CN113342873B - A Population Analysis Unit Division Method Based on Urban Morphology and Convergence Mode - Google Patents

Abstract

The invention discloses a population analysis unit division method based on urban form and aggregation mode, comprising the following steps: S1. Taking into account the spatial structure of the city and the fine-grained needs of population analysis, and orienting to the functional heterogeneity of the micro-scale population activity area to fit the city Natural form, divide the basic analysis unit; S2. Consider the spatial distribution characteristics of urban population flow in local convergence and the long-term stability characteristics that depend on the dynamic change of entrances and exits, and construct a population convergence preference model; S3. Use the microstructural elements in the basic analysis unit to divide FPAZ (Fine Population Analysis Zone), which is suitable for expressing and analyzing population distribution and variation characteristics. The present invention divides the FPAZ suitable for expressing the characteristics of population distribution and change based on the urban morphological elements and the population convergence preference model by considering the urban morphological characteristics and the population convergence mode, which is helpful for the analysis and mining of the population distribution and the spatial-temporal mode of the variation, thereby further supporting Delicate management of urban population.

Description

A population analysis unit division method based on urban morphology and aggregation mode

technical field

The present application belongs to the field of urban geographic intelligent computing, and designs a refined population analysis unit division method based on urban morphology and population aggregation mode.

Background technique

The population analysis unit is the basic spatial unit for the spatio-temporal operation, rule mining and analysis results visualization of urban population analysis. Static population distribution or dynamic population change information is displayed through the population analysis unit, and spatiotemporal analysis is carried out based on the population analysis unit. The shape, size, and spatial continuity of the population analysis unit directly determine the fineness of population information expression and population analysis. result accuracy. Therefore, an appropriate population analysis unit division method is crucial for effectively expressing urban population activity information, analyzing the spatiotemporal patterns of population changes, and supporting urban population-related applications.

In long-term urban population studies, a variety of population analysis units have been used or divided. Based on the expression of population distribution and change information, in order to meet the conditions and application requirements of population analysis unit division, the main division methods are divided into three directions.

Population analysis unit division based on spatial scale requirements. The common macro- and meso-population analysis units in cities are traditional administrative division units, including municipal districts and streets. These spatial units can directly match census data and other types of government statistical data. They are not only standard verification units for population distribution data, but also It provides intuitive scientific basis data feedback for government management and policy planning. For the microscopic scale, buildings are natural microscopic elements in cities. Some researchers use building vector surface elements directly as population analysis units, which can express spatially fine-grained population distribution information or carry out refined population distribution pattern analysis. A population analysis unit division method based on application scenario requirements. Considering the dynamic change and flow of the population, the typical application scenario is urban traffic. Related researchers divide the traffic analysis zone (TAZ) by using the main road in the urban road network, and use it to analyze the city on the basis of conforming to the meso-scale urban form. Traffic characteristics of the population. On this basis, the related data of traffic stations and vehicles are further combined to analyze the interaction mode between urban population and traffic facilities or the operation mode of vehicles. Based on the needs of traffic scenarios, this type of unit expresses the characteristics of population distribution and changes, and provides effective population information for urban traffic planning or optimization. A population analysis unit division method based on research data. Geographic grids are currently the most popular and convenient unit of population analysis divided for matching data quality. The expression of population information at different spatial resolutions is realized by manually establishing regular polygons of different sizes. It can be directly matched with raster data such as land use data and nighttime light data that are popular in urban population research. With the development of new sensors, part of the research centers on individual spatiotemporal marker data and related urban service facility data combined with mathematical theory to divide Thiessen polygons as population analysis units, such as smart card swiping data and mobile phone signaling data. Scientific statistics and spatial representation of population spatiotemporally labeled data.

In general, the current population analysis unit division methods are mainly driven by research data and application requirements, and there is a lack of human unit division methods that consider the expression and analysis accuracy of population distribution and change characteristics. Especially at the micro-scale, the results of the three types of population analysis units are prone to different problems such as differences in urban form, low continuity of spatial units, and insufficient spatial resolution, resulting in inaccurate expression of population information and poor scene versatility. , it cannot support precise and accurate urban population analysis and application of analysis results for a long time.

Therefore, the present invention constructs an FPAZ (Fine Population Analysis Zone, Fine Population Analysis Zone) suitable for characterizing and analyzing population distribution and variation characteristics at a micro scale according to the characteristics of urban population flow and convergence combined with urban morphological elements.

SUMMARY OF THE INVENTION

Aiming at the problem of lack of micro-population analysis units suitable for expressing population distribution and change characteristics, the present invention provides a refined population analysis unit division method based on urban morphology and population aggregation mode. According to the characteristics of urban population flow and convergence combined with urban morphological elements, an FPAZ suitable for characterizing and analyzing population distribution and change characteristics at the micro-scale is constructed.

The present invention provides a refined population analysis unit division method based on urban morphology and population aggregation mode, comprising the following steps:

S1. Taking into account the fine-grained needs of urban spatial structure and population analysis, extracting urban arterial roads, water systems, and other elements that divide and compose urban spatial areas, face the functional heterogeneity of micro-scale population activity areas and fit the urban natural form, based on geometric attributes Extract road and water system polygons with spatial topological features, and divide the remaining morphological elements into basic analysis units through polygonal topology processing;

S2. Considering the spatial distribution characteristics of urban population flow in local convergence and the long-term stability characteristics that depend on the dynamic changes of entrances and exits, construct a population convergence preference model with local entrance and exit elements as the core of spatial division, and use semantic dictionary matching and population flow simulation methods to extract population Aggregate key entry and exit elements;

S3. Using the microstructural elements in the basic analysis unit, the main entrance and exit elements extracted by the comprehensive population aggregation preference model use the spatial clustering method to divide the FPAZ suitable for expressing and analyzing the characteristics of population distribution and change, that is, the fine population analysis area.

Further, the specific implementation manner of extracting road polygons in step S1 is as follows;

S11. Single-level road polygon extraction

In the divided area R, select the main road set Li with the highest grade, _i ={1, 2, 3...r}, the larger the r, the lower the grade; Spatial topology processing constructs a set of spatial units U _i = {ut _i1 , ut _i2 , ut _i3 ...... ut _in }, then obtains the area uta of all units in U _i , and constructs a histogram of area value distribution, according to the mutation of the number of histogram groups Point to determine the minimum area threshold minArea, for any ut _ip (p=[1,2...n]) If the area uta _ip is less than minArea and the cell does not contain other features other than road facility types, the cell is marked as a road polygon; Conversely, if the uta _ip is greater than the threshold minArea, but only contains spatial features related to road facilities, the unit is also marked as a road polygon;

S12, multi-level road polygon extraction

Filter the road polygons on the basis of U _i , and select the main road L _i+1 of the secondary level, and merge them into a spatial unit set U _i+1 ={ut _(i+1)1 , ut _(i +1 ) through the spatial topology operation ₎₂ , ut _(i+1)1 ...... ut _(i+1)m }; and then repeat step S11 on the basis of U _i+1 , until the main roads of all levels are merged, and the road polygon extraction is completed.

Further, in step S2, using semantic dictionary matching and population flow simulation methods to extract the main entrance and exit elements that play a key role in population aggregation is as follows;

S21. Construction of population flow path network

First, considering the geographic characteristics of population flow convergence, select a geographic location inside or outside the divided area R, assuming a virtual population flow starting point Op _j , j is the starting point selected according to the geographic location; At the target location, the main urban road network is used as the main network for population flow; at the same time, only the influence of geographical distance factors on population movement is considered, and the Dijkstra shortest path algorithm is used to construct a population flow path network set EP _j = {ep _j1 ,ep _j2 …ep _jk };

S22. Calculation of similarity of inlet and outlet flow paths

Extract the road intersections of the main road network, express the population flow path network as the ID sequence of the path intersection elements, and then use the difflib algorithm to calculate the path similarity rs simulated by each entrance and exit in the same basic analysis unit, and combine the similarity threshold tv Carry out main entrance and exit marking, and set tv according to the walking sensitive distance value of urban population to infrastructure point elements, combined with the average length of the network point sequence of the population flow path;

S23. Main entrance and exit marking based on path similarity

Calculate the similarity rs of the corresponding flow paths ep _jv and ep _jw for any two entrances and exits in any basic analysis unit, v=[1,2...k], w=[1,2...k], v≠w, If rs is less than the threshold tv, the two entry and exit elements are marked as the main entry and exit; if rs is greater than the threshold tv, the two entry and exit elements are first marked as the same cluster; perform the above operations on each entry and exit element in turn to obtain a cluster set CTR={ ct ₁ , ct ₂ , ct ₃ ...... ct _x }, then for the entrance and exit elements in any cluster ct _g , _g = [1, 2... The geographical distance dis of the corresponding path of each entry and exit element in the distance set DS={dis _g1 , dis _g2 , dis _g3 ……dis _gy }, and the sequence SQ={s _g1 , s _g2 , s _g3 is obtained by sorting the DS ...s _gy }, select the entry and exit elements corresponding to the median value of the sequence SQ to be marked as the main entry and exit.

Further, the difflib algorithm described in step S22 is based on the LCS problem, as shown in formula (1), combined with the idea of dynamic programming, as shown in formula (2), and the improved sequence difference calculation method of the gestalt matching algorithm; formula ( 1), X _m is the sequence X of length m, Y _n is the sequence Y of length n, LCS(X _m , Y _n ) is the longest common subsequence of the sequences X and Y, in formula (2), c[b][d] is used to record the length of the longest common subsequence of sequence X and sequence Y, and b and d are the lengths of sequence X and sequence Y respectively;

Further, the steps of dividing the FPAZ in step S3 are as follows:

1) Extraction of microstructural elements

Extract the microstructural elements in the basic analysis unit, including internal roads and artificial lakes, and combine the line elements with the boundary of the basic analysis unit to convert the line elements into surface elements to obtain a unit set M;

2) Unit classification based on main entrance and exit elements

According to the set of main entrance and exit elements in the basic analysis unit E={e ₁ , e ₂ , e ₃ ...... e _h }, the entrance and exit classification of each unit of M is set as EntryC, and if a unit contains a unique main entrance and exit element, the unit type EntryC is marked is the ID attribute of the element; on the contrary, if a unit contains more than one or no main entrance and exit elements, calculate the Euclidean distance from the centroid of the unit to each main entrance and exit element, and obtain the shortest distance minEntry of the entry and exit element ID attribute to mark the unit class EntryC;

3) FPAZ division

Based on the classification results of the main entrance and exit, spatially fuse the units with the same category EntryC in M, and merge all the fused units to obtain the FPAZ in the basic analysis unit.

Further, the value of tv is 0.95.

The beneficial effects produced by the invention are as follows: an FPAZ suitable for expressing the characteristics of population distribution and change at a microscopic scale is divided by using urban morphological elements and a population aggregation mode. Taking defining the spatial scale and the basic division range of the population analysis unit as the starting point, the basic analysis unit is divided by the urban morphological elements of the large and medium scales, and the entrance and exit elements that have a major impact on the population aggregation change are obtained by the population aggregation preference model. The basic analysis unit is further divided into FPAZ by synthesizing the microstructural elements. This method not only considers the impact of urban morphological characteristics on population distribution and change, but also considers the distribution change characteristics and long-term stability characteristics of urban population flow aggregation itself, which makes FPAZ more suitable for analyzing the spatiotemporal patterns of population distribution and changes.

Description of drawings

FIG. 1 is a technical flow chart of an embodiment of the present invention.

Detailed ways

The technical solutions of the present invention will be further described below with reference to the accompanying drawings and embodiments.

The technical solution and detailed modeling process of the present invention are described below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

As shown in FIG. 1 , the fine-grained population analysis unit division method based on urban morphology and population aggregation mode according to an embodiment of the present invention includes the following steps:

S1. Taking into account the fine-grained needs of urban spatial structure and population analysis, extract the elements that divide and form urban spatial areas such as urban trunk roads and water systems, and face the functional heterogeneity of micro-scale population activity areas to fit the urban natural form, based on geometric attributes Extract road and water system polygons with spatial topological features, and divide the remaining morphological elements into basic analysis units through polygonal topology processing.

S2. Considering the spatial distribution characteristics of urban population flow in local convergence and the long-term stability characteristics that depend on the dynamic changes of entrances and exits, construct a population convergence preference model with local entrance and exit elements as the core of spatial division, and use semantic dictionary matching and population flow simulation methods to extract population Aggregate key entry and exit elements.

S3. Using the microstructural elements in the basic analysis unit, the main entrance and exit elements extracted by the comprehensive population aggregation preference model are divided into FPAZs suitable for expressing and analyzing the characteristics of population distribution and change by using the spatial clustering method.

The basic idea of the present invention is: based on the understanding of urban population distribution and change process, the pattern of population flow and convergence in the city is similar to the characteristics of the formation process of water flow to lakes in nature. The population moves through the multi-level transportation network, driven by different activity behaviors, and finally gathers in local spatial areas through fixed entrances and exits. These gathering areas form irregular spatial units through the restrictions of urban morphological elements such as roads and water networks. Therefore, the present invention takes defining the spatial scale and the basic division range of the population analysis unit as the starting point, uses the urban form elements of large and medium scales to divide the basic analysis unit, and further uses the population aggregation preference model to obtain entrances and exits that have a major impact on population aggregation changes. FPAZ is obtained by further dividing the basic analysis unit by synthesizing the microstructure elements.

Compared with other methods of dividing population analysis units, the key creation point of the present invention is to consider the characteristics of population distribution and change, and also take into account the urban form and fine-grained space requirements, and provide a micro-scale population suitable for analyzing population distribution and change patterns. Analysis unit division method.

In another specific embodiment of the present invention:

A refined population analysis unit division method based on urban morphology and population aggregation mode proposed by the present invention applies this method to a certain city, and the specific steps are as follows:

Step 1. Consider the basic analysis unit division of urban form

The morphological elements in the city also have a direct restrictive effect on dividing urban spatial areas, forming finer-grained spatial units, and changing the flow and aggregation patterns of the population. As a transportation facility for urban population movement and passage, urban roads are the most basic urban natural form elements, and have an important impact on the division of building plots and the formation of urban residents' living boundaries.

1) Extraction of macroscopic and mesoscopic urban morphological elements and construction of spatial topological units

Firstly, the urban form elements are used to define the spatial scope and scale, and the basic analysis units are divided. According to the city-to-road network data and water network data, a total of about 250,000 main road line elements above the county level and about 800 polygonal elements of the main water system were extracted in a city, and a total of about 50,000 spatial topological units were constructed.

2) Multi-level road polygon extraction and filtering

For the micro-scale population activity gathering area, it fits the natural form of the city, and filters multi-level road polygons, water system polygons and green polygons. Considering the geometric attribute characteristics of road polygons, since TAZ is suitable for mesoscopic or macroscopic spatial analysis, the area of road polygons is generally smaller than that of TAZ; considering the spatial topology characteristics of road polygons, roads are basic urban traffic elements, and road polygons only contain road facilities. related spatial elements. Therefore, the filtering steps of the road polygon are as follows:

(1) Single-level road polygon extraction

In the divided area R, select the arterial road set Li with the highest level (the larger _i ={1, 2, 3...r}r, the lower the level), merge the area boundary B, and convert it into the space of area elements through line elements Topological processing constructs a set of space units U _i = {ut _i1 , ut _i2 , ut _i3 ...... ut _in }, where n is the maximum number of space units in the set U _i , and then obtains the area uta of all units in U _i , the construction area The value distribution histogram, the minimum area threshold minArea is determined according to the mutation point of the histogram group number. For any ut _ip (p=[1,2...n]) if the area uta _ip is less than minArea and the cell contains no features other than road facility type, the cell is marked as a road polygon. Conversely, if the uta _ip is greater than the threshold minArea, but only contains spatial features related to road facilities, the cell is also marked as a road polygon. According to the catalogue of urban road facilities, the relevant elements of road facilities are extracted from POI data of traffic and storage types, as shown in Table 1.

Table 1 Traffic and storage type POI data for filtering micro-scale road polygons

(2) Multi-level road polygon extraction

Filter the road polygons on the basis of U _i , and select the main road L _i+1 of the secondary level, and merge them into a spatial unit set U _i+1 ={ut _(i+1)1 , ut _(i +1 ) through the spatial topology operation ₎₂ , ut _(i+1)1 ......ut _(i+1)m }, where m is the maximum number of space units in the set U _i+1 . Then repeat step (1) on the basis of U _i+1 until the main roads of all levels are merged, and the extraction of road polygons is completed. Based on the above steps, about 30,800 road polygons and water system polygons were filtered in a city, and about 19,000 basic analysis units were finally obtained.

Step 2, Population Convergence Preference Model

At the urban micro-scale, the stability of population distribution changes is not only affected by urban morphological factors but also related to population behavior. The distribution and changes of population in local areas are limited by urban form factors and mainly rely on entrances and exits for flow. On this basis, the urban population chooses preferred entrances and exits driven by factors such as the purpose of activities, distance and other factors that affect mobile behavior, and then a long-term stable spatial convergence phenomenon occurs due to the similarity of behaviors, and then a space with stable population distribution changes is formed with the entrances and exits as the core. unit.

1) Definition and extraction of entry and exit elements

The definition of the entrance and exit in the present invention is a fixed channel for the inflow and outflow of the population in a local area. According to the attribute characteristics, the entrance and exit elements that match the attribute description are obtained by constructing a semantic dictionary, wherein the semantic dictionary mainly includes the directory such as orientation, grade, serial number, and entrance and exit description. As shown in table 2.

Table 2. Entry and exit element extraction semantic dictionary

Taking into account the spatial characteristics, the internal roads directly connected to the boundary are used for spatial superposition to obtain the entrance and exit elements that are consistent with the spatial characteristics. According to the semantic dictionary, about 45,000 entrance and exit elements are matched from the POI data of a city; at the same time, the intersection points of the internal road and the boundary of the basic analysis unit are extracted by combining the attributes of "parking lot connecting road" and "POI connecting road" in the road network classification directory, and finally a total of Obtained about 53,000 entrance and exit elements.

2) Construction of population aggregation preference model

Taking into account the preference characteristics of population flow on the choice of entrance and exit, the entrance and exit elements of different levels and spatial locations have differences in population flow and aggregation. Population aggregation mode, main entrances and exits with large population flow and large differences in spatial characteristics from other entrance and exit elements have a significant impact on population aggregation in local areas.

First, taking into account the semantic attribute features of the entry and exit elements, the entry and exit elements are obtained from the matching results with the semantic dictionary, so the main entry and exit elements can be extracted according to the semantic dictionary catalog. The steps are as follows:

(1) Considering the low-flow characteristics of pedestrian entrances and exits, not the main nodes of population flow and local area aggregation, the entrance and exit elements that only allow pedestrians to pass through are filtered according to the "pedestrian" vocabulary in the "category" directory of the semantic dictionary. In a city, about 8,000 entrance and exit elements are filtered by this method, and about 39,000 entrance and exit elements are left.

(2) Considering the location and classification of the entrance and exit elements, according to the similar semantic words such as "main" and "front" in the "orientation and classification" directory of the semantic dictionary, the entrance and exit elements with the main position in the spatial position are extracted, and marked as the main entrance and exit.

(3) Considering the level similarity characteristics of entrance and exit elements, entrance and exit elements with the same name but different serial numbers have similar population preference effects. Therefore, according to the specific serial number in the "number" directory of the semantic dictionary, the entrance and exit elements with the same entrance and exit names are extracted and marked as the main entrance and exit. Based on the above steps, a total of about 160 main entrances and exits were marked in a city.

Secondly, taking into account the shape, area and spatial distribution characteristics of the entrance and exit elements of the basic analysis unit, the main entrance and exit elements are further marked by using the population flow simulation method to mark. The labeling steps based on the population simulation method are as follows:

(1) Construction of population flow path network

First, consider the geographic characteristics of population flow convergence, select a geographic location inside or outside the divided area R, and assume a virtual population flow starting point Op _j (j is the starting point selected according to the geographic location, at most 4 simulation points are needed to complete the simulation , so in this embodiment, j=1, 2, 3, 4) and take all the set of entrance and exit elements in the area as the target location, and take the urban arterial road network as the main network of population flow. At the same time, only considering the influence of geographical distance on population movement, the Dijkstra shortest path algorithm is used to construct a population flow path network set EP _j = {ep _j1 ,ep _j2 ...ep _jk }, where k is the maximum number of flow paths in the set EP _j .

In this example, two traffic stations, the railway station and the airport, located in the east and west directions of a city were selected as the starting point of the population flow simulation, 39,000 entrance and exit elements were used as the end points, and the roads above the county level were used as the simulation network for population flow. The shortest path search algorithm simulates and constructs about 78,000 population flow paths.

(2) Calculation of similarity of inlet and outlet flow paths

Extract the road intersections of the main road network, express the population flow path network as the ID sequence of the path intersection elements, and then use the difflib algorithm to calculate the path similarity rs simulated by each entrance and exit in the same basic analysis unit. The difflib algorithm is based on the LCS problem (as shown in formula (1)), combined with the idea of dynamic programming (as shown in formula (2)) and the improved sequence difference calculation method of the cloze matching algorithm. In formula (1), X _m is the sequence X of length m, Y _n is the sequence Y of length n, LCS(X _m , Y _n ) is the longest common subsequence of the sequences X and Y, and max represents the maximum value. value. In formula (2), c[b][d] is used to record the length of the longest common subsequence of sequence X and sequence Y, and b and d are the lengths of sequence X and sequence Y, respectively.

Then, the main entrance and exit marking is carried out in combination with the similarity threshold tv. In this embodiment, tv is set to 0.95 according to the walking sensitive distance value of the urban population to the infrastructure point elements of 400m and the average length of the network point sequence of the population flow path.

(3) Main entrance and exit marking based on path similarity

Calculate the similarity rs of the corresponding flow paths ep _jv and ep _jw (v=[1,2...k],w=[1,2...k],v≠w) for any two entrances and exits in any basic analysis unit , if rs is less than the threshold tv, the two entrance and exit elements are marked as the main entrance and exit respectively; if rs is greater than the threshold tv, the two entrance and exit elements are first marked as the same cluster. Perform the above operations on each entry and exit element in turn to obtain a cluster set CTR={ct ₁ , ct ₂ , ct ₃ ...... ct _x }, where x is the maximum number of clusters in the set CTR. Then, for the entrance and exit elements in any cluster ct _g (g=[1,2...x]), considering the representative features of the center of the spatial cluster, calculate the geographic distance dis of the corresponding path of each entrance and exit element in the cluster ct _g to obtain the distance. Set DS={dis _g1 , dis _g2 , dis _g3 ……dis _gy }, and obtain sequence SQ={s _g1 , s _g2 , s _g3 …… s _gy } by sorting DS, where y is the distance set DS and The maximum number of sorted sequences SQ, and the entry and exit elements corresponding to the median value of the selected sequence SQ are marked as the main entry and exit. In this embodiment, about 6000 basic analysis units including more than two main entrances and exits are obtained after calculation based on the above steps.

Step 3, FPAZ division based on microscopic elements

Based on the main entrance and exit elements obtained by the population convergence preference model, the basic analysis unit is further divided to obtain FPAZ that is suitable for expressing population distribution and change information at the micro-scale. Considering the influence of the microstructural elements in the basic analysis unit on the spatial extent of population aggregation, the division steps of FPAZ are as follows:

1) Extraction of microstructural elements

Extract the microstructure elements in the basic analysis unit, such as internal roads, artificial lakes, etc., and combine the line elements with the boundary of the basic analysis unit to convert the line elements into surface elements to obtain the unit set M.

2) Unit classification based on main entrance and exit elements

According to the set of main entry and exit elements in the basic analysis unit E={e ₁ , e ₂ , e ₃ ...... e _h }, h is the maximum number of main entry and exit elements in the set E, and the entry and exit classification of each unit of M is set as EntryC, if When a unit contains a unique main entrance and exit element, mark the unit category EntryC as the ID attribute of the element; on the contrary, if a unit contains more than one or no main entrance and exit elements, calculate the Euclidean distance from the centroid of the unit to each main entrance and exit element, The entry and exit element ID attribute of the shortest distance minEntry among them is obtained and the unit type EntryC is marked.

3) FPAZ division

Based on the classification results of the main entrance and exit, spatially fuse the units with the same category EntryC in M, and merge all the fused units to obtain the FPAZ in the basic analysis unit.

In this embodiment, based on the calculation result of a city's population aggregation preference model, 6000 basic analysis units are further divided based on the above steps, and a total of about 39,000 FPAZs are obtained in combination with the original basic analysis units.

Based on the characteristics of urban population flow and convergence, the invention defines the spatial scale and the spatial division range with urban macro-meso form elements, filters multi-level road polygons for micro-scale requirements, and constructs a basic analysis unit. The population convergence preference model is used to mark the entrance and exit elements that have a major impact on population convergence, and combined with the microstructural elements in the basic analysis unit to further divide the FPAZ that is suitable for expressing the characteristics of population distribution and change. The research on the division method of the micro-scale population analysis unit in the present invention is helpful for the analysis and mining of the population distribution and the changing spatiotemporal pattern, thereby further supporting the refined management of the urban population.

The specific embodiments described herein are merely illustrative of the spirit of the invention. Those skilled in the art to which the present invention pertains can make various modifications or additions to the described specific embodiments or substitute in similar manners, but will not deviate from the spirit of the present invention or go beyond the definitions of the appended claims range.

Claims (5)

1. a population analysis unit division method based on urban form and convergence mode, is characterized in that, comprises the following steps: S1. According to the fine-grained needs of urban spatial structure and population analysis, extract the elements that segment and compose urban spatial areas, including urban trunk roads and water systems, oriented to the functional heterogeneity of micro-scale population activity areas and fit the urban natural form, based on geometric Attributes and spatial topological features extract road and water polygons, and divide the remaining morphological elements into basic analysis units through polygonal topology processing; S2. Considering the spatial distribution characteristics of urban population flow in local convergence and the long-term stability characteristics that depend on the dynamic changes of entrances and exits, construct a population convergence preference model with local entrance and exit elements as the core of spatial division, and use semantic dictionary matching and population flow simulation methods to extract population Aggregate key entry and exit elements; In step S2, using semantic dictionary matching and population flow simulation methods to extract the main entrance and exit elements that play a key role in population aggregation is as follows; S21. Construction of population flow path network First, considering the geographic characteristics of population flow convergence, select a geographic location inside or outside the divided area R, assuming a virtual population flow starting point Op _j , j is the starting point selected according to the geographic location; At the target location, the main urban road network is used as the main network for population flow; at the same time, only the influence of geographical distance factors on population movement is considered, and the Dijkstra shortest path algorithm is used to construct a population flow path network set EP _j = {ep _j1 ,ep _j2 …ep _jk }, where k is the maximum number of flow paths in set EP _j ; S22. Calculation of similarity of inlet and outlet flow paths Extract the road intersections of the main road network, express the population flow path network as the ID sequence of the path intersection elements, and then use the difflib algorithm to calculate the path similarity rs simulated by each entrance and exit in the same basic analysis unit, and combine the similarity threshold tv Carry out main entrance and exit marking, and set tv according to the walking sensitive distance value of urban population to infrastructure point elements, combined with the average length of the network point sequence of the population flow path; S23. Main entrance and exit marking based on path similarity Calculate the similarity rs of the corresponding flow paths ep _jv and ep _jw for any two entrances and exits in any basic analysis unit, v=[1,2...k],w=[1,2...k], v≠w, If rs is less than the threshold tv, the two entry and exit elements are marked as the main entry and exit; if rs is greater than the threshold tv, the two entry and exit elements are first marked as the same cluster; perform the above operations on each entry and exit element in turn to obtain a cluster set CTR={ ct ₁ , ct ₂ , ct ₃ ...... ct _x }, where x is the maximum number of clusters in the set CTR; then for the entry and exit elements in any cluster ct _g , g = [1,2... The center has a representative feature. Calculate the geographic distance dis of the corresponding path of each entry and exit element in the cluster ct _g to obtain the distance set DS={dis _g1 , dis _g2 , dis _g3 ......dis _gy }, and obtain the sequence by sorting the DS SQ={s _g1 , s _g2 , s _g3 ...... s _gy }, where y is the maximum number of the distance set DS and the sorting sequence SQ, and the entrance and exit elements corresponding to the median value of the sequence SQ are selected to be marked as the main entrance and exit; S3. Using the microstructural elements in the basic analysis unit, the main entrance and exit elements extracted by the comprehensive population aggregation preference model use the spatial clustering method to divide the FPAZ suitable for expressing and analyzing the characteristics of population distribution and change, that is, the fine population analysis area. 2. a kind of population analysis unit division method based on urban form and convergence mode as claimed in claim 1, is characterized in that: the concrete implementation mode of extracting road polygon in step S1 is as follows; S11. Single-level road polygon extraction In the divided area R, select the main road set Li with the highest grade, _i ={1, 2, 3...r}, the larger the r, the lower the grade; Spatial topology processing constructs a set of space units U _i = {ut _i1 , ut _i2 , ut _i3 ...... ut _in }, where n is the maximum number of space units in the set U _i , and then obtains the area uta of all units in U _i , and constructs The area value distribution histogram, the minimum area threshold minArea is determined according to the mutation point of the histogram group number, for any ut _ip , p=[1,2...n], if the area uta _ip is less than minArea and the unit does not contain other than road facility types If the uta _ip is greater than the threshold minArea, but only contains spatial features related to road facilities, the unit is also marked as a road polygon; S12, multi-level road polygon extraction Filter the road polygons on the basis of U _i , and select the main road L _i+1 of the secondary level, and merge them into a spatial unit set U _i+1 ={ut _(i+1)1 , ut _(i +1 ) through the spatial topology operation ₎₂ , ut _(i+1)1 ......ut _(i+1)m }, where m is the maximum number of space units in the set U _i+1 ; then repeat step S11 on the basis of U _i+1 , until Merge all levels of arterial roads to complete the extraction of road polygons. 3. a kind of population analysis unit division method based on urban form and convergence mode as claimed in claim 1, is characterized in that: the difflib algorithm described in step S22 is based on LCS problem, as shown in formula (1), combined with dynamic The planning idea, as shown in formula (2), and the improved sequence difference calculation method of the cloze matching algorithm; in formula (1), X _m is the sequence X of length m, Y _n is the sequence Y of length n, LCS(X _m , Y _n ) is the longest common subsequence of sequences X and Y, and max means taking the maximum value. In formula (2), c[b][d] is used to record the longest common subsequence of sequence X and sequence Y The length of the common subsequence, b and d are the lengths of sequence X and sequence Y respectively;

4. a kind of population analysis unit dividing method based on urban form and convergence pattern as claimed in claim 1 is characterized in that: the dividing step of FPAZ in step S3 is as follows; 1) Extraction of microstructural elements Extract the microstructural elements in the basic analysis unit, including internal roads and artificial lakes, and combine the line elements with the boundary of the basic analysis unit to convert the line elements into surface elements to obtain a unit set M; 2) Unit classification based on main entrance and exit elements According to the set of main entry and exit elements in the basic analysis unit E={e ₁ , e ₂ , e ₃ ...... e _h }, h is the maximum number of main entry and exit elements in the set E, and the entry and exit classification of each unit of M is set as EntryC, if When a unit contains a unique main entrance and exit element, mark the unit category EntryC as the ID attribute of the element; on the contrary, if a unit contains more than one or no main entrance and exit elements, calculate the Euclidean distance from the centroid of the unit to each main entrance and exit element, Obtain the entry and exit element ID attribute of the shortest distance minEntry and mark the unit category EntryC; 3) FPAZ division Based on the classification results of the main entrance and exit, spatial fusion is performed on the units with the same category EntryC in M, and all fused units are merged to obtain the FPAZ in the basic analysis unit. 5 . The method for dividing population analysis units based on urban morphology and aggregation mode according to claim 1 , wherein the value of tv is 0.95. 6 .

Images