A method for reconstructing activities and travel chains of travelers based on mobile phone signaling data

Abstract

The application provides a kind of traveller activity-travel chain reconstruction method based on mobile phone signaling data, comprising the following steps: input original mobile phone signaling data and carry out data pretreatment to mobile phone signaling data;Potential stop point and displacement point are judged by the improved ST-DBSCAN density clustering algorithm applied to pretreated mobile phone signaling data, and potential stop point sequence and potential displacement point sequence are obtained;Travel OD is identified using OD identification algorithm based on dynamic threshold, and travel information such as O point position and stay start and end time, D point position and stay start and end time and travel duration is extracted;According to the stay information of travel O point and D point, activity duration, activity location, activity start time, activity end time activity information is extracted;Residence and work place are identified according to effective stop point information in user multi-day trajectory;The activity type of user is inferred according to the distance between user stop point position and residence and work place;According to identified travel information and activity information, activity-travel chain of traveller is generated in sequence.

Description

Technical Field

[0001] This invention belongs to the field of traffic big data, and in particular relates to a method for reconstructing traveler activities and travel chains based on mobile phone signaling data. Background Technology

[0002] Mobile signaling data is a new type of big data source. Compared with traditional data, mobile signaling data has advantages such as high user coverage, real-time performance, completeness, and full spatiotemporal coverage of travel, which helps to comprehensively explore the travel and activity characteristics of urban users. Currently, the technology for processing and analyzing mobile signaling data is not yet perfect. Mobile signaling data is large in volume and suffers from problems such as ping-pong data and drifting data. Existing methods for processing mobile signaling data often ignore these issues, resulting in low accuracy in identifying user travel and activity information, and a discrepancy between the data and actual user travel patterns. Summary of the Invention

[0003] Purpose of the invention: One of the purposes of this invention is to provide a method for reconstructing traveler activity-travel chain based on mobile phone signaling data. This method can use mobile phone signaling data to identify travel and activity information, thereby reconstructing the traveler's activity-travel chain within a day, providing data support and decision support for urban transportation system management and planning.

[0004] Technical Solution: To achieve the above objectives, this invention proposes a method for reconstructing traveler activities and travel chains based on mobile phone signaling data. This method includes the following steps:

[0005] (1) Input the original mobile phone signaling data and perform data preprocessing on the original mobile phone signaling data. The preprocessing includes data cleaning, ping-pong data processing and drift data processing.

[0006] (2) Based on the preprocessed mobile signaling data, the ST-DBSCAN density clustering algorithm is applied to determine potential stationary points and displacement points, and to obtain the potential stationary point sequence and the potential displacement point sequence.

[0007] (3) Based on the potential station sequence obtained in step (2), the OD identification algorithm based on dynamic threshold is used to identify travel OD and extract travel information of travelers by setting dynamic thresholds according to the traveler's base station dwell time, travel time between base stations and travel distance between base stations respectively.

[0008] (4) Extract activity information such as activity duration, activity location and activity start time by identifying the stay information of the travel point O and D. The activity duration is obtained by subtracting the stay end time of the O or D point from the stay start time. The activity location corresponds to the O and D point locations. The activity start time corresponds to the O or D point stay start time.

[0009] (5) Identify the place of residence and workplace based on the valid station information in the user's multi-day trajectory;

[0010] (6) Infer the user's activity type based on the distance of the user's stop location from their residence and workplace;

[0011] (7) Based on the identified travel information and activity information, the traveler activity-travel chain is reconstructed in sequence.

[0012] Preferably, the specific method of step (1) is as follows:

[0013] (11) After extracting the original mobile phone signaling data, the original data is cleaned. Data cleaning refers to correcting the missing or erroneous data in the original data.

[0014] (12) Input the cleaned mobile signaling data. First, merge adjacent duplicate base stations. If the base stations recorded in the two adjacent mobile signaling data in the data are the same, then merge the two adjacent records. After merging, the base station stay start time of the data record is the base station stay start time of the previous data before merging, and the base station stay end time of the data record after merging is the base station stay end time of the next data before merging.

[0015] (13) Ping-pong data processing: After merging adjacent duplicate base stations, calculate the traveler's dwell time at each base station. For each mobile signaling data record, iteratively judge its temporal and spatial characteristics: Let the base station dwell time threshold be t during the ping-pong data processing. stay If the time the record stays at the current base station is less than t stay The dwell time at both the previous and subsequent base stations is greater than t. stay If the preceding and following base stations are identical, then the record is considered ping-pong data, marked, and deleted. After deleting ping-pong data, new adjacent duplicate base station information may appear. The process of merging adjacent duplicate base stations and removing ping-pong data is repeated multiple times until no current base station has a dwell time less than t. stay The dwell time at both the previous and subsequent base stations is greater than t. stay If the previous base station and the next base station are the same, the iteration stops and the ping-pong data processing is complete.

[0016] (14) Drift data processing: First, calculate the distance between each signaling record and the preceding and following base stations, and then calculate the transfer speed of each signaling record with respect to the preceding and following base stations. Let the distance threshold between base stations during drift data processing be d. shift The inter-base station transfer speed threshold is v shift If the distance between the current base station and the base stations before and after it is greater than d shift And the transfer speed between base stations is greater than vshift If the record is found to be drift data, it is deleted. After removing the drift data, adjacent duplicate base stations are merged and the traveler's stay time at each base station is re-acquired.

[0017] Preferably, the method for step (2) is as follows:

[0018] (21) Sort the preprocessed mobile signaling data of each user according to time;

[0019] (22) Set the ST-DBSCAN algorithm parameters, set ε1 as the spatial threshold, ε2 as the time threshold, MinPts as the minimum number of individuals in the cluster threshold, and Δε as the clustering threshold. Set the parameter values ​​according to the travel and stay characteristics of travelers.

[0020] (23) For each user's mobile phone signaling data sequence, input the ST-DBSCAN algorithm to perform clustering until all data points in the data sequence have been processed and obtain the clustering results. Points within the cluster are potential stationary points; points outside the cluster are potential displacement points.

[0021] Preferably, step (3) specifically includes the following steps:

[0022] (31) Record the potential stationary sequence obtained in step (2) For the potential stop sequence of traveler i during the day, T represents the corresponding base station coverage area. travel D serves as the baseline threshold for travel time between traveler base stations. travel T serves as the baseline threshold for the travel distance between traveler base stations. stay Using the base station dwell time of travelers as a baseline threshold, travel origin-destination (OD) identification is performed as follows: Calculate the traveler i at each potential base station. Duration of stay like Then remove the k-th record from the base station trajectory sequence, where k∈[1,n1]∩N * Based on this, by iterating through each potential stop data point for all travelers, a preliminary effective stop sequence is identified.

[0023] (32) Based on the preliminary effective station sequence Continue calculating the traveler i at two adjacent stations Travel time between and travel distance like or Then remove the j-th record from the base station trajectory sequence, where j∈[1,n2]∩N * It then iterates through each preliminary valid checkpoint data point for each traveler to identify the valid checkpoint sequence.

[0024] (33) After obtaining the effective station sequence, the processed effective station sequence for traveler i in one day remember For travelers, the base station is the starting point of their journey. As the final base station for a trip, connecting The origin and destination (OD) of a traveler's trip;

[0025] (34) After obtaining the travel OD, obtain the location of point O, the start and end time of the stay at point O, the location of point D, the start and end time of the stay at point D, and the travel time by subtracting the start time of the stay at point D from the end time of the stay at point O.

[0026] Preferably, the specific method for step (4) is as follows:

[0027] Activity information such as duration, location, start time, and end time is extracted by identifying the stop information at points O and D. Accordingly, the activity duration is obtained by subtracting the start time from the end time at point O or D; the activity location corresponds to the location at point O and point D; the activity start time corresponds to the start time at point O or point D; and the activity end time corresponds to the end time at point O or point D.

[0028] Preferably, the specific method for step (5) is as follows:

[0029] Based on the effective location data from travelers' multi-day trajectories, let the threshold for the duration of stay at the workplace be T. work If traveler i stays at the base station for a maximum of [number] hours during the preset daytime period on a weekday, And at the base station The length of stay at the workplace is greater than the workplace stay threshold T. work If the base station is considered to be the base station corresponding to traveler i's workplace, then traveler i is considered to have stayed at the base station for the longest period of time during the night. Then the base station is considered to be the base station corresponding to the residence of traveler i.

[0030] Preferably, the specific method for step (6) is as follows:

[0031] If the distance between the user's activity point and the base station corresponding to the work location is less than the first preset value, and the activity time is within the preset daytime period, it is considered to be a work activity.

[0032] If the distance between a user's activity point and the base station corresponding to their residence is less than the second preset value, and the activity time falls within a preset nighttime period, it is considered a home activity.

[0033] If a user's activity does not meet either of the above two conditions, then the user's activity will be set to another activity.

[0034] Preferably, the method for step (7) is as follows:

[0035] Based on the traveler's O-point location, O-point stay end time, D-point location, D-point stay start time and trip time identified in step (3), the traveler's travel chain is formed by corresponding to the travel origin, travel start time, travel destination, travel end time and travel duration. Based on the activity location, activity start time, activity end time and activity duration extracted in step (4), and combined with the activity type identified in step (6), the traveler's activity chain is formed. The activity chain and travel chain are combined and reconstructed in chronological order to generate the traveler's activity-travel chain.

[0036] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0037] Compared with existing technologies, this invention has the following advantages: Considering the characteristics of mobile phone signaling data, this invention proposes a method for reconstructing traveler activities and travel chains based on mobile phone signaling data. Compared with traditional traveler activity research based on traditional data sources such as questionnaire data and card swipe data, the method proposed in this invention has advantages such as full spatiotemporal coverage, wide coverage of the population, and high accuracy. This invention provides a traveler activity and travel chain reconstruction process that includes mobile phone signaling data preprocessing, potential stopover and displacement point identification, OD identification, travel information identification, residence and workplace identification, activity information identification, and daily activity and travel chain generation. It emphasizes a standardized system framework, is transferable to different cities and application scenarios with different data, allows for customized input parameters, has high compatibility, and can reconstruct user activity and travel chains in different scenarios, which is of great significance for urban traffic planning and management. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0039] Figure 2 A flowchart of a table tennis data processing algorithm;

[0040] Figure 3 A flowchart of the drift data processing algorithm;

[0041] Figure 4 Example of an activity-travel chain. Detailed Implementation

[0042] The technical solution of the present invention will be further described below with reference to the accompanying drawings and implementation examples.

[0043] The present invention provides a method for reconstructing the daily activities and travel chain of travelers based on mobile phone signaling data. The flowchart is shown below. Figure 1 Specifically, it includes the following steps:

[0044] (1) Input the original mobile phone signaling data and perform data preprocessing on the original mobile phone signaling data, including data cleaning, ping-pong data processing and drift data processing;

[0045] (2) Based on the preprocessed mobile signaling data, the improved ST-DBSCAN density clustering algorithm is applied to determine potential stationary points and displacement points, and to obtain potential stationary point sequences and potential displacement point sequences.

[0046] (3) Based on the potential station sequence obtained in step (2), the OD identification algorithm based on dynamic threshold is used to identify travel OD by setting dynamic thresholds according to the traveler's base station dwell time, travel time between base stations and travel distance between base stations, and extract travel information.

[0047] (4) Extract activity information such as activity duration, activity location, activity start time, and activity end time by identifying the stay information of the travel point O and point D. The activity duration is obtained by subtracting the stay end time of point O or point D from the stay start time. The activity location corresponds to the location of point O and point D. The activity start time corresponds to the stay start time of point O or point D. The activity end time corresponds to the stay end time of point O or point D.

[0048] (5) Identify the place of residence and workplace based on the valid station information in the user's multi-day trajectory;

[0049] (6) Infer the user's activity type based on the distance of the user's stop location from their residence and workplace;

[0050] (7) Based on the identified travel information and activity information, the traveler activity-travel chain is reconstructed in sequence.

[0051] Furthermore, the specific implementation steps for inputting the original mobile phone signaling data and performing data preprocessing in step (1) are as follows:

[0052] The raw mobile signaling data contains spatiotemporal sequence information of user movement. It is large in volume and contains issues such as ping-pong data and drift data. Therefore, after extracting the raw mobile signaling data, it is necessary to perform data cleaning and remove ping-pong and drift data. The flowchart for the ping-pong data processing algorithm is shown below. Figure 2 Flowchart of drift data processing algorithm (for reference) Figure 3 The following is an example of ping-pong data and drift data processing.

[0053]

[0054] Ping Pong Data Processing Example

[0055]

[0056] Drift data processing example

[0057] (11) First, the raw mobile signaling data is cleaned. Data cleaning mainly refers to processing obviously missing or erroneous data in the raw data to prevent them from affecting the subsequent data analysis results. The problems targeted by data cleaning include: missing data, duplicate data, abnormal data, and irrelevant data. Missing data refers to data with incomplete fields; duplicate data refers to the same data appearing repeatedly; abnormal data refers to data with missing fields or incorrect formats caused by the collection and storage process; irrelevant data refers to redundant data that is not related to the analysis content (such as some irrelevant fields). The above data may interfere with subsequent data calculation and analysis, so it is necessary to remove such data.

[0058] (12) Ping-Pong Data Processing: Input the cleaned mobile signaling data. First, merge adjacent duplicate base stations. If two adjacent mobile signaling data records the same base station, merge the two records. The base station dwell start time of the merged data record is the base station dwell start time of the previous data record before merging, and the base station dwell end time of the merged data record is the base station dwell end time of the next data record before merging. After merging adjacent duplicate base stations, calculate the traveler's dwell time at each base station. For each mobile signaling data record, its time and spatial characteristics are iteratively judged. In this example, a base station dwell time threshold t is set during the ping-pong data processing. stay =3min. If the dwell time of a record at the current base station is less than 3 minutes, but the dwell time at the previous and next base stations is greater than 3 minutes, and the previous and next base stations are the same, then the record is considered ping-pong data, marked, and deleted. After deleting ping-pong data, new adjacent duplicate base station information may appear. Adjacent duplicate base station merging and ping-pong data removal are repeated multiple times until no record has a dwell time of less than t at the current base station. stay The dwell time at both the previous and subsequent base stations is greater than t. stay If the previous base station and the next base station are the same, the iteration stops and the ping-pong data processing is complete.

[0059] (13) Drift Data Processing: Input the mobile signaling data after removing the ping-pong data. First, calculate the distance between each signaling record and the preceding and following base stations, and calculate the transfer speed of each signaling record with the preceding and following base stations. In this example, a threshold d for the distance between base stations is set during the drift data processing. shift =3km, inter-base station transfer speed threshold v shift=120km / h. If the distance between the current base station and the base stations before and after it is greater than 3km and the transfer speed between base stations is greater than 120km / h, then the record is determined to be drift data and deleted. After removing the drift data, adjacent duplicate base stations are merged and the traveler's stay time at each base station is recalculated.

[0060] Furthermore, step (2) applies the improved ST-DBSCAN density clustering algorithm to determine potential stationary points and displacement points, thereby obtaining a sequence of potential stationary points and a sequence of potential displacement points. The specific implementation steps are as follows:

[0061] (21) First, for each user, the preprocessed mobile signaling data is sorted according to time.

[0062] (22) Set the ST-DBSCAN algorithm parameters. Set ε1 as the spatial threshold, ε2 as the time threshold, MinPts as the minimum number of individuals in the cluster threshold, and Δε as the clustering threshold. Set reasonable parameter values ​​according to the travel and stay characteristics of travelers.

[0063] (23) For each user's mobile signaling data sequence, input the ST-DBSCAN algorithm for clustering until all data points in the data sequence have been processed. Obtain the clustering results; the set of points within a cluster is the potential stationary point, and the points outside the cluster (i.e., noise points) are the potential displacement points.

[0064] Furthermore, step (3) involves identifying travel origin (OD) based on the potential station sequence identified in step (2). The specific implementation process is as follows:

[0065] (31) Input the potential stationary point sequence obtained in step (2), and denote it as... For the potential stop sequence of traveler i during the day, T represents the corresponding base station coverage area. travel D serves as the baseline threshold for travel time between traveler base stations. travel T serves as the baseline threshold for the travel distance between traveler base stations. stay Using the base station dwell time of travelers as a baseline threshold, travel origin-destination (OD) identification is performed as follows: Calculate the traveler i at each potential base station. Duration of stay like Then remove the k-th record from the base station trajectory sequence, where k∈[1,n1]∩N * Based on this, by iterating through each potential stop data point for all travelers, a preliminary effective stop sequence is identified.

[0066] (32) Based on the preliminary effective station sequence Continue calculating the traveler i at two adjacent stations Travel time between and travel distance like or Then remove the j-th record from the base station trajectory sequence, where j∈[1,n2]∩N * It then iterates through each preliminary valid checkpoint data point for each traveler to identify the valid checkpoint sequence.

[0067] (33) After obtaining the effective station sequence, the processed effective station sequence for traveler i in one day remember For travelers, the base station is the starting point of their journey. As the final base station for a trip, connecting The origin and destination (OD) of a traveler's trip;

[0068] (34) After obtaining the travel OD, obtain the location of point O, the start and end time of the stay at point O, the location of point D, the start and end time of the stay at point D, and the travel time by subtracting the start time of the stay at point D from the end time of the stay at point O.

[0069] Furthermore, step (4) extracts activity information such as activity duration, activity location, activity start time, and activity end time by identifying the stay information at points O and D. Accordingly, the activity duration can be obtained by subtracting the stay end time at point O or D from the stay start time, the activity location corresponds to the location at point O and point D, the activity start time corresponds to the stay start time at point O or point D, and the activity end time corresponds to the stay end time at point O or point D.

[0070] Furthermore, step (5) identifies the place of residence and place of work based on the valid station data identified in step (3). The specific implementation process is as follows:

[0071] Identify residence and workplace based on valid location information from users' multi-day trajectory data: Set a threshold value of T for the duration of stay at the workplace based on the valid location data from the user's multi-day trajectory data. work If traveler i stays at a base station for the longest period during weekday daytime hours (8:00 AM – 10:00 PM) is And at the base station The length of stay at the workplace is greater than the workplace stay threshold T. work If the base station is the one corresponding to traveler i's workplace, then the base station is considered to be the one where traveler i stays the longest during the nighttime period (22:00–8:00). Then the base station is considered to be the base station corresponding to the residence of traveler i.

[0072] Furthermore, step (6) infers the user's activity type based on the residential and workplace information identified in step (5) according to the distance of the user's stop location from their residential and workplace locations. The specific process is as follows:

[0073] If the distance between the user's activity point and the base station corresponding to the work location is less than the first preset value, and the activity time is within the preset daytime period, it is considered to be a work activity.

[0074] If the distance between a user's activity point and the base station corresponding to their residence is less than the second preset value, and the activity time falls within a preset nighttime period, it is considered a home activity.

[0075] If a user's activity does not meet either of the above two conditions, then the user's activity will be set to another activity.

[0076] Furthermore, in step (7), based on the traveler's O-point location, O-point stay end time, D-point location, D-point stay start time, and trip time identified in step (3), corresponding to the travel origin, travel start time, travel destination, travel end time, and travel duration, a traveler's travel chain is formed; based on the activity location, activity start time, activity end time, and activity duration extracted in step (4), combined with the activity type identified in step (6), a traveler's activity chain is formed; combining the activity chain and travel chain, a traveler's activity-travel chain is reconstructed in chronological order to realize the reconstruction of the traveler's daily activity-travel chain. An example of an activity-travel chain is shown below. Figure 4 As shown.

Claims

1. A method for reconstructing traveler activity-travel chain based on mobile phone signaling data, characterized in that, Includes the following steps: (1) Input the original mobile phone signaling data and perform data preprocessing on the original mobile phone signaling data. The preprocessing includes data cleaning, ping-pong data processing and drift data processing. (2) Based on the preprocessed mobile signaling data, the ST-DBSCAN density clustering algorithm is applied to determine potential stationary points and displacement points, and to obtain the potential stationary point sequence and the potential displacement point sequence. (3) Based on the potential station sequence obtained in step (2), the OD identification algorithm based on dynamic threshold is used to identify travel OD and extract travel information of travelers by setting dynamic thresholds according to the traveler's base station dwell time, travel time between base stations and travel distance between base stations respectively. (4) Extract activity information such as activity duration, activity location and activity start time by identifying the stay information of the travel point O and D. The activity duration is obtained by subtracting the stay end time of the O or D point from the stay start time. The activity location corresponds to the O and D point locations. The activity start time corresponds to the O or D point stay start time. (5) Identify the place of residence and workplace based on the valid station information in the user's multi-day trajectory; (6) Infer the user's activity type based on the distance of the user's stop location from their residence and workplace; (7) Based on the identified travel and activity information, reconstruct and generate the traveler's activity-travel chain in sequence; Step (3) specifically includes the following steps: (31) Record the potential stationary point sequence obtained in step (2) For travelers Potential stationary sequence throughout the day This represents the corresponding base station coverage area. This serves as the baseline threshold for travel time between base stations for travelers. This serves as the baseline threshold for the travel distance between traveler base stations. Using the base station dwell time of travelers as a basic threshold, travel origin-destination (OD) identification is performed as follows: Calculate the traveler's origin-destination (OD) time. At each potential outpost Duration of stay ,like Then remove the first one from the base station trajectory sequence. 1 record Based on this, by iterating through each potential stop data point for all travelers, a preliminary effective stop sequence is identified. ; (32) Based on the preliminary effective station sequence Continue calculating travelers At two adjacent outposts Travel time between and travel distance ,like or Then remove the first one from the base station trajectory sequence. 1 record ; It then iterates through each preliminary valid checkpoint data point for each traveler to identify the valid checkpoint sequence. ; (33) After obtaining the valid station sequence, for travelers Effective station sequence after processing throughout the day ,remember For travelers The starting point of a trip, As the final base station for a trip, connecting This constitutes a traveler's origin-destination (OD) journey; (34) After obtaining the travel OD, obtain the location of point O, the start and end time of the stay at point O, the location of point D, the start and end time of the stay at point D, and the travel time by subtracting the start time of the stay at point D from the end time of the stay at point O.

2. The method for reconstructing traveler activity-travel chain based on mobile phone signaling data according to claim 1, characterized in that, The specific method for step (1) is as follows: (11) After extracting the original mobile phone signaling data, the original data is cleaned. Data cleaning refers to correcting the missing or erroneous data in the original data. (12) Input the cleaned mobile signaling data. First, merge adjacent duplicate base stations. If the base stations recorded in the two adjacent mobile signaling data are the same, then merge the two adjacent records. After merging, the base station dwell start time of the data record is the base station dwell start time of the previous data before merging, and the base station dwell end time of the data record after merging is the base station dwell end time of the next data before merging. (13) Ping-pong data processing: After merging adjacent duplicate base stations, calculate the traveler's dwell time at each base station. For each mobile signaling data record, iteratively judge its temporal and spatial characteristics: Let the base station dwell time threshold during ping-pong data processing be... If the time the record stays at the current base station is less than The dwell time at both the previous and subsequent base stations is greater than If the preceding and following base stations are identical, then the record is considered ping-pong data, marked, and deleted. After deleting ping-pong data, new adjacent duplicate base station information may appear. The process of merging adjacent duplicate base stations and removing ping-pong data is repeated multiple times until no current base station has a dwell time of less than [a certain duration]. The dwell time at both the previous and subsequent base stations is greater than [a certain value]. If the previous base station and the next base station are the same, the iteration stops and the ping-pong data processing is complete. (14) Drift data processing: First, calculate the distance between each signaling record and the preceding and following base stations, and then calculate the transfer speed of each signaling record with respect to the preceding and following base stations. Assume the distance threshold between base stations during drift data processing is... The inter-base station transfer speed threshold is If the distance between the current base station and the base stations before and after it is greater than... And the transfer speed between base stations is greater than If the record is found to be drift data, it is deleted. After removing the drift data, adjacent duplicate base stations are merged and the traveler's stay time at each base station is re-acquired.

3. The method for reconstructing traveler activity-travel chain based on mobile phone signaling data according to claim 2, characterized in that, The method for step (2) is as follows: (21) Sort the preprocessed mobile signaling data of each user according to time; (22) Set the ST-DBSCAN algorithm parameters, set Spatial threshold, As a time threshold, The threshold for the minimum number of individuals within a cluster. The clustering threshold is set by setting parameter values ​​based on the travel and stay characteristics of travelers. (23) For each user's mobile phone signaling data sequence, input the ST-DBSCAN algorithm to perform clustering until all data points in the data sequence have been processed and obtain the clustering results. Points within the cluster are potential stationary points; points outside the cluster are potential displacement points.

4. The method for reconstructing traveler activity-travel chain based on mobile phone signaling data according to claim 1, characterized in that, The specific method for step (4) is as follows: Activity information such as activity duration, activity location, activity start time, and activity end time is extracted by identifying the stay information at points O and D. The activity duration is obtained by subtracting the stay start time from the stay end time at point O or D. The activity location corresponds to the location at point O and point D. The activity start time corresponds to the stay start time at point O or point D. The activity end time corresponds to the stay end time at point O or point D.

5. The method for reconstructing traveler activity-travel chain based on mobile phone signaling data according to claim 1, characterized in that, The specific method for step (5) is as follows: Based on valid location data from travelers' multi-day travel trajectories, a threshold for the duration of stay at the workplace is set as follows: If travelers The maximum time a base station can stay during the preset daytime hours on weekdays is [time period]. And at the base station The length of stay exceeds the workplace stay threshold. If so, then the base station is considered to be for travelers. The workplace corresponds to the base station; if the traveler The longest time a base station can stay at night is [time period]. If so, then the base station is considered to be for travelers. The location of residence corresponds to the base station.

6. The method for reconstructing traveler activity-travel chain based on mobile phone signaling data according to claim 1, characterized in that, The specific method for step (6) is as follows: If the distance between the user's activity point and the base station corresponding to the work location is less than the first preset value, and the activity time is within the preset daytime period, it is considered to be a work activity. If the distance between a user's activity point and the base station corresponding to their residence is less than the second preset value, and the activity time falls within a preset nighttime period, it is considered a home activity. If a user's activity does not meet either of the above two conditions, then the user's activity will be set to another activity.

7. The method for reconstructing traveler activity-trip chain based on mobile phone signaling data according to claim 1, characterized in that, The method for step (7) is as follows: Based on the traveler's location at point O, end time of stay at point O, location at point D, start time of stay at point D, and travel time identified in step (3), the traveler's travel chain is formed by corresponding to the travel origin, travel start time, travel destination, travel end time, and travel duration. Based on the activity location, activity start time, activity end time, and activity duration extracted in step (4), and combined with the activity type identified in step (6), the traveler's activity chain is formed. The activity chain and travel chain are combined and reconstructed in chronological order to generate the traveler's activity-travel chain.

Images