A community life circle identification method and device based on resident travel behavior

Abstract

The application relates to the technical field of urban intelligent planning, and provides a community life circle identification method and device based on resident travel behaviors, which comprises the following steps: based on obtained preset regional grid data and built-up area range residential land data, extracting residential land grids and determining grid sharing indexes between the residential land grids; using a preset clustering algorithm to perform clustering analysis on the grid sharing indexes between the residential land grids, obtaining a plurality of life circle grid groups sharing activity distribution; for each life circle grid group, according to the number of life circle surface domains in the grid group and the distance between the surface domains, splitting the life circle grid group to obtain an initial community life circle; matching and fusing the initial community life circle with urban plots to obtain a final community life circle. The life circle boundary generated by the application is more in line with the actual activity space distribution characteristics of residents, and the spatial precision and dynamic adaptability of the life circle are improved.

Description

Technical Field

[0001] This invention relates to the field of urban intelligent planning technology, and in particular to a method and apparatus for identifying community living circles based on residents' travel behavior. Background Technology

[0002] As urban development shifts from incremental expansion to optimizing existing resources, meeting residents' daily needs and improving their quality of life have become crucial issues. Community living circle identification technology, as a planning method that comprehensively considers multiple scales and residents' needs, analyzes residents' travel behavior to achieve a more natural and coherent way of delineating community living circles, providing strong technical support for urban planning.

[0003] Existing community living circle planning techniques largely rely on static data on current facilities (such as points of interest) or land use. Although some methods attempt to incorporate dynamic resident activity logs, they still struggle to overcome the inherent limitations of urban geography, resulting in imprecise delineation of community living circle boundaries. For example, existing community-based methods delineate living circles by setting a service radius for the central community, failing to fully consider the spatial distribution characteristics of residents' actual activities. Furthermore, the living circle boundaries strictly adhere to community boundaries, ignoring heterogeneous phenomena within communities. Another example is the simulation measurement method for community living circles, which uses planned spatial plots as a base and constructs gravity models using resident behavioral data to delineate community living circles. However, within the plot-based analytical framework, it still suffers from insufficient spatial granularity and poor flexibility. These shortcomings in dynamic adaptability and spatial accuracy directly impact the accuracy and practicality of community living circle planning. Summary of the Invention

[0004] This invention provides a method and apparatus for identifying community living circles based on residents' travel behavior, addressing the shortcomings of existing community living circle planning technologies in terms of dynamic adaptability and spatial accuracy. It enables the scientific delineation of community living circle boundaries and resolves the problem of inconsistent community living circle delineation. The technical solution proposed by this invention is as follows: In a first aspect, the present invention provides a method for identifying community living circles based on residents' travel behavior, comprising: Based on the acquired pre-defined regional grid data and residential land data within the built-up area, residential land grids are extracted; The grid sharing index between residential land grids is determined based on residents' near-home travel behavior data on residential land grids; wherein, the residents' near-home travel behavior data refers to travel records of living needs within a preset range around the residents' residences; A pre-defined clustering algorithm was used to perform cluster analysis on the grid sharing index among the residential land grids, resulting in multiple living circle grid clusters with shared activity distributions. Traverse each grid in the said living circle grid group, correct the living circle category of the grid according to the living circle category of the adjacent grids, and merge the grids to obtain multiple living circle regions; For each community grid cluster, the community grid cluster is split into initial community communities based on the number of community areas in the community grid cluster and the distance between the community areas. The initial community living circle is matched and integrated with urban plots to obtain the final community living circle.

[0005] Optionally, the step of extracting residential land grids based on the acquired preset area grid data and residential land data within the built-up area includes: Acquire grid data for the preset area and residential land data within the built-up area; The residential land grid is obtained by intersecting the preset area grid data with the residential land data within the built-up area.

[0006] Optionally, the data on residents' near-home travel behavior on the residential land grid is obtained through the following methods: Acquire resident travel behavior data, identify the residence of each sample based on the resident travel behavior data, and filter data whose residence is located within the residential land grid to obtain initial activity data; The initial activity data is filtered based on preset filtering conditions. Non-commuting stay records outside the home are filtered out, and the travel records of residents' living needs within a preset range around the residents' residences are retained to obtain the residents' near-home travel behavior data on the residential land grid.

[0007] Optionally, the residents' home-bound travel behavior data on the residential land grid includes the distance between two residential land grids, the number of times each residential land grid accesses each shared grid, and the number of shared grids; Based on residents' near-home travel behavior data within residential land grids, a grid sharing index is determined between residential land grids, including: The grid sharing index between residential land grids is determined based on the distance between two residential land grids, the number of times each residential land grid accesses each shared grid, and the number of shared grids; where a shared grid is a grid that is accessed by both residential land grids.

[0008] Optionally, the step of using a preset clustering algorithm to perform cluster analysis on the grid sharing index among various residential land grids yields multiple living circle grid clusters with shared activity distributions, including: The Leiden algorithm was used to perform cluster analysis on the shared index of each grid to obtain multiple living circles; By removing living circles whose population is less than a preset population threshold or whose area is less than a preset area threshold, the living circle grid clusters with shared activity distribution are obtained.

[0009] Optionally, the step of traversing each grid in the living circle grid cluster, correcting the living circle category of the grid according to the living circle categories of its neighboring grids, and merging the grids to obtain multiple living circle regions, including: Traverse each grid in the said living circle grid group, take that grid as the current grid, and perform the operation of updating the living circle category of the current grid to obtain the multiple living circle areas; wherein, the method of updating the living circle category of the current grid is as follows: Obtain the living circle category of the current grid, and the living circle categories of the first preset number of neighboring grids located in the current grid; If the living circle category of the first preset number of adjacent grids is the same and different from the living circle category of the current grid, then the living circle category of the current grid is updated to the living circle category of the adjacent grids; If the living circle categories of the first preset number of adjacent grids are different, then obtain the living circle categories of the second preset number of adjacent grids of the current grid; Sort the occurrence frequency of the living circle categories of the second preset number of adjacent grids. If the first two living circle categories have the same occurrence frequency, and one of the adjacent grids has the same living circle category as the current grid, then keep the current living circle category of the current grid unchanged; otherwise, update the living circle category of the current grid to the living circle category with the highest occurrence frequency. The second preset quantity is greater than the first preset quantity.

[0010] Secondly, the present invention also provides a community living circle identification device based on residents' travel behavior, comprising the following modules: The extraction module is used to extract residential land grids based on the acquired preset area grid data and residential land data within the built-up area. The calculation module is used to determine the grid sharing index between residential land grids based on residents' near-home travel behavior data on the residential land grid; wherein, the residents' near-home travel behavior data is the travel records of living needs within a preset range around the residents' residences; The clustering module is used to perform clustering analysis on the grid sharing index between various residential land grids using a preset clustering algorithm, and to obtain multiple living circle grid clusters with shared activity distribution; The traversal module is used to traverse each grid in the living circle grid group, correct the living circle category of the grid according to the living circle category of the adjacent grids, and merge the grids to obtain multiple living circle regions; The splitting module is used to split each living circle grid group into an initial community living circle based on the number of living circle areas in the living circle grid group and the distance between the living circle areas. The matching module is used to match and merge the initial community living circle with urban plots to obtain the final community living circle.

[0011] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the community living circle identification method based on residents' travel behavior as described in the first aspect above.

[0012] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the community living circle identification method based on residents' travel behavior as described in the first aspect above.

[0013] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the community living circle identification method based on residents' travel behavior as described in the first aspect above.

[0014] Based on the above technical solution, the beneficial effects of the present invention compared with the prior art are as follows: This invention provides a method and apparatus for identifying community living circles based on residents' travel behavior. By utilizing residents' near-home travel behavior data, it reflects residents' actual living needs and activity patterns, making the delineation of community living circles more closely aligned with residents' real lives, thereby improving the dynamic adaptability of planning. By acquiring pre-defined regional grid data and residential land data within the built-up area, the study area is divided into multiple grids, achieving refined data management. This gridding process helps to more accurately capture the spatial distribution characteristics of residents' activities, improving the spatial accuracy of community living circle delineation. Based on activity data and residents' near-home travel behavior data from each residential land grid, a grid sharing index is calculated between residential land grids. This index reflects the degree of correlation between residents' activities in different grids, providing an important basis for subsequent cluster analysis and further improving the accuracy of spatial delineation. A pre-defined clustering algorithm is used to perform cluster analysis on the grid sharing index, generating multiple living circle grid clusters with shared activity distributions. Then, by traversing each grid and correcting and merging them according to the living circle categories of adjacent grids, multiple coherent living circle areas are obtained. This process effectively solves the problem of inconsistent community living circle delineation, ensuring that residents' activities within each living circle have high correlation and consistency. After obtaining the initial community living circles, this method further optimizes them based on the spatial distance between the areas of each living circle within the same circle. This helps ensure that each living circle is more spatially compact and rational. Finally, by matching the initial community living circles with urban plots, the final community living circle division results are obtained. The boundaries of the living circles are adjusted to match the shape of the plots cut by roads, achieving urban morphological adaptation of the boundaries. This effectively solves the shortcomings of existing community living circle planning technologies in terms of dynamic adaptability and spatial accuracy, and achieves the scientific delineation of community living circle boundaries and resolves the problem of inconsistent delineation.

[0015] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the community living circle identification method based on residents' travel behavior provided by the present invention.

[0019] Figure 2 This is a comparative diagram showing the living circle before and after the modification provided by this invention.

[0020] Figure 3 This is a schematic diagram of the community living circle identification device based on residents' travel behavior provided by the present invention.

[0021] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0023] This invention presents a community living circle identification method based on residents' travel behavior. Based on quantified data (i.e., grid sharing index) and a defined network clustering algorithm, it performs cluster analysis on residents' near-home travel behavior data to scientifically delineate living circles. Based on the tightly arranged grid structure, it iterative calculations for each grid solve the problem of interface discontinuity in community living circle delineation, achieving organic integration of living circles. (Refer to...) Figure 1 As shown, the method includes the following: S110. Based on the acquired preset area grid data and residential land data within the built-up area, extract the residential land grid.

[0024] In the community living circle identification method based on residents' travel behavior, the core input data consists of pre-set regional grid data, residential land data within the built-up area, and residents' travel behavior data, each carrying different spatial and semantic information. Their detailed descriptions are as follows: Preset regional grid data refers to dividing the study area (such as a city or urban area) into regular geographic grid units (such as 100m×100m or 200m×200m square grids), with each grid serving as the smallest unit of analysis. The data format is a vector grid (such as GeoJSON or Shapefile). Preset regional grid data provides the basic unit for spatial analysis, replacing traditional zoning (such as streets or communities) and avoiding interference from traditional zoning in the analysis of residents' activities. For example, if studying a certain urban area, it can be divided into 2,000,000 100m×100m grids, with each grid assigned a unique ID (such as G001, G002).

[0025] Residential land data within built-up areas refers to land parcels or grids within urban built-up areas primarily used for residential purposes. This data includes land use codes (e.g., R1 represents a type of residential land) to limit the analysis scope, excluding non-built-up areas (e.g., farmland, mountains) and non-residential areas (e.g., industrial land, water areas). It identifies core residential areas, ensuring that the living circle is based on residents' actual residences. For example, all "residential land" (e.g., R1, R2, R3 types) are extracted from urban land use data and mapped onto preset grids. If more than 50% of the area within a grid is residential land, it is marked as a residential land grid.

[0026] S120. Determine the grid sharing index between residential land grids based on residents' near-home travel behavior data on residential land grids.

[0027] Resident travel behavior data refers to spatiotemporal data recording residents' daily travel activities, reflecting their movement trajectories from their residences to various activity locations (such as work, shopping, and leisure). Travel data should distinguish between weekdays and weekends, and daytime and nighttime, to capture the different lifestyle patterns at different times. Resident travel behavior data includes origin (O), destination (D), and travel frequency or duration. Origin (O) and destination (D) are represented by residential grid IDs or coordinates, while travel frequency or duration refers to, for example, the number of times a week a resident visits a particular shopping mall and the length of time spent there. Resident travel behavior data is used to quantitatively calculate the activity correlation between grids (such as the grid sharing index). It reveals the actual range of facilities and services used by residents, rather than the theoretical service radius. Resident travel behavior data can be mobile phone signaling data or location-based service (LBS) data. Mobile phone signaling data is used to extract users' commuting trajectories from their residences to their workplaces through base station positioning. Mobile phone signaling data originates from anonymized mobile location data provided by operators (which requires de-identification processing).

[0028] The aforementioned resident travel behavior data refers to travel records within a pre-defined area surrounding residents' residences to meet their daily needs. First, resident residences are identified based on anonymized mobile phone signaling data: the location where users spend the most time at night over a three-month period is used as their residence. Simultaneously, travel data within a pre-defined area (e.g., 2 kilometers) around these residences is collected for subsequent analysis. In the data processing stage, residential land use data within the built-up area is used to filter out eligible residential land grids. For each residential land grid, residents' travel behavior characteristics are statistically analyzed, including frequently visited activity grids, visit frequency, and dwell time. This travel behavior data is mapped to the corresponding "residential land grid - activity grid" relationship, ultimately constructing a sharing matrix reflecting residents' activity patterns. This matrix is ​​used for calculating the grid sharing index and performing cluster analysis in subsequent living circle identification.

[0029] Grid sharing indices (such as the Jaccard similarity coefficient) measure the degree of overlap in the activities of residents in two residential grids. The higher the index, the greater the overlap in the activity spaces of the residents in the two grids, indicating that they may belong to the same living area.

[0030] S130. Use a preset clustering algorithm to perform cluster analysis on the grid sharing index among the residential land grids to obtain multiple living circle grid clusters with shared activity distribution.

[0031] This step generates a preliminary classification of living areas. The aforementioned living area grid clusters are sets of grids with shared activity distribution characteristics identified from residential land grids through a pre-defined clustering algorithm based on residents' travel behavior.

[0032] A pre-defined clustering algorithm was used to perform cluster analysis on the grid sharing index among residential land grids. Based on the magnitude of the grid sharing index, the algorithm grouped residential land grids with similar sharing characteristics into one category, thus obtaining multiple residential circle grid clusters with shared activity distributions. The grids within these clusters showed high similarity and correlation in residents' travel behaviors, indicating that residents' activities between these grids were frequent and close.

[0033] Within a residential grid cluster, residents' travel activities exhibit significant sharing characteristics across different residential grids. Residents move frequently between these grids, sharing similar commercial, leisure, and service facilities. For example, a residential grid cluster may contain multiple residential grids, and residents of these grids often frequent a large shopping mall within the cluster, indicating that the shopping mall is a shared commercial facility for residents within the cluster.

[0034] The aforementioned pre-defined clustering algorithm can be a network clustering algorithm such as the Louvain algorithm or the Leiden algorithm. Taking the Leiden algorithm as an example, residential land grids are used as nodes, and grid sharing indices are used as edge weights to construct a weighted graph. Through iterative optimization (partitioning, aggregation, and refinement), the grids are clustered into communities with the highest modularity (i.e., living circle grid clusters). It can adapt to nonlinear relationships and can identify multi-center, low-overlapping community structures.

[0035] S140. Traverse each grid in the said living circle grid group, correct the living circle category of the grid according to the living circle category of the adjacent grids, and merge the grids to obtain multiple living circle areas.

[0036] A preliminary classification of habitats has been generated, but issues exist where habitats contain grids from neighboring habitats, and where grid categories are mixed at the boundaries of different habitats. Specifically, this manifests as jagged or illogical boundary divisions. To resolve this issue, corrections are needed. The correction method is as follows: A systematic traversal method is adopted, such as visiting each grid within a community grid group one by one according to the row and column order. For example, one can start from the grid in the upper left corner of the group and traverse in order from left to right and from top to bottom to ensure that no grid is missed. During the traversal, an identifier is created for each grid to record its current community category information for subsequent correction and operation.

[0037] For each grid being traversed, its neighboring grids are identified, such as the four adjacent grids above, below, left, and right. Based on the grid's living circle category and the living circle categories of its neighboring grids, a decision is made as to whether to correct the grid's living circle category and whether to merge it, thereby reassigning the grid to suitable neighboring living circles. This smooths living circle boundaries and improves geographical continuity. The process of traversing each grid in a living circle grid cluster continues, correcting the category and performing merging operations as described above, until all grids within the cluster that meet the merging conditions have been merged. This ultimately results in multiple living circle areas, each representing an area with similar characteristics in residents' travel behaviors and shared activities. These areas can serve as the basic units for subsequent community living circle segmentation and planning. GIS software can be used to visualize the living circle areas, allowing for a direct observation of their spatial distribution and boundary extent.

[0038] S150. For each living circle grid cluster, the living circle grid cluster is split into an initial community living circle based on the number of living circle areas in the living circle grid cluster and the distance between the living circle areas.

[0039] If a living circle grid cluster contains only one living circle region, it is not split. If it contains multiple living circle regions, the straight-line distance between the centroids (the average of the coordinates of all points within the region) of two living circle regions is calculated using the Euclidean distance formula in a Cartesian coordinate system.

[0040] For all living circle areas within a living circle grid cluster, the distance between each pair is calculated. Based on the actual characteristics of the living circles and residents' travel habits, a reasonable distance threshold is set. When the distance between two living circle areas exceeds this threshold, they are considered spatially independent and are thus split into two independent living circles to determine the final initial community living circle split. The area information contained in each community living circle is recorded, including area number and geographical location. The setting of the distance threshold can refer to factors such as urban living circle planning standards, the coverage radius of service facilities, and spatial continuity, combined with actual surveys and urban spatial scale for comprehensive judgment.

[0041] S160. Match and integrate the initial community living circle with the urban plots to obtain the final community living circle.

[0042] The spatial overlap is obtained by calculating the area of ​​the intersection between the initial community living circle and the urban plot and dividing it by the area of ​​the plot.

[0043] Matching is based on spatial overlap, with a fixed overlap threshold set according to actual needs and experience. For example, an overlap threshold of 50% can be set; if an urban plot overlaps with an initial community living circle by more than 50%, the plot is considered a successful match. The overlap threshold is dynamically adjusted to consider the characteristics of different areas. For instance, in urban centers, where plots are more finely divided and functions are more complex, the threshold can be appropriately lowered (e.g., 30%); while in suburban areas, where plots are larger and functions are relatively simpler, the threshold can be increased (e.g., 60%).

[0044] The calculated spatial overlap data is linked with the attribute data of the initial community living areas and urban plots to form a comprehensive data table containing information such as living area number, plot number, overlap degree, plot use, and plot area. Based on the set matching rules, the comprehensive data table is filtered to find living area-plot combinations that meet the matching conditions. The matching results are visualized on a map, using different colors or symbols to represent successfully matched initial community living areas and urban plots. The boundaries of the successfully matched urban plots are merged with the boundaries of the initial community living areas to obtain the final boundary of the community living area. For example, in GIS software, the "merge" tool can be used to merge multiple polygons into one polygon, forming a new living area boundary. This adjusts the living area boundary to match the shape of the plots cut by roads, achieving urban morphological adaptation of the boundary.

[0045] This invention provides a community living circle identification method based on residents' travel behavior, effectively addressing the shortcomings of existing community living circle planning technologies in terms of dynamic adaptability and spatial accuracy. It also achieves the scientific delineation of community living circle boundaries and resolves the problem of inconsistent delineation. Specifically, this invention utilizes residents' travel behavior data, particularly their near-home travel behavior data (i.e., travel records of living needs within a preset range around residents' residences). This data reflects residents' actual living needs and activity patterns, making the delineation of community living circles more closely aligned with residents' real lives, thereby improving the dynamic adaptability of the planning. As residents' travel behavior changes, this method can adjust the delineation of community living circles accordingly, ensuring that the planning remains consistent with actual needs. By acquiring preset area grid data and residential land data within the built-up area, the study area is divided into multiple grids, achieving refined data management. This grid-based processing helps to more accurately capture the spatial distribution characteristics of residents' activities, improving the spatial accuracy of community living circle delineation. Based on the activity data and near-home travel behavior data of each residential land grid, a grid sharing index between residential land grids is calculated. This index reflects the degree of correlation between residents' activities across different grids, providing an important basis for subsequent cluster analysis and further improving the accuracy of spatial division. A preset clustering algorithm is used to perform cluster analysis on the shared index of each grid, generating multiple community circle grid clusters with shared activity distributions. Then, by traversing each grid and correcting and merging them according to the community circle categories of adjacent grids, multiple coherent community circle areas are obtained. This process effectively solves the problem of inconsistent community circle delineation, ensuring high correlation and consistency of residents' activities within each community circle. After obtaining the corrected community circles (i.e., the initial community community circles mentioned above), the method further optimizes them based on the spatial distance between the areas of each community circle within the same community circle. This helps ensure that each community circle is more compact and rational in space. Finally, by calculating the spatial overlap between the initial community community circles and urban plots, and matching the initial community community circles with urban plots, the final community community circle division result is obtained. The community circle boundaries are adjusted to match the shape of the plots cut by roads, achieving urban morphological adaptation of the boundaries. This step ensures that the delineation of community living areas is in line with the overall layout of urban planning, thereby improving the feasibility and practicality of the plan.

[0046] In an optional embodiment, the extraction of residential land grids based on the acquired preset area grid data and built-up area residential land data described in S110 above includes: S1101. Obtain the preset area grid data and the residential land data within the built-up area.

[0047] The pre-defined regional grid data serves as the foundation for spatial division, defining the spatial framework of the study area. The residential land data within the built-up area defines the specific location and extent of residential land within the built-up area.

[0048] S1102. Intersect the preset area grid data with the residential land data within the built-up area to obtain the residential land grid. The purpose of this step is to limit the research scope to residential land within the built-up area to eliminate data interference from irrelevant areas.

[0049] This invention, based on multi-source data fusion and filtering, extracts residents' near-home travel behavior data from each residential land grid within the built-up area to support urban planning and community living circle division. The residents' near-home travel behavior data from the residential land grids mentioned in S120 above is obtained through the following methods: S1201. Obtain resident travel behavior data, identify the residence of each sample based on the resident travel behavior data, and filter the data whose residence is located within the residential land grid to obtain initial activity data.

[0050] Residents' travel behavior data records key information such as residents' travel trajectories, travel times, and travel purposes, and is an important basis for analyzing residents' activity patterns.

[0051] Based on anonymized resident travel behavior data, the residence of each sample was identified by accumulating more than 11 hours of nighttime stay per month and the longest stay location between 8 pm and 6 am. Then, data was filtered to include only residential land grids within the built-up area, resulting in the initial activity data mentioned above. Data samples relevant to the study area were then selected from the massive amount of travel behavior data.

[0052] The residential location identification algorithm employs a multi-stage processing flow: First, it calculates the cumulative dwell time of each user at each base station during the nighttime period (20:00-6:00), and selects the base station where the user spends the longest as a candidate residential location; then, it applies the DBSCAN spatial clustering algorithm, setting a neighborhood radius of 300 meters to eliminate positioning errors, and finally determines the user's actual residential location. This algorithm can effectively handle problems such as signal drift and improve positioning accuracy.

[0053] S1202. Based on preset filtering conditions, the initial activity data is filtered to filter out non-commuting stay records outside the home and retain the travel records of living needs within a preset range around the residents' residences to obtain the residents' near-home travel behavior data on the residential land grid.

[0054] The above-mentioned preset filtering conditions include the first filtering condition and the second filtering condition.

[0055] First, the initial activity data is filtered once based on the first filtering condition to remove noisy data or data that does not meet the research requirements, resulting in filtered activity data.

[0056] The first preset filtering conditions include, but are not limited to, abnormal travel time and excessively long travel distance. The filtering process is as follows: The basic information of each data sample in the initial activity data is checked for stability and completeness. For example, it checks whether basic information such as travel records is missing or abnormal, filtering data samples with unstable basic information. A province information inclusion check is performed to ensure that each data sample contains province information. Samples lacking province information are considered invalid and filtered out. The travel records of each data sample are analyzed to calculate the number of travel days per month. Samples with fewer than a specified number of travel days per month (e.g., 4 days) are considered to have insufficient travel behavior to reflect daily activity patterns and are therefore considered invalid samples and filtered out.

[0057] Filtering out invalid samples with unstable basic information, lacking provincial information, or with too few travel days helps reduce the interference of noisy data on subsequent analysis, improving the accuracy and reliability of the results. Through screening and filtering, subsequent analysis can focus on valid samples that truly reflect residents' daily activity patterns and travel behaviors, thereby improving the relevance and practicality of the research.

[0058] Then, the activity data after the first filtering is further filtered based on the second filtering condition to obtain activity data for each residential land grid within the built-up area. The second filtering condition can filter out non-commuting stay records outside the home, retaining travel records for daily needs within a preset range around the residents' residences. Activity data closely related to residents' daily life needs is extracted to support subsequent analysis and research.

[0059] Commuting behavior recognition uses an improved TF-IDF algorithm, which analyzes the regular travel characteristics during weekday morning and evening rush hours. Specifically, travel records departing between 7:00-9:00 and returning between 17:00-19:00 on weekdays are excluded if the frequency is more than 3 times per week and continues for more than 4 weeks.

[0060] This invention effectively removes noisy and non-research-requirement data through multi-round data filtering, improving the accuracy and reliability of the data. Limiting the research scope to residential land within built-up areas avoids data interference from irrelevant regions, making the research results more focused and targeted. By extracting activity data from each residential land grid within the built-up area, it is possible to deeply analyze residents' daily travel behavior, activity spatial distribution, and other characteristics, providing a scientific basis for urban planning and community living circle division.

[0061] Taking a city as an example within the aforementioned preset area, the process of extracting residents' near-home travel behavior data from each residential land grid within the built-up area is explained: 1. Obtain grid data for a city, which forms the basic spatial division of the city. Simultaneously, obtain residential land data within the built-up area, which details the specific location and boundaries of residential areas.

[0062] 2. Perform a spatial intersection operation between the grid data and the residential land data to identify grids that are wholly or partially located within residential land. The result of this step is a set of grids that only contains residential land within the built-up area.

[0063] 3. Mobile phone signaling data contains a large amount of residents' daily travel trajectory information, but the raw data often contains noise and errors. The mobile phone signaling data needs to be cleaned, including removing duplicate data and handling missing values. Ping-pong error (frequent switching between adjacent base stations due to signal fluctuations) needs to be corrected; this can be reduced through smoothing algorithms or time window analysis. Data drift (significant deviation between the positioning result and the actual location) needs to be identified and corrected; location data can be calibrated by combining geospatial information and base station distribution. Out-of-city records need to be verified and filtered to ensure that the data reflects residents' activities within a specific city, thus obtaining the aforementioned data on residents' near-home travel behavior.

[0064] 4. Based on anonymized resident near-home travel behavior data, analyze the nighttime stay of each sample. Locations where residents spend more than 11 hours cumulatively each month at night (8 PM to 6 AM) are generally considered to be residents' primary residences. Among multiple such locations, the location with the longest stay is selected as the residence for each sample. The identified residences are compared with a previously obtained grid containing only residential land within built-up areas. Only data samples whose residences fall within these grids are retained, resulting in the initial activity data described above.

[0065] 5. Further filter the retained samples, removing those with unstable basic information (such as frequent changes in contact information or address). Ensure that each sample includes province information (in this case, a specific city), and filter samples lacking province information. Analyze the number of travel days for each sample, filtering out samples with fewer than 4 travel days per month, as these samples may not accurately reflect residents' daily travel patterns, resulting in filtered activity data.

[0066] 6. Conduct in-depth analysis of the activity data after the first filtering, including travel time, travel distance, and travel purpose. Identify non-commuting out-of-home records, which are usually unrelated to residents' daily commuting behavior and may be records of leisure, entertainment, shopping, or other activities. Filter out non-commuting out-of-home records, retaining only travel records within a two-kilometer radius of residents' residences. These records are more likely to reflect travel undertaken by residents to meet daily needs (such as shopping, medical treatment, education, etc.), ultimately obtaining residents' near-home travel behavior data for each residential grid.

[0067] In an optional embodiment, this invention analyzes residents' near-home travel behavior data to assess the degree of overlap in activity ranges between two residential land grids, measuring the closeness of interaction between residential areas, and proposes a sharing index, specifically defined as the grid sharing index. The aforementioned near-home travel behavior data for residential land grids includes the distance between two residential land grids, the number of times each residential land grid visits each shared grid, and the number of shared grids. A shared grid refers to a grid jointly visited by two residential land grids, and the number of visits reflects the degree of interaction between residents' activities between the two residential land grids.

[0068] The determination of the grid sharing index between residential land grids based on residents' near-home travel behavior data on residential land grids, as described in S120 above, includes: The grid sharing index between residential land grids is determined based on the distance between two residential land grids, the number of times each residential land grid accesses each shared grid, and the number of shared grids.

[0069] The calculation of the grid sharing index requires comprehensive consideration of the distance between two residential land grids, the number of times each residential land grid visits a shared grid, and the number of shared grids. Specifically, a calculation formula can be designed to incorporate these three factors to reflect the degree of sharing between two residential land grids. For example, a weighting coefficient can be set, assigning different weights based on the distance between the two residential land grids; the closer the distance, the greater the weight, indicating closer interaction between the two residential land grids. Simultaneously, the number of times each residential land grid visits a shared grid also needs to be considered; a higher number indicates more frequent sharing activities between the two residential land grids. A shared grid is a grid that is jointly visited by both residential land grids. Finally, the number of shared grids is also an important consideration; a larger number indicates a wider shared activity area between the two residential land grids. Through the above calculation process, a specific grid sharing index value can be obtained, which is used to quantify the degree of sharing between two residential land grids. The higher the index value, the more frequent the sharing activities and the higher the degree of interaction between the two residential land grids.

[0070] The grid sharing index is defined as:

[0071] In the formula, Two residential land grids The grid sharing index, This refers to the number of grids that are shared between two residential land grids, i.e., the number of the aforementioned shared grids. Residential land grid Visit the The number of times a shared grid is used. Residential land grid Visit the The number of times a shared grid is used. for The distance between two residential land grids.

[0072] The more grids shared between two residential land grids, the higher the frequency of joint access, and the closer the distance, the higher the grid sharing index.

[0073] The grid sharing index of this invention provides a quantitative assessment tool that intuitively presents the degree of sharing among residential land grids. Through the calculation and analysis of the grid sharing index, cluster analysis of residents' near-home travel behavior data can be performed based on quantitative data, scientifically delineating living circles and more accurately defining the scope of community living circles.

[0074] In an optional embodiment, the Leiden algorithm is an improvement on the Louvain algorithm used to discover network structures with tight internal connections and sparse external connections, avoiding the creation of arbitrarily poorly connected communities. Its basic principle is to optimize network modularity by introducing multi-level partitioning and node movement strategies to ensure that the final clustering result maintains high internal connectivity while effectively reducing cross-community interference. Compared to traditional clustering methods, the Leiden algorithm can more accurately identify highly interactive areas in urban space, avoiding fragmented or isolated living circle units. This invention uses the Leiden algorithm to perform clustering analysis on the sharing index of each grid (the result of step S120). The above-mentioned step S130, which uses a preset clustering algorithm to perform clustering analysis on the grid sharing index between residential land grids, yields multiple living circle grid clusters with shared activity distributions, including: S1301. Use the Leiden algorithm to perform cluster analysis on the shared index of each grid to obtain multiple living circles.

[0075] In this process, each residential land grid is considered a node, and the sharing index between grids is considered the connection weight between nodes, constructing a weighted graph. The Leiden algorithm is applied to perform cluster analysis on the weighted graph, and the community structure in the network is identified by optimizing modularity. Modularity is an indicator of the quality of community structure, reflecting the tightness of connections between nodes within a community and the sparsity of connections between nodes between communities. The process of applying the Leiden algorithm to perform cluster analysis on the weighted graph is as follows: First, based on the grid sharing index Construct a symmetric sharing degree matrix, where the matrix elements represent the similarity between residential land grids. If Higher indicates residential land grid and If they have high similarity, they should be grouped into the same cluster.

[0076] Next, initialization is performed: the parameters of the Leiden algorithm are determined, such as the resolution parameter (which affects the coarseness of clustering), the maximum number of iterations, and the convergence criteria. Initial partitioning is performed, treating each grid as an independent cluster (i.e., initially, each grid is a potential living space).

[0077] Next, the Leyton algorithm is used for iteration: Calculate the modularity gain: For each residential grid, calculate the modularity gain when moving it to the cluster of its neighboring grids based on the sharing degree matrix mentioned above. This gain is used to evaluate the tightness of connections within the community and is generally taken as [0.3, 0.7]. By comparing the modularity gains of different movement methods, the optimal movement strategy can be determined.

[0078] Modularity The calculation formula is:

[0079] in, and These are residential land grids. and The degree (i.e., the number of grids connected to them). It represents the total number of edges in the weighted graph. It is an indicator function, when the residential land grid... and The value is 1 if the elements belong to the same cluster, and 0 otherwise. These represent residential land grids. and Clustering categories.

[0080] Before and after each node move, calculate the modularity before the move using the formula described above. and the degree of modularity after movement Calculate the modularity after the move. Modularity before the move The difference yields the modularity gain. .

[0081]

[0082] Optimal move selection: Based on the calculated modularity gain, select the grid move that maximizes the overall modularity. If multiple moves have the same maximum gain, one of them can be selected (e.g., based on randomness or other heuristics).

[0083] Update partitions: Perform the selected grid move to update the cluster partitions. After each move, recalculate the modularity of the affected clusters.

[0084] Check for convergence: Check if the convergence conditions are met (e.g., the change in modularity is less than a certain threshold, or the maximum number of iterations has been reached). If not, return to the step of calculating the modularity gain; if so, proceed to the next step.

[0085] Once the algorithm converges, the resulting clustering partitions represent the identified living areas. Because the Leiden algorithm performs clustering analysis based solely on the graph topology constructed using the sharing index and does not introduce geographical adjacency constraints, enclaves may occur spatially. To enhance the spatial coherence of living areas, subsequent spatial continuity checks and necessary post-processing of the clustering results are required.

[0086] S1302. Eliminate living circles with fewer than a preset population threshold or smaller than a preset area threshold to obtain living circle grid clusters with shared activity distribution. This ensures the rationality and practicality of living circles and avoids overly scattered or small-scale living circles from interfering with subsequent analysis.

[0087] This invention, based on the Leiden algorithm, generates more reasonable clustering results by considering the weights between nodes and optimizing modularity. This helps to reveal more potential subgroup structures, making the division of life zones more refined and accurate. The Leiden algorithm ensures that all life zone grid clusters are internally connected, meaning that the grids within each life zone grid cluster are interconnected through certain connections. This characteristic makes the life zone structure more stable, which is beneficial for subsequent analysis and applications.

[0088] This invention uses the Leiden algorithm to perform cluster analysis on the sharing index of each grid, and sets the maximum area of ​​the cluster to 50 grids (equivalent to a living circle with a radius of 1km). Areas with fewer than 100 residents or less than 10 grids are removed, resulting in multiple living circle grid clusters, with each living circle averaging approximately a circle with a radius of 1km. Within these living circle grid clusters, residents have similar home-based travel options, leading to more frequent sharing of facilities.

[0089] In an optional embodiment, S130 above generates a preliminary living circle classification, but there are cases where living circles contain neighboring living circle grids and where grid categories are mixed at the boundaries of different living circles. This invention further refines the living circle classification through a consistency and statistical correction mechanism based on adjacent grids. S140 above describes traversing each grid in the living circle grid cluster, correcting the living circle category of that grid according to the living circle categories of its neighboring grids, and merging the grids to obtain multiple living circle regions, including: Traverse each grid in the said living circle grid group, take that grid as the current grid, and perform the operation of updating the living circle category of the current grid to obtain the multiple living circle areas; wherein, the method of updating the living circle category of the current grid is as follows: Obtain the living circle category of the current grid, and the living circle categories of the first preset number of neighboring grids located in the current grid; If the living circle category of the first preset number of adjacent grids is the same and different from the living circle category of the current grid, then the living circle category of the current grid is updated to the living circle category of the adjacent grids; If the living circle categories of the first preset number of adjacent grids are different, then obtain the living circle categories of the second preset number of adjacent grids of the current grid; Sort the occurrence frequency of the living circle categories of the second preset number of adjacent grids. If the first two living circle categories have the same occurrence frequency, and one of the adjacent grids has the same living circle category as the current grid, then keep the current living circle category of the current grid unchanged; otherwise, update the living circle category of the current grid to the living circle category with the highest occurrence frequency. The second preset quantity is greater than the first preset quantity.

[0090] This invention determines whether the habitat category of a first preset number of adjacent grids is consistent with that of the current grid, and compares them. If the adjacent grids have the same category but are different from the current grid, a correction is made. This mechanism ensures that when adjacent areas have similar characteristics or functions, the current grid can be correctly classified into the corresponding habitat, thereby optimizing the overall coherence of habitat boundaries. When the habitat categories of the first preset number of adjacent grids are inconsistent, a second preset number of adjacent grids (and the second preset number is greater than the first preset number) is further considered. By counting the occurrences of these adjacent grid habitat categories, and based on the sorting results and specific rules (such as the first two categories having the same occurrence count and one of them being the same as the current grid category), a correction is made. This mechanism can handle more complex situations, such as when there may be a mixture of multiple habitat categories in the boundary area. Through statistics and sorting, the habitat category that best matches the characteristics of the current grid can be found.

[0091] In the delineation of living zones, boundary areas often experience category conflicts or ambiguities. The aforementioned correction process ensures that the living zone category of the boundary grid remains consistent with its adjacent grids, thereby reducing category conflicts and enhancing the coherence of the delineation. By correcting the living zone category of the boundary grid, functional synergy and integration between adjacent areas can be promoted, improving the overall efficiency and effectiveness of the region.

[0092] This correction process considers various scenarios, including whether adjacent grids have the same or different categories, and whether they appear the same or different times, thus handling diverse and complex situations. Clear rules and procedures ensure the algorithm's stability and robustness, reducing the risk of errors or anomalous results. The process allows adjustment of the first and second preset quantities based on actual conditions to accommodate different scales and complexities of habitat partitioning. Flexible parameter settings enable fine-grained partitioning and correction of habitats with different regions and characteristics, improving the algorithm's adaptability and flexibility.

[0093] Figure 2 This is a comparative diagram showing the distribution of living circles before and after the adjustment. The left diagram shows the distribution of living circles before the adjustment, and the right diagram shows the distribution of living circles after the adjustment. Figure 2 Different colors represent different living area categories, and white grids represent undivided blank areas. The above correction process is illustrated using examples of four and eight preset quantities, respectively: Iterate through each grid element and check its living circle category in each of the four main directions (up, down, left, right). If the living circle category is the same in all four directions but different from the current grid's living circle category, update the current grid's living circle category to match the category in all four directions; refer to... Figure 2As shown, taking a blue grid element in the green area as an example, its living circle category is the same in all four directions, all being green, and different from the living circle category of the blue grid element. Therefore, the blue grid element is updated to the same category in all four directions, i.e., green. If the living circle categories in the four directions are different, the frequency of occurrence of the living circle categories in the eight directions (up, down, left, right, upper left, upper right, lower left, lower right) is sorted. If the top two living circle categories in the sort have the same frequency, and one of them is equal to the living circle category of the current grid, the current living circle category remains unchanged; otherwise, the current grid's living circle category is updated to the category with the highest frequency to correct for mixed situations and optimize the continuity of living circle boundaries. Taking a blue grid element in the yellow area as an example, its living circle categories are different in all four directions. The frequency of occurrence of the living circle categories in the eight directions is sorted. The top two living circle categories in the sort are yellow and blue, respectively. Since yellow and blue have different frequency of occurrence, the current grid's living circle category is updated to the category with the highest frequency, i.e., yellow. By comparing the left and right images, it can be seen that the correction process effectively eliminated scattered spatial noise and filled boundary gaps, making the originally jagged and fragmented boundaries of the living area more continuous and smooth.

[0094] In an optional embodiment, after generating corrected living circles through preliminary clustering, the present invention optimizes the living circles by comprehensively considering spatial distance and the distribution of non-construction land, generating community living circle division results. The above-described S150, for each living circle grid cluster, splits the living circle grid cluster to obtain an initial community living circle based on the number of living circle areas in the living circle grid cluster and the distance between the living circle areas, including: S1501. Calculate the distance between two living circle areas in the same living circle grid group. When the distance exceeds the preset distance threshold, divide the two living circle areas into two independent living circles.

[0095] For all residential land grids within the same corrected living circle, calculate the Euclidean distance between each pair to determine the maximum spatial span of the living circle (i.e., the distance between the two furthest grids).

[0096] If the maximum spatial distance exceeds the preset distance threshold (e.g., 5 kilometers, which can be adjusted according to the city size), it indicates that the living circle may contain spatially discontinuous resident activity units and needs to be split into two independent living circles.

[0097] Example: If a modified living circle covers two residential areas, A and B, and the furthest grid distance between area A and area B is 5.6 kilometers (exceeding the preset distance threshold), then the two living circle areas will be divided into two independent living circles.

[0098] By synthesizing the results of the decomposition, a final, continuous, and behaviorally consistent community living circle boundary is generated, resulting in the initial community living circle. This ensures that the decomposition results are both spatially reasonable and reflect the actual activity patterns of residents.

[0099] This invention calculates the distance between two living circle areas within the same living circle grid cluster and automatically splits them when the distance exceeds a preset threshold, ensuring the geographical continuity and service accessibility of living circles and achieving refined division of community living circles.

[0100] The community living circle identification device based on residents' travel behavior provided by the present invention will be described below. The community living circle identification device based on residents' travel behavior described below can be referred to in correspondence with the community living circle identification method based on residents' travel behavior described above.

[0101] The community living circle identification device based on residents' travel behavior provided by this invention refers to... Figure 3 As shown, it includes: Extraction module 210 is used to extract residential land grids based on the acquired preset area grid data and residential land data within the built-up area; The calculation module 220 is used to determine the grid sharing index between residential land grids based on residents' near-home travel behavior data on the residential land grid; wherein, the residents' near-home travel behavior data is the travel records of living needs within a preset range around the residents' residences; Clustering module 230 is used to perform cluster analysis on the grid sharing index between each residential land grid using a preset clustering algorithm to obtain multiple living circle grid clusters with shared activity distribution; Traversal module 240 is used to traverse each grid in the living circle grid group, correct the living circle category of the grid according to the living circle category of the adjacent grids, and merge the grids to obtain multiple living circle areas; The splitting module 250 is used to split each living circle grid group into an initial community living circle based on the number of living circle areas in the living circle grid group and the distance between the living circle areas. Matching module 260 is used to match and merge the initial community living circle with urban plots to obtain the final community living circle.

[0102] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4As shown, the electronic device may include a processor 310, a communications interface 320, a memory 330, and a communication bus 340. The processor 310, communications interface 320, and memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions from the memory 330 to execute a community living circle identification method based on residents' travel behavior.

[0103] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0104] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the community living circle identification method based on residents' travel behavior provided by the above methods.

[0105] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the community living circle identification method based on residents' travel behavior provided by the above methods.

[0106] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0107] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying community living circles based on residents' travel behavior, characterized in that, include: Based on the acquired pre-defined regional grid data and residential land data within the built-up area, residential land grids are extracted; The grid sharing index between residential land grids is determined based on residents' near-home travel behavior data on residential land grids; wherein, the residents' near-home travel behavior data refers to travel records of living needs within a preset range around the residents' residences; A pre-defined clustering algorithm was used to perform cluster analysis on the grid sharing index among the residential land grids, resulting in multiple living circle grid clusters with shared activity distributions. Traverse each grid in the said living circle grid group, correct the living circle category of the grid according to the living circle category of the adjacent grids, and merge the grids to obtain multiple living circle regions; For each community grid cluster, the community grid cluster is split into initial community communities based on the number of community areas in the community grid cluster and the distance between the community areas. The initial community living circle is matched and integrated with urban plots to obtain the final community living circle.

2. The method for identifying community living circles based on residents' travel behavior according to claim 1, characterized in that, The process of extracting residential land grids based on the acquired preset area grid data and residential land data within the built-up area includes: Acquire grid data for the preset area and residential land data within the built-up area; The residential land grid is obtained by intersecting the preset area grid data with the residential land data within the built-up area.

3. The method for identifying community living circles based on residents' travel behavior according to claim 1, characterized in that, Data on residents' near-home travel behavior in residential land grids is obtained through the following methods: Acquire resident travel behavior data, identify the residence of each sample based on the resident travel behavior data, and filter data whose residence is located within the residential land grid to obtain initial activity data; The initial activity data is filtered based on preset filtering conditions. Non-commuting stay records outside the home are filtered out, and the travel records of residents' living needs within a preset range around the residents' residences are retained to obtain the residents' near-home travel behavior data on the residential land grid.

4. The method for identifying community living circles based on residents' travel behavior according to claim 1, characterized in that, The data on residents' near-home travel behavior on the residential land grid includes the distance between two residential land grids, the number of times each residential land grid accesses each shared grid, and the number of shared grids; Based on residents' near-home travel behavior data within residential land grids, a grid sharing index is determined between residential land grids, including: The grid sharing index between residential land grids is determined based on the distance between two residential land grids, the number of times each residential land grid accesses each shared grid, and the number of shared grids; where a shared grid is a grid that is accessed by both residential land grids.

5. The method for identifying community living circles based on residents' travel behavior according to claim 1, characterized in that, The method uses a preset clustering algorithm to perform clustering analysis on the grid sharing index among residential land grids, resulting in multiple living circle grid clusters with shared activity distributions, including: The Leiden algorithm was used to perform cluster analysis on the shared index of each grid to obtain multiple living circles; By removing living circles whose population is less than a preset population threshold or whose area is less than a preset area threshold, the living circle grid clusters with shared activity distribution are obtained.

6. The method for identifying community living circles based on residents' travel behavior according to claim 1, characterized in that, The process involves traversing each grid in the living area grid cluster, correcting the living area category of that grid based on the living area categories of its neighboring grids, and then merging the grids to obtain multiple living area regions, including: Traverse each grid in the said living circle grid group, take that grid as the current grid, and perform the operation of updating the living circle category of the current grid to obtain the multiple living circle areas; wherein, the method of updating the living circle category of the current grid is as follows: Obtain the living circle category of the current grid, and the living circle categories of the first preset number of neighboring grids located in the current grid; If the living circle category of the first preset number of adjacent grids is the same and different from the living circle category of the current grid, then the living circle category of the current grid is updated to the living circle category of the adjacent grids; If the living circle categories of the first preset number of adjacent grids are different, then obtain the living circle categories of the second preset number of adjacent grids of the current grid; Sort the occurrence frequency of the living circle categories of the second preset number of adjacent grids. If the first two living circle categories have the same occurrence frequency, and one of the adjacent grids has the same living circle category as the current grid, then keep the current living circle category of the current grid unchanged; otherwise, update the living circle category of the current grid to the living circle category with the highest occurrence frequency. The second preset quantity is greater than the first preset quantity.

7. A community living circle identification device based on residents' travel behavior, characterized in that, include: The extraction module is used to extract residential land grids based on the acquired preset area grid data and residential land data within the built-up area. The calculation module is used to determine the grid sharing index between residential land grids based on residents' near-home travel behavior data on the residential land grid; wherein, the residents' near-home travel behavior data is the travel records of living needs within a preset range around the residents' residences; The clustering module is used to perform clustering analysis on the grid sharing index between various residential land grids using a preset clustering algorithm, and to obtain multiple living circle grid clusters with shared activity distribution; The traversal module is used to traverse each grid in the living circle grid group, correct the living circle category of the grid according to the living circle category of the adjacent grids, and merge the grids to obtain multiple living circle regions; The splitting module is used to split each living circle grid group into an initial community living circle based on the number of living circle areas in the living circle grid group and the distance between the living circle areas. The matching module is used to match and merge the initial community living circle with urban plots to obtain the final community living circle.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the community living circle identification method based on residents' travel behavior as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the community living circle identification method based on residents' travel behavior as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the community living circle identification method based on residents' travel behavior as described in any one of claims 1 to 6.

Images