Vehicle type-based track identification method and system based on mobile phone signaling data

Abstract

The invention discloses a vehicle type track identification method and system based on mobile phone signaling data, and the method comprises the steps: obtaining the mobile phone signaling data from an operator, and generating a track point sequence of a vehicle; on the basis of the space distance and the time interval between the track points, abnormal track points are obtained through judgment and removed, time-space interpolation complementation is carried out on long-time breakpoints, and sorting is carried out according to time; vehicle track statistical characteristics are extracted, wherein the vehicle track statistical characteristics comprise the total moving distance, the total parking frequency and the total parking time; the parking points are mapped to passenger transport, freight transport and mixed parking areas, passenger transport and freight transport feature scores are calculated for vehicles in the mixed parking areas respectively, and passenger transport or freight transport is judged after comprehensive scoring; for the vehicles which are judged to be passenger transport, the buses are identified based on the track repetition rate and the maximum number of mobile phones connected in the vehicles, then travel time and space regularity indexes are calculated, private cars and taxis are distinguished, and the taxis include online taxi-hailing cars; and outputting a classification result of each vehicle type and a corresponding complete track sequence.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent transportation, and particularly relates to a method and system for identifying vehicle trajectories by vehicle type based on mobile phone signaling data. Through the analysis and processing of mobile phone signaling data, accurate identification of vehicle trajectories of different vehicle types in complex fusion areas is achieved. Background Art

[0002] In the development process of intelligent transportation systems, accurate identification and analysis of vehicle trajectories are the key basis for realizing many applications such as traffic flow optimization, intelligent navigation, and traffic safety management. In complex fusion areas, such as commercial areas, transportation hubs, and port areas in the city center, the traffic flow is large, the vehicle types are diverse, and the driving trajectories are complex. Traditional vehicle trajectory identification methods face many challenges.

[0003] Existing vehicle trajectory identification methods mostly rely on specific sensor devices, such as GPS locators, in-vehicle OBD devices, etc. These methods not only require additional hardware devices to be installed on the vehicle, resulting in high costs, but also cannot effectively identify vehicles without such devices installed. In addition, some trajectory identification methods based on video surveillance are limited by the coverage of surveillance devices and environmental factors (such as weather, light, etc.), and it is difficult to achieve comprehensive and accurate vehicle trajectory identification in complex fusion areas.

[0004] Mobile phones, as widely used mobile devices in people's daily lives, generate signaling data that contains rich location and movement information. However, there is relatively little research on identifying vehicle trajectories by vehicle type using mobile phone signaling data at present, and there are problems such as low identification accuracy and inability to effectively process data in complex fusion areas.

[0005] Existing technologies mostly use the activity chain reconstruction technology for people's travel patterns based on mobile phone signaling data (such as: CN202210208596.9, CN202311199605.3), and there is relatively little research on vehicles based on mobile phone signaling. The invention patent CN202311711150.9 mentions a method for estimating the average number of passengers in a section of vehicles based on mobile phone signaling data, proving the effectiveness of mobile phone signaling for aggregate traffic travel identification. However, the existing technologies have not solved the problem of identifying vehicle trajectories by vehicle type based on usage, especially for private cars, taxis, freight vehicles, etc., which are difficult to distinguish only by the number of passengers in the vehicle. Summary of the Invention

[0006] The object of the present invention is to overcome the problems of high cost, limited coverage, low recognition accuracy, etc. existing in the vehicle trajectory recognition method in the prior art, and to provide a method and system for recognizing vehicle trajectories by vehicle type based on mobile phone signaling data. The present invention utilizes mobile phone signaling data to achieve efficient and accurate recognition of vehicle trajectories of different vehicle types such as private cars, buses, taxis (online car-hailing vehicles), and freight vehicles in complex fusion areas, and particularly provides data support for the subsequent large-scale popularization and application environment of intelligent connected vehicle-road collaboration and dynamic traffic management by vehicle type.

[0007] The object of the present invention is achieved by the following technical solutions:

[0008] A method for recognizing vehicle trajectories by vehicle type based on mobile phone signaling data, comprising:

[0009] S1. Trajectory data generation: Obtain mobile phone signaling data from the operator, and generate a sequence of trajectory points of the vehicle based on the mobile phone signaling data;

[0010] S2. Data cleaning and completion: Based on the spatial distance and time interval between trajectory points, determine and eliminate abnormal trajectory points, perform spatio-temporal interpolation and completion on long-time breakpoints, and sort them by time;

[0011] S3. Extract statistical features of vehicle trajectories: Divide trajectory segments according to base station handover data, determine parking points based on speed and stay time thresholds, and obtain statistical features of vehicle trajectories, including total moving distance, total number of stops, and total stop time;

[0012] S4. Passenger and freight attribute determination: Map the parking points to passenger, freight, and mixed parking areas, calculate the passenger and freight feature scores for the vehicles in the mixed parking area respectively, and determine passenger or freight after comprehensive scoring;

[0013] S5. Passenger vehicle type classification: For the vehicles determined to be passenger vehicles, first identify buses based on the trajectory repetition rate and the maximum number of connected mobile phones in the vehicle, and then calculate the travel time and spatial regularity indexes to distinguish private cars and taxis, where taxis include online car-hailing vehicles;

[0014] S6. Result output: Output the classification results of each vehicle type and the corresponding complete sequence of trajectory points.

[0015] Further, step S1 includes:

[0016] S101. Mobile phone signaling data is used to record the communication information between the user equipment and the base station, including the International Mobile Subscriber Identity (IMSI), base station ID, timestamp, signal strength, longitude and latitude information, and event type information, and the event types include location update, base station handover, call establishment, and call release;

[0017] S102, retaining location update, base station switching, base station ID, timestamp and latitude and longitude information, and filtering out incomplete or abnormal mobile phone signaling data;

[0018] S103: Query the location information of the base station according to the base station ID to form a base station location table ; For mobile phone signaling data Converted to location information, ;in, is the ith data in the mobile phone signaling data; Indicates the timestamp of the i-th mobile phone signaling data, that is, the specific time of the mobile phone signaling data;

[0019] S104, each Base station location As a track point , the trajectory point sequence of the generated vehicle is: ,in, , n represents the total number of trajectory points.

[0020] Further, step S2 includes:

[0021] S201, set judgment conditions: if , then remove the trajectory points ;in, Represents trajectory points and The spatial distance is calculated as follows: Among them, R is the radius of the earth, which is 6371km; represents the time interval between adjacent trajectory points, , Indicates the upper limit of reasonable speed; , Respectively represent the timestamps of the i-th and i+1-th mobile phone signaling data; , n represents the total number of trajectory points;

[0022] S202: If there are long breakpoints longer than 10 minutes in the trajectory point sequence, the breakpoints are completed by time interpolation and base station position interpolation. The interpolation formula is as follows:

[0023] ;

[0024] in, and Respectively represent the base station position and timestamp of the completed interpolation;

[0025] S203, the trajectory point sequence T is sorted by time Sorting;

[0026] 。

[0027] Further, step S3 includes:

[0028] S301. Identify continuous trajectory segments based on the base station handover data of mobile phone signaling data. , where each trajectory segment is defined as the path of the vehicle during continuous driving or stationary period, and the expression is as follows:

[0029] ;

[0030] Among them, , m represents the upper limit of the number of different trajectory segments into which the vehicle is divided;

[0031] If the time interval between two consecutive trajectory points , then they are divided into different segments;

[0032] S302. Calculate the speed of the vehicle between adjacent trajectory points . If and the parking time , then it is determined as a parking point; use S to represent the set of parking points, represents the spatial distance between trajectory points and ; takes 1 - 3 m / s, takes 5 - 10 min;

[0033] S303. Extract trajectory statistical features, including the total moving distance ; the total number of parking times , which represents the total number of trajectory points that meet the parking conditions in the trajectory point sequence; the total parking time , which represents the total stay time between parking points.

[0034] Further, step S4 includes:

[0035] S401. Analyze the parking location of the vehicle based on mobile phone signaling data, and divide the parking area into the following three categories:

[0036] Passenger parking area , for passenger vehicles, including station and airport parking areas;

[0037] Freight parking area , for freight vehicles, including port loading and unloading areas and logistics centers;

[0038] Mixed parking area , an area that allows both passenger and freight vehicles to dock, including the parking lot of the integrated transportation hub;

[0039] S402. Classify the parking location of the vehicle into one of three parking areas using the base station location information P:

[0040] If , the vehicle attribute is a passenger vehicle;

[0041] If , the vehicle attribute is a freight vehicle;

[0042] If , enter the judgment process for the mixed parking area;

[0043] S403. Judgment of vehicle attributes in the mixed parking area: If , it is a freight vehicle; if , it is a passenger vehicle; represents the freight feature score, which is calculated based on the overlapping degree of the parking area with the freight-related area and the parking duration ratio . The calculation formula is as follows:

[0044] ;

[0045] ;

[0046] ;

[0047] represents the freight feature score, which is calculated based on the overlapping degree of the parking area with the passenger-related area and the short-term parking times ratio . The calculation formula is as follows:

[0048] ;

[0049] ;

[0050] ;

[0051] Among them, and respectively represent the number of times the vehicle stops at the freight and passenger-related areas; represents the total driving time of the vehicle; represents the total parking time of the vehicle; represents the number of short-term parking times of the vehicle; and respectively represent the weights for balancing and in the freight score; and respectively represent the weights for balancing and The weight coefficient in the passenger transportation score is the total number of stops; short time means less than 5 minutes;

[0052] S404. Calculate the parking area and score of each vehicle, and output the attribute classification result of the vehicle. For the specific classification method of passenger vehicle models, go to step S5;

[0053] The freight vehicles are used for freight transportation, including container trucks and logistics vehicles;

[0054] The passenger vehicles are used for passenger transportation, including private cars, buses and taxis, and taxis include online car-hailing vehicles.

[0055] Furthermore, step S5 includes:

[0056] S501. Bus identification, including:

[0057] S5011. Extract the trajectory sequence T of the vehicle from the mobile phone signaling data, ; if , it is determined that the trajectory is fixed, and go to S5012 for continued determination; where represents the trajectory repetition rate, that is, the matching degree between the vehicle trajectory and the number of historical trajectory segments. The calculation method is as follows:

[0058] ;

[0059] represents the number of trajectory segments matched by the vehicle, represents the number of historical trajectory segments; represents the trajectory repetition rate threshold, taking 0.8;

[0060] S5012. Statistically calculate the maximum number of mobile phones Q that the vehicle is connected to the base station at the same time during the time period when the trajectory sequence T is located; if , the corresponding vehicle is identified as a bus; where represents the passenger number threshold; taking 7 - 10;

[0061] S502: Classification of private cars and taxis, including:

[0062] S5021. Calculate the travel regularity, including travel time regularity, travel space regularity and travel regularity index;

[0063] First, calculate the vehicle travel time regularity , The lower it is, the higher the travel time regularity;

[0064] ;

[0065] where Represents the average value of all travel times; Represents the timestamp of the i-th mobile signaling data record; N represents the number of timestamps of mobile signaling data records;

[0066] Secondly, calculate the travel spatial regularity , using the starting and ending positions of the vehicle, calculate the spatial aggregation degree through the density clustering algorithm DBSCAN;

[0067] ;

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] Among them, and respectively represent the i'-th and j-th clustering clusters of the starting point and the ending point; and are the number of clustering clusters; and represent the number of points in the i'-th and j-th clustering clusters; and represent the total number of points of the starting point and the ending point; represents the proportion of the starting point clustering cluster i'; represents the proportion of the ending point clustering cluster j;

[0073] Finally, calculate the travel regularity index ; Comprehensive travel time regularity and travel spatial regularity , the calculation method of the travel regularity index is as follows:

[0074] ;

[0075] Among them, and are weight factors, used to balance the importance of time and spatial regularity, and are taken as 0.4 and 0.6 respectively;

[0076] S5022, Classification and determination rules: If , it is determined as a private car, if , it is determined as a taxi, among which, represents the regularity threshold, is taken as 0.65 - 0.80;

[0077] S503. Output the vehicle type classification results of passenger vehicles according to the trajectory repetition rate and travel regularity index.

[0078] Further, the specific steps of step S6 are as follows:

[0079] S601. Output the set of freight vehicles, and the judgment criteria are ;

[0080] S602. Output the set of passenger vehicles, and the judgment criteria are ;

[0081] S603. Output the set of bus vehicles, and the judgment criteria are , and ;

[0082] S604. Output the set of private cars, and the judgment criteria are and ;

[0083] S605. Output the set of taxis, and the judgment criteria are and ;

[0084] where represents the freight feature score, represents the freight feature score, represents the trajectory repetition rate, represents the trajectory repetition rate threshold, represents the passenger number threshold, represents the travel regularity index, represents the regularity threshold.

[0085] The present invention also provides a vehicle type-based trajectory recognition device based on mobile phone signaling data. Based on the above vehicle type-based trajectory recognition method, it includes:

[0086] A data acquisition module for obtaining and standardizing mobile phone signaling data from a communication operator;

[0087] A data processing module for generating a sequence of trajectory points of a vehicle based on the mobile phone signaling data; determining and removing abnormal trajectory points based on the spatial distance and time interval between trajectory points, performing spatio-temporal interpolation to complete long-time breakpoints, and sorting by time;

[0088] A module for extracting statistical features of vehicle trajectories for dividing trajectory segments according to base station handover data, determining parking points based on speed and stay time thresholds, and obtaining statistical features of vehicle trajectories, including total moving distance, total number of stops, and total stop time;

[0089] The vehicle attribute determination module maps parking spots to passenger, freight and mixed parking areas, calculates passenger and freight characteristic scores for vehicles in mixed parking areas, and determines passenger or freight after comprehensive scoring;

[0090] Passenger vehicle classification module, which is used to identify buses based on the trajectory repetition rate and the maximum number of mobile phones connected in the vehicle for vehicles identified as passenger vehicles, and then calculate the travel time and spatial regularity indicators to distinguish between private cars and taxis, including online-hailing taxis;

[0091] The result output module is used to integrate and output the classification results of each vehicle type and the corresponding complete trajectory sequence, and supports visual reports or external interface calls.

[0092] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the vehicle type trajectory recognition method based on mobile phone signaling data are implemented.

[0093] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the vehicle type trajectory recognition method based on mobile phone signaling data are implemented.

[0094] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0095] 1. Low cost and wide coverage: By using the existing operator's mobile phone signaling, massive location data can be obtained without installing additional hardware on the vehicle; the mobile phone signaling coverage rate is high, and full coverage identification can be achieved for vehicles without GPS devices and on-board OBD.

[0096] 2. High-precision identification: Through multi-level cleaning and spatiotemporal interpolation completion, base station switching noise is eliminated to improve trajectory integrity and accuracy; refined parking point extraction and zoning scoring can effectively distinguish between passenger and freight vehicles and reduce misjudgments.

[0097] 3. Multi-vehicle classification capability: Accurately identify buses based on trajectory repetition rate and number of in-vehicle connections; distinguish between private cars and taxis based on temporal and spatial regularity indicators, taking into account travel time concentration and start-end distribution characteristics.

[0098] 4. Both real-time and batch processing: The data collection module supports real-time stream and historical batch processing, which can meet the needs of daily monitoring and post-analysis; the result output module provides an interface design to facilitate docking with the intelligent connected vehicle-road collaborative platform and traffic management system.

[0099] 5. Support for refined traffic management: Provide vehicle-type flow details for traffic planning to help alleviate congestion on private car sections, layout bus lanes, and control freight transport time periods; accurately monitor the trajectories of illegal operation vehicles, freight overloads, and abnormal routes to assist security and law enforcement. BRIEF DESCRIPTION OF THE DRAWINGS

[0100] Figure 1 It is a flowchart of the method of the present invention;

[0101] Figure 2 It is a schematic diagram of vehicle-type trajectory recognition based on mobile phone signaling data in this embodiment;

[0102] Figure 3 It is a specific flowchart of step S1 in the method of the present invention;

[0103] Figure 4 It is a specific flowchart of step S4 in the method of the present invention;

[0104] Figure 5 It is a specific flowchart of step S5 in the method of the present invention;

[0105] Figure 6 The application architecture diagram of the system provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0106] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0107] In this embodiment, the Lingang Road is taken as an object to elaborate on the technology of the present invention. The Lingang Road has long been in a state of mixed passenger and freight traffic, with significant safety risks in road traffic and serious traffic congestion problems, especially during the morning and evening rush hours. The present invention identifies the vehicle types and estimates based on mobile phone signaling big data, providing data support for the optimization design of dynamic traffic organization in the Lingang Road networked environment, such as: the optimization design of schemes for dedicated passenger and freight lanes, bus lanes, and dedicated lanes for commercial passenger vehicles.

[0108] See Figure 1 , the method flowchart of the present invention is as follows:

[0109] S1. Trajectory data generation: Obtain mobile phone signaling data from the operator and generate a sequence of vehicle trajectory points; see Figure 3 , specifically:

[0110] S101. Data acquisition and analysis: Mobile phone signaling data comes from operator records and is used to record the communication information between user equipment and base stations, including: IMSI (International Mobile Subscriber Identity), base station ID (Cell ID), timestamp (Timestamp), event type (Event Type), signal strength (RSSI), latitude (Latitude), longitude (Longitude). Event types include: location update, base station switching, call establishment, and call release.

[0111] In order to illustrate the receiving and processing process of mobile phone signaling data, Figure 2 The schematic diagram of vehicle type trajectory recognition based on mobile phone signaling data in this embodiment is given. The mobile phone signaling data signal is received through the 4G / 5G base station, and then the data is transmitted and processed through the cloud platform. The information recorded in the mobile phone signaling data is shown in Table 1.

[0112] Table 1 Example of mobile phone signaling data collection format

[0113] IMSI Cell ID Timestamp Event Type RSSI Latitude Longitude 460011234567890 101234 2024 / 12 / 30 9:00 Location Update -78dBm 39.904200 116.4074 460011234567890 101235 2024 / 12 / 30 9:02 Handover -70dBm 39.906100 116.4107 460011234567890 101236 2024 / 12 / 30 9:05 Location Update -65dBm 39.909600 116.4143 460021234567891 101237 2024 / 12 / 30 9:06 Handover -80dBm 39.912300 116.4178 460021234567891 101238 2024 / 12 / 30 9:08 Call Setup -72dBm 39.915100 116.4214 460031234567892 101239 2024 / 12 / 30 9:10 Location Update -75dBm 39.917900 116.4249 460031234567892 101240 2024 / 12 / 30 9:12 Handover -68dBm 39.920500 116.4285 460041234567893 101241 2024 / 12 / 30 9:15 Location Update -74dBm 39.923100 116.4321 460041234567893 101242 2024 / 12 / 30 9:18 Handover -67dBm 39.925800 116.4356 460051234567894 101243 2024 / 12 / 30 9:20 Call Setup -71dBm 39.928600 116.4392

[0114] S102, data screening: only retain signaling data related to the location, remove data irrelevant to the trajectory, and filter out incomplete or abnormal data records.

[0115] S103, base station positioning mapping: query the location information (latitude and longitude information) of the base station according to the base station ID (Cell ID) to form a base station location table ; For mobile phone signaling data Converted to location information, .in, is the i-th record in the mobile phone signaling data; Indicates the timestamp of the i-th record, that is, the specific time of the record.

[0116] S104, trajectory point generation: Base station location As a track point The trajectory point sequence is: ,in, , n represents the total number of trajectory points.

[0117] S2. Data cleaning and completion: Based on the spatial distance and time interval between trajectory points, abnormal trajectory points are determined and removed, and long-term breakpoints are interpolated and completed in time and space, and sorted by time; specifically:

[0118] S201, data cleaning:

[0119] If the distance between certain trajectory points is extremely large and the time interval is extremely short (possibly due to base station handover errors or multipath effects), it is determined as an abnormal point and eliminated. The judgment condition is: If , then eliminate the point . Among them, represents and 's spatial distance, usually calculated based on the spherical distance formula, and the calculation formula is as follows: ; where R is the radius of the earth, usually taking a value of 6371 km; represents the time interval between adjacent trajectory points , represents the reasonable speed upper limit. Preferably, takes a value of 120 km / h.

[0120] S202. Trajectory breakpoint completion: If there is a breakpoint in the trajectory for a long time (more than 10 minutes), it is completed through time interpolation and base station position interpolation. The interpolation formula is as follows:

[0121] ;

[0122] Among them, and respectively represent the position and timestamp of the completion interpolation.

[0123] S203. Time series sorting: Sort the trajectory point sequence T by time ;

[0124] .

[0125] S3. Extract vehicle trajectory statistical features: Divide the trajectory segments according to the base station handover data, determine the parking points based on the speed and stay time thresholds, and obtain the vehicle trajectory statistical features, including the total moving distance, total parking times, and total parking time; specifically:

[0126] S301. Trajectory segment division: Identify continuous trajectory segments according to the base station handover data of the mobile phone signaling. Each trajectory segment is defined as the path of the vehicle during continuous driving or stationary periods, and the expression is as follows:

[0127] ;

[0128] Among them, , m represents the upper limit of the number of different trajectory segments into which the vehicle is divided; the trajectory segment division rule is: If the time interval between two consecutive trajectory points, then they are divided into different segments. is preferably taken as 5 - 10 minutes.

[0129] S302. Parking point identification. Calculate the speed of the vehicle between adjacent trajectory points , if and the parking time , then it is determined as a parking point. Let S represent the set of parking points, that is, the set of points with speed close to zero and minimal position change in the trajectory segment.

[0130] Preferably, in this embodiment , it is appropriate to take a value of 1 - 3 m / s.

[0131] S303. Trajectory statistical feature extraction: including the total moving distance ; the total number of parking times : the total number of points satisfying the parking condition in the trajectory point sequence; the total parking time : the total stay time between parking points. represents the total time of the trajectory segment, which is the time of the last trajectory point minus the time of the first trajectory point.

[0132] S4. Passenger and freight attribute determination: Map the parking points to the passenger, freight, and mixed parking areas, calculate the passenger and freight feature scores for the vehicles in the mixed parking area respectively, and determine passenger or freight after comprehensive scoring; as Figure 4 shown, specifically:[[]]

[0133] S401. Parking area division. Based on the mobile phone signaling data, analyze the parking location of the vehicle and divide the parking area into the following three categories:[[]]

[0134] Passenger parking area : Dedicated to passenger vehicles (such as station and airport parking areas, etc.);[[]]

[0135] Freight parking area : Dedicated to freight vehicles (such as port loading and unloading areas, logistics centers, etc.);[[]]

[0136] Mixed parking area : An area where both passenger and freight vehicles are allowed to park (such as the parking lot of an integrated transportation hub).[[]]

[0137] S402. Parking area determination. Use the base station location information P to classify the parking location of the vehicle into one of the above areas:[[]]

[0138] If , then the vehicle attribute is a passenger vehicle;[[]]

[0139] If , then the vehicle attribute is a freight vehicle;[[]]

[0140] If , enter the judgment process for the mixed parking area.[[]]

[0141] S403. Judgment of vehicle attributes in the mixed parking area. If , it is a freight vehicle; if , it is a passenger vehicle;

[0142] represents the freight feature score, which is calculated according to the overlapping degree between the parking area and the freight-related area and the proportion of parking duration The calculation formula is as follows:

[0143] ;

[0144] ;

[0145] ;

[0146] represents the freight feature score, which is calculated according to the overlapping degree between the parking area and the passenger-related area and the proportion of the number of short-term parking times The calculation formula is as follows:

[0147] ;

[0148] ;

[0149] ;

[0150] Among them, and respectively represent the number of times the vehicle stops in the freight and passenger-related areas; represents the total driving time of the vehicle; represents the total parking time of the vehicle; represents the number of short-term (less than 5 minutes) parking times of the vehicle; and respectively represent the balance and weight coefficients in the freight score; and respectively represent the balance and weight coefficients in the passenger score. Preferably, in this embodiment and are preferably taken as 0.7, and are taken as 0.3.

[0151] S404. Calculate the parking area and score of each vehicle, and output the vehicle attribute classification result. For the specific classification method of the passenger vehicle model, go to step S5.

[0152] Freight vehicles: mainly used for cargo transportation, such as container trucks, logistics vehicles, etc.

[0153] Passenger vehicles: mainly used for passenger transportation, such as private cars, buses, taxis (online car-hailing), etc.

[0154] S5, Classification of passenger vehicle types: For vehicles determined to be passenger vehicles, first identify buses based on the trajectory repetition rate and the maximum number of connected mobile phones inside the vehicle, and then calculate the travel time and spatial regularity indicators to distinguish private cars from taxis, where taxis include online car-hailing; see Figure 5 , specifically:

[0155] S501, Bus identification. The significant features of buses include running on fixed routes with periodicity or repetition, and usually having a large number of passengers inside the vehicle, resulting in a large number of mobile phones simultaneously connected to the base station. Specifically include:

[0156] S5011, Calculation of trajectory repetition rate. Extract the trajectory sequence T of the vehicle from the signaling data, ; if , it is determined that the trajectory is fixed, and go to S5012 for further determination.

[0157] Among them, represents the trajectory repetition rate, that is, the matching degree between the vehicle trajectory and the number of historical trajectory segments, and the calculation method is as follows:

[0158] ;

[0159] represents the number of trajectory segments matched by the vehicle, represents the number of historical trajectory segments, represents the trajectory repetition rate threshold, preferably, it is advisable to take 0.8.

[0160] S5012, Statistics of the upper limit of the number of people inside the vehicle. Statistically, within the time period where the trajectory sequence T is located, the maximum number of mobile phones Q that the vehicle is simultaneously connected to the base station; if , the corresponding vehicle is identified as a bus. Among them, represents the passenger number threshold, preferably, the value is preferably 7 - 10.

[0161] S502, Classification of private cars and taxis (online car-hailing). The main difference between private cars and taxis (online car-hailing) lies in the travel regularity. Private cars usually have higher spatio-temporal regularity, while the travel behavior of taxis (online car-hailing) is more random. Specifically include:

[0162] S5021, Calculation of travel regularity.

[0163] First, calculate the travel time regularity of the vehicle , the lower this value is, the higher the regularity of the travel time.

[0164] ;

[0165] Among them, represents the average value of all travel times, represents the timestamp of the i-th mobile signaling data record; N represents the number of timestamps of the mobile signaling data records.

[0166] Secondly, calculate the travel spatial regularity . Using the starting and ending positions of the vehicles, calculate the degree of spatial aggregation through the density clustering algorithm DBSCAN, The closer it is to 1, the more concentrated the distribution of the starting and ending positions is, and the higher the spatial regularity; The closer it is to 0, the more dispersed the distribution of the starting and ending positions is, and the lower the spatial regularity.

[0167] ;

[0168] ;

[0169] ;

[0170] ;

[0171] ;

[0172] Among them, and respectively represent the i'-th and j-th clustering clusters of the starting and ending points; and are the number of clustering clusters; and represent the number of points in the i'-th and j-th clustering clusters; and represent the total number of points of the starting and ending points; represents the proportion of the i'-th clustering cluster of the starting point; represents the proportion of the j-th clustering cluster of the ending point;

[0173] Finally, calculate the travel regularity index . Combining the travel time regularity and the travel spatial regularity , the calculation method of the travel regularity index is as follows:

[0174] ;

[0175] Among them, and is a weighting factor used to balance the importance of temporal and spatial regularities. Preferably, and are preferably set to 0.4 and 0.6 respectively.

[0176] S5022: Classification determination rule.

[0177] If , it is determined as a private car. If , it is determined as a taxi (online car-hailing). Among them, represents the regularity threshold. Preferably, is preferably set to 0.65 - 0.80.

[0178] S503. Calculate based on the trajectory repetition rate and travel regularity index, and output the vehicle type classification result of passenger vehicles.

[0179] S6. Result output: Output the classification results of each vehicle type and the corresponding complete trajectory sequences. The specific steps are as follows:

[0180] S601. Output the set of freight vehicles, and the judgment criterion is ;

[0181] S602. Output the set of passenger vehicles, and the judgment criterion is ;

[0182] S603. Output the set of bus vehicles, and the judgment criterion is , and ;

[0183] S604. Output the set of private cars, and the judgment criterion is and ;

[0184] S605. Output the set of taxis (online car-hailing), and the judgment criterion is and .

[0185] Among them represents the freight feature score, represents the freight feature score, represents the trajectory repetition rate, represents the trajectory repetition rate threshold, represents the passenger number threshold, represents the travel regularity index, represents the regularity threshold.

[0186] Based on the same inventive concept, the embodiments of the present invention provide a specific implementation manner of a vehicle type classification trajectory recognition system based on mobile phone signaling data that can implement the vehicle type classification trajectory recognition method. SeeFigure 6 , which specifically includes the following content:

[0187] The data acquisition module, as the input source of the entire system, is responsible for obtaining mobile phone signaling data from communication operators. The signaling data includes information such as base station ID, timestamp, signal strength (RSSI), etc., which reflects the communication process between the mobile phone and the base station. Based on this data, the location information and time information of the vehicle can be extracted. The core function of this module is to standardize complex and raw signaling data and ensure the integrity and accuracy of data acquisition. At the same time, this module supports real-time data stream acquisition and batch processing of historical data, providing input support for subsequent modules.

[0188] The data processing module is used to extract information such as base station ID and timestamp from the mobile phone signaling data and map them to base station location information, generating a sequence of vehicle trajectory points; based on the spatial distance and time interval between trajectory points, determine and eliminate abnormal trajectory points, perform spatio-temporal interpolation to complete long-time breaks, and sort by time. The processed data needs to have high quality and continuity, thus providing reliable basic data for subsequent attribute determination and classification.

[0189] The module for extracting statistical features of vehicle trajectories is used to divide trajectory segments according to base station handover data, determine parking points based on speed and stay time thresholds, and obtain statistical features of vehicle trajectories, including total moving distance, total number of stops, and total parking time;

[0190] The vehicle attribute determination module distinguishes whether the vehicle is a passenger vehicle or a freight vehicle by analyzing the parking area and behavior characteristics of the vehicle. According to the parking area (passenger parking area , freight parking area , mixed parking area ) where the parking point p is located, initially determine the attribute of the vehicle. If the parking point is in the mixed parking area, then further calculate the freight feature score and the passenger feature score , and combine the parking time and parking area distribution of the vehicle to finally determine the vehicle attribute. This module is the key part of the system to achieve passenger and freight classification, providing input for subsequent passenger vehicle type classification.

[0191] The passenger vehicle type classification module further divides the vehicles that have been determined to be passenger vehicles into buses, private cars, and taxis (online car-hailing). First, identify buses through the analysis of the trajectory repetition rate and the number of passengers Q in the vehicle. Then, for non-bus vehicles, through the travel regularity index , comprehensively considering the time regularity (standard deviation of time distribution ) and spatial regularity (aggregation of starting and ending points ), distinguish private cars and taxis (online car-hailing vehicles). This module uses a multi-dimensional feature analysis method to ensure the accuracy of classification.

[0192] Result output module. The result output module is the terminal part of the system, which is used to integrate the classification results and output the attribute category and trajectory data of each vehicle. The output content includes the classification of each vehicle (freight / passenger), the specific type of passenger vehicles (bus, private car or taxi), and the complete trajectory sequence T corresponding to the vehicle. In addition, this module supports formatting the output results into a visual report or interface for external system calls, facilitating analysis and decision-making by traffic management departments or logistics operators. This module can also provide statistical analysis functions, such as: accuracy evaluation of classification results and dynamic monitoring of traffic flow.

[0193] Preferably, the embodiments of the present application also provide a specific implementation manner of an electronic device capable of implementing all steps in the above-described method for identifying vehicle type trajectories based on mobile phone signaling data. The electronic device specifically includes the following:

[0194] A processor, a memory, a communications interface, and a bus;

[0195] Among them, the processor, the memory, and the communications interface communicate with each other through the bus; the communications interface is used to implement information transmission between related devices such as server-side devices, metering devices, and user-side devices.

[0196] The processor is used to call the computer program in the memory. When the processor executes the computer program, it implements all steps in the above-described method for identifying vehicle type trajectories based on mobile phone signaling data.

[0197] The embodiments of the present application also provide a computer-readable storage medium capable of implementing all steps in the above-described method for identifying vehicle type trajectories based on mobile phone signaling data. A computer program is stored on the computer-readable storage medium. When the computer program is executed by the processor, it implements all steps in the above-described method for identifying vehicle type trajectories based on mobile phone signaling data.

[0198] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the hardware + program type embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0199] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the acts or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0200] Although this application provides method operation steps such as in the embodiments or flowcharts, it may include more or fewer operation steps based on routine or non-creative labor. The order of steps recited in the embodiments is only one way among many orders of performing the steps and does not represent the sole order of execution. When implemented in an actual device or client product, it may be executed in the order of the method shown in the embodiments or the figures or in parallel (e.g., in an environment with parallel processors or multithreaded processing).

[0201] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.

[0202] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one Figure 1 one process or multiple processes and / or blocks Figure 1 specified in one block or multiple blocks.

[0203] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are performed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 one process or multiple processes and / or blocks Figure 1 specified in one block or multiple blocks.

[0204] The present invention is not limited to the embodiments described above. The above description of the specific embodiments is intended to describe and illustrate the technical solutions of the present invention. The above specific embodiments are merely illustrative and not restrictive. Without departing from the spirit of the present invention and the scope protected by the claims, those of ordinary skill in the art can make many specific changes in form under the inspiration of the present invention, and these all fall within the protection scope of the present invention.

Claims

1. A method for identifying vehicle-type trajectories based on mobile phone signaling data, characterized in that, Including: S1. Trajectory data generation: Obtain mobile phone signaling data from the operator, and generate a sequence of trajectory points of the vehicle based on the mobile phone signaling data; S2. Data cleaning and completion: Based on the spatial distance and time interval between trajectory points, determine and remove abnormal trajectory points, perform spatio-temporal interpolation and completion on long-time breakpoints, and sort them by time; S3. Extract statistical features of vehicle trajectories: Divide trajectory segments according to base station handover data, determine parking points based on speed and stay time thresholds, and obtain statistical features of vehicle trajectories, including total moving distance, total number of stops, and total parking time; S4. Passenger and freight attribute determination: Map parking points to passenger, freight, and mixed parking areas, calculate passenger and freight feature scores for vehicles in the mixed parking area respectively, and determine passenger or freight after comprehensive scoring; S5. Passenger vehicle type classification: For vehicles determined to be passenger vehicles, first identify buses based on trajectory repetition rate and the maximum number of mobile phones connected inside the vehicle, and then calculate travel time and spatial regularity indicators to distinguish private cars and taxis, where taxis include online car-hailing services; S6. Result output: Output the classification results of each vehicle type and the corresponding complete sequence of trajectory points.

2. The method for identifying vehicle type-based trajectories according to claim 1, wherein Step S1 includes: S101. Mobile phone signaling data is used to record communication information between user equipment and base stations, including International Mobile Subscriber Identity (IMSI), base station ID, timestamp, signal strength, longitude and latitude information, and event type information. Event types include location update, base station handover, call establishment, and call release; S102. Retain location update, base station handover, base station ID, timestamp, and longitude and latitude information, and filter out incomplete or abnormal mobile phone signaling data; S103. Query the location information of the base station according to the base station ID to form a base station location table ; For the mobile phone signaling data Convert it into location information, ; Among them, is the i-th data in the mobile phone signaling data; represents the timestamp of the i-th mobile phone signaling data, that is, the specific time of this mobile phone signaling data; S104. Take each base station location as a trajectory point , and generate a sequence of trajectory points of the vehicle as: , where , and n represents the total number of trajectory points.

3. The method for identifying vehicle-type specific trajectories based on mobile phone signaling data according to claim 1, wherein Step S2 includes: S201. Set judgment conditions: If , then exclude the trajectory point ; where represents the spatial distance between the trajectory points and , and the calculation formula is as follows: where R is the radius of the earth, and the value is 6371 km; represents the time interval between adjacent trajectory points, , represents the upper limit of reasonable speed; , represent the timestamps of the i-th and i + 1-th mobile signaling data respectively; , and n represents the total number of trajectory points; S202. If there is a long-time breakpoint greater than 10 minutes in the sequence of trajectory points, complete it through time interpolation and base station location interpolation. The interpolation formula is as follows: ； Among them, and respectively represent the base station location and timestamp for completing interpolation; S203. Sort the trajectory point sequence T by time ; 。 4. The vehicle type-based trajectory recognition method based on mobile phone signaling data according to claim 1, wherein Step S3 includes: S301. Identify continuous trajectory segments based on the base station handover data of mobile phone signaling data , where each trajectory segment is defined as the path of a vehicle during continuous driving or stationary periods, and the expression is as follows: ； Among them, , m represents the upper limit of the number of different trajectory segments into which the vehicle is divided; If the time interval between two consecutive trajectory points , they are divided into different segments; S302. Calculate the speed of the vehicle between adjacent trajectory points If and the parking time , it is determined as a parking point; Let S represent the set of parking points, denote the spatial distance between the trajectory points and ; Take 1 - 3 m / s, Take 5 - 10 min; S303. Extract trajectory statistical features, including the total moving distance ; the total number of stops , representing the total number of trajectory points in the trajectory point sequence that meet the parking conditions; the total parking time , representing the total stay time between parking points.

5. The method for identifying vehicle type-specific trajectories based on mobile phone signaling data according to claim 1, wherein, Step S4 includes: S401. Based on mobile phone signaling data, analyze the parking location of the vehicle, and divide the parking area into the following three categories: Passenger parking area , for passenger vehicles, including station and airport parking areas; Freight parking area , for freight vehicles, including port loading and unloading areas, logistics centers; Mixed parking area An area that allows both passenger and freight vehicles to stop, including the parking lot of an integrated transportation hub; S402. Use the base station location information P to classify the parking location of the vehicle into one of the three parking areas: If , then the vehicle attribute is a passenger vehicle; If , the vehicle attribute is a freight vehicle; If , enter the judgment process for the mixed parking area; S403. Vehicle attribute judgment in the mixed parking area: If , it is a freight vehicle; if , it is a passenger vehicle; represents the freight feature score, which is calculated according to the overlap degree between the parking area and the freight-related area and the proportion of parking duration . The calculation formula is as follows: ； ； ； Indicates the freight feature score, which is calculated based on the overlapping degree between the parking area and the passenger-related area and the proportion of short-term parking times The calculation formula is as follows: ； ； ； Among them, and respectively represent the number of times the vehicle stops in the areas related to freight and passenger transportation; represents the total driving time of the vehicle; represents the total parking time of the vehicle; represents the number of short-term parking times of the vehicle; and respectively represent the weights in the freight score for balancing and ; and respectively represent the weights in the passenger score for balancing and ; is the total number of parking times; short-term means less than 5 minutes. S404. Calculate the parking area and score of each vehicle, and output the attribute classification result of the vehicle. For the specific classification method of passenger vehicle types, go to step S5; The freight vehicles are used for freight transportation, including container trucks and logistics vehicles; The passenger vehicles are used for passenger transportation, including private cars, buses, and taxis, where taxis include online car-hailing services.

6. The method for identifying vehicle type-based trajectories according to claim 1, wherein Step S5 includes: S501. Bus identification, including: S5011. Extract the trajectory sequence T of the vehicle from the mobile phone signaling data, ; If , it is determined that the trajectory is fixed, and go to S5012 for continued determination; where represents the trajectory repetition rate, that is, the matching degree between the vehicle trajectory and the number of historical trajectory segments. The calculation method is as follows: ； Indicates the number of trajectory segments matching the vehicle, Indicates the number of historical trajectory segments; Indicates the threshold of trajectory repetition rate, taking 0.8; S5012. Statistically calculate the maximum number of mobile phones Q that the vehicle is simultaneously connected to the base station during the time period when the trajectory sequence T is located; if , then identify the corresponding vehicle as a bus; where represents the passenger number threshold; take 7 - 10. S502: Classification of private cars and taxis, including: S5021. Calculate travel regularity, including travel time regularity, travel spatial regularity, and travel regularity index; First, calculate the regularity of vehicle travel time , The lower the value, the higher the regularity of travel time; ； Among them, represents the average value of all travel times; represents the timestamp of the i-th mobile signaling data record; N represents the number of timestamps of mobile signaling data records; Secondly, calculate the regularity of the travel space , and use the starting and ending positions of the vehicle to calculate the degree of spatial aggregation through the density clustering algorithm DBSCAN; ； ； ； ； ； Among them, and respectively represent the i'-th and j-th clustering clusters of the starting point and the ending point; and are the number of clustering clusters; and represent the number of points in the i'-th and j-th clustering clusters; and represent the total number of points of the starting point and the ending point; represents the proportion of the starting point clustering cluster i'; represents the proportion of the ending point clustering cluster j; Finally, the travel regularity index is calculated. ; Comprehensive travel time regularity and travel space regularity The calculation method of the travel regularity index is as follows: ； Among them, and are weighting factors used to balance the importance of temporal and spatial regularities, and take 0.4 and 0.6 respectively; S5022, Classification and determination rules: If , it is determined as a private car. If , it is determined as a taxi. Among them, represents the regularity threshold, takes values from 0.65 to 0.80; S503. According to the trajectory repetition rate and travel regularity index, output the vehicle type classification result of passenger vehicles.

7. The method for identifying vehicle type-based trajectories according to claim 1, wherein The specific steps of step S6 are as follows: S601. Output a set of freight vehicles, and the evaluation criteria are ; S602. Output the passenger vehicle set, and the evaluation criteria are ; S603. Output the set of bus vehicles, and the evaluation criteria are , and ; S604. Output the set of private cars, and the evaluation criteria are and ; S605. Output the set of taxi vehicles, and the evaluation criteria are and ; Among them represents the freight feature score represents the freight feature score represents the trajectory repetition rate represents the trajectory repetition rate threshold represents the passenger number threshold represents the travel regularity index represents the regularity threshold 8. A vehicle-type-based trajectory recognition device based on mobile phone signaling data, based on the vehicle-type-based trajectory recognition method according to any one of claims 1-7, characterized in that, Including: A data acquisition module, used to obtain and standardize mobile phone signaling data from the communication operator; A data processing module, configured to generate a sequence of trajectory points of a vehicle based on mobile phone signaling data; determine and remove abnormal trajectory points based on the spatial distance and time interval between trajectory points, perform spatio-temporal interpolation to complete long-time breakpoints, and sort them by time; A vehicle trajectory statistical feature extraction module, configured to divide trajectory segments according to base station handover data, determine stop points based on speed and residence time thresholds, and obtain vehicle trajectory statistical features, including total moving distance, total number of stops, and total stop time; A vehicle attribute determination module, which maps stop points to passenger, freight, and mixed parking areas, calculates passenger and freight feature scores for vehicles in the mixed parking area respectively, and determines passenger or freight after comprehensive scoring; A passenger vehicle type classification module, for vehicles determined to be passenger vehicles, first identify buses based on trajectory repetition rate and the maximum number of mobile phones connected inside the vehicle, and then calculate travel time and spatial regularity indicators to distinguish private cars and taxis, where taxis include online car-hailing services; A result output module, configured to integrate and output the classification results of each vehicle type and the corresponding complete trajectory sequence, and support visualization reports or external interface calls.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method for identifying vehicle type trajectories based on mobile phone signaling data according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for identifying vehicle type trajectories based on mobile phone signaling data according to any one of claims 1 to 7.

Images