Resident travel activity mode identification method and system based on mobile phone signaling data
Through the resident travel activity pattern recognition method based on mobile phone signaling data, the activity nodes and reference points are identified, the space-time intensity and time of going out are calculated, and the problem of insufficient data quality and correlation in traditional traffic modeling is solved, and more accurate residents are provided Analysis of travel characteristics and traffic demand.
Patent Information
- Application Number
- CN202410872662.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-01
- Publication Date
- 2024-11-08
AI Technical Summary
Traditional urban transportation modeling relies on residents' travel survey data, has data quality deviations, long update cycles, high costs and difficult organization, and ignores the multi-dimensional relationship between travel activities, making it difficult to reflect residents' real transportation needs and travel characteristics.
Using the method of resident travel activity pattern recognition based on mobile phone signaling data, by obtaining mobile phone signaling data and base station data, identifying active nodes and reference points, calculating the activity time and space intensity, outing time and daily activity count, and constructing a user travel attribute data set. Calculate the similarity and iteratively calculate the attraction and attribute degree to realize user clustering and obtain residents' travel activity modes.
Improve the real and objective data acquisition of residents' travel activities, overcome the problems of poor accuracy and high cost of traditional survey methods, can effectively reflect the multi-dimensional relationship between activities and travel, provide more accurate traffic demand analysis, and avoid traditional models Catalog error.
Smart Images

Figure CN118921628A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of user travel classification calculation, and relates to a method and system for identifying resident travel activity patterns based on mobile phone signaling data. Background Art
[0002] Transportation is a basic, leading and strategic industry in the national economy, an important service industry and an important part of the modern economic system. It is an important support for building a new development pattern and a solid guarantee for serving the people's better life and promoting common prosperity. New requirements have been put forward for urban and transportation planning, and it is necessary to evaluate the lifestyle and travel characteristics of residents before and after construction through planning models.
[0003] However, traditional urban traffic modeling mainly relies on residents' travel surveys to obtain data. The data quality has long been criticized. Subjective memories cause data deviations and errors. The data update cycle is long and the survey organization is cumbersome. In addition, the current "four-stage" traffic model is mainly designed to meet the travel demand forecast required for large-scale road construction at that time. It takes a single trip (Trip) as an independent analysis unit, ignoring the derivative demand characteristics of travel generated by daily activities, and it is difficult to reflect the multi-dimensional relationship between activities, activities and trips, and trips. Summary of the invention
[0004] The purpose of the present invention is to solve the problem that in the prior art, in order to meet the travel demand forecast required for large-scale road construction at that time, a single trip (Trip) is mainly used as an independent analysis unit, ignoring the derivative demand characteristics of travel generated by daily activities, making it difficult to reflect the multi-dimensional correlation between activities, activities and trips, and trips and trips, and unable to obtain the real traffic needs and travel characteristics of residents. A method and system for identifying residents' travel activity patterns based on mobile phone signaling data are provided.
[0005] In order to achieve the above object, the present invention adopts the following technical solutions:
[0006] A method for identifying resident travel activity patterns based on mobile phone signaling data comprises the following steps:
[0007] Obtain mobile phone signaling data and base station data on residents’ travel activities;
[0008] Obtain residents' travel activity nodes and activity benchmarks based on mobile phone signaling data and base station data;
[0009] Based on the activity nodes and activity benchmarks of residents' travel, the spatiotemporal intensity of residents' activities, the time spent outside the activity, and the number of daily activities are calculated;
[0010] A data set of user travel attributes is constructed based on the spatiotemporal intensity of residents' activities, the time they go out for activities, and the number of daily activities. The similarity of travel attributes between users is calculated based on the data set of user travel attributes. The user's attractiveness and belongingness are iteratively calculated based on the similarity until the number of iterations reaches the set number or the iterations meet the set convergence parameters. The user clustering center is obtained based on the attractiveness and belongingness at the current moment, and the recognition result of the residents' travel activity pattern is obtained based on the user clustering center.
[0011] A further improvement of the present invention is:
[0012] The activity nodes and activity reference points for obtaining residents' travel include:
[0013] The mobile phone signaling data continuously recorded by the same base station is merged to identify the base station short stop S in turn. i {MSID,Lon i ,Lat i ,T i S ,T i E};
[0014] When two consecutive base stations are briefly parked at S i With S i+1 Satisfy Δt between i,i+1 ≤τ and Δx i,i+1 When ≤d, two consecutive base station short stops are merged to obtain the active node A. i , where Δt i,i+1 represents the time difference of short dwell between consecutive base stations, Δx i,i+1 Indicates the distance difference between consecutive base station short dwells;
[0015] Based on the obtained activity nodes, the activity node with the longest duration during the period of 0:00-8:00 of the user is defined as the activity reference point.
[0016] The spatiotemporal intensity of residents' activities is calculated by the following formula:
[0017]
[0018] In the formula, I i represents the temporal and spatial intensity of user i’s daily travel activities, and Represent the start time and the leave time of the jth activity node of user i, Dis(A i,j ,A i,h ) is the distance between the jth activity node of user i and the activity reference point;
[0019] The activity out-of-office time is calculated by the following formula:
[0020]
[0021] Where D i The time spent outside by user i. h ( j ) refers to the activity reference point sequence number of user i, A n It is the user's last activity node of the day, and H represents the activity benchmark point;
[0022] The daily activity number is calculated by the following formula:
[0023]
[0024] Where P i is the number of daily activities of user i, A i,j is the j-th active node of user i.
[0025] The data set of user travel attributes is X = {x 1 ,x 2 ,…,xi,…,xn}, where x i Represents multi-dimensional travel feature information {I i ,D i ,P i} the i-th user.
[0026] The calculating of user similarity comprises:
[0027]
[0028] Among them, I i represents the temporal and spatial intensity of user i’s daily travel activities; D i P is the time spent outside the home by user i; i The number of daily user activities.
[0029] Iterative calculation of user attraction and belonging based on similarity includes:
[0030] First, initialize the user's attraction and belongingness, so that both attraction and belongingness are zero matrices isomorphic to the similarity matrix S(i,k), where attraction R(i,k) is used to describe the degree to which user k identifies user i as the cluster center, and belongingness A(i,k) is used to describe the degree to which user i identifies user k as its cluster center;
[0031] Based on the similarity S(i,k), the user attraction R(i,k) and the attribution A(i,k) are continuously updated iteratively.
[0032] The iterative calculation of the user's attractiveness R(i,k) includes:
[0033]
[0034] R(i,k) new =(1-λ)R(i,k) t +λR(i,k)
[0035] The iterative calculation of the user's attribution degree A(i,k) includes:
[0036]
[0037] A(i,k) new =(1-λ)A(i,k) t +λA(i,k)
[0038] Among them, S(i,k) represents the user similarity, and λ is the attenuation coefficient to avoid iterative oscillation.
[0039] A resident travel activity pattern recognition system based on mobile phone signaling data, comprising:
[0040] A data acquisition module, used to obtain mobile phone signaling data and base station data of residents' travel activities;
[0041] Obtain residents' travel activity nodes and activity benchmarks based on mobile phone signaling data and base station data;
[0042] Based on the activity nodes and activity benchmarks of residents' travel, the spatiotemporal intensity of residents' activities, the time spent outside the activity, and the number of daily activities are calculated;
[0043] A data set of user travel attributes is constructed based on the spatiotemporal intensity of residents' activities, the time they go out for activities, and the number of daily activities. The similarity of travel attributes between users is calculated based on the data set of user travel attributes. The user's attractiveness and belongingness are iteratively calculated based on the similarity until the number of iterations reaches the set number or the iterations meet the set convergence parameters. The user clustering center is obtained based on the attractiveness and belongingness at the current moment, and the recognition result of residents' travel activities is obtained based on the user clustering center.
[0044] A terminal device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of any one of the methods of the present invention when executing the computer program.
[0045] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of any method described in the present invention.
[0046] Compared with the prior art, the present invention has the following beneficial effects:
[0047] The present invention discloses a method for identifying residents' travel activity patterns based on mobile phone signaling data. The method obtains the real and objective activity data of residents through mobile phone signaling data and base station data. The data has a wider coverage range and can overcome the defects of poor accuracy, high cost and great organizational difficulty of traditional resident survey data. The individual travel information is identified by identifying the activity nodes and activity reference points of residents' travel, and based on this, the activity time weighting is considered on the basis of the activity space index to establish three activity characteristic calculation indicators, namely, activity spatiotemporal intensity, activity outing time and daily activity number, which effectively reflect the spatiotemporal characteristics of residents' activities, thereby improving the recognition of individual activity intensity attributes. Then, clustering calculation and analysis are performed on this basis to identify and classify residents' activity patterns. The method explores the individual travel activity rules from the perspective of individual activities, can grasp the traffic travel demand in essence, avoid the aggregation error of the traditional "four-stage" traffic planning model, can obtain the demand characteristics of residents' daily activities, and obtain the multi-dimensional correlation between activities and activities, activities and travel, and travel and travel, thereby providing a theoretical basis for optimizing the real traffic needs of residents in the later stage. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.
[0049] Figure 1 A comparison diagram of data preprocessing effects of an embodiment of the present invention (where a represents the original trajectory data of the user; b represents the trajectory data after preprocessing of the user);
[0050] Figure 2 A spatial distribution diagram of active nodes and active reference points according to an embodiment of the present invention;
[0051] Figure 3 This is a diagram of the user activity characteristic index identification result of an embodiment of the present invention (where a is the user activity intensity result; b is the user activity outing time result; c is the user daily activity number result diagram);
[0052] Figure 4 This is the user travel mode classification result of the embodiment of the present invention. DETAILED DESCRIPTION
[0053] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.
[0054] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0055] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.
[0056] In the description of the embodiments of the present invention, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. indicate an orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed when in use, it is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.
[0057] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", which does not mean that the structure must be completely horizontal, but can be slightly tilted.
[0058] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0059] The present invention is further described in detail below in conjunction with the accompanying drawings:
[0060] See also Figure 1 The present invention discloses a method for identifying residents' travel activity patterns based on mobile phone signaling data. A two-stage base station sequence inference merging method is established based on mobile phone signaling data to identify travel information such as activity nodes and activity reference points therein. On this basis, a resident activity spatiotemporal intensity index system based on activity reference points is established. Relying on the identified individual activity intensity characteristic attributes, a multidimensional clustering algorithm is established to divide people with different travel patterns, analyze the travel activity patterns of urban residents, and establish a complete set of resident activity intensity recognition and behavior pattern recognition systems.
[0061] The specific steps include:
[0062] Step 1: Obtain mobile phone signaling data and base station data of residents’ travel activities;
[0063] The travel time and space feature data obtained from China Unicom, China Mobile and other communication companies include signaling data (user travel time and location information) and communication base station data (base station longitude and latitude and service information). Mobile phone signaling data is a passive trace record of the communication services provided by operators to users, triggered by network activities such as user communication activities and service area switching. The data mainly contains information such as the user's unique identification number MSID, timestamp TIMESTAMP, location area number LAC, and cellular cell code CELLID. The base station data also contains LAC and CELLID information, and retains attribute information such as the longitude and latitude of the base station. According to the signaling data and the LAC and CELLID fields of the base station data, the longitude and latitude coordinate information is matched for each signaling data, and finally the mobile phone positioning data shown in Table 1 is obtained.
[0064] Table 1 Example of coordinate matching results of mobile phone signaling data
[0065]
[0066] Step 2: Obtain the activity nodes and activity benchmarks of residents' travel based on mobile phone signaling data and base station data;
[0067] For large sample user data, this method defines a two-stage base station sequence inference merging method to realize user activity node identification:
[0068] First, the signaling data recorded continuously by the same base station are merged, and the base station short-stay S is identified in turn. i {MSID,Lon i ,Lat i ,T i S ,T i E}, where each base station short stay contains the user ID, longitude and latitude coordinates, and start and end time, ultimately forming a continuous base station short stay sequence S = {S1 ,S 2 ,…,S n};
[0069] Secondly, short-term parking S i Merge, when two consecutive base stations are briefly parked S i With S i+1 Satisfy Δt between i,i+1 ≤τ and Δx i,i+1 ≤d, merge them into the same active node A i ; where Δt i,i+1 is the time difference of short dwelling of consecutive base stations, which is given by Calculated, Δx i,i+1 is the distance difference between consecutive base station short stops, which is obtained by substituting the longitude and latitude coordinates of consecutive base station short stops into the Haversine spherical distance formula. τ and d are the contact discrimination thresholds for determining whether consecutive base station short stops are for the same active node. It is generally recommended to select 30min and 400m.
[0070] The above merging is performed in sequence until there is no base station short-stay meeting the above conditions in the base station sequence, then the merging process ends, and the output result is the active node sequence.
[0071] Identification of active reference points includes:
[0072] In order to measure the intensity characteristics of residents' travel activities and further explore the travel activity patterns of the population, it is also necessary to identify the activity reference point of each user based on the activity node identification results, which is defined as the activity node with the longest duration during the user's 0:00-8:00 period.
[0073] Step 3: Calculate residents’ activity spatiotemporal intensity, out-of-town time, and daily activity number based on their travel activity nodes and activity benchmarks;
[0074] In order to measure the spatiotemporal characteristics of individual travel activities, it is proposed to construct a spatiotemporal index that comprehensively reflects the travel intensity characteristics of users based on the identification results of activity nodes and activity benchmark points, and propose three indicators: spatiotemporal intensity of activity, time spent outside the activity, and number of daily activities.
[0075] Step 3.1: Spatiotemporal intensity of activity
[0076] Residents' travel activities are a multidimensional attribute set composed of various elements such as time and space. Characterization from the perspective of activity space alone is not enough to reflect the characteristics of residents' activities in essence. As shown in formula (1), this method proposes the quantification of residents' activity intensity weighted by activity time, which can effectively improve the defect of insufficient representation of time attributes in activity space and more effectively reflect the spatiotemporal intensity of residents' travel activities.
[0077]
[0078] In the formula, I i represents the temporal and spatial intensity of user i’s daily travel activities, and Respectively represent the start time and the end time of the jth activity of user i, Dis(A i,j ,A i,h ) is the distance between the jth active node of user i and the reference point.
[0079] Step 3.2: Active time out
[0080] The activity out time is the sum of the time difference between the user's two benchmark activities, which is calculated in two cases:
[0081] First case:
[0082] When the user's last activity is a benchmark activity, the cumulative time difference between the benchmark activities of the day is calculated normally.
[0083] Second case:
[0084] When the user's last activity is not a benchmark activity, the latest benchmark activity of the day is found, and the cumulative time difference between benchmark activities before the benchmark activity is calculated. The time difference between the end time of the last benchmark activity and 0:00 of the next day is added to this result, and this is used as the user's activity out-of-town time of the day. The specific calculation method is shown in formula (2).
[0085]
[0086] Where D i is the time of user i’s outing, h(j) is the sequence number of user i’s activity reference point, A n It is the user's last activity node of the day, and H represents the activity benchmark point.
[0087] Step 3.3: Daily activity count
[0088] The number of daily activities is defined as the number of user activity nodes in a single day. The daily activity frequency of users reflects the quantitative characteristics of the node structure of individual activity chains. The calculation method is shown in formula (3):
[0089]
[0090] Where P i is the number of daily activities of user i, A i,j is the j-th active node of user i.
[0091] Step 4: Construct a data set of user travel attributes based on the spatiotemporal intensity of residents' activities, the time of going out for activities, and the number of daily activities. Calculate the similarity of travel attributes between users based on the data set of user travel attributes. Iterate and calculate the user's attractiveness and belongingness based on the similarity until the number of iterations reaches the set number or the iteration meets the set convergence parameters. Obtain the user cluster center based on the attractiveness and belongingness at the current moment, and obtain the recognition result of the resident travel activity pattern based on the user cluster center.
[0092] Affinity Propagation (AP) clustering algorithm is used to cluster users with different travel attribute characteristics to obtain travel groups with different activity modes. The basic implementation steps are as follows:
[0093] Step 4.1: Based on the spatiotemporal intensity of residents' activities, time spent outside, and number of daily activities in step 3, construct a user travel attribute dataset:
[0094] Input user travel attribute dataset X = {x 1 ,x 2 ,…,xi,…,xn}, where x i Represents multi-dimensional travel feature information {I i ,D i ,P i} the i-th user.
[0095] Step 4.2: Calculate user similarity Its value is the negative Euclidean distance between the three major travel attributes of each user.
[0096] The diagonal S(i,i) value of the similarity matrix represents the degree to which the i-th user is the cluster center. In order to avoid the S(i,i) calculated by the similarity formula being 0, S(i,i) is generally replaced by the median of S(i,k) without prior knowledge.
[0097] Step 4.3: Initialize attraction R(i,k) and attribution A(i,k), both of which are zero matrices isomorphic to the similarity matrix S(i,k). Attraction is used to describe the degree to which user k is suitable as the cluster center of user i, and attribution is used to describe the degree to which user i chooses user k as its cluster center.
[0098] Step 4.4: Based on the user similarity S(i,k), the user attraction R(i,k) and the attribution A(i,k) are continuously updated iteratively, where λ is the attenuation coefficient to avoid iterative oscillation. ;
[0099] The attraction R(i,k) is iteratively calculated according to formulas (4) and (5):
[0100]
[0101] R(i,k) new =(1-λ)R(i,k) t +λR(i,k) (5)
[0102] The attribution degree A(i,k) is iteratively calculated according to formulas (6) and (7):
[0103]
[0104] A(i,k) new =(1-λ)A(i,k) t +λA(i,k) (7)
[0105] Step 4.5: Repeat the above iterative process until the algorithm converges or reaches the maximum number of iterations, stop updating and generate a series of high-quality user cluster centers.
[0106] Step 4.6: According to the cluster center results, all users are matched to the cluster center with the closest attributes, forming and outputting result clusters representing users with different travel characteristics, including users with low daily activity number, low activity time outside, and low activity intensity; users with high daily activity number, high activity time outside, and low activity intensity; users with high daily activity number, high activity time outside, and high activity intensity.
[0107] The method disclosed in the present invention uses mobile phone signaling data as passive collection, and has individual spatiotemporal accompanying data with wide coverage and low collection cost, which can overcome the defects of traditional resident survey data such as poor accuracy, high cost, and difficulty in organization. Secondly, exploring the laws of individual travel activities from the perspective of individual activities can essentially grasp the transportation travel needs and avoid the aggregation errors of the traditional "four-stage" traffic planning model. In addition, based on the activity space index, the activity time weighting is considered to establish a characterization index that effectively reflects the comprehensive spatiotemporal intensity of residents' activities, and innovatively introduces the idea of pattern division from the activity chain research to the activity intensity characteristic analysis, which can provide effective method support for grasping the laws of urban residents' activities.
[0108] This embodiment also discloses a practical case
[0109] A case analysis was conducted using Guiyang's signaling data for a certain day to identify the characteristics of residents' activity intensity and travel activity patterns, as follows:
[0110] Step 1: Mobile phone signaling, base station data collection and coordinate matching
[0111] The communication positioning data used in this case is provided by China Unicom, which is divided into two parts: mobile phone signaling data and communication base station data. Tables 2 and 3 show the original mobile phone signaling data and original base station data after removing the communication event information, respectively. The coordinates of longitude and latitude are matched to the mobile phone signaling data through the LAC and CELLID fields, and irrelevant spatiotemporal attribute fields are removed to obtain the signaling data with the final matched longitude and latitude coordinates as shown in Table 4.
[0112] Table 2 Mobile phone signaling data sample
[0113]
[0114] Table 3. Example of base station data in Guiyang
[0115]
[0116]
[0117] Table 4. Mobile phone signaling data matching coordinate results
[0118]
[0119] like Figure 1 The following is a comparison chart of the effects before and after data cleaning. Figure 1 (a) is the user’s original travel trajectory, Figure 1 (b) is the user trajectory data after cleaning the mobile phone signaling drift data and ping-pong data. It can be clearly seen that after the data preprocessing step, the user's stay period data has obvious characteristics, and the noise-free data interferes with the judgment of travel and stay status, which is expected to greatly improve the recognition effect of travel endpoints.
[0120] Step 2: Active Node Identification
[0121] See also Figure 2 By establishing a two-stage base station sequence inference merging method, τ and d are set to 30min and 400m respectively to identify all the user's activity nodes in turn, and select the activity node with the longest duration from 00:00 to 8:00 from the user's activity nodes as the activity benchmark point. The activity benchmark points and activity nodes basically cover the entire city. This result is used as the basis for calculating the activity feature index in step 3.
[0122] Step 3: Calculation of individual activity spatiotemporal characteristic indicators
[0123] See also Figure 3 , which is the calculation result of the three characteristic indicators of residents’ activities. Figure 3(a) is the distribution of user activity spatiotemporal intensity results, among which 86.45% of users’ activity spatiotemporal intensity is less than 50 km·h, 95.55% of users’ activity spatiotemporal intensity is less than 150 km·h, and most users’ activity spatiotemporal intensity is stably distributed in a fixed range. However, there are still a small number of users whose activity spatiotemporal intensity is greater than 150 km·h, and even a very small number of users whose activity spatiotemporal intensity reaches above 1000 km·h. This is also consistent with the actual travel characteristics of citizen users. There are a small number of users who travel long distances and for a long time, such as taxi drivers and food delivery riders, in the city, which will result in a higher calculated value of activity spatiotemporal intensity. Figure 3 (b) is the distribution of user activity time. The activity time is distributed in the range of 0h-20h. In the range of 1h-17h, the number of users included in each hour is basically the same, accounting for about 3.50% of the total number of users. Among them, the proportion of users in the 0h-1h range is significantly higher, accounting for 19.17% of the total number of users, indicating that there are more people at home on that day. There are fewer users in the 17h-20h range, and the sum of users in the three ranges only accounts for about 2.51% of the total number of users, which may include a small number of overtime workers and people who spend a long time entertaining. Figure 3 (c) is the distribution result of the number of daily activities of users. Most users have 1 to 13 daily activities, among which the number of people who have 6 to 8 activities is the highest, accounting for 32.58% of the total number of people, which is the main representative activity structure form of users on that day.
[0124] Step 4: Travel activity mode segmentation
[0125] Figure 4 The results of user travel activity pattern classification based on the clustering of three major travel characteristic indicators. In order to avoid unsatisfactory clustering results caused by the dimensional differences between different indicators, the Z-Score standardization is used to pre-process the input data before clustering, and then the clustering algorithm is input in turn to obtain the clustering results of user travel patterns. From the clustering result diagram, it can be found that all users are divided into three categories according to the clustering results. From the analysis of activity characteristics, the three categories of people are: ① users with low daily activity number, low activity time and low activity intensity. Such users are mainly distributed in the suburbs and surrounding counties. The transportation convenience in the area is low, and users do not generate activity trips or have few activity trips; ② users with high daily activity number, high activity time and low activity intensity. Such users are mainly distributed in the core urban area. They are the activity trips of urban residents, which are characterized by multiple frequencies and long-term outings, but the activity locations are close to the activity benchmark points; ③ users with high daily activity number, high activity time and high activity intensity. Such users are relatively few and are mainly distributed around the core areas of surrounding districts and counties and near railway stations. There is a certain attraction for transportation travel, and the trips generated are mostly long-distance trips.
[0126] This embodiment also discloses a system for identifying resident travel activity patterns based on mobile phone signaling data, including:
[0127] A data acquisition module, used to obtain mobile phone signaling data and base station data of residents' travel activities;
[0128] Obtain residents' travel activity nodes and activity benchmarks based on mobile phone signaling data and base station data;
[0129] Based on the activity nodes and activity benchmarks of residents' travel, the spatiotemporal intensity of residents' activities, the time spent outside the activity, and the number of daily activities are calculated;
[0130] A data set of user travel attributes is constructed based on the spatiotemporal intensity of residents' activities, the time they go out for activities, and the number of daily activities. The similarity of travel attributes between users is calculated based on the data set of user travel attributes. The user's attractiveness and belongingness are iteratively calculated based on the similarity until the number of iterations reaches the set number or the iterations meet the set convergence parameters. The user clustering center is obtained based on the attractiveness and belongingness at the current moment, and the recognition result of residents' travel activities is obtained based on the user clustering center.
[0131] A schematic diagram of a terminal device provided in an embodiment of the present invention. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are implemented.
[0132] The computer program may be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to accomplish the present invention.
[0133] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.
[0134] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
[0135] The memory may be used to store the computer program and / or module, and the processor implements various functions of the terminal device by running or executing the computer program and / or module stored in the memory and calling the data stored in the memory.
[0136] If the module / unit integrated in the terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.
[0137] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A method for identifying residents' travel activity patterns based on mobile phone signaling data, characterized in that: The following steps are involved: Obtain mobile phone signaling data and base station data on residents’ travel activities; Obtain residents' travel activity nodes and activity benchmarks based on mobile phone signaling data and base station data; Based on the activity nodes and activity benchmarks of residents' travel, the spatiotemporal intensity of residents' activities, the time spent outside the activity, and the number of daily activities are calculated; A data set of user travel attributes is constructed based on the spatiotemporal intensity of residents' activities, the time they go out for activities, and the number of daily activities. The similarity of travel attributes between users is calculated based on the data set of user travel attributes. The user's attractiveness and belongingness are iteratively calculated based on the similarity until the number of iterations reaches the set number or the iterations meet the set convergence parameters. The user clustering center is obtained based on the attractiveness and belongingness at the current moment, and the recognition result of the residents' travel activity pattern is obtained based on the user clustering center.
2. A method for identifying resident travel activity patterns based on mobile phone signaling data according to claim 1, characterized in that: The activity nodes and activity reference points for obtaining residents' travel include: The mobile phone signaling data continuously recorded by the same base station is merged to identify the base station short stop S in turn. i {MSID,Lon i ,Lat i ,T i S ,T i E }; When two consecutive base stations are briefly parked at S i With S i+1 Satisfy Δt between i,i+1 ≤τ and Δx i,i+1 When ≤d, two consecutive base station short stops are merged to obtain the active node A. i , where Δt i,i+1 represents the time difference of short dwell between consecutive base stations, Δx i,i+1 Indicates the distance difference between consecutive base station short dwells; Based on the obtained activity nodes, the activity node with the longest duration during the period of 0:00-8:00 of the user is defined as the activity reference point.
3. The method for identifying resident travel activity patterns based on mobile phone signaling data according to claim 1, characterized in that: The spatiotemporal intensity of residents' activities is calculated by the following formula: In the formula, I i represents the temporal and spatial intensity of user i’s daily travel activities, and Represent the start time and the leave time of the jth activity node of user i, Dis(A i,j ,A i,h ) is the distance between the jth activity node of user i and the activity reference point; The activity out-of-office time is calculated by the following formula: Where D i The time spent outside by user i. h ( j ) refers to the activity reference point sequence number of user i, A n It is the user's last activity node of the day, and H represents the activity benchmark point; The daily activity number is calculated by the following formula: Where P i is the number of daily activities of user i, A i,j is the j-th active node of user i.
4. The method for identifying resident travel activity patterns based on mobile phone signaling data according to claim 1, characterized in that: The data set of user travel attributes is X = {x1, x2, ..., x i ,…,x n }, where x i Represents multi-dimensional travel feature information {I i ,D i ,P i } the i-th user.
5. A method for identifying resident travel activity patterns based on mobile phone signaling data according to claim 4, characterized in that: The calculating of user similarity comprises: Among them, I i represents the temporal and spatial intensity of user i’s daily travel activities; D i P is the time spent outside the home by user i; i The number of daily user activities.
6. A method for identifying resident travel activity patterns based on mobile phone signaling data according to claim 5, characterized in that: Iterative calculation of user attraction and belonging based on similarity includes: First, initialize the user's attraction and belongingness, so that both attraction and belongingness are zero matrices isomorphic to the similarity matrix S(i,k), where attraction R(i,k) is used to describe the degree to which user k identifies user i as the cluster center, and belongingness A(i,k) is used to describe the degree to which user i identifies user k as its cluster center; Based on the similarity S(i,k), the user attraction R(i,k) and the attribution A(i,k) are continuously updated iteratively.
7. The method for identifying resident travel activity patterns based on mobile phone signaling data according to claim 5, characterized in that: The iterative calculation of the user's attractiveness R(i,k) includes: R(i,k) new =(1-λ)R(i,k) t +λR(i,k) The iterative calculation of the user's attribution degree A(i,k) includes: A(i,k) new =(1-λ)A(i,k) t +λA(i,k) Among them, S(i,k) represents the user similarity, and λ is the attenuation coefficient to avoid iterative oscillation.
8. A system for identifying resident travel activity patterns based on mobile phone signaling data, characterized in that: include: A data acquisition module, used to obtain mobile phone signaling data and base station data of residents' travel activities; Obtain residents' travel activity nodes and activity benchmarks based on mobile phone signaling data and base station data; Based on the activity nodes and activity benchmarks of residents' travel, the spatiotemporal intensity of residents' activities, the time spent outside the activity, and the number of daily activities are calculated; A data set of user travel attributes is constructed based on the spatiotemporal intensity of residents' activities, the time they go out for activities, and the number of daily activities. The similarity of travel attributes between users is calculated based on the data set of user travel attributes. The user's attractiveness and belongingness are iteratively calculated based on the similarity until the number of iterations reaches the set number or the iterations meet the set convergence parameters. The user clustering center is obtained based on the attractiveness and belongingness at the current moment, and the recognition result of residents' travel activities is obtained based on the user clustering center.
9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.
