User mobility clustering analysis method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses a user mobility clustering analysis method, which comprises the following steps: acquiring communication signaling data of a user and operator MR data, the communication signaling data of the user comprising a user mobile phone number, a base station cell code and access time of a signaling event; the service provider MR data comprises a service cell, an adjacent cell, a time slice and received signal strength of the user; determining flow position data of the user in a preset area according to the communication signaling data of the user and the MR data of an operator; rasterizing the flow position data to obtain rasterized flow position data; and carrying out clustering processing on the rasterized flow position data to obtain a high-density flow area.

Description

Technical Field

[0001] The present application belongs to the field of Internet technology, and in particular, relates to a user mobility clustering analysis method, device, electronic device and storage medium. Background Art

[0002] With the rapid advancement of urbanization in modern society, the increasing density and enhancement of population mobility have become an important feature and trend of modern society. This high-density population mobility not only reflects the level of urbanization development, but also has a profound impact on social economy, transportation system, resource allocation, etc. Therefore, in-depth exploration and control of the laws and characteristics of population mobility are of great significance for the optimal allocation of national and regional human resources, balanced social and economic development, and optimization of transportation systems.

[0003] In the context of high-density population mobility, the precise division of high-density mobility areas has become an important basis for population mobility analysis. The precise division of high-density mobility areas can ensure the control of the trajectory and characteristics of foreign population mobility, and provide scientific data basis for emergency management of social emergencies and allocation of social resources.

[0004] At present, traditional approaches to high-density flow areas mainly include static manual methods such as prior knowledge, manual division or demographics. Manual division relies on the professional knowledge, historical past and personal experience of relevant persons in charge in terms of geographical environment positioning, social hot spots, commercial activities, etc. Although the cost is low, it cannot actively, in real time and dynamically grasp the changes in population flow and hot spots. Therefore, it has the disadvantages of lack of objectivity, quantifiability, timeliness and adaptability. Demographic methods, such as GPS positioning and traffic flow, usually have problems such as high latency caused by data collection and processing, high data costs caused by technical equipment, and data privacy and security. Therefore, social development has put forward the requirements of high precision, high efficiency and high adaptability for the division of high-density flow areas.

[0005] Therefore, how to improve the accuracy and efficiency of high-density flow area division is a problem that needs to be solved urgently. Summary of the invention

[0006] The embodiments of the present application provide a user mobility clustering analysis method, apparatus, device and storage medium, which can improve the accuracy and efficiency of high-density mobility area division.

[0007] In a first aspect, an embodiment of the present application provides a user mobility clustering analysis method, the method comprising:

[0008] Obtain the user's communication signaling data and operator MR data. The user's communication signaling data includes: user mobile phone number, base station cell code and access time of signaling event; operator MR data includes: user's service cell, neighboring cell, time slice and received signal strength;

[0009] Determine the user's mobile location data within a preset area based on the user's communication signaling data and the operator's MR data;

[0010] rasterizing the flow position data to obtain rasterized flow position data;

[0011] The rasterized flow position data are clustered to obtain high-density flow areas.

[0012] In a second aspect, an embodiment of the present application provides a user mobility clustering analysis device, the user mobility clustering analysis device comprising:

[0013] The acquisition module is used to acquire the user's communication signaling data and operator MR data. The user's communication signaling data includes: user mobile phone number, base station cell code and access time of signaling event; the operator MR data includes: user's service cell, neighboring cell, time slice and received signal strength;

[0014] A determination module, used to determine the mobile location data of the user in a preset area based on the user's communication signaling data and the operator's MR data;

[0015] A processing module, used for performing rasterization processing on the flow position data to obtain rasterized flow position data;

[0016] The clustering module is used to perform clustering processing on the rasterized flow position data to obtain high-density flow areas.

[0017] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the method in the first aspect or any possible implementation of the first aspect.

[0018] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the method in the first aspect or any possible implementation of the first aspect is implemented.

[0019] In an embodiment of the present application, the user's communication signaling data and operator MR data are obtained, and the user's communication signaling data includes: the user's mobile phone number, the base station cell code and the access time of the signaling event; the operator MR data includes: the user's service cell, neighboring cell, time slice and received signal strength; according to the user's communication signaling data and the operator MR data, the user's mobile position data in the preset area is determined; by obtaining the user's communication signaling data and the operator MR data, the user's mobile position in the preset area can be accurately located; the mobile position data is rasterized to obtain rasterized mobile position data; through rasterization, the continuous geographic space is divided into discrete grid units to simplify the data structure and facilitate subsequent clustering processing; the rasterized mobile position data is clustered to obtain a high-density mobile area; through clustering, high-density mobile areas with a large number of users and frequent location changes can be quickly and accurately identified, helping operators optimize network resource allocation and improve service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 is a flow chart of a user mobility clustering analysis method provided by an embodiment of the present application;

[0022] Figure 2 It is a grid clustering schematic diagram provided in an embodiment of the present application;

[0023] Figure 3 It is a grid clustering schematic diagram provided in an embodiment of the present application;

[0024] Figure 4 It is a schematic diagram of a Delaunay triangulation provided in an embodiment of the present application;

[0025] Figure 5 It is a schematic diagram of a high-density flow area provided in an embodiment of the present application;

[0026] Figure 6 It is a structural schematic diagram of a user mobility clustering analysis device provided in an embodiment of the present application;

[0027] Figure 7 It is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present application and are not configured to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0029] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0030] The following is a brief introduction to the technical terms involved in this application.

[0031] Signaling Information: In mobile communication systems, it is used to transmit user information and ensure the protocol control signals between devices and networks required for normal communication. Signaling events usually include: power on / off, voice calls, text messages, location updates, handovers, attachment and detachment generated by Internet access, etc.

[0032] MR (Measurement Report): The original interaction information between the user's mobile phone and the base station, including the base station service cell number, signal strength, longitude and latitude of the user's mobile phone during service retention.

[0033] Grid: Divide the space into square grids, and assign each grid a unique code and attribute value that meets the business requirements scenario to represent the grid entity.

[0034] Triangulation positioning: Calculates the precise positioning of users based on the base station location, coverage direction, RSRP, TA, user distance and other information in MR

[0035] Fingerprint algorithm: Create a fingerprint database to store information such as the signal strength and location coordinates of scattered points, calculate the similarity through the KNN algorithm, and obtain the user's precise positioning

[0036] Canopy clustering algorithm: A coarse clustering algorithm that specifies the t1 and t2 hyperparameters and the distance metric function to quickly cluster all points into clusters. Its advantages are fast and accurate.

[0037] DBSCAN clustering algorithm: A density-based clustering algorithm that finds the largest set of density-connected points as a cluster by defining neighborhoods and core objects.

[0038] Figure 1 It is a flow chart of a user mobility clustering analysis method provided in an embodiment of the present application.

[0039] like Figure 1 As shown, the user mobility cluster analysis method may include steps 110 to 140, and the method is applied to a user mobility cluster analysis device, as shown below:

[0040] Step 110, obtaining the user's communication signaling data and the operator's MR data, the user's communication signaling data includes: the user's mobile phone number, base station cell code and access time of the signaling event; the operator's MR data includes: the user's serving cell, neighboring cell, time slice and received signal strength;

[0041] Communication signaling data and operator MR data are data generated by users during the communication process, including the user's location information and communication status. By obtaining this data, you can understand the user's location and communication status at different time points.

[0042] The user's communication signaling data and operator MR data are extracted from the operator's database. The communication signaling data includes the user's mobile phone number, base station cell code and access time of the signaling event. The operator MR data includes the user's service cell, neighboring cell, time slice and received signal strength.

[0043] Step 120, determining the user's mobile location data within a preset area based on the user's communication signaling data and the operator's MR data;

[0044] Flow location data: data on the user's location changes within a preset area.

[0045] By analyzing the communication signaling data and the operator's MR data, the user's location changes in the preset area can be determined. The communication signaling data provides the time and location of the user's access to the base station, and the operator's MR data provides the user's signal strength and neighboring area information at different time points.

[0046] According to the base station cell code and access time in the communication signaling data, the user's location at different time points is determined. Combined with the service cell and neighboring cell information in the operator's MR data, the user's location change is further refined. These location change data are integrated to obtain the user's mobile location data in the preset area.

[0047] Step 130, rasterizing the flow position data to obtain rasterized flow position data;

[0048] Rasterization processing: Divide the continuous geographic space into discrete grid units to facilitate data analysis and processing.

[0049] Rasterization divides the continuous geographic space into discrete grid cells, which is convenient for data analysis and processing. By mapping the flow location data into grid cells, the data structure can be simplified, which is convenient for subsequent clustering processing. The preset area is divided into multiple grid cells, each of which represents a geographic area.

[0050] The user's flow location data is mapped to the corresponding grid unit, and the number of users and location changes in each grid unit are counted. The gridded flow location data is obtained, and each grid unit contains the number of users and location change information.

[0051] Step 140 , clustering the rasterized flow position data to obtain a high-density flow area.

[0052] Clustering: Group similar data points into different clusters to identify high-density areas. Clustering: Group similar data points into different clusters to identify high-density areas. Through the clustering algorithm, the rasterized flow position data can be divided into different clusters, each cluster representing a high-density flow area.

[0053] Clustering algorithms (such as K-means, DBSCAN, etc.) are used to cluster the rasterized flow location data. The grid cells are divided into different clusters according to the number of users and location change information. Each cluster represents a high-density flow area, identifying areas with a large number of users and frequent location changes.

[0054] By acquiring the user's communication signaling data and the operator's MR data, the user's mobile location in the preset area can be accurately located. Through rasterization processing, the continuous geographic space is divided into discrete grid units to simplify the data structure and facilitate subsequent clustering processing. Through clustering processing, high-density mobile areas with a large number of users and frequent location changes can be identified, helping operators optimize network resource allocation and improve service quality.

[0055] The following are the instructions:

[0056] In a possible embodiment, step 120 includes:

[0057] Analyze communication signaling data to obtain the user's residence time at each base station;

[0058] According to the length of time users stay at each base station, effective mobility stay sample data is filtered from the communication signaling data;

[0059] Determine the user's location data based on the operator's MR data;

[0060] The user's positioning data is associated with the effective mobility residence sample data to obtain the user's mobility location data in the preset area.

[0061] The communication signaling data contains the time and location information of the user's access to the base station. By parsing the communication signaling data, the user's residence time at each base station can be calculated. The time and location information of the user's access to the base station are extracted from the communication signaling data. The user's residence time at each base station is calculated, that is, the time interval from the user's access to the base station to the user's departure from the base station.

[0062] The user's dwell time can reflect the user's stay at a base station. By filtering sample data with longer dwell time, the user's effective mobility dwell sample data can be obtained. Set a dwell time threshold and filter out sample data with a dwell time exceeding the threshold. These sample data represent the user's long stay at a base station and have high mobility.

[0063] The operator's MR data contains information about the user's serving cell, neighboring cells, time slices, received signal strength, etc. By analyzing the operator's MR data, the user's location data at different time points can be determined.

[0064] Extract the user's service cell, neighboring cell, time slice, received signal strength and other information from the operator's MR data. Based on this information, determine the user's location data at different time points, that is, the user's geographical location.

[0065] Since the operator's MR data contains MR records of wireless measurement information such as service cells, neighboring cells, TA, RSRP, etc. during the service maintenance process, after parsing the original MR XML file, based on the latitude and longitude of the main base station, matching neighboring cells, TA, RSRP, AGPS and other information, triangulation positioning, fingerprint positioning, and machine learning algorithms are used to complete the position backfill and achieve user fine positioning. The user's MR fine positioning location data collected in the end is associated with the user's mobile phone number and base station cell number in the effective resident base station information table, and the fine latitude and longitude positioning of the user under the base station cell can be obtained, and the location data of mobile users can be integrated.

[0066] By associating the user's location data with the effective mobility residence sample data, the user's mobile location data in the preset area can be obtained. It reflects the user's location changes and stay at different time points. By matching the user's location data with the effective mobility residence sample data, the user's location and stay at different time points are associated. The user's mobile location data in the preset area is obtained, that is, the user's location changes and stay duration at different time points.

[0067] In a possible embodiment, step 120, based on the residence time of the user in each base station, screening effective mobility residence sample data from the communication signaling data includes:

[0068] Eliminate the records whose dwell time is less than the preset dwell time to obtain effective mobility dwell sample data, where the effective mobility dwell sample data includes the effective dwell records of the user at each base station; or

[0069] Determine the user's activity pattern at each base station based on the user's residence time at each base station;

[0070] According to the activity pattern of the user at each base station, the base station with the longest residence time is determined as the user's permanent base station, and a label of the permanent base station is generated;

[0071] According to the label of the permanent base station, the user's residence record at the permanent base station is removed to obtain the effective mobility residence sample data.

[0072] Records with shorter dwell time may represent short stays or error data of users. By eliminating records with dwell time shorter than the preset dwell time, effective mobile dwell sample data of users can be obtained.

[0073] Set a preset dwell time, such as 30 minutes. Remove records with a dwell time shorter than the preset dwell time from the communication signaling data. Obtain effective mobility dwell sample data, which includes effective dwell records of users at each base station.

[0074] The dwell time of a user can reflect the activity pattern of the user at a certain base station. By analyzing the dwell time of the user, the activity pattern of the user at each base station can be determined. Analyze the dwell time of the user at each base station and identify the regularity of the user's stay at different base stations. Determine the activity pattern of the user at each base station, such as frequent stays, long stays, etc.

[0075] The base station with the longest dwell time usually represents the user's permanent residence. By determining the user's permanent base station, a label of the permanent base station can be generated. According to the user's activity pattern at each base station, find the base station with the longest dwell time. Determine the base station as the user's permanent base station and generate a label of the permanent base station.

[0076] The residence records of permanent base stations may represent the daily activities of users, rather than their mobility activities. By eliminating the residence records of permanent base stations, we can obtain the effective mobility residence sample data of users.

[0077] According to the label of the permanent base station, the user's residence record in the permanent base station is removed from the communication signaling data to obtain effective mobility residence sample data, which includes the user's effective mobility residence record in each base station.

[0078] For example, the signaling data provided by the communication device includes the user's mobile phone number, the current base station cell code CELLKEY, and the access time when the signaling event starts. By integrating the signaling data of 2G, 3G, 4G, and 5G, the same user retains one base station connection information per second, and comprehensive and time-coherent user signaling raw data can be obtained. Based on the signaling data, the initial time and residence time of each user accessing the current base station can be counted, thereby forming a user trajectory data table, as shown in Table 1.

[0079] The account period of the user trajectory data table can be adjusted according to actual needs and scenarios. Here, the daily account period is used, that is, the daily user trajectory data is used, and the final regional division result is updated every day. If the update frequency of the flow regional division result is required to be higher, it can be done at the hourly level. For example, if it is updated every three hours, the user trajectory data can be aggregated into an account period of every three hours.

[0080] Table 1

[0081]

[0082] Based on the user trajectory data in Table 1, the total residence time of the user in each base station under all the base stations he has visited on that day can be counted, and the effective mobility residence sample can be defined based on the total residence time. If the total residence time is too short, it is considered that the user has passed by the grid only a few times and briefly, and is not counted in the effective mobility times.

[0083] Therefore, the user-base station with a total dwell time greater than α seconds (α=900) is regarded as a valid dwell sample, and the results are summarized in Table 2. The value of α can be adjusted according to the billing period of the user trajectory data. The adjustment rule is that the higher the billing period frequency, the smaller the value of α.

[0084] Billing frequency: Billing frequency refers to the frequency of updating user track data. For example, if user track data is updated once a day, the billing frequency is once a day; if it is updated once a week, the billing frequency is once a week.

[0085] High billing frequency: If the billing frequency is very high, such as multiple updates per day, the data changes quickly and the α value is small. If the billing frequency is low, such as once a week or month, the data changes slowly and the α value is large.

[0086] Table 2

[0087]

[0088] In the acquisition of the above effective user residence information, no distinction is made between user mobility residence base stations and user permanent residence base stations, such as night residence and daytime work place. In order to accurately mine the mobile area, it is necessary to delete the user's residence samples at the base station where he / she resides, and only retain the effective mobility residence samples. The specific implementation process is as follows:

[0089] Based on the original data of user signaling, the base station code and residence time of the user during the daytime working hours (9:00-17:00) and the night time (0:00-6:00) can be counted, so as to summarize the base station cell with the longest residence time of the user during the preset week as the user's permanent residence, forming a weekly table of user permanent residence location labels, as shown in Table 3.

[0090] Table 3

[0091]

[0092] Deleting the user's permanent base station information in Table 3 from Table 2 ensures that the number of visits of the user at his permanent location does not participate in the process of dividing the mobile area.

[0093] At the same time, users with empty permanent location codes in the user permanent location tag weekly table are users without permanent location affiliation, which can be simply understood as short-term users from other places coming to Chongqing. If the business scenario only needs to mine the high-density flow area for the user group from other places coming to Chongqing, the effective resident base station information table of users without permanent location affiliation can be filtered out from the effective resident user base station data (Table 2) according to Table 3. If the business scenario needs to retain users with permanent location affiliation, that is, retain local users, this step can be omitted.

[0094] Therefore, by analyzing the communication signaling data and the operator's MR data, the user's mobile location in the preset area can be accurately located; by screening the sample data with a longer residence time, the user's effective mobility residence sample data can be obtained to improve the accuracy of data analysis. The user's positioning data is associated with the effective mobility residence sample data to obtain the user's mobile location data in the preset area, helping operators optimize network resource allocation and improve service quality.

[0095] Therefore, by eliminating records with shorter residence time and residence records of permanent base stations, we can obtain users' effective mobility residence sample data and improve the accuracy of data analysis. By analyzing users' residence time, we can determine users' activity patterns at each base station and help operators better understand user behavior. By determining users' permanent base stations, we can generate labels for the permanent base stations to facilitate subsequent data processing and analysis.

[0096] In a possible embodiment, step 130 includes:

[0097] The flow location data are divided into multiple grids, each with a side length of 50 meters;

[0098] The number of user grid residence times in each grid is counted, and a grid statistical data table is generated. The grid statistical data table includes rasterized flow position data. The grid statistical data table is used to reflect the flow frequency and density of each grid.

[0099] Rasterization divides the continuous geographic space into discrete grid cells for easy data analysis and processing. By dividing the mobile location data into multiple grids, the data structure can be simplified for subsequent statistics and analysis. The preset area is divided into multiple grids, each with a side length of 50 meters. The user's mobile location data is mapped to the corresponding grid cells.

[0100] By counting the number of user grid residences in each grid, the flow frequency and density of the grid can be reflected. By generating a grid statistical data table, the flow situation of each grid can be intuitively reflected.

[0101] Count the number of user grid dwell times in each grid, that is, the number of times the user stays in each grid. Generate a grid statistical data table, which includes rasterized flow location data, the number of dwell times in each grid, and other information. The grid statistical data table is used to reflect the flow frequency and density of each grid, helping operators better understand user behavior.

[0102] In order to improve the efficiency of the clustering algorithm and optimize memory, the longitude and latitude data of mobile users are rasterized to obtain the grid ID, the longitude and latitude of the center of the grid, and the longitude and latitude of all resident users under the grid. The time complexity of the clustering algorithm is O(n 2 ), the clustering algorithm based on tens of millions of user groups has the problems of high time cost and slow result update.

[0103] Through rasterization processing, the density clustering process based on tens of millions of users is converted into a distance process based on hundreds of thousands of grids, and the overall clustering efficiency is improved.

[0104] The grid size determines the accuracy of the final clustering area. If the grid is too large, the final area division will cover a larger area, causing some low-density areas to be included in the high-density area division range, reducing its practicality in business scenarios; while if the grid is too small, the algorithm efficiency will be reduced, and the effect of overly detailed area division will not be significantly improved in actual application. Therefore, the assignment of the grid size needs to be based on the final application scenario and business understanding. Here, the grid side length is set to 50 meters.

[0105] After rasterization, the number of effective user stays in each grid can be counted. The size of the statistic reflects the flow frequency and density of the current grid. The data table after statistical integration is shown in Table 4, which is used as the final data table for subsequent clustering.

[0106] Table 4

[0107]

[0108] Therefore, by dividing the flow location data into multiple grids, the data structure can be simplified, which is convenient for subsequent statistics and analysis. By counting the number of user grid residences in each grid, the flow frequency and density of the grid can be reflected, helping operators to better understand user behavior. By generating a grid statistical data table, the flow situation of each grid can be intuitively reflected, which is convenient for subsequent data processing and analysis.

[0109] In a possible embodiment, step 140 includes:

[0110] According to the number of grid residence times, high flow density grids are selected from the grid statistical data table; the flow density of the flow position data in the high flow density grid is greater than the preset density;

[0111] Perform Canopy clustering calculation on the high flow density grid to obtain the high flow density grid in each Canopy cluster;

[0112] Perform DBSCAN clustering calculation on the high flow density grids in each Canopy cluster to obtain the effective clustering area;

[0113] The effective clustering area is subjected to Delaunay triangulation to obtain the high-density flow area.

[0114] The number of grid dwell times can reflect the flow density of the grid. By screening grids with a grid dwell time greater than a preset density, high flow density grids can be obtained. Grids with a grid dwell time greater than a preset density are screened from the grid statistical data table. The flow density of the flow position data in these grids is greater than the preset density, representing a high flow density area.

[0115] The grid center longitude, grid center latitude, and number of residences in Table 4 are parameter indicators that determine the final regional division. The grid center longitude and center latitude represent the geographic location dimension characteristics of the clustering area, while the number of residences represents the density dimension characteristics of the clustering area. Traditional density-based clustering algorithms can effectively use distance measurement functions to perform density clustering based on geographic location characteristics, but there is a problem that it is impossible to accurately quantify the distance between geographic location dimension characteristics and density dimension characteristics during the clustering process.

[0116] In this regard, the embodiment of the present application performs a grid pre-screening step before clustering, thereby indirectly adding the density index to the clustering process. First, the grid set is divided into high-density grids and low-density grids according to the number of grid residence times. Grids with a residence time greater than or equal to the density threshold θ1 are defined as high-flow density grids, and those with a residence time less than θ1 are defined as low-flow density grids. They are screened out from the grid set and do not participate in subsequent clustering. The assignment of θ1 needs to be combined with actual business analysis and data analysis. Here, θ1 is set to 50. The specific implementation effect is as follows: Figure 2 shown.

[0117] The grid pre-screening step makes up for the insufficiency of density-based clustering algorithms that cannot take into account both grid geographic location and grid density indicators, ensuring that the final clustering area meets the dual conditions of geographical proximity and high density. At the same time, the initial screening of grids can avoid the problem that the clustering algorithm cannot divide the effective area due to the connection of all grids, and at the same time improve the subsequent clustering efficiency by reducing the number of grids that need to participate in clustering.

[0118] Canopy clustering is a rough clustering method used to preliminarily divide a data set. Through Canopy clustering, high flow density grids can be divided into different clusters to facilitate subsequent fine clustering. Canopy clustering calculations are performed on high flow density grids to obtain high flow density grids within each Canopy cluster. Each Canopy cluster represents a preliminary high flow density area.

[0119] DBSCAN clustering is a density-based clustering method that can identify high-density areas. Through DBSCAN clustering, the high flow density grid within each Canopy cluster can be further subdivided to obtain valid clustering areas. DBSCAN clustering calculations are performed on the high flow density grid within each Canopy cluster. Valid clustering areas are obtained, which represent high-density flow areas.

[0120] The result set output by the Canopy algorithm is used for secondary DBSCAN density clustering. The grid set under each Canopy cluster is clustered again to divide independent high-density areas. In order to ensure that the grids in the same cluster are adjacent, the radius distance parameter ε of the DBSCAN algorithm is limited to the straight-line distance between the grid centers.

[0121] Considering that the grid is 50 meters by 50 meters, and the straight-line distance between the center points of two adjacent grids is 50 meters, the lower limit of ε is about 71 meters and the upper limit is 100 meters. Here, ε is set to 80 meters. In addition, in order to avoid the phenomenon that the final clustering area is too scattered, the minimum sample size parameter K of the algorithm is set to 3 to ensure that each clustering result cluster contains at least 3 grids. The secondary DBSCAN clustering effect is as follows: Figure 3 shown.

[0122] Delaunay triangulation is a method of dividing a point set into a triangular mesh, which can generate a minimum convex polygon that contains all points. Through Delaunay triangulation, the effective clustering area can be divided into high-density flow areas. Delaunay triangulation is performed on the effective clustering area to generate a Delaunay triangulation network. The outer boundary constitutes a minimum convex polygon that contains all grid center points in the area, representing the high-density flow area.

[0123] Therefore, by screening grids whose grid residence times are greater than the preset density, high flow density grids can be obtained to help operators identify high flow density areas; through Canopy clustering, high flow density grids can be divided into different clusters to facilitate subsequent fine clustering; through DBSCAN clustering, the high flow density grids within each Canopy cluster can be further subdivided to obtain effective clustering areas, representing high-density flow areas; through Delaunay triangulation, effective clustering areas can be divided into high-density flow areas, helping operators optimize network resource allocation and improve service quality.

[0124] In a possible embodiment, performing Canopy clustering calculation on the high flow density grid to obtain the high flow density grid in each Canopy cluster includes:

[0125] Set the radius of the cluster area and the distance of the cluster area;

[0126] According to the radius and distance of clustering area, Canopy clustering calculation is performed on the high flow density grid to obtain the high flow density grid in each Canopy cluster.

[0127] The radius of the cluster area and the distance of the cluster area are two key parameters in the Canopy clustering algorithm. The radius (T1) is used to define whether a grid belongs to a Canopy cluster, and the distance (T2) is used to define whether a grid belongs to the core area of ​​a Canopy cluster.

[0128] Set the radius of the cluster area (T1) and the distance of the cluster area (T2). The values ​​of T1 and T2 can be adjusted according to actual conditions, usually T1>T2.

[0129] Canopy clustering is a rough clustering method used to preliminarily divide the data set. Through Canopy clustering, high flow density grids can be divided into different clusters, which is convenient for subsequent fine clustering.

[0130] Canopy clustering is performed on high flow density grids using the radius (T1) and distance (T2) of the clustering region. Each grid is compared with an existing Canopy cluster. If the distance between the grid and the center of a Canopy cluster is less than T1, the grid is added to the Canopy cluster. If the distance between the grid and the center of a Canopy cluster is less than T2, the grid is marked as the core area of ​​the Canopy cluster. A high flow density grid is obtained within each Canopy cluster, and each Canopy cluster represents a preliminary high flow density area.

[0131] The first Canopy clustering is performed on the selected grid set. The main purpose of clustering is to avoid the situation where the high-density mobility areas in the city are concentrated and adjacent to each other, which leads to the fact that the final clustering effect has low reference significance and business practicality. The algorithm needs to input two hyperparameters, t1 and t2.

[0132] The t1 parameter value determines the upper limit of the radius length of the final clustering area, which can be adjusted based on the business, actual implementation scenario (such as the size of the city area, the scale of managed resources, etc.), and the final clustering effect. Here it is set to 250 meters and t2 is set to 200 meters.

[0133] In order to calculate the actual geographic distance between grids and ensure that each grid is clustered based on the geographic distance that conforms to the business logic, it is necessary to implement a distance calculation function based on longitude and latitude to replace the default Euclidean distance function in the algorithm. The specific implementation steps of the longitude and latitude distance function are as follows:

[0134]

[0135] Among them, latitude and longitude are used to represent the longitude and latitude coordinates of any point on the surface of the earth.

[0136] After clustering, the Canopy clustering label table 5 is as follows:

[0137] Table 5

[0138]

[0139] Through Canopy clustering, high flow density grids can be divided into different clusters for subsequent fine clustering. By setting the radius and distance of the clustering area, the results of Canopy clustering can be controlled to help operators better understand user behavior.

[0140] Through Canopy clustering, a high flow density grid can be generated within each Canopy cluster, representing the preliminary high-density flow area, helping operators optimize network resource allocation and improve service quality.

[0141] After completing the secondary DBSCAN clustering, it is necessary to determine whether each divided area is an area with high traffic density, that is, to determine whether the sum of the flow residence times of all grids in the area is greater than a certain threshold. The value of the threshold can be adjusted according to factors such as business scenarios and urban management resources. Here, the threshold is taken as 500 (θ2 = 500), that is, the area with a total effective residence time greater than 500 times is the final effective clustering area.

[0142] Specifically, for each clustering area, all grids in the area are traversed, and the number of flow residence times in each grid is accumulated. The calculated sum is compared with the preset threshold. If the sum is greater than the threshold, the area is a high flow density area, that is, the final effective clustering area.

[0143] By filtering and screening, we can ensure that the final output areas are all high-density flow areas. The final output data structure is shown in Table 6 below.

[0144] Table 6

[0145]

[0146] Table 7

[0147]

[0148] In a possible embodiment, Delaunay triangulation is performed on the effective clustering area to obtain a high-density flow area, including:

[0149] Delaunay triangulation is performed on the effective clustering area to obtain a Delaunay triangulation network; the outer boundary of the Delaunay triangulation network constitutes a minimum convex polygon containing all grid center points in the effective clustering area;

[0150] According to the grid longitude and latitude on the outer boundary of the Delaunay triangulation network, the longitude and latitude of the outer boundary vertices of each high-density flow area are determined to obtain the high-density flow area.

[0151] Delaunay triangulation is a method that divides a point set into a triangular mesh and generates the smallest convex polygon that contains all the points. Through Delaunay triangulation, the effective clustering area can be divided into high-density flow areas.

[0152] Delaunay triangulation is performed on the valid clustering area to generate a Delaunay triangulation network. The outer boundary of the Delaunay triangulation network constitutes a minimum convex polygon containing all grid center points in the valid clustering area.

[0153] The grid longitude and latitude on the outer boundary of the Delaunay triangulation can determine the longitude and latitude of the outer boundary vertices of the high-density flow area. By determining the longitude and latitude of the outer boundary vertices, the boundary of the high-density flow area can be obtained.

[0154] According to the grid longitude and latitude on the outer boundary of the Delaunay triangulation network, the longitude and latitude of the outer boundary vertices of each high-density flow area are determined. These vertex longitudes and latitudes are connected to form the outer boundary of the high-density flow area. High-density flow areas are obtained, which represent high-density flow areas with a large number of users and frequent location changes.

[0155] like Figure 4 As shown in FIG. 1 , after obtaining the final cluster region labels and all grid positions in the region, Delaunay triangulation is performed on each region, and the outer boundary of the formed Delaunay triangulation network constitutes a minimum convex polygon that includes all grid center points in the region.

[0156] Through Delaunay triangulation, the effective clustering area can be divided into high-density flow areas, helping operators optimize network resource allocation and improve service quality. By determining the longitude and latitude of the grid on the outer boundary of the Delaunay triangulation, the longitude and latitude of the outer boundary vertices of the high-density flow area can be determined, helping operators better understand user behavior. Through Delaunay triangulation, the minimum convex polygon containing all grid center points in the effective clustering area can be generated to represent the high-density flow area, helping operators optimize network resource allocation and improve service quality.

[0157] By selecting the grid longitude and latitude on the outer boundary, the longitude and latitude of the outer boundary vertices of each divided area can be obtained, and the effect is as follows: Figure 5 As shown in the figure, the dots in the figure are the grids in the region, and the black squares are the grids at the vertices of the region's boundaries. The steps of regional boundary positioning can enable relevant business departments to realize regional monitoring and control deployment more conveniently and efficiently.

[0158] In an embodiment of the present application, the user's communication signaling data and operator MR data are obtained, and the user's communication signaling data includes: the user's mobile phone number, the base station cell code and the access time of the signaling event; the operator MR data includes: the user's service cell, neighboring cell, time slice and received signal strength; according to the user's communication signaling data and the operator MR data, the user's mobile position data in the preset area is determined; by obtaining the user's communication signaling data and the operator MR data, the user's mobile position in the preset area can be accurately located; the mobile position data is rasterized to obtain rasterized mobile position data; through rasterization, the continuous geographic space is divided into discrete grid units to simplify the data structure and facilitate subsequent clustering processing; the rasterized mobile position data is clustered to obtain a high-density mobile area; through clustering, high-density mobile areas with a large number of users and frequent position changes can be identified, which helps operators optimize network resource allocation and improve service quality.

[0159] Based on the above Figure 1 The user mobility clustering analysis method shown in the embodiment of the present application also provides a user mobility clustering analysis device, such as Figure 6 As shown, the user mobility clustering analysis device 600 may include:

[0160] The acquisition module 610 is used to acquire the user's communication signaling data and the operator's MR data. The user's communication signaling data includes: the user's mobile phone number, base station cell code and the access time of the signaling event; the operator's MR data includes: the user's serving cell, neighboring cell, time slice and received signal strength;

[0161] The determination module 620 is used to determine the mobile location data of the user in the preset area according to the communication signaling data of the user and the operator MR data;

[0162] The processing module 630 is used to perform rasterization processing on the flow position data to obtain rasterized flow position data;

[0163] The clustering module 640 is used to perform clustering processing on the rasterized flow position data to obtain a high-density flow area.

[0164] In a possible embodiment, the determination module 620 is specifically configured to:

[0165] Analyze communication signaling data to obtain the user's residence time at each base station;

[0166] According to the length of time users stay at each base station, effective mobility stay sample data is filtered from the communication signaling data;

[0167] Determine the user's location data based on the operator's MR data;

[0168] The user's positioning data is associated with the effective mobility residence sample data to obtain the user's mobility location data in the preset area.

[0169] In a possible embodiment, the determination module 620 is specifically configured to:

[0170] Eliminate the records whose dwell time is less than the preset dwell time to obtain effective mobility dwell sample data, where the effective mobility dwell sample data includes the effective dwell records of the user at each base station; or

[0171] Determine the user's activity pattern at each base station based on the user's residence time at each base station;

[0172] According to the activity pattern of the user at each base station, the base station with the longest residence time is determined as the user's permanent base station, and a label of the permanent base station is generated;

[0173] According to the label of the permanent base station, the user's residence record at the permanent base station is removed to obtain the effective mobility residence sample data.

[0174] In a possible embodiment, the processing module 630 is specifically configured to:

[0175] The flow location data are divided into multiple grids, each with a side length of 50 meters;

[0176] The number of user grid residence times in each grid is counted, and a grid statistical data table is generated. The grid statistical data table includes rasterized flow position data. The grid statistical data table is used to reflect the flow frequency and density of each grid.

[0177] In a possible embodiment, the clustering module 640 is specifically configured to:

[0178] According to the number of grid residence times, high flow density grids are selected from the grid statistical data table; the flow density of the flow position data in the high flow density grid is greater than the preset density;

[0179] Perform Canopy clustering calculation on the high flow density grid to obtain the high flow density grid in each Canopy cluster;

[0180] Perform DBSCAN clustering calculation on the high flow density grids in each Canopy cluster to obtain the effective clustering area;

[0181] The effective clustering area is subjected to Delaunay triangulation to obtain the high-density flow area.

[0182] In a possible embodiment, the clustering module 640 is specifically configured to:

[0183] Set the radius of the cluster area and the distance of the cluster area;

[0184] According to the radius and distance of clustering area, Canopy clustering calculation is performed on the high flow density grid to obtain the high flow density grid in each Canopy cluster.

[0185] In a possible embodiment, the clustering module 640 is specifically configured to:

[0186] Delaunay triangulation is performed on the effective clustering area to obtain a Delaunay triangulation network; the outer boundary of the Delaunay triangulation network constitutes a minimum convex polygon containing all grid center points in the effective clustering area;

[0187] According to the grid longitude and latitude on the outer boundary of the Delaunay triangulation network, the longitude and latitude of the outer boundary vertices of each high-density flow area are determined to obtain the high-density flow area.

[0188] In an embodiment of the present application, the user's communication signaling data and operator MR data are obtained, and the user's communication signaling data includes: the user's mobile phone number, the base station cell code and the access time of the signaling event; the operator MR data includes: the user's service cell, neighboring cell, time slice and received signal strength; according to the user's communication signaling data and the operator MR data, the user's mobile position data in the preset area is determined; by obtaining the user's communication signaling data and the operator MR data, the user's mobile position in the preset area can be accurately located; the mobile position data is rasterized to obtain rasterized mobile position data; through rasterization, the continuous geographic space is divided into discrete grid units to simplify the data structure and facilitate subsequent clustering processing; the rasterized mobile position data is clustered to obtain a high-density mobile area; through clustering, high-density mobile areas with a large number of users and frequent position changes can be identified, which helps operators optimize network resource allocation and improve service quality.

[0189] Figure 7 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application is shown.

[0190] The electronic device may include a processor 401 and a memory 402 storing computer program instructions.

[0191] Specifically, the processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0192] The memory 402 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 402 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 402 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 402 is a non-volatile solid-state memory. In a specific embodiment, the memory 402 includes a read-only memory (ROM). In appropriate cases, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM) or a flash memory or a combination of two or more of these.

[0193] The processor 401 implements any one of the user mobility clustering analysis methods in the embodiments shown in the figure by reading and executing the computer program instructions stored in the memory 402.

[0194] In one example, the electronic device may further include a communication interface 404 and a bus 410. Figure 4 As shown, the processor 401, the memory 402, and the communication interface 404 are connected via a bus 410 and communicate with each other.

[0195] The communication interface 404 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0196] Bus 410 includes hardware, software or both, and the parts of electronic equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 410 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0197] The electronic device can execute the user mobility clustering analysis method in the embodiment of the present application, thereby realizing the combination Figure 1 Describe the user mobility clustering analysis method.

[0198] In addition, in combination with the user mobility clustering analysis method in the above embodiment, the present application embodiment can provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by the processor, the Figure 1 A clustering analysis method for user mobility.

[0199] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0200] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0201] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0202] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A user mobility clustering analysis method, characterized in that: include: Obtain the user's communication signaling data and operator MR data. The user's communication signaling data includes: user mobile phone number, base station cell code and access time of signaling event; operator MR data includes: user's service cell, neighboring cell, time slice and received signal strength; Determine the user's mobile location data within a preset area based on the user's communication signaling data and the operator's MR data; rasterizing the flow position data to obtain rasterized flow position data; The rasterized flow position data are clustered to obtain high-density flow areas.

2. The method according to claim 1, characterized in that Determining the mobile location data of the user in the preset area according to the communication signaling data of the user and the MR data of the operator includes: Analyze communication signaling data to obtain the user's residence time at each base station; According to the length of time users stay at each base station, effective mobility stay sample data is filtered from the communication signaling data; Determine the user's location data based on the operator's MR data; The user's positioning data is associated with the effective mobility residence sample data to obtain the user's mobility location data in the preset area.

3. The method according to claim 2, characterized in that The method of screening effective mobility residence sample data from communication signaling data according to the residence time of the user in each base station includes: Eliminate the records whose dwell time is less than the preset dwell time to obtain effective mobility dwell sample data, where the effective mobility dwell sample data includes the effective dwell records of the user at each base station; or Determine the user's activity pattern at each base station based on the user's residence time at each base station; According to the activity pattern of the user at each base station, the base station with the longest residence time is determined as the user's permanent base station, and a label of the permanent base station is generated; According to the label of the permanent base station, the user's residence record at the permanent base station is removed to obtain the effective mobility residence sample data.

4. The method according to claim 1, characterized in that The rasterizing the flow position data to obtain the rasterized flow position data includes: The flow location data is divided into multiple grids, each with a side length of 50 meters; The number of user grid residence times in each grid is counted, and a grid statistical data table is generated. The grid statistical data table includes rasterized flow position data. The grid statistical data table is used to reflect the flow frequency and density of each grid.

5. The method according to claim 4, characterized in that The clustering process is performed on the rasterized flow position data to obtain a high-density flow area, including: According to the number of grid residence times, high flow density grids are selected from the grid statistical data table; the flow density of the flow position data in the high flow density grid is greater than the preset density; Perform Canopy clustering calculation on the high flow density grid to obtain the high flow density grid in each Canopy cluster; Perform DBSCAN clustering calculation on the high flow density grids in each Canopy cluster to obtain the effective clustering area; The effective clustering area is subjected to Delaunay triangulation to obtain the high-density flow area.

6. The method according to claim 5, characterized in that The Canopy clustering calculation is performed on the high flow density grid to obtain the high flow density grid in each Canopy cluster, including: Set the radius of the cluster area and the distance of the cluster area; According to the radius and distance of clustering area, Canopy clustering calculation is performed on the high flow density grid to obtain the high flow density grid in each Canopy cluster.

7. The method according to claim 5, characterized in that The effective clustering area is subjected to Delaunay triangulation to obtain a high-density flow area, including: Delaunay triangulation is performed on the effective clustering area to obtain a Delaunay triangulation network; the outer boundary of the Delaunay triangulation network constitutes a minimum convex polygon containing all grid center points in the effective clustering area; According to the grid longitude and latitude on the outer boundary of the Delaunay triangulation network, the longitude and latitude of the outer boundary vertices of each high-density flow area are determined to obtain the high-density flow area.

8. A user mobility clustering analysis device, characterized in that: The user mobility clustering analysis device comprises: The acquisition module is used to acquire the user's communication signaling data and operator MR data. The user's communication signaling data includes: user mobile phone number, base station cell code and access time of signaling event; the operator MR data includes: user's service cell, neighboring cell, time slice and received signal strength; A determination module, used to determine the mobile location data of the user in a preset area based on the user's communication signaling data and the operator's MR data; A processing module, used for performing rasterization processing on the flow position data to obtain rasterized flow position data; The clustering module is used to perform clustering processing on the rasterized flow position data to obtain high-density flow areas.

9. An electronic device, characterized in that: The electronic device comprises: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the user mobility clustering analysis method as described in any one of claims 1-7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the user mobility clustering analysis method as described in any one of claims 1-7 is implemented.

Images