A method for clustering rural networks

Abstract

The present application relates to a kind of village network clustering method, comprising the following steps: mobile phone signaling data acquisition, check and preliminary processing;Based on mobile phone signaling data to build village overall network;Network clustering calculation, obtain village community structure.The present application is by using mobile phone signaling to identify mobile phone user's full purpose travel and commuting travel data across administrative village represent the flow connection between village, then reference the analysis method of complex network to build village network, then use Louvain community discovery algorithm to carry out clustering calculation, finally obtain several communities and the network structure inside community, for the spatial organization and development positioning optimization of our country village settlement, village system, it provides an efficient method for the optimization of land space resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of network clustering, in particular to a village and town network clustering method. BACKGROUND

[0002] With the continuous advancement of urbanization, the relationship between rural areas and cities becomes close, and gradually participates in the process of urban industrial division and production specialization; the functional differentiation between different village and town settlements and the blurring of the village and town system hierarchy are becoming more and more obvious, and the traditional spatial organization logic based on the central place theory is no longer completely applicable to village and town settlements. At present, the analysis technology for village and town settlements is relatively blank, and the small sample size, difficulty in obtaining and poor timeliness of traditional data lead to the fact that village and town planning often lacks scientific data support, so there is unreasonable division of villages and towns, and therefore there are problems such as mismatch of spatial resources, imbalance of public infrastructure and service supply and demand. SUMMARY

[0003] Therefore, the present application provides a village and town network clustering method, which effectively solves the problem that scientific and accurate data support cannot be provided for new village and town system planning, and provides a new technical method for village and town planning, village and town spatial layout planning and construction.

[0004] In order to achieve the above-mentioned purpose, the present application provides a village and town network clustering method, which comprises the following steps:

[0005] S1, obtaining, checking and preliminary processing of mobile phone signaling data;

[0006] S2, constructing a village and town overall network based on mobile phone signaling data;

[0007] S3, network clustering calculation to obtain village and town community structure.

[0008] Preferably, the obtaining, checking and preliminary processing of mobile phone signaling data comprises the following steps:

[0009] S101, dividing the obtained mobile phone signaling data into two types of data, i.e. full-purpose travel and commuting travel, according to the location and duration of mobile phone user stay;

[0010] S102, the bottom map data adopts a shp format file with administrative villages as basic spatial units, and generally selects the township area of the city and county and excludes the county town and urban area;

[0011] S103, the population identified by the signal base station is divided into the population of the administrative village where the spatial position of the signal base station is located, and the obtained mobile phone signaling data is preliminarily processed by the PostgreSQL database management system and the ArcGIS platform to obtain the data of population travel connection between administrative villages;

[0012] S104, regression analysis is performed on the statistical permanent population number of each county, township and street within the research scope and the residence population number obtained through the mobile phone signaling, and the goodness of fit R 2 The higher the R, the more accurate and reliable the mobile phone signaling data is; the mobile phone signaling population identification rate of each county, township and street is obtained by dividing the residence population number obtained through the mobile phone signaling within the research scope by the statistical permanent population number of each county, township and street; if the identification rate is too low, the accuracy and authenticity of the final calculation result will be affected.

[0013] Preferably, the construction of the overall network of villages and towns based on the mobile phone signaling data comprises the following steps:

[0014] S201, the center point of an administrative village is taken as a node of the network, and the population travel connection between administrative villages is taken as an edge; specifically, the frequency of all-purpose travel is taken as the weight of the edge, and the number of commuting travelers is taken as the weight of the edge, and the number of two-way flow is accumulated respectively, thereby establishing two undirected networks, and the formula is as follows:

[0015] W ij =C ij +C ji

[0016] Wherein, i and j are nodes, W ij is the weight of the edge between node i and node j, C ij is the number of travelers from node i to node j, and C ji is the number of travelers from node j to node i;

[0017] S202, the weight of the edge is assigned to the connection between the village and town points on the ArcGIS platform, so as to realize the visualization of the village and town network.

[0018] Preferably, the network clustering calculation adopts the Louvain(BGLL) algorithm, and the Louvain(BGLL) algorithm comprises the following steps:

[0019] S301, the processed village and town network data is substituted into the calculation process of the Louvain algorithm network clustering of the igraph package;

[0020] S302, each vertex is first regarded as an independent community, and each vertex is traversed in a certain order, and the change of the modularity of the vertex m moving to its neighbor vertex n is△Q; if△Q>0, the vertex m is moved to the administrative village of the vertex with the largest change; if△Q<0, the vertex m remains unchanged; the process is repeated until the movement of each vertex cannot increase the modularity, and a new community structure after the first iteration is obtained;

[0021] S303, regarding each community obtained in step S302 as a new independent vertex, repeating the process of step S302 to start a new round of iteration; after each iteration, the communities generated in the previous iteration are merged, the number of communities is reduced, and the iteration is ended until the nodes of the whole graph no longer change and the modularity no longer increases;

[0022] S304, visualizing the community structure and the network structure inside the community obtained by clustering on the ArcGIS platform.

[0023] The beneficial effects of the present application are:

[0024] (1) The present application uses the full-purpose travel and commuting travel data of mobile phone users across administrative villages recognized by mobile phone signaling to represent the flow connection between villages and towns, compared with other traditional data, the mobile phone signaling data acquisition method is relatively simple, wide coverage, large sample quantity and real-time sustainable, and the huge database provides accurate data for clustering calculation.

[0025] (2) The present application adopts Louvain(BGLL) algorithm, which is a heuristic algorithm based on modularity optimization. The algorithm has two layers of iteration. The outer iteration is a bottom-up condensation method, and the inner iteration is a condensation method plus a switching strategy, which avoids the shortcoming that two nodes cannot be separated once they are merged in a simple condensation method.

[0026] (3) The analysis result of the network clustering of the present application is relatively reliable, can clearly and intuitively reflect the spatial structure characteristics of the village and town system, is conducive to supporting the development planning and spatial optimization of the village and town system under the background of national spatial planning, and provides favorable help for the practice of public infrastructure planning such as rural community life circle. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 The flowchart of the present application. DETAILED DESCRIPTION

[0028] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific embodiments, structures, features and effects according to the present application are described in detail as follows in combination with the drawings and preferred embodiments.

[0029] The present embodiment discloses a village and town network clustering method, comprising the following steps:

[0030] S1, obtaining, verifying and preliminarily processing the mobile phone signaling data;

[0031] S2, constructing the overall network of villages and towns based on the mobile phone signaling data;

[0032] S3, network clustering calculation, obtaining the community structure of villages and towns.

[0033] The mobile phone user cross-administrative village full-purpose travel and commuting travel data of mobile phone signaling recognition represent the human flow connection between villages. The acquisition, verification and preliminary processing of mobile phone signaling data include the following steps:

[0034] S101, according to the location and duration of the mobile phone user stay, the acquired mobile phone signaling data is divided into full-purpose commuting and commuting travel data;

[0035] S102, the base map data adopts a shp format file with administrative village as the basic spatial unit, generally selects the township area of the city and county and excludes the county and city area;

[0036] S103, the population identified by the signal base station is divided into the population of the administrative village where the spatial position of the signal base station is located. On this basis, the acquired mobile phone signaling data is preliminarily processed by the PostgreSQL database management system and the ArcGIS platform to obtain the population travel connection data between administrative villages;

[0037] S104, regression analysis is performed on the statistical permanent population number of each district, county, township and street in the research range and the resident population number obtained through mobile phone signaling, and the goodness of fit R 2 is higher, indicating that the mobile phone signaling data is more accurate and reliable. The mobile phone signaling population recognition rate of each district, county, township and street is obtained by dividing the statistical permanent population number of each district, county, township and street in the research range by the resident population number obtained through mobile phone signaling. If the recognition rate is too low, it will affect the accuracy and authenticity of the final calculation result. There is no base station distribution in some administrative villages, which almost does not exist in China and can be ignored or appropriately merged before calculation.

[0038] The construction of the overall network of villages and towns based on mobile phone signaling data includes the following steps:

[0039] S201, taking the center point of the administrative village as the node of the network and the population travel connection between administrative villages as the edge; specifically, first taking the frequency of full-purpose travel as the weight of the edge, and then taking the number of commuting travel as the weight of the edge, respectively accumulating the number of two-way flow, thereby establishing two undirected networks, the formula is as follows:

[0040] W ij ij = C ij ij + C ji

[0041] Where i and j are nodes, W ij is the weight of the edge between node i and node j, C ij is the number of travel from node i to node j, and C ji is the number of travel from node j to node i;

[0042] S202, assign the weight of the edge to the connection between the village and town points on the ArcGIS platform to realize the visualization of the village and town network.

[0043] The network clustering calculation adopts the Louvain (BGLL) algorithm, which comprises the following steps:

[0044] S301, substitute the processed village and town network data into the igraph package and the calculation process of the network clustering of the Louvain algorithm;

[0045] The network data must be processed to be free of multiple edges and cyclic edges, wherein the multiple edges refer to two or more edges with the same vertex, for example, there are more than one edge with the same vertex in the data The cyclic edges refer to two or more edges with opposite vertices, for example, there are two edges with opposite vertices in the data and If both the multiple edges and the cyclic edges exist, the weight values should be combined in advance.

[0046] For the same set of data, if the order of the rows of the data is changed, the difference in the iteration process may occur due to the change in the order of the starting points of the clustering, but the change has little effect on the final iteration result.

[0047] S302, regard each vertex as an independent community, traverse each vertex in a certain order, and the change of the modularity of the vertex m moving to the neighbor vertex n is △Q; if △Q>0, the vertex m moves to the community of the vertex with the largest change; if △Q<0, the vertex m remains unchanged; repeat the process until the movement of each vertex cannot increase the modularity, and obtain the new community structure after the first iteration;

[0048] The increase of the modularity is to increase the edges in the sub-community and reduce the edges outside, that is, to be more aggregated.

[0049] S303, regard each community obtained in step S302 as a new independent vertex, repeat the process of step S302 to start a new round of iteration; after each iteration, the communities generated in the previous iteration are combined, the number of communities is reduced, and the iteration is ended until the nodes of the whole graph no longer change and the modularity no longer increases.

[0050] Each iteration will get several different, new community structure, the iteration is more and more, the community will be merged, so the number of communities obtained is less, the nodes contained in the community are more, the modularity of the village network clustering based on the commuting travel connection after the first iteration is more than 0.7, and the modularity after the second iteration is more than 0.85, which is significantly improved compared with the first iteration, indicating that the community has high aggregation degree, the third iteration relative to the second iteration modularity is relatively stable at this time, indicating that the community structure formed is relatively stable, and the result is more reliable.

[0051] S304, the clustering obtained community structure, the network structure in the community is visualized on the ArcGIS platform.

[0052] The village network clustering technology is mainly applied to the fields of village planning, village spatial layout planning and construction; the application range of the village network clustering technology can be determined according to the population scale of administrative villages / communities, the economic and social development level, the natural geographical conditions and the distribution of base stations and the quality of mobile phone data acquisition, and can cover the entire township range of a city or county, and the research space unit is at the level of administrative villages / communities; in the application process, the overall and local structure characteristics of the village network from the perspective of human flow are considered, the spatial organization and development positioning optimization of the village settlement and the village system are realized, and the optimization configuration of land space resources is realized.

[0053] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content without departing from the scope of the technical solution of the present application, and any brief introduction, modification, equivalent change and modification of the above embodiments according to the technical essence of the present application are still within the scope of the technical solution of the present application.

Claims

1. A method of clustering rural networks, characterized by, It comprises the following steps: S1, obtaining, checking and preliminary processing of mobile phone signaling data; S101, dividing the obtained mobile phone signaling data into two categories of data of full-purpose travel and commuting travel according to the places and time lengths of mobile phone user stays; S102, using a shp format file with administrative villages as basic spatial units for bottom map data, selecting town areas of a city domain and a county domain and excluding parts of county towns and city areas; S103, dividing the population identified by signal base stations into the population of administrative villages where the spatial positions of the signal base stations are located, and preliminarily processing the obtained mobile phone signaling data through a PostgreSQL database management system and an ArcGIS platform to obtain data of population travel connections between administrative villages; S104, regression analysis is performed on the statistical permanent population number of each district, county, township and street within the research scope and the residence population number obtained through mobile phone signaling, and the goodness of fit R 2 The higher the value is, the more accurate and reliable the mobile phone signaling data is. The mobile phone signaling population identification rate of each district, county, township and street is obtained by dividing the residence population number obtained through mobile phone signaling within the research scope by the statistical permanent population number of each district, county, township and street. If the identification rate is too low, the accuracy and authenticity of the final calculation result will be affected. S2, constructing a whole network of villages and towns based on mobile phone signaling data; S3, network clustering calculation to obtain a community structure of villages and towns.

2. The method of claim 1, wherein, The step of constructing a whole network of villages and towns based on mobile phone signaling data comprises the following steps: S201, taking the center points of administrative villages as nodes of the network and taking population travel connections between administrative villages as edges; specifically, taking the frequencies of full-purpose travel as weights of the edges, and taking the numbers of commuting travel as weights of the edges, respectively, adding the numbers of two-way flows to obtain two undirected networks, and the formula is as follows: W ij = C ij + C ji where i and j are nodes, W ij is the weight of the edge between node i and node j, C ij is the number of trips from node i to node j, C ji is the number of trips from node j to node i; S202, assigning weights of edges to lines between village and town points on an ArcGIS platform to realize visualization of the network of villages and towns.

3. The method of claim 1, wherein, The network clustering calculation adopts a Louvain(BGLL) algorithm, and the Louvain(BGLL) algorithm comprises the following steps: S301, substituting the processed network data of villages and towns into a calculation process of Louvain algorithm network clustering of an igraph package; S302, first taking each vertex as an independent community, traversing each vertex in a certain order, and moving vertex m to the vertex n with the largest change in modularity; if △Q>0, moving vertex m to the community of the vertex with the largest change; if △Q<0, keeping vertex m unchanged; repeating the process until the movement of each vertex cannot increase the modularity, and obtaining a new community structure after the first iteration; S303, taking each community obtained in step S302 as a new independent vertex, and repeating the process of step S302 to start a new round of iteration; after each iteration, the communities generated in the previous iteration are combined, and the number of communities is reduced until the nodes of the whole graph no longer change and the modularity no longer increases, and the iteration is ended; S304, visualizing the community structure obtained by clustering and the network structure inside the community on an ArcGIS platform.

Images