.

[0014] Furthermore, after comparing the standardized deviation with a preset anomaly detection threshold, the method further includes: If the absolute value of the standardized deviation is less than the anomaly determination threshold, the dynamic change is superimposed on the weighted spatial integration degree in the preset static population distribution benchmark model to deduce the real-time population density of each grid cell.

[0015] In this way, regardless of whether traffic conditions are significantly abnormal, the system dynamically fine-tunes itself based on real-time data, avoiding the blind spot of updating the model only when "abnormal" conditions occur. This continuous calibration mechanism enhances the model's ability to capture the smooth daily flow of urban population, thereby improving the overall continuity and accuracy of population density projection.

[0016] Another embodiment of the present invention provides a system for extrapolating urban population density, comprising: The first module is used to perform spatial syntactic analysis on each grid unit based on historical urban road network data to obtain the spatial syntactic integration degree of each grid unit, perform kernel density analysis based on the interest point data of each functional category to obtain the urban functional density of each grid unit, fuse the spatial syntactic integration degree and the urban functional density to obtain the weighted spatial integration degree of each grid unit, and construct a static population distribution benchmark model based on the weighted spatial integration degree. The second module is used to acquire traffic congestion data for each road segment in the city and calculate the congestion index of each grid unit at the current time based on the traffic congestion data. The third module is used to calculate the standardized deviation between the congestion index and the historical congestion benchmark. Based on the standardized deviation and the preset proportional coefficient, the dynamic change representing the effective integration degree of each grid unit is calculated. The historical congestion benchmark is calculated based on historical congestion data, and the proportional coefficient is obtained by training with historical traffic anomaly data and population change data. The fourth module is used to superimpose the dynamic changes onto the weighted spatial integration degree in the static population distribution benchmark model to deduce the real-time population density of each grid cell.

[0017] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the steps of the urban population density extrapolation method as described in the present invention.

[0018] Another embodiment of the present invention provides a computer-readable storage medium item, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to perform the steps of the urban population density extrapolation method as described in the present invention. Attached Figure Description

[0019] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating one embodiment of the urban population density extrapolation method provided in this application; Figure 2 This is a flowchart illustrating one embodiment of steps S201 to S206 provided in this application; Figure 3 This is a flowchart of the static benchmark model construction provided in this application; Figure 4 This is a schematic diagram illustrating the calculation process of the weighted spatial integration degree provided in this application; Figure 5 This is a flowchart illustrating one embodiment of steps S301 to S306 provided in this application; Figure 6 This is the dynamic simulation flowchart provided in this application; Figure 7 This is a flowchart illustrating another embodiment of the urban population density extrapolation method provided in this application; Figure 8 This is a schematic diagram of the structure of one embodiment of the urban population density projection system provided in this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0023] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0024] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0025] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0026] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0027] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0028] Projecting urban population density distribution is a crucial foundation for advancing smart city construction and modernizing urban governance. Existing technologies primarily rely on traditional demographic methods using administrative units or fixed grids, spatial syntax models based on spatial topology analysis, and various static population distribution models. However, these methods struggle to guarantee the accuracy of the projections.

[0029] See Figure 1 To accurately extrapolate urban population density, an embodiment of the present invention provides a method for extrapolating urban population density, comprising steps S101 to S104: Step S101: Based on historical urban road network data, perform spatial syntactic analysis on each grid unit to obtain the spatial syntactic integration degree of each grid unit. Based on the interest point data of each functional category, perform kernel density analysis to obtain the urban functional density of each grid unit. Combine the spatial syntactic integration degree and the urban functional density to obtain the weighted spatial integration degree of each grid unit. Based on the weighted spatial integration degree, construct a static population distribution benchmark model. Please refer to Figure 2 In some embodiments, the static population distribution benchmark model is trained based on urban road network data, population data of each grid unit, point of interest data, and regional interest surface data, including steps S201 to S206; wherein, the flowchart of the static benchmark model construction is as follows. Figure 3 As shown; Step S201: Obtain the historical urban road network data, population data of each grid unit, interest point data, and regional interest surface data; In some embodiments, urban road network vector data (e.g., in Shapefile format) is obtained from natural resources or surveying departments, covering roads of different levels, and topology checks and repairs are performed to ensure network connectivity. Topology processing includes operations such as removing duplicate lines, connecting dead-end roads, and correcting geometric errors, providing high-quality input data for spatial syntax analysis; population data of grid units registered with the road network space is obtained from statistical departments or through big data channels such as mobile signaling, serving as ground truth labels for model training; urban points of interest (POI) data is obtained through map service provider APIs or open data platforms, and core types such as residential (e.g., residential communities, apartments), employment (e.g., companies, office buildings, factories), and commercial (e.g., shopping malls, restaurants, supermarkets) are filtered according to their attributes, and the POI data is cleaned to remove duplicate and invalid data to ensure data quality; regional interest surface vector data of functional areas such as commercial areas, residential areas, and industrial parks are obtained from urban planning departments, and a multi-level classification system covering functional type, density scale, and temporal characteristics is established.

[0030] It should be noted that population statistics usually come from population censuses, mobile signaling data, or big data population estimation results, which represent a relatively accurate baseline for population distribution.

[0031] It should be noted that the established three-level AOI classification system is as follows: the first level is classified according to the main function (residential, commercial, industrial, public service, etc.); the second level is classified according to the functional density and scale (high-density commercial, low-density residential, etc.); and the third level is classified according to the time activity characteristics (daytime, nighttime, all-day).

[0032] Step S202: Based on the historical urban road network data and the regional interest surface data, calculate the spatial syntactic integration degree of each grid cell; It should be noted that the original road network vector data used in this invention is real urban road data (such as Shapefile format) obtained from natural resources or surveying departments, which serves as the basic input for spatial syntactic analysis. However, when traditional spatial syntactic analysis processes urban road networks containing large functional areas (such as residential communities, industrial parks, and commercial complexes), the roads within these areas introduce a large number of invalid topological nodes and short road segments, leading to two problems: first, it interferes with the calculation accuracy of the global topology, causing the integration index to deviate from the true spatial accessibility; second, it significantly increases computational complexity, which is detrimental to large-scale urban road network analysis. Therefore, this invention does not discard the original road network vector data, but rather uses AOI (Area of ​​Interest) data as prior knowledge to aggregate road networks located within the same functional area. The essence of this aggregation operation is to replace all road nodes and road segments within an AOI with a representative node without affecting the region's external connectivity. The final "aggregated road network" used for spatial syntactic analysis has all its original line segments derived from the original road network vector data, only with topological simplification. In other words, AOI aggregation is a reasonable simplification of the original road network, rather than a data replacement. If AOI data does not exist in a certain area, the road network of the corresponding area remains in its original state and is not aggregated. Therefore, the spatial syntactic analysis of this invention is always based on an optimized version of the original road network vector data, and there is a clear inheritance and mapping relationship between the two.

[0033] In some embodiments, step S202 includes: aggregating the historical urban road network data and the regional interest surface data to obtain an aggregated road network; performing spatial syntactic analysis on the aggregated road network to obtain the global integration degree of each line segment unit in the urban road network; and distributing the global integration degree of each line segment unit to the corresponding grid units through spatial connectivity or weighted averaging to obtain the spatial syntactic integration degree of each grid unit. Specifically, a road network aggregation algorithm is executed based on the historical urban road network data and the regional interest surface data, which is formally represented as follows: ,in, Historical urban road network data, For regional interest surface data, This is the output aggregated road network. Subsequently, the obtained aggregated road network... Importing the data into spatial syntactic analysis software (such as DepthmapX or sDNA), the topology computation radius is set to the total number of line segments (usually denoted as n). Global integration is calculated for each line segment unit in the aggregated road network to obtain an index value quantifying its centrality and reachability potential within the entire network. During the calculation, an appropriate topology computation radius needs to be set; typically, n or the total number of line segments is used as the radius for global integration to capture spatial centrality within the entire network. Finally, to perform correlation analysis with population grid cells, the global integration value at the line segment level needs to be assigned to a unified grid cell, thereby obtaining a representative value for the spatial syntactic integration of each grid cell. .

[0034] In some embodiments, the specific operation steps of the road network aggregation algorithm include: First, acquiring all nodes and road segments in the historical urban road network data and loading all AOI area data; then, traversing each AOI area to identify its spatial boundary, and finding all original road network nodes and road segments completely located within the AOI area boundary from the historical urban road network data. Next, for each AOI, determining a center point A located within the area that represents its overall spatial position by calculating its area-weighted center point, etc., this center point will serve as the unique representative node of that AOI in the aggregated road network. Then, using the boundary of each AOI, performing spatial cutting operations on the external roads intersecting the AOI area boundary in the original historical road network data, creating new nodes at the intersection points as external connection points. Simultaneously, deleting all original road segments and nodes within the AOI area from the original historical road network data; then, creating a new connection line for each created external connection point, directly connecting it to the center representative point A of that AOI. Thus, the aggregated road network is output. All the nodes and edges that are retained are directly derived from the original historical urban road network data, without introducing any external fictitious roads.

[0035] Through the above steps, each functional area (AOI) is aggregated into a representative single node, through which all external traffic accesses the network. This significantly simplifies the road network topology while maintaining the region's external connectivity, effectively improving the efficiency and accuracy of subsequent spatial syntactic analysis. It should be noted that all nodes and edges retained in the aggregated road network are directly or indirectly derived from the historical urban road network data. The aggregation operation does not introduce any external fictitious roads; it merely simplifies the topology of the original road network.

[0036] It should be noted that the allocation method can use a weighted average of the lengths of line segments within the grid to ensure that the spatial syntax index can accurately reflect the accessibility characteristics of the grid.

[0037] Step S203: Perform kernel density analysis on the point of interest data to obtain the urban function density of each functional category in each grid cell, wherein the functional categories include at least residential, employment and commercial categories; Further, step S203 includes: classifying the point of interest data according to its functional type to obtain a subset of points of interest for each functional category; performing kernel density analysis on the subset of points of interest for each functional category to calculate the residential density, employment density, and commercial density within each grid cell. Specifically, firstly, the acquired raw point of interest data is cleaned and classified to remove duplicate records and invalid data points, ensuring data quality; then, based on the functional type field in its attribute information, it is divided into subsets of points of interest for core functional categories such as residential, employment, and commercial. For example, the residential subset may include point elements such as residential communities and apartments, the employment subset may include companies, office buildings, and factories, and the commercial subset may include shopping malls, restaurants, and supermarkets. Next, kernel density analysis is performed on the subset of points of interest for each functional category to obtain the residential density corresponding to each grid cell. Employment density Commercial density .

[0038] It should be noted that kernel density estimation is a nonparametric estimation method used to calculate the density distribution of point features in the surrounding space. Its basic idea is to treat each point of interest as a density diffusion source, and calculate the density contribution of that point at any location in the surrounding space using a specified kernel function (such as a Gaussian kernel function) and a search radius (i.e., bandwidth). The contributions of all points are then superimposed in space to form a continuous density surface.

[0039] It should be noted that the bandwidth parameter is the key to kernel density analysis and needs to be optimized according to the urban spatial scale and grid size. The Silverman criterion or cross-validation method is usually used to determine the optimal bandwidth in order to balance the smoothness of density estimation and the degree of detail preservation.

[0040] It should be noted that for a specific functional density value (such as residential density) within each grid cell... This can be obtained by overlaying the density surface with the grid cell boundary and extracting or calculating the average density value within the grid range.

[0041] Step S204: Calculate the regional functional strength of each grid cell based on the regional interest surface data; In some embodiments, for a particular function type, its functional strength within a grid cell is calculated using the following formula: , where the molecule is the first Interest surfaces and grids for this type of region The overlapping area, with the denominator being the grid. area, The weight of the interest surface in the area (such as the grade of the commercial area, the plot ratio of the residential area, etc.) can be determined by a comprehensive evaluation method based on its grade, plot ratio, functional mixing degree and other attributes. This indicator quantifies the land cover intensity and importance of this type of function within the grid.

[0042] Step S205: By integrating the spatial syntactic integration degree, the urban functional density, and the regional functional intensity, the weighted spatial integration degree of each grid unit is calculated. In some embodiments, step S205 includes: using the entropy weighting method to determine the residential weight coefficient, employment weight coefficient, and commercial weight coefficient based on the degree of variation in urban functional density of each functional category; calculating the interest enhancement factor for the corresponding functional category based on the regional functional intensity of each grid unit; and calculating the weighted spatial integration degree of each grid unit by weighting and fusing the spatial syntactic integration degree, residential weight coefficient, employment weight coefficient, commercial weight coefficient, urban functional density corresponding to each functional category, and the interest enhancement factor. Specifically, firstly, the functional density data of various types are standardized to make them dimensionless and comparable, resulting in index values; and the information entropy of each index value is calculated; the information utility value is calculated based on the information entropy, and the information utility value of each index is normalized to obtain the residential weight coefficient. Employment-related weighting coefficients And business category weighting coefficient Subsequently, based on the regional functional strength of each grid cell... Calculate the interest enhancement factor for the corresponding functional category. To further enhance the influence of functional points located in high-intensity functional areas, the calculation formula is as follows: ;in This is an adjustment coefficient (usually between 0.5 and 1.0); a larger value indicates a stronger enhancement effect on the AOI data. The weighted spatial integration degree of each grid cell is calculated using a weighted fusion formula. The calculation formula is: ,in, Including residential categories Employment and business .

[0043] It should be noted that the entropy weight method is an objective weighting method based on information entropy. It determines the weight of each indicator (in this case, the three types of functional density data) according to its own degree of variation. The greater the degree of variation (i.e., the smaller the information entropy), the higher the weight of the indicator, thus avoiding the influence of subjective factors.

[0044] It should be noted that the weighted spatial integration degree integrates "spatial accessibility potential" and "functional attractiveness," considering both the topological characteristics of the space and the influence of urban functional distribution, thus providing a more comprehensive feature representation for subsequent modeling. The core innovation of this formula lies in the fact that POIs located within high-intensity functional areas receive higher weights, more accurately reflecting actual functional attractiveness. A schematic diagram of the weighted spatial integration degree calculation process is shown below. Figure 4 As shown.

[0045] Step S206: Using the population data as the dependent variable and the weighted spatial integration degree as the independent variable, train a geographically weighted regression model to obtain a static population distribution benchmark model.

[0046] In some embodiments, known population data for each grid cell are used. As the dependent variable, the weighted spatial integration degree calculated above is used. As the core independent variable, a geographically weighted regression model is trained, wherein the model takes the following form: ,in, It is a grid cell The center coordinates, and These are local regression coefficients that vary with spatial geographic location. The error term is represented by [missing term]. The model is fitted using weighted least squares, with Gaussian or double-squared functions typically chosen for the spatial weights. The bandwidth is optimized using the AICc criterion or cross-validation. After training, the coefficient surface and goodness-of-fit index are obtained, thus forming the static population distribution baseline model.

[0047] It should be noted that the static population distribution baseline model, corrected for empirical data, is obtained through fitting. This model not only includes... and The coefficient surface also includes model evaluation metrics (such as...) ,Adjustment The AICc value and residual spatial distribution are used to evaluate model quality and identify outlier regions.

[0048] By integrating road network topology (spatial syntactic integration), urban functional distribution (POI kernel density), and regional characteristics (AOI functional intensity), a weighted spatial integration is constructed, and a static benchmark model is trained based on geographic weighted regression. This model accurately quantifies the impact of spatial structure and functional attractiveness on population distribution, providing a high-precision foundation for dynamic projection and significantly improving the accuracy of population density projection.

[0049] Step S102: Obtain traffic congestion data for each road segment in the city, and calculate the congestion index of each grid unit at the current time based on the traffic congestion data; In some embodiments, firstly, by accessing the application programming interface provided by a map service provider's API (such as Amap or Baidu Maps Traffic Cloud), the traffic data service is called cyclically at fixed time intervals (such as 15 minutes) to obtain the real-time congestion index of each road segment within the city. The acquired congestion data consists of structured information including road segment identifiers and real-time congestion indices. Subsequently, through data parsing and spatial mapping processing, the road segment-level congestion index is... The congestion index for each grid cell is obtained by summarizing the data through spatial connectivity or interpolation methods to pre-divided grid cells. .

[0050] It should be noted that API calls must comply with the service provider's access frequency limits and data usage agreements, and usually require prior application for developer permissions and configuration of identity authentication keys.

[0051] It should be noted that the aggregation method takes into account road segment length, road grade, and spatial relationship with the grid to ensure that the congestion index can accurately reflect the traffic conditions within the grid.

[0052] Step S102: Calculate the standardized deviation between the congestion index and the historical congestion benchmark. Based on the standardized deviation and the preset proportional coefficient, calculate the dynamic change representing the effective integration degree of each grid unit. The historical congestion benchmark is calculated based on historical congestion data, and the proportional coefficient is obtained by training with historical traffic anomaly data and population change data. In some embodiments, the historical congestion benchmark is calculated based on historical congestion data. Specifically, for each grid cell, historical congestion data of the same period is selected as the calculation basis. The historical congestion data of the same period refers to the historical time period with the same time attributes as the current projection time, including the same day of the week and the same time period. Then, based on the current time, the historical congestion index corresponding to the same time period for multiple consecutive weeks in the past is extracted to construct the historical congestion index set of the grid cell.

[0053] It should be noted that, in order to improve the timeliness and adaptability of the benchmark, methods such as exponentially weighted moving averages can be introduced during the calculation to give more weight to recent data in the historical data. In addition, different day types such as weekdays, weekends, and holidays can be distinguished, and corresponding historical benchmark models can be established and maintained to accurately capture the periodic patterns of traffic activities under different date attributes.

[0054] In some embodiments, the proportional coefficient is obtained through training with historical traffic anomaly data and population change data. Specifically, a training sample set is formed by collecting a series of known traffic anomaly events with clear start and end times and their corresponding population distribution changes during those periods. The traffic anomaly events include, but are not limited to, significant traffic congestion fluctuations caused by large-scale events, major road construction closures, sudden traffic accidents, or severe weather. The population change data comes from mobile phone signaling, public transportation card swipe records, or surveys and statistics of specific areas, and can reflect the actual changes in population distribution before and after the event. Then, based on this sample set, the standardized deviation of the traffic anomaly events is used as the input feature, and the change in spatial effective integration inferred from the actual observed population distribution changes is used as the target label. A quantitative relationship model between the two is established using linear regression or machine learning methods. The proportional coefficient is obtained by minimizing the error between the predicted value and the true value.

[0055] In practice, large-scale construction projects with long-term impacts, such as the opening of commercial districts, hospital expansions, and new school constructions, can be selected. Based on the traffic flow changes and population distribution evolution data over a long period before and after their completion, parameter calibration can be performed to ensure that the proportional coefficient can robustly reflect the degree of impact of traffic condition changes on spatial accessibility and population distribution.

[0056] Please refer to Figure 5 In some embodiments, step S102 includes steps S501 to S504; Step S501: Obtain historical traffic congestion data for the same period in each grid cell, and calculate the average value and standard deviation of the historical congestion index based on the traffic congestion data; In some embodiments, when it is necessary to calculate a historical baseline for the current simulation time, the system extracts a specific subset of data from the historical congestion index set for each grid cell as the basis for calculation; then, it sums the congestion index values ​​in all historical congestion index sets and divides them by the number of data points (N) to obtain the average congestion index of that grid cell in the historical congestion period. Secondly, calculate the deviation of each congestion index value from the above average, take the sum of squares, and then divide by the number of data points. The variance is obtained, and finally, the square root of the variance is taken to obtain the standard deviation of the congestion index of that grid cell in the same historical period. .

[0057] In some embodiments, the average congestion index This reflects the grid's typical congestion level during the same historical period, calculated using the following formula: ; In the formula, To set All congestion index values Add, The number of data points.

[0058] In some embodiments, the standard deviation of the congestion index The normal fluctuation range of its congestion level can be quantified by the following formula: .

[0059] Step S502: Calculate the difference between the congestion index of each grid cell and the average value, and divide the difference by the standard deviation to obtain the standardized deviation; In some embodiments, for each grid cell, its real-time congestion index at the current moment is obtained; this real-time congestion index is then compared with the average congestion index. Subtracting the two values ​​yields the deviation; subsequently, this deviation is divided by the standard deviation of the congestion index corresponding to that grid cell. The standardized deviation value is obtained. The calculation formula is: .

[0060] It should be noted that the standardized deviation value It is a dimensionless statistical quantity, the magnitude of which reflects the degree of deviation of the current congestion state from the historical normal. A positive value indicates that it is more congested than the historical normal, and a negative value indicates that it is smoother than the historical normal. The larger the absolute value, the more significant the deviation.

[0061] Step S503: Compare the standardized deviation with a preset anomaly determination threshold. If the absolute value of the standardized deviation is greater than the anomaly determination threshold, the corresponding grid cell is determined to be in a traffic anomaly state. In some embodiments, firstly, an anomaly detection threshold is preset; then, the standardized deviation calculated for each grid cell is... The absolute value of the normalization deviation is compared with the threshold; if the absolute value of the normalization deviation of a certain grid cell is greater than the threshold, that is... If the grid cell is in an abnormal traffic state at the current moment, it indicates that its traffic conditions have changed significantly from the historical norm. This change may be caused by factors such as emergencies, large-scale events, or traffic control.

[0062] It should be noted that the anomaly detection threshold is usually set based on the characteristics of a normal distribution. For example, setting the threshold to 2 or 2.5 corresponds to a confidence level of approximately 95% or 99%, respectively.

[0063] In some embodiments, after comparing the standardized deviation with a preset anomaly detection threshold, the method further includes: if the absolute value of the standardized deviation is less than the anomaly detection threshold, then the dynamic change is superimposed on the weighted spatial integration degree in a preset static population distribution benchmark model to deduce the real-time population density of each grid cell. Specifically, if the standardized deviation of the grid cell is less than the preset anomaly detection threshold, then the dynamic change is superimposed on the weighted spatial integration degree in a preset static population distribution benchmark model to deduce the real-time population density of each grid cell. The absolute value is less than or equal to the anomaly detection threshold, that is... If the traffic condition is within the normal historical fluctuation range, then the anomaly-based inversion calculation is not triggered separately. Instead, the calculated standardized deviation (whether positive or negative) is directly used as a fine-tuning factor and substituted into the subsequent extrapolation formula. That is, when extrapolating real-time population density, the standardized deviation will still be proportionally converted into a small effective integration degree change and superimposed on the weighted spatial integration degree of the static baseline model, thereby achieving smooth and continuous real-time extrapolation of population density for all grids (including areas with normal traffic conditions).

[0064] In this way, regardless of whether traffic conditions are significantly abnormal, the system dynamically fine-tunes itself based on real-time data, avoiding the blind spot of updating the model only when "abnormal" conditions occur. This continuous calibration mechanism enhances the model's ability to capture the smooth daily flow of urban population, thereby improving the overall continuity and accuracy of population density projection.

[0065] Step S504: Multiply the standardized deviation corresponding to the grid cell in the abnormal state by the scaling factor to obtain the dynamic change in the effective integration degree of the corresponding grid cell.

[0066] In some embodiments, for mesh cells in an abnormal state, their corresponding standardized deviations are... With a preset scaling factor Multiply by this to calculate the dynamic change in the effective integration degree of the grid cell. The calculation formula is: Among them, the proportionality coefficient It is a constant obtained through training with historical data, which represents the quantitative conversion relationship between the standardized anomaly degree of traffic congestion and the change in spatial effective accessibility (reflected in weighted integration degree).

[0067] By establishing a dynamic anomaly monitoring mechanism, the degree to which real-time traffic deviates from historical norms (standardized deviation) is quantified as a change in spatial accessibility. This enables the model to perceive and respond to the actual impact of dynamic events such as traffic accidents and large-scale events on population distribution, thereby achieving dynamic correction based on the static baseline model and significantly improving the accuracy and timeliness of population density projection in spatiotemporal changing scenarios.

[0068] Step S103: The dynamic change is superimposed on the weighted spatial integration degree in the static population distribution benchmark model to deduce the real-time population density of each grid cell.

[0069] In some embodiments, the dynamic change is superimposed on the weighted spatial integration degree in a preset static population distribution benchmark model to deduce the real-time population density of each grid cell. Specifically, this involves obtaining the dynamic change in the effective integration degree of each grid cell caused by real-time traffic anomalies. Then, it is integrated with the baseline weighted spatial integration degree in the static population distribution baseline model. By superimposing the values, we obtain an updated value that reflects the effective spatial reachability at the current moment. Subsequently, this updated value is substituted into the pre-trained static baseline geographic weighted regression (GWR) model to calculate and derive the real-time population density estimate for each grid cell. The calculation formula is as follows: ; in, It is provided by the static datum model and varies with the grid center coordinates. The changing local regression coefficient surface, where the essence of the formula is to use the "space-function-population" local quantitative relationship established by the GWR model to map the effective integration degree that incorporates dynamic traffic impacts to the real-time population density estimate of the corresponding grid. The dynamic extrapolation flowchart is as follows: Figure 6 As shown.

[0070] It should be noted that after obtaining the population density projection results, post-processing is required, including outlier filtering, spatial smoothing, and range constraints, to ensure the rationality and stability of the results.

[0071] In some embodiments, after obtaining the real-time population density of each grid cell, it is necessary to extrapolate the results. The system visualizes heatmaps on electronic maps to create dynamic projections of urban population. The visualization uses color gradients to represent population density, supports multiple color schemes and transparency settings, and allows for historical retrospective and comparative analysis. A timeline control allows users to view population distribution changes at different points in time, and animations showcasing population flow patterns are also supported.

[0072] It should be noted that the present invention also includes receiving hypothetical conditional input from users (such as "Build a new subway station in location A" or "Road B is closed due to construction"), and simulating spatial syntactic indices under these conditions. and congestion The scenario simulation engine analyzes potential changes in population distribution and invokes the aforementioned model chain to predict possible future population distribution scenarios. Based on spatial analysis methods and machine learning predictions, the engine can assess the impact of different planning schemes on population distribution, providing data support for planning decisions.

[0073] It should be noted that the population size and changes over different time periods can be used to determine the population size of different types, such as the resident population and the working population. Then, the population and buildings can be correlated.

[0074] Compared with the prior art, the present invention has the following significant advantages: 1. Significantly improved accuracy: It creatively integrates the "topological potential" of spatial syntax with the "urban functional attractiveness" represented by POI and performs local correction (GWR) using real population data, fundamentally solving the problem of pure spatial models being divorced from socio-economic context.

[0075] 2. Significantly improved computational efficiency: Through AOI aggregation, road network aggregation greatly reduces the number of nodes and edges, thereby reducing the computational complexity of spatial syntax, making it particularly suitable for large-scale urban road network analysis.

[0076] 3. Achieved dynamic perception capabilities: For the first time, a method for inverting spatial effectiveness using high-frequency real-time traffic data was proposed, establishing a dynamic inference chain of "traffic flow - spatial effectiveness - population distribution," enabling near real-time, low-cost monitoring of urban population activities. The system can update the city-wide population projection results within 5-15 minutes, overcoming the fundamental bottlenecks of slow and costly traditional population data updates, and providing possibilities for emergency response and real-time decision-making.

[0077] 4. In-depth utilization of AOI types: By establishing a multi-dimensional AOI classification system and extracting deep features, the semantic value of AOI data was fully explored. This improved the accuracy of daytime population projection in commercial areas and nighttime population projection in residential areas.

[0078] 5. High level of intelligence and automation: By integrating intelligent agent technology, the entire process from data access, cleaning, calculation, and inference to report generation is automated, significantly reducing the need for manual intervention. The system can automatically identify abnormal patterns, generate analysis reports, and provide decision-making suggestions, greatly lowering the technical barrier to entry and improving the practicality of the technology and the efficiency of decision support.

[0079] 6. High feasibility and wide applicability: The data sources upon which the solution relies (road network, POI, congestion index) are becoming increasingly open and standardized, with a clear technical roadmap and modular design, facilitating practical development, deployment, and operational use. The system can adapt to the application needs of cities of different sizes, providing effective technical support from small and medium-sized cities to megacities.

[0080] 7. Supports multi-scenario applications: It not only provides current population distribution data but also supports scenario simulation and predictive analysis, offering decision support for multiple fields such as urban planning, traffic management, business site selection, and public safety. The system is highly scalable, easily integrating new data sources and algorithm models to continuously improve its simulation capabilities and application scope.

[0081] 8. Significant Economic Benefits: Compared with traditional population surveys and statistical methods, this invention significantly reduces data acquisition and updating costs while improving analytical efficiency. Once invested in, it can sustainably provide population distribution projection services with extremely low marginal costs, demonstrating significant economic and social value.

[0082] This invention provides a high-frequency, high-spatial-resolution dynamic data source for simulation by acquiring and calculating the traffic congestion index of each grid unit in real time. This directly reflects the current operating status of the urban road network, which is a prerequisite for capturing the real-time flow of population caused by changes in traffic conditions. By comparing the current congestion index with a benchmark value obtained from historical data statistics, a standardized deviation is calculated. Combined with a proportional coefficient trained using historical data, the abstract traffic flow anomaly is precisely quantified into a dynamic change in the spatial effective integration degree, thereby achieving an objective and quantitative characterization of the impact of external events on population distribution. Finally, by superimposing the above dynamic changes onto a pre-trained static population distribution benchmark model based on multi-source data, a "static benchmark + dynamic correction" simulation framework is creatively constructed. The static model ensures an accurate depiction of the stable relationship between urban spatial structure and functional attractiveness under normal conditions, while the dynamic correction introduces spatial accessibility disturbances caused by real-time traffic anomalies. The combination of the two allows the simulation results to reflect both the long-term laws of urban population distribution and the sensitive response to short-term dynamic changes, thus significantly improving the spatiotemporal accuracy and real-time performance of population density simulation.

[0083] For easier understanding, please refer to Figure 7 The static baseline model construction process on the left starts with multi-source static data (road network, AOI, POI, grid population), and through steps such as AOI aggregation, spatial syntax calculation, and functional strength analysis, it ultimately trains a geographic weighted regression model to form a high-precision population distribution baseline. The real-time data stream on the right represents the dynamic operation of the system: the dynamic inference engine accesses real-time traffic data, calculates historical baselines and monitors anomalies, drives the dynamic inversion model, and thus infers real-time population density; the results are finally delivered to the visualization and intelligent application layer, outputting heat maps, analysis reports, and scenario simulations to support decision-making. The entire process achieves a closed loop from data to insight.

[0084] like Figure 8As shown, based on the above method embodiments, corresponding apparatus embodiments are provided; One embodiment of the present invention provides a population density projection system for cities, comprising: The first module is used to perform spatial syntactic analysis on each grid unit based on historical urban road network data to obtain the spatial syntactic integration degree of each grid unit, perform kernel density analysis based on the interest point data of each functional category to obtain the urban functional density of each grid unit, fuse the spatial syntactic integration degree and the urban functional density to obtain the weighted spatial integration degree of each grid unit, and construct a static population distribution benchmark model based on the weighted spatial integration degree. The second module is used to acquire traffic congestion data for each road segment in the city and calculate the congestion index of each grid unit at the current time based on the traffic congestion data. The third module is used to calculate the standardized deviation between the congestion index and the historical congestion benchmark. Based on the standardized deviation and the preset proportional coefficient, the dynamic change representing the effective integration degree of each grid unit is calculated. The historical congestion benchmark is calculated based on historical congestion data, and the proportional coefficient is obtained by training with historical traffic anomaly data and population change data. The fourth module is used to superimpose the dynamic changes onto the weighted spatial integration degree in the static population distribution benchmark model to deduce the real-time population density of each grid cell.

[0085] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the urban population density extrapolation method provided by any of the above-described method embodiments of the present invention.

[0086] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0087] Based on the above-mentioned embodiments of the urban population density extrapolation method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the urban population density extrapolation method of any embodiment of the present invention.

[0088] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0089] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0090] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0091] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the urban population density extrapolation method described in any of the above-described method embodiments of the present invention.

[0092] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0093] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for extrapolating urban population density, characterized in that, include: Spatial syntactic analysis is performed on each grid cell based on historical urban road network data to obtain the spatial syntactic integration degree of each grid cell. Kernel density analysis is performed on the interest point data of each functional category to obtain the urban functional density of each grid cell. The spatial syntactic integration degree and the urban functional density are fused to obtain the weighted spatial integration degree of each grid cell. A static population distribution benchmark model is constructed based on the weighted spatial integration degree, including: acquiring the historical urban road network data, population data of each grid cell, the interest point data, and regional interest surface data; and based on the historical urban road network data and the regional interest surface data... The spatial syntactic integration degree of each grid cell is calculated; kernel density analysis is performed on the point of interest data to obtain the urban function density of each functional category in each grid cell, wherein the functional categories include at least residential, employment, and commercial categories; the regional function intensity of each grid cell is calculated based on the regional interest surface data; the weighted spatial integration degree of each grid cell is calculated by fusing the spatial syntactic integration degree, the urban function density, and the regional function intensity; a geographic weighted regression model is trained with the population data as the dependent variable and the weighted spatial integration degree as the independent variable to obtain a static population distribution baseline model; Obtain traffic congestion data for each road segment in the city, and calculate the congestion index of each grid unit at the current time based on the traffic congestion data; The standardized deviation of the congestion index from the historical congestion benchmark is calculated. The standardized deviation of the grid cell in the traffic anomaly state is compared with a preset proportional coefficient to calculate the dynamic change of the effective integration of each grid cell. The historical congestion benchmark is calculated based on historical congestion data. The proportional coefficient is trained by historical traffic anomaly data and population change data. The traffic anomaly state refers to the absolute value of the standardized deviation being greater than a preset anomaly judgment threshold. The dynamic changes are superimposed on the weighted spatial integration degree in the static population distribution benchmark model to deduce the real-time population density of each grid cell.

2. The method for extrapolating urban population density according to claim 1, characterized in that, The spatial syntactic analysis of each grid cell based on historical urban road network data yields the spatial syntactic integration degree of each grid cell, including: Based on the aggregation processing of the historical urban road network data and the regional interest surface data, an aggregated road network is obtained; Spatial syntactic analysis is performed on the aggregated road network to obtain the global integration degree of each line segment unit in the urban road network; The global integration degree of each line segment unit is distributed to the corresponding grid unit through spatial connection or weighted average to obtain the spatial syntactic integration degree of each grid unit.

3. The method for extrapolating urban population density according to claim 1, characterized in that, The urban functional density includes residential density, employment density, and commercial density. The kernel density analysis of the point-of-interest data to obtain the urban functional density of each grid cell includes: Based on the functional type of the interest point data, a subset of interest points for each functional category is obtained; Kernel density analysis is performed on the subsets of interest points for each functional category to calculate the residential density, employment density, and business density within each grid cell.

4. The method for extrapolating urban population density according to claim 1, characterized in that, The weighted spatial integration degree of each grid cell is calculated by integrating the spatial syntactic integration degree, the urban functional density, and the regional functional intensity, including: Using the entropy weight method, the weight coefficients for residential, employment, and commercial categories are determined based on the degree of variation in urban functional density for each functional category. Based on the regional functional strength of each grid cell, calculate the interest enhancement factor for the corresponding functional category; The weighted spatial integration degree of each grid unit is calculated by weighting and fusing the spatial syntactic integration degree, the residential category weight coefficient, the employment category weight coefficient, the commercial category weight coefficient, the urban functional density corresponding to each functional category, and the interest enhancement factor.

5. The method for extrapolating urban population density according to claim 1, characterized in that, The historical congestion benchmark includes a mean and a standard deviation. The calculation of the standardized deviation between the congestion index and the historical congestion benchmark, based on the standardized deviation and a preset scaling factor, yields a dynamic change representing the effective integration degree of each grid cell, including: Obtain historical traffic congestion data for the same period in each grid cell, and calculate the average value and standard deviation of the historical congestion index based on the traffic congestion data; Calculate the difference between the congestion index of each grid cell and the average value, and divide the difference by the standard deviation to obtain the standardized deviation; The standardized deviation is compared with a preset anomaly detection threshold. If the absolute value of the standardized deviation is greater than the anomaly detection threshold, the corresponding grid cell is determined to be in a traffic anomaly state. The standardized deviation corresponding to the grid cell in an abnormal traffic state is multiplied by the proportional coefficient to obtain the dynamic change in the effective integration degree of the corresponding grid cell.

6. The method for extrapolating urban population density according to claim 1, characterized in that, After comparing the standardized deviation with a preset anomaly detection threshold, the method further includes: If the absolute value of the standardized deviation is less than the anomaly determination threshold, the dynamic change is superimposed on the weighted spatial integration degree in the static population distribution benchmark model to deduce the real-time population density of each grid cell.

7. A system for projecting urban population density, characterized in that, include: The first module is used to perform spatial syntactic analysis on each grid unit based on historical urban road network data to obtain the spatial syntactic integration degree of each grid unit; perform kernel density analysis on interest point data of each functional category to obtain the urban functional density of each grid unit; fuse the spatial syntactic integration degree and the urban functional density to obtain the weighted spatial integration degree of each grid unit; and construct a static population distribution benchmark model based on the weighted spatial integration degree. This includes: acquiring the historical urban road network data, population data of each grid unit, the interest point data, and regional interest surface data; and using the historical urban road network data and the regional interest surface... The data is used to calculate the spatial syntactic integration degree of each grid unit; kernel density analysis is performed on the point of interest data to obtain the urban function density of each functional category in each grid unit, wherein the functional categories include at least residential, employment, and commercial; the regional function intensity of each grid unit is calculated based on the regional interest surface data; the spatial syntactic integration degree, the urban function density, and the regional function intensity are fused to calculate the weighted spatial integration degree of each grid unit; a geographic weighted regression model is trained with the population data as the dependent variable and the weighted spatial integration degree as the independent variable to obtain a static population distribution baseline model; The second module is used to acquire traffic congestion data for each road segment in the city and calculate the congestion index of each grid unit at the current time based on the traffic congestion data. The third module is used to calculate the standardized deviation between the congestion index and the historical congestion benchmark. The standardized deviation of the grid cell in the traffic anomaly state is compared with a preset proportional coefficient to calculate the dynamic change of the effective integration degree of each grid cell. The historical congestion benchmark is calculated based on historical congestion data, and the proportional coefficient is trained by historical traffic anomaly data and population change data. The traffic anomaly state refers to the absolute value of the standardized deviation being greater than a preset anomaly judgment threshold. The fourth module is used to superimpose the dynamic changes onto the weighted spatial integration degree in the static population distribution benchmark model to deduce the real-time population density of each grid cell.

8. A terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the urban population density extrapolation method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, include: A stored computer program, wherein, when the computer program is executed, the device containing the computer-readable storage medium controls the steps of the urban population density extrapolation method as described in any one of claims 1-6 to perform the steps.

Images