CN111653093B - Urban trip mode comprehensive identification method based on mobile phone signaling data - Google Patents

Urban trip mode comprehensive identification method based on mobile phone signaling data Download PDF

Info

Publication number
CN111653093B
CN111653093B CN202010475924.2A CN202010475924A CN111653093B CN 111653093 B CN111653093 B CN 111653093B CN 202010475924 A CN202010475924 A CN 202010475924A CN 111653093 B CN111653093 B CN 111653093B
Authority
CN
China
Prior art keywords
travel
mode
time
trip
user
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010475924.2A
Other languages
Chinese (zh)
Other versions
CN111653093A (en
Inventor
万紫吟
陆振波
刘娟
夏井新
余启航
李�瑞
张改
张念启
张静芬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Ruiqi Intelligent Transportation Technology Industry Research Institute Co ltd
Original Assignee
Nanjing Ruiqi Intelligent Transportation Technology Industry Research Institute Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing Ruiqi Intelligent Transportation Technology Industry Research Institute Co ltd filed Critical Nanjing Ruiqi Intelligent Transportation Technology Industry Research Institute Co ltd
Priority to CN202010475924.2A priority Critical patent/CN111653093B/en
Publication of CN111653093A publication Critical patent/CN111653093A/en
Application granted granted Critical
Publication of CN111653093B publication Critical patent/CN111653093B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0125—Traffic data processing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2462—Approximate or statistical queries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G06F16/285—Clustering or classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/024—Guidance services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/029—Location-based management or tracking services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/20—Services signaling; Auxiliary data signalling, i.e. transmitting data via a non-traffic channel
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • H04W4/42—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for mass transport vehicles, e.g. buses, trains or aircraft

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Probability & Statistics with Applications (AREA)
  • Analytical Chemistry (AREA)
  • Chemical & Material Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Fuzzy Systems (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Remote Sensing (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention discloses a comprehensive urban trip mode identification method based on mobile phone signaling data, which is characterized in that for mobile phone signaling data generated all day by day in a target city, stopping points and all-day trip ODs of each person are identified, all underground rail transit trip modes are accurately identified through a subway special base station, then compared with Gord GPS real navigation data on the basis of extracting characteristic parameters of the mobile phone data, and the traffic modes of resident trip are identified by using an unsupervised machine learning algorithm. In order to improve the identification accuracy, the invention adopts three correction modes to respectively correct the identification result (road network correction, user personal attribute correction, road network and user personal attribute combined correction), and provides an optimization algorithm to compare the three correction modes to obtain the optimal correction result.

Description

Urban trip mode comprehensive identification method based on mobile phone signaling data
Technical Field
The invention relates to the field of traffic planning, in particular to the field of traffic demand prediction.
Background
The traffic demand prediction is the basis of urban traffic planning, and the prediction of the urban traffic demand accurately has important significance for reasonably managing and controlling an urban traffic system. In the existing traffic demand prediction methods (such as traffic distribution models), travel demand estimation under different travel modes is highly dependent. However, given its highly complex nature, accurately predicting the needs of different approaches has certain difficulties. This is because temporal/spatial fluctuations in the operation of the transport system and the flow of traffic are undetectable, and therefore, efficient identification of the mode of transportation is critical to current technology.
In the past, the traditional method for acquiring traffic information such as travel modes through a traditional investigation means has many defects, such as high investigation cost and low sampling rate, in recent years, mobile intelligent devices are rapidly developed, and how to acquire travel modes from track data containing resident mobile position information, such as GPS data and mobile phone signaling data, becomes a hot point of attention in the traffic field.
In the prior art, Zhang et al discloses a method (CN201510452430.1) for comprehensively distinguishing resident travel modes based on mobile phone signaling data, a method for constructing prior probability by matching and combining a travel mode subchain and a GIS (geographic information system) net is adopted, the subchain belonging to a motor vehicle travel mode is respectively matched with a rail transit net and a bus net of the GIS to distinguish rail transit travel and conventional bus travel, and then the construction of the prior probability for the remaining possible travel modes is distinguished through three attributes of average speed, maximum speed and travel time length. The method for judging the rail transit trip and the bus trip by GIS wire net matching ignores the car trip condition overlapped with the rail wire net and the bus wire net, and is easy to cause misjudgment; in the subsequent process of establishing the prior probability, only three attributes of the average speed, the maximum speed and the travel time length are considered, and since the position updating of the mobile phone signaling data is based on the base station, certain errors exist in the obtained travel characteristics and the actual situation, erroneous judgment is easily caused by only considering the three attributes.
Zhang et al also discloses a traffic mode discrimination method (CN201910076104.3) of semi-supervised SVM based on mobile phone signaling data, after carrying out mode recognition on part of trips through a manual recognition process, training a semi-supervised SVM classifier by using the marked samples and the unmarked samples together. Because the method adopts partial manual judgment process, the selection of experts and the organization of worker marks are difficult in actual operation, the time consumption is long, and the result is influenced by subjective factors.
Yangwei et al discloses a trip mode identification method (CN201710693960.4) based on big data machine learning, which comprises the steps of collecting mobile phone signaling data of a training sample investigation object and acceleration detection equipment data, analyzing data fluctuation characteristics, obtaining speed and acceleration fluctuation characteristic data as a prediction input value, taking a trip mode as an output value, training a machine learning algorithm, and finally selecting an algorithm with high precision to divide the trip mode. According to the method, the investigation object is required to be matched with and collect the data of the acceleration detection equipment, the implementation difficulty is high due to the influence of multiple factors such as manpower and material resources in the process, the number of samples is limited, the data quality is influenced by the quality of the acceleration detection equipment and the use degree of the investigation object, and the mode identification that the algorithm model obtained by training the data collected by one-time investigation is applied to all the trip data is not reasonable.
Disclosure of Invention
Aiming at the problems in the background art, the invention provides a comprehensive urban trip mode identification method based on mobile phone signaling data.
The technical scheme is as follows:
a comprehensive urban trip mode identification method based on mobile phone signaling data comprises the following specific steps:
s1, identifying a user parking point according to the mobile phone signaling data to obtain a travel OD;
s2, preprocessing the travel OD, extracting travel characteristic parameters and deleting invalid ODs;
s3, identifying a subway trip OD according to the information of the special subway base station;
s4, dividing walking and long-distance car travel ODs according to the travel distance and the average travel speed to obtain travel OD division results with remarkable characteristics, and performing S5 on the remaining unidentified travel ODs;
s5, according to coordinates of travel origin-destination points, crawling travel distances and travel time consumption of four types of transportation modes, namely walking, bicycles, buses and cars in a route planning API, identifying OD travel modes according to deviation values of mobile phone data and route data, and performing S6 on the remaining unidentified travel ODs;
s6, performing mode division on the remaining unidentified travel ODs by using a fuzzy K-means clustering algorithm;
s7, correcting the partially misjudged row mode according to the urban road and the public traffic network, comprising the following steps:
s71, establishing a corresponding relation between the service range of the mobile phone 4G base station and an urban road network;
s72, matching the mobile phone signaling data travel OD with the urban road network to obtain all road sections and road section attributes contained in the travel;
s73, correcting matching results of the elevated road section, the high speed road section and the road section of the overlapped part of the tunnel and the ground and the OD according to travel characteristics;
s74, correcting the travel mode of the non-motor vehicle based on different road types;
if the identified travel mode is a non-motor vehicle, the traffic mode is modified to be a car according to the OD of the road section which only allows the motor vehicle to travel, such as a high speed road, an overhead road, a tunnel and the like, contained in the travel route obtained through the steps;
s75, using the division result as initial data, performing three types of corrections to the initial data:
correction 1: correcting the bus trip mode based on the trip time and the urban bus network; the scheme fully considers the setting of different grades of roads in different urban roads at the supply side, the control regulation of different local specific road sections and the influence of the real bus network arrangement condition on the travel mode, and corrects the identification result of the traffic mode according to the control regulation.
And (3) correction 2: correcting the travel mode based on the personal attribute and the personal travel behavior characteristic; according to the scheme, the identified travel OD traffic mode is further corrected based on the user population type, the private vehicle travel chain closing principle and the user travel behavior characteristics, and the accuracy of mode division results is improved.
And (3) correction: and sequentially correcting the bus trip mode based on the trip time and the urban bus network and the trip mode based on the personal attribute and the personal trip behavior characteristic.
For the division result of each correction scheme, dividing the ODs adopting the same trip mode into a group, counting k groups, and marking the set of each group of data as CiI ═ 1,2, …, k; calculating the departure time start _ time, the travel distance, the travel time consumption move _ time, the travel average speed, the 85-quantile speed _85 of the travel whole-course speed and the variation coefficient cv of the travel whole-course speed of each travel OD as characteristic parameters of each OD, and respectively calculating the DVI index of the division result of the three schemes:
Figure BDA0002515873100000031
wherein k denotes the number of result categories divided by way, Ci(i-1, 2, …, k) represents a set of OD data of the same format, and x representsi/xjA six-dimensional feature vector composed of the 6 feature parameters for each piece of OD data;
and selecting a scheme corresponding to the result with the larger DVI index as a final correction result.
In order to improve the identification accuracy, the invention adopts three correction modes to respectively correct the identification result (road network correction, user personal attribute correction, road network and user personal attribute combined correction), and provides an optimization algorithm to compare the three correction modes to obtain the optimal correction result.
Preferably, after the step S1 cleans the obtained mobile phone signaling data, the user stopping point is identified based on the base station stopping time and the service radius, and when the user stays within the service radius threshold radius _ range for a time period exceeding the stopping time threshold min _ state _ time with a certain base station as the center, the base station is used as the stopping point of the user, and then the travel OD is obtained according to the travel stopping point. In practice, radius _ range can take 800-1500 m, and min _ state _ time can take 40 min.
Preferably, the step S2 is to extract the travel characteristics of each obtained travel OD, and includes: the starting point longitude lng, the starting point latitude lat, the end point longitude lng, the end point latitude lat, the starting time start _ time, the travel distance, the travel time consumption _ time, the travel average speed, the 85 quantile speed _85 of the travel whole-course speed and the variation coefficient cv of the travel whole-course speed; and screening and deleting invalid ODs with the row distance smaller than the effective travel distance threshold value min _ dis, the travel time move _ time smaller than the effective travel time threshold value min _ time and the travel average speed larger than the average travel speed maximum value max _ speed of the urban transportation. In the implementation, min _ dis is selected to be 1km, min _ time is selected to be 5min, and max _ speed is selected to be 120 km/h.
Preferably, in step S4, the method is divided according to the trip distance and the trip average speed of the trip OD, and the specific steps are as follows:
s41, if the distance exceeds the long-distance travel OD of the long-distance travel threshold long _ dis in the city, dividing the travel into travel in a car mode;
and S42, if the average travel speed is less than or equal to the walking speed threshold walk _ speed and the travel distance is less than or equal to the travel OD of the walking travel distance threshold walk _ dis, dividing the travel into walking type travel.
In the implementation, long _ dis can be 30 km-100 km, walk _ speed can be 8km/h, and walk _ dis can be 3 km.
Preferably, in step S5, the OD travel mode is identified by comparing the travel characteristics of the OD with the characteristic similarity of travel planned in different climbing modes, and the specific steps are as follows:
s51, respectively crawling the travel distance GD _ dis and travel time consumption GD _ time of routes planned in a mode of Gauder walking, bicycles, buses and cars by taking travel characteristic starting point longitude lng, starting point latitude lat, ending point longitude lng, ending point latitude lat and starting time start _ time as parameters for each unidentified travel OD;
s52, calculating the total deviation GD _ diff of each travel OD and the traffic mode height plan in the step 4:
i) abs (travel time consumption move time-high planned time GD _ time)/travel time consumption move time
ii) a planned distance deviation from the certain traffic pattern of high, GD _ dis _ diff, math
iii) a total deviation from the planned route of the high German traffic mode GD _ diff ═ a deviation from the planned time of the planned route of the high German traffic mode GD _ time _ diff α + a deviation from the planned distance of the planned route of the high German traffic mode GD _ dis _ diff (1- α), wherein α is a weight consumed during the planned travel of the high German;
in the implementation, alpha can be 0.6-0.9, walk _ dis can be 3km, and max _ diff can be 0.05-0.15.
S53, comparing and judging the traffic mode with the lowest OD deviation GD _ diff;
s54, if the walking mode deviation GD _ diff is the lowest and is smaller than the deviation threshold max _ diff, and the travel distance is smaller than the walking travel distance threshold walk _ dis, judging that the walking mode is walking;
s55, if the riding mode deviation GD _ diff is the lowest and is smaller than the deviation threshold value max _ diff, judging that the bicycle is a bicycle;
and S56, if the deviation GD _ diff of the car mode is the lowest and is smaller than the deviation threshold value max _ diff, determining that the car is the car.
Preferably, step S6 is to divide the remaining ODs to be recognized by using a fuzzy K-means machine learning algorithm, and the specific steps are as follows:
s61, selecting 10000 samples randomly, and training the samples based on an objective function J (U, C), a fuzzification degree m and a cluster number K by using a fuzzy K-means algorithm to obtain a clustering center C ═ C1,c2,…,cK}；
S62, calculating the distance between each center and the residual samples to obtain a membership matrix Ui＝{ui1,ui2,…,uiK}；
S63, sequentially defining the traffic mode of each cluster according to the average speed parameter speed of each cluster center, wherein the speed of each cluster is car, bus, electric vehicle, bicycle and walking from high to low.
In the implementation process, the main control panel is provided with a plurality of control panels,
Figure BDA0002515873100000051
m is 1.2-3.5, and K is 2-5.
Preferably, in step S75, the step of correcting the bus trip mode based on the trip time and the urban public transportation network includes:
s711, establishing a corresponding relation between the service range of the mobile phone 4G base station and an urban road network;
s712, matching the mobile phone signaling data travel OD with the urban road network to obtain all road sections and road section attributes contained in the travel;
s713, correcting matching results of the section of the overhead, the high speed and the part of the tunnel coinciding with the ground and the OD according to the travel characteristics;
s714, correcting the travel mode of the non-motor vehicle based on different road types;
if the identified travel mode is a non-motor vehicle, the traffic mode is modified to be a car according to the OD of the road section which only allows the motor vehicle to travel, such as a high speed road, an overhead road, a tunnel and the like, contained in the travel route obtained through the steps;
s715, correcting the bus trip mode based on the trip time and the urban bus network, specifically:
s7151, matching the urban public transport network with an urban road network to form a road node sequence corresponding to each public transport line, and storing the operation time of the line in attribute information;
s7152, identifying the travel mode as an OD of bus travel, and modifying the traffic mode into a car if the corresponding route passes through the road section without a bus line;
s7153, identifying the travel mode as the OD of the bus travel, and if the departure time start _ time is not within the corresponding bus route operation time period, modifying the traffic mode into a car;
s716, correcting the travel mode identification result based on the current urban road management and control regulation, specifically:
s7161, traversing the control rule and the limit vehicle limit information corresponding to each road section through which each travel route obtained in S73 is matched with each travel OD;
s7162, if the road section passing by is a bus lane, modifying the traffic mode into a bus;
s7163, if the road section passing through is a one-way road, or the left/right turn of the vehicle is forbidden through the intersection, the corresponding track still turns left/right, and the identified travel mode is motor vehicle travel, the traffic mode is corrected to be the electric vehicle;
and S7164, if the passing road section is a control road section which is specially forbidden in a certain traffic mode and the identified travel mode is the traffic mode, correcting the traffic mode into the traffic mode which is allowed to pass through the road section in the current city.
In step S711, specifically:
s7111, generating an undirected graph from a real urban road network in a node, road section and road section attribute format, and storing road section number roadID, road section name roadName, road section type roadType, road section length, control rule, vehicle limit and other information in the road section attribute;
s7112, establishing a service range buffer area by taking each base station as a center, establishing a mapping relation with road nodes in a service range, and calculating the distance from the base station to each node to be used as a weight weighted node;
s7113, establishing a service range buffer area by taking each base station as a center, establishing a mapping relation with the road sections in the service range, and calculating the vertical distance from the base station to each road section to be used as a weight weighted edge.
In step S712, the specific steps are:
s7121, sequencing each travel OD base station track sequence according to starting time start _ time;
s7122, according to the corresponding relation of the base station sections obtained in the S71, counting the cumulative frequency roadFreq and the cumulative weight roadWeight of all sections covered by the trip OD, and sequencing from high to low;
s7123, sequentially aiming at each base station in the travel OD track sequence, obtaining all road nodes and weight weightNodes corresponding to the base station according to the corresponding relation of the base station road nodes obtained in S711, sequencing all nodes according to the sequence of the road section frequency roadFreq, the road section weight roadWeight and the node weight weightNode, and marking the road node with the first priority as the corresponding road node of the base station, wherein the road nodes correspond to the base station according to the corresponding road section frequency roadFreq and the weight roadWeight information obtained in S722;
and S7124, circularly matching all base stations in the track sequence to obtain a sequence of road nodes, sequentially connecting the nodes according to the road network undirected graph generated in S711, calling a shortest path function to supplement a missing road section aiming at non-adjacent nodes, and finally obtaining a complete travel path on the road network.
In step S713, specifically:
s7131, screening special road sections such as an overhead road section, a high speed road section and a tunnel according to the road section type roadType, and respectively obtaining a corresponding base station sequence of each overhead road section, high speed road section and tunnel according to the corresponding relation of the base station road sections obtained in the S711;
s7132, screening out paths containing overhead, high speed and sections of which the tunnels are overlapped with the ground in matching result paths, and calculating the average travel speed of the travel OD on the corresponding road;
s7133, calculating the longest similar subsequence of the travel OD base station track and the corresponding elevated, high-speed and tunnel, and marking the ratio of the number of the base stations of the longest similar subsequence to the total number of the base stations of the corresponding base station sequence of the corresponding elevated, high-speed and tunnel as the similarity of the special road section;
and S7134, if the average travel speed of the travel OD on the corresponding road is greater than the speed threshold value of the corresponding special road section, and the similarity of the special road section is greater than the similarity threshold value, judging that the travel route is an elevated road section, a high speed road section and a tunnel, and otherwise, judging that the travel route is a ground road section.
Preferably, in step S75, the modifying the travel mode based on the personal attribute and the personal travel behavior characteristic includes:
s721, obtaining mobile phone signaling data of a target city for one month continuously, and identifying the travel mode of each travel OD every day according to the steps;
s722, counting the number of days of stay in the target city in the residence time period of the user and identifying the residence according to the stay time;
s723, judging the population type of the user according to the activity condition of the user in the target city in one month;
s724, correcting the mode division of the travel OD of the user based on the population type of the user, specifically:
s7241, for a user with a population type of a transit population, if the user has a travel OD which is identified as a mode other than car travel, modifying the traffic mode into a car;
s7242, for a user with a short-term business travel population as a population type, if the user has a travel mode identified as an OD for the electric vehicle to travel, modifying the traffic mode into a car;
s725, correcting the travel mode based on the private transportation travel chain closing principle, specifically:
s7251, if the starting point or the ending point of the trip is the residence of the user identified in step S72, determining that the type of the trip is a home-based trip, regarding the trip OD of the user every day;
s752, if one of the paired home-based travel ODs identified by the user is a private vehicle, modifying the other home-based travel OD into the same private vehicle;
s726, judging the common user of the motorized traffic mode according to the travel mode of the user for one month;
s727, correcting the daily trip mode division of the user based on the trip behavior characteristics, specifically:
s7271, clustering the travel OD of each user based on unit day;
s7272, aiming at the electric vehicle common users, if the electric vehicle travels in all travel OD on the same day, correcting the OD travel mode that all travel distances are smaller than an electric vehicle travel distance threshold ebike _ distance on the same day into the electric vehicle;
s737, for the car-used user, if the home-based travel mode on the same day is a car, all travel OD modes other than walking travel on the same day are modified to be cars.
Wherein, step S722 specifically is:
s7221, aiming at the full-sample one-month mobile phone signaling time-space data, screening data appearing in the living time period, and clustering each user based on a staying base station;
s7222, counting the total time live _ state _ time of stay of the residence time period of the user at each base station and the days live _ days of the residence time period of the base station in sample data;
s7223, calculating the average stay time avg _ live _ stay of each base station in the residence time period of the user;
s7224, identifying the base station with the longest total dwell time live _ state _ time in the user dwell period and the average dwell time avg _ live _ state being greater than the dwell time threshold min _ live _ time as the user dwell base station.
Step S723 specifically is:
s7231, counting the average stay time avg _ stay and the total stay days stay _ days of the user in the target city every day;
s7232, if the average stay duration avg _ state of the user is smaller than the city stay time threshold value state _ time _ min, judging that the user is the cross-border population;
s7233, if the average stay duration avg _ stay of the user is larger than the threshold value of the urban stay time stay _ time _ min and the total stay days stay _ days is smaller than the threshold value of the urban stay days stay _ days _ min, determining that the user is the short-term business travel population.
In practice, the city stay time threshold value, say _ time _ min, may take from 2hr to 5hr, and the city stay days threshold value, say _ days _ min, may take from 7 to 15 days.
Step S726 specifically includes:
s7261, counting the number of trips of the middle-sized and small-sized automobiles in all trips OD, the number of trips of the bus, the number of trips of the electric vehicle and the total number of times of motorized trips of the user in one month;
s7262, respectively calculating the proportion of cars, buses and electric vehicles in motorized trips;
s7263, if the proportion of one mode is higher than the motor vehicle trip threshold value, defining the user as the common user of the motorized mode.
In implementation, the motor vehicle trip threshold value can be 70% -90%.
The invention has the advantages of
The method is based on mobile phone signaling data, and is used for identifying the traffic mode of full-sample travel generated in one day of a target city by combining the division of obvious travel characteristics, the judgment of travel path characteristic deviation degrees in different modes of Goods GPS navigation planning and an unsupervised fuzzy K-means clustering machine learning algorithm on the basis of judging underground rail traffic travel by utilizing a subway special base station.
In order to improve the identification accuracy, the invention adopts three correction modes to respectively correct the identification result (road network correction, user personal attribute correction, road network and user personal attribute combined correction), and provides an optimization algorithm to compare the three correction modes to obtain the optimal correction result.
Drawings
FIG. 1 is a flow chart of the urban trip mode comprehensive identification method
FIG. 2 is a diagram of the proportion of the resident travel modes obtained by the actual survey in the embodiment
FIG. 3 is a diagram showing the results of the correction based on the correction 1 scheme in the example
FIG. 4 is a diagram showing the results of the correction based on the correction 2 scheme in the example
FIG. 5 is a diagram showing the results of the correction based on the correction 3 scheme in the example
FIG. 6 is a travel distance distribution diagram of different transportation modes in the result of the correction 3
FIG. 7 is a travel speed distribution diagram of different transportation modes in the result of the correction 3
FIG. 8 is a diagram of the daily traffic distribution of the middle-sized and small cars in the corrected result of 3
Detailed Description
The invention is further illustrated by the following examples, without limiting the scope of the invention:
the mobile phone signaling data refers to a series of control instructions generated by the mobile communication network for actively or passively, regularly or irregularly keeping contact with the mobile terminal of the mobile phone user, and includes fields such as mobile phone identification codes, timestamps, event types, base station numbers, base station longitude and latitude, number attributions and the like, and includes the spatio-temporal information of the all-day operation track of each user, as shown in the following table:
dt msid start_time start_ci start_lng start_lat end_time end_ci end_lng end_lat
20190522 1 20190522000000 85132041 120.9892 31.4025 20190522000001 85132057 120.9892 31.4025
20190522 1 20190522000001 85132057 120.9892 31.4025 20190522000037 85132032 120.9892 31.4025
20190522 1 20190522000037 85132032 120.9892 31.4025 20190522000055 2.33E+08 120.9892 31.4025
20190522 1 20190522000055 85132032 120.9892 31.4025 20190522000143 85132057 120.9892 31.4025
20190522 1 20190522000143 85132057 120.9892 31.4025 20190522000244 85132032 120.9892 31.4025
20190522 1 20190522000244 85132032 120.9892 31.4025 20190522000246 85221124 120.9844 31.40311
table 1 handset signaling data example
Taking the mobile phone signaling data of 5 months in Kunshan city, Jiangsu province as an example, the steps S1 to S6 identify the travel mode of each travel OD every day, and the results are shown in the following table:
Figure 2
table 2 example of travel OD mode division results
Step S75, using the above-mentioned division result as initial data, and performing three kinds of corrections to the initial data:
correction 1: correcting the bus trip mode based on the trip time and the urban bus network; the scheme fully considers the setting of different grades of roads in different urban roads at the supply side, the control regulation of different local specific road sections and the influence of the real bus network arrangement condition on the travel mode, and corrects the identification result of the traffic mode according to the control regulation.
And (3) correction 2: correcting the travel mode based on the personal attribute and the personal travel behavior characteristic; according to the scheme, the identified travel OD traffic mode is further corrected based on the user population type, the private vehicle travel chain closing principle and the user travel behavior characteristics, and the accuracy of mode division results is improved.
And (3) correction: and sequentially correcting the bus trip mode based on the trip time and the urban bus network and the trip mode based on the personal attribute and the personal trip behavior characteristic.
For the division result of each correction scheme, dividing the ODs adopting the same trip mode into a group, counting k groups, and marking the set of each group of data as Ci，i＝1，2，...，k；Calculating the departure time start _ time, the travel distance, the travel time consumption move _ time, the travel average speed, the 85-quantile speed _85 of the travel whole-course speed and the variation coefficient cv of the travel whole-course speed of each travel OD as characteristic parameters of each OD, and respectively calculating the DVI index of the division result of the three schemes:
Figure BDA0002515873100000091
wherein k denotes the number of result categories divided by way, Ci(i 1, 2.. k) represents a set of OD data in the same manner, xi/xjA six-dimensional feature vector composed of the 6 feature parameters for each piece of OD data;
and selecting a scheme corresponding to the result with the larger DVI index as a final correction result.
Examples of the results are shown in the following table:
Figure BDA0002515873100000101
table 3DVI index evaluation results example
According to the DVI index, the correction result of the correction 3 scheme is optimal, and the result of the correction 3 scheme is analyzed and found to be the highest in accuracy according to comparison of the results of the three correction schemes shown in fig. 2-5 (fig. 2 is a proportion structure diagram of the travel modes of residents obtained by actual investigation in the embodiment as a standard, fig. 3 is a diagram of the result after correction based on the correction 1 scheme in the embodiment, fig. 4 is a diagram of the result after correction based on the correction 2 scheme in the embodiment, and fig. 5 is a diagram of the result after correction based on the correction 3 scheme in the embodiment). Fig. 6 to 8 respectively analyze three travel indexes of travel distance distribution of different transportation modes, travel speed distribution of different transportation modes, and car travel all-day traffic distribution, and it is known that the travel indexes of the travel modes of the correction 3 scheme are reasonable.
Therefore, correction 3 is selected as the final recognition scheme.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.

Claims (8)

1. A comprehensive urban trip mode identification method based on mobile phone signaling data is characterized by comprising the following specific steps:
s1, identifying a user parking point according to the mobile phone signaling data to obtain a travel OD;
s2, preprocessing the travel OD, extracting travel characteristic parameters and deleting invalid ODs;
s3, identifying a subway trip OD according to the information of the special subway base station;
s4, dividing walking and long-distance car travel ODs according to the travel distance and the average travel speed to obtain travel OD dividing results with obvious characteristics, and performing S5 on the remaining unidentified travel ODs;
s5, climbing the travel distance and travel time consumption of the four types of transportation ways, namely walking, bicycles, buses and cars in the route API planned by the Hide navigation according to the coordinates of the travel origin-destination points, identifying the travel way of the OD according to the deviation value of the mobile phone data and the Hide data, and performing S6 on the rest unidentified travel ODs;
s6, performing mode division on the remaining unidentified travel ODs by using a fuzzy K-means clustering algorithm;
s7, correcting the partially misjudged row mode according to the urban road and the public traffic network, comprising the following steps:
s71, establishing a corresponding relation between the service range of the mobile phone 4G base station and the urban road network;
s72, matching the mobile phone signaling data travel OD with the urban road network to obtain all road sections and road section attributes contained in the travel;
s73, correcting matching results of the elevated road section, the high speed road section and the road section of the overlapped part of the tunnel and the ground and the OD according to travel characteristics;
s74, correcting the travel mode of the non-motor vehicle based on different road types;
aiming at the OD of the road section which only allows the motor vehicle to travel at high speed, high frame and tunnel in the travel path obtained through the steps, if the identified travel mode is a non-motor vehicle, the traffic mode is corrected to be a car;
and S75, taking the division result as initial data, and respectively correcting the initial data in three ways:
correction 1: correcting the bus trip mode based on the trip time and the urban bus network;
and (3) correction 2: correcting the travel mode based on the personal attribute and the personal travel behavior characteristic;
and (3) correction: the method comprises the steps that a public transport trip mode based on trip time and an urban public transport network is sequentially used, and the trip mode is corrected based on individual attributes and individual trip behavior characteristics;
for the division result of each correction scheme, dividing the ODs adopting the same trip mode into a group, counting k groups, and marking the set of each group of data as CiI ═ 1,2, …, k; calculating the departure time start _ time, the travel distance, the travel time consumption move _ time, the travel average speed, the 85-quantile speed _85 of the travel whole-course speed and the variation coefficient cv of the travel whole-course speed of each travel OD as characteristic parameters of each OD, and respectively calculating the DVI index of the division result of the three schemes:
Figure FDA0003624739300000021
wherein k denotes the number of result categories divided by way, CiDenotes a set of OD data of the same type, i 1,2, …, k, xi/xjA six-dimensional feature vector composed of the 6 feature parameters for each piece of OD data;
and selecting a scheme corresponding to the result with the larger DVI index as a final correction result.
2. The method according to claim 1, wherein step S1 is to identify a user parking point based on a base station parking time and a service radius after the obtained mobile phone signaling data is cleaned, and when a user stays within a service radius threshold radius _ range for a time period exceeding a parking time threshold min _ state _ time with a certain base station as a center, the base station is used as the parking point of the user, and a travel OD is obtained according to the travel parking point.
3. The method according to claim 1, wherein the step S2 is for each obtained travel OD, extracting travel features thereof, and includes: the starting point longitude lng, the starting point latitude lat, the end point longitude lng, the end point latitude lat, the starting time start _ time, the travel distance, the travel time consumption _ time, the travel average speed, the 85 quantile speed _85 of the travel whole-course speed and the variation coefficient cv of the travel whole-course speed; and screening and deleting invalid ODs with the distance of the row smaller than an effective travel distance threshold value min _ dis, the travel time move _ time smaller than an effective travel time threshold value min _ time and the average travel speed greater than the maximum average travel speed max _ speed of the urban transportation.
4. The method according to claim 1, wherein step S4 divides the mode according to a travel distance and a travel average speed of the travel OD, and comprises the following specific steps:
s41, if the distance exceeds the long-distance travel OD of the long-distance travel threshold long _ dis in the city, dividing the travel into travel in a car mode;
and S42, if the average travel speed is less than or equal to the walking speed threshold walk _ speed and the travel distance is less than or equal to the travel OD of the walking travel distance threshold walk _ dis, dividing the travel into walking type travel.
5. The method according to claim 1, wherein step S5 identifies the OD travel mode by comparing the travel characteristics of the OD with the characteristic similarity of travel planned in different climbing heights, and the specific steps are as follows:
s51, respectively crawling the travel distance GD _ dis and travel time consumption GD _ time of routes planned in a mode of Gauder walking, bicycles, buses and cars by taking travel characteristic starting point longitude lng, starting point latitude lat, ending point longitude lng, ending point latitude lat and starting time start _ time as parameters for each unidentified travel OD;
s52, calculating the total deviation GD _ diff of each travel OD and the 4 kinds of traffic mode height plans:
i) and a certain traffic mode planning time offset degree GD _ time _ diff of Gagde is Math
ii) distance offset GD _ dis _ diff from the planned distance of the high-end transportation mode, which is equal to (math
iii) a total deviation GD _ diff from the Gaode transportation mode plan, a time deviation GD _ time _ diff alpha + from the Gaode transportation mode plan distance deviation GD _ dis _ diff (1-alpha), wherein alpha is a weight of time consumption of the Gaode planned travel;
s53, comparing and judging the traffic mode with the lowest OD deviation GD _ diff;
s54, if the walking mode deviation GD _ diff is the lowest and is smaller than the deviation threshold max _ diff, and the travel distance is less than the walking travel distance threshold walk _ dis, judging that the walking mode is walking;
s55, if the riding mode deviation GD _ diff is the lowest and is smaller than the deviation threshold value max _ diff, the bicycle is judged to be a bicycle;
and S56, if the deviation GD _ diff of the car mode is the lowest and is smaller than the deviation threshold value max _ diff, determining that the car is the car.
6. The method of claim 1, wherein the step S6 is performed by dividing the remaining ODs to be identified by using a fuzzy K-means machine learning algorithm, and comprises the following steps:
s61, selecting 10000 samples randomly, and training the samples based on an objective function J (U, C), the fuzzification degree m and the cluster number K by using a fuzzy K-means algorithm to obtain a clustering center C (C)1,c2,…,cK}；
S62, calculating the distance between each center and the residual samples to obtain a membership matrix Ui＝{ui1,ui2,…,uiK}；
S63, sequentially defining the traffic mode of each cluster according to the average speed parameter speed of each cluster center, wherein the speed of each cluster is car, bus, electric vehicle, bicycle and walking from high to low.
7. The method according to claim 1, wherein in step S75, the step of correcting the bus trip mode based on the trip time and the urban public transportation network comprises:
s711, establishing a corresponding relation between the service range of the mobile phone 4G base station and an urban road network;
s712, matching the mobile phone signaling data travel OD with the urban road network to obtain all road sections and road section attributes contained in the travel;
s713, correcting matching results of the section of the overhead, the high speed and the part of the tunnel coinciding with the ground and the OD according to the travel characteristics;
s714, correcting the travel mode of the non-motor vehicle based on different road types;
aiming at the OD of the road section which only allows the motor vehicle to travel at high speed, high frame and tunnel in the travel path obtained through the steps, if the identified travel mode is a non-motor vehicle, the traffic mode is corrected to be a car;
s715, correcting the bus trip mode based on the trip time and the urban bus network, specifically:
s7151, matching the urban public transport network with an urban road network to form a road node sequence corresponding to each public transport line, and storing the operation time of the line in attribute information;
s7152, identifying the travel mode as an OD of bus travel, and if the corresponding route passes through the road section and has no bus line, modifying the traffic mode into a car;
s7153, identifying the travel mode as the OD of the bus travel, and if the departure time start _ time is not within the corresponding bus route operation time period, modifying the traffic mode into a car;
s716, correcting the travel mode identification result based on the current urban road management and control regulation, specifically:
s7161, traversing the control rule and the limit vehicle limit information corresponding to each road section through which each travel route obtained in S73 is matched with each travel OD;
s7162, if the road section passing by is a bus lane, modifying the traffic mode into a bus;
s7163, if the road section passing through is a one-way road, or the left/right turn of the vehicle is forbidden through the intersection, the corresponding track still turns left/right, and the identified travel mode is motor vehicle travel, the traffic mode is corrected to be the electric vehicle;
and S7164, if the passing road section is a control road section which is specially forbidden in a certain traffic mode and the identified travel mode is the traffic mode, correcting the traffic mode into the traffic mode which is allowed to pass through the road section in the current city.
8. The method according to claim 1, wherein in step S75, the modifying the travel style based on the personal attributes and the personal travel behavior characteristics comprises:
s721, obtaining mobile phone signaling data of a target city for one month continuously, and identifying the travel mode of each travel OD every day according to the steps;
s722, counting the number of days of stay in the target city in the residence time period of the user and identifying the residence according to the stay time;
s723, judging the population type of the user according to the activity condition of the user in the target city in one month;
s724, modifying the mode division of the user' S travel OD based on the population type of the user, specifically:
s7241, for a user with a population type of a transit population, if the user has a travel OD which is identified as a mode other than car travel, modifying the traffic mode into a car;
s7242, for a user with a short-term business travel population as a population type, if the user has a travel mode identified as an OD for the electric vehicle to travel, modifying the traffic mode into a car;
s725, correcting the travel mode based on the private transportation travel chain closing principle, specifically:
s7251, if the starting point or the ending point of the trip is the residence of the user identified in step S72, determining that the type of the trip is a home-based trip, regarding the trip OD of the user every day;
s752, if one of the paired home-based travel ODs identified by the user is a private vehicle, modifying the other home-based travel OD into the same private vehicle;
s726, judging the common user of the motorized traffic mode according to the travel mode of the user for one month;
s727, correcting the daily trip mode division of the user based on the trip behavior characteristics, specifically:
s7271, clustering the travel OD of each user based on unit day;
s7272, aiming at the electric vehicle common users, if the electric vehicle travels in all travel OD on the same day, correcting the OD travel mode that all travel distances are smaller than an electric vehicle travel distance threshold ebike _ distance on the same day into the electric vehicle;
s737, for the car-used user, if the home-based travel mode on the same day is a car, all travel OD modes other than walking travel on the same day are modified to be cars.
CN202010475924.2A 2020-05-29 2020-05-29 Urban trip mode comprehensive identification method based on mobile phone signaling data Active CN111653093B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010475924.2A CN111653093B (en) 2020-05-29 2020-05-29 Urban trip mode comprehensive identification method based on mobile phone signaling data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010475924.2A CN111653093B (en) 2020-05-29 2020-05-29 Urban trip mode comprehensive identification method based on mobile phone signaling data

Publications (2)

Publication Number Publication Date
CN111653093A CN111653093A (en) 2020-09-11
CN111653093B true CN111653093B (en) 2022-06-17

Family

ID=72352727

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010475924.2A Active CN111653093B (en) 2020-05-29 2020-05-29 Urban trip mode comprehensive identification method based on mobile phone signaling data

Country Status (1)

Country Link
CN (1) CN111653093B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112542045B (en) * 2020-12-01 2021-09-24 江苏欣网视讯软件技术有限公司 Method and system for identifying traffic travel mode based on mobile phone signaling
CN112801181B (en) * 2021-01-27 2022-08-05 南京邮电大学 Urban signaling traffic flow user classification and prediction method, storage medium and system
CN113407906B (en) * 2021-05-13 2024-02-02 东南大学 Method for determining traffic travel distribution impedance function based on mobile phone signaling data
CN113256987B (en) * 2021-06-30 2021-10-29 深圳市城市交通规划设计研究中心股份有限公司 Multi-source data fusion-based resident trip chain generation method and co-multiplication query method
CN114141008B (en) * 2021-10-29 2023-05-12 东南大学 Shared public transportation service area selection method based on mobile phone signaling data
CN117292546B (en) * 2023-10-23 2024-05-07 重庆交通大学 Traffic travel mode OD calculation method based on mobile phone GPS data

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20030089059A (en) * 2002-05-16 2003-11-21 에스케이 텔레콤주식회사 Apparatus and service method for navigating of car driving or walking by using mobile phone
JP2008146249A (en) * 2006-12-07 2008-06-26 Nippon Telegraph & Telephone West Corp Probe data analysis system
CN102799897A (en) * 2012-07-02 2012-11-28 杨飞 Computer recognition method of GPS (Global Positioning System) positioning-based transportation mode combined travelling
CN103606279A (en) * 2013-11-27 2014-02-26 中国航天系统工程有限公司 Road trip mode distinguishing method and system based on smart phone
CN104751631A (en) * 2015-03-13 2015-07-01 同济大学 Method of judging mode of transportation of train chain based on GPS (Global Positioning System) positioning and fuzzy theory
CN105101092A (en) * 2015-09-01 2015-11-25 上海美慧软件有限公司 Mobile phone user travel mode recognition method based on C4.5 decision tree
CN106327000A (en) * 2015-06-30 2017-01-11 阿里巴巴集团控股有限公司 Method and system for identifying trip mode
CN106448173A (en) * 2016-11-28 2017-02-22 东南大学 Method for classifying long-distance travel transportation types based on data of mobile phones
CN108171974A (en) * 2017-12-27 2018-06-15 东南大学 A kind of traffic trip mode discrimination method based on cellular triangulation location data
CN109272032A (en) * 2018-09-05 2019-01-25 广州视源电子科技股份有限公司 Travel mode identification method and device, computer equipment and storage medium
CN109561386A (en) * 2018-11-23 2019-04-02 东南大学 A kind of Urban Residential Trip activity pattern acquisition methods based on multi-source location data
CN109727452A (en) * 2019-01-08 2019-05-07 江苏交科能源科技发展有限公司 Trip proportion accounting method based on mobile phone signaling data

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3091498A1 (en) * 2015-05-07 2016-11-09 TrueMotion, Inc. Motion detection system for transportation mode analysis

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20030089059A (en) * 2002-05-16 2003-11-21 에스케이 텔레콤주식회사 Apparatus and service method for navigating of car driving or walking by using mobile phone
JP2008146249A (en) * 2006-12-07 2008-06-26 Nippon Telegraph & Telephone West Corp Probe data analysis system
CN102799897A (en) * 2012-07-02 2012-11-28 杨飞 Computer recognition method of GPS (Global Positioning System) positioning-based transportation mode combined travelling
CN103606279A (en) * 2013-11-27 2014-02-26 中国航天系统工程有限公司 Road trip mode distinguishing method and system based on smart phone
CN104751631A (en) * 2015-03-13 2015-07-01 同济大学 Method of judging mode of transportation of train chain based on GPS (Global Positioning System) positioning and fuzzy theory
CN106327000A (en) * 2015-06-30 2017-01-11 阿里巴巴集团控股有限公司 Method and system for identifying trip mode
CN105101092A (en) * 2015-09-01 2015-11-25 上海美慧软件有限公司 Mobile phone user travel mode recognition method based on C4.5 decision tree
CN106448173A (en) * 2016-11-28 2017-02-22 东南大学 Method for classifying long-distance travel transportation types based on data of mobile phones
CN108171974A (en) * 2017-12-27 2018-06-15 东南大学 A kind of traffic trip mode discrimination method based on cellular triangulation location data
CN109272032A (en) * 2018-09-05 2019-01-25 广州视源电子科技股份有限公司 Travel mode identification method and device, computer equipment and storage medium
CN109561386A (en) * 2018-11-23 2019-04-02 东南大学 A kind of Urban Residential Trip activity pattern acquisition methods based on multi-source location data
CN109727452A (en) * 2019-01-08 2019-05-07 江苏交科能源科技发展有限公司 Trip proportion accounting method based on mobile phone signaling data

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于手机信令和导航数据的出行方式识别方法;杜亚朋 等;《计算机应用研究》;20180831;第35卷(第8期);第2311-2314页 *

Also Published As

Publication number Publication date
CN111653093A (en) 2020-09-11

Similar Documents

Publication Publication Date Title
CN111653097B (en) Urban trip mode comprehensive identification method based on mobile phone signaling data and containing personal attribute correction
CN111653093B (en) Urban trip mode comprehensive identification method based on mobile phone signaling data
CN111653094B (en) Urban trip mode comprehensive identification method based on mobile phone signaling data and containing road network correction
CN111653096B (en) Urban trip mode identification method based on mobile phone signaling data
CN109145954B (en) Network taxi appointment travel safety evaluation method and system based on multi-source time-space data
CN108761509B (en) Automobile driving track and mileage prediction method based on historical data
CN112133090A (en) Multi-mode traffic distribution model construction method based on mobile phone signaling data
CN111862606B (en) Illegal operating vehicle identification method based on multi-source data
CN103646560B (en) The extracting method in taxi wheelpath experimental knowledge path
Veloso et al. Sensing urban mobility with taxi flow
CN110836675B (en) Decision tree-based automatic driving search decision method
CN112036757B (en) Mobile phone signaling and floating car data-based parking transfer parking lot site selection method
CN103680130B (en) A kind ofly obtain lead the way expert's method of region based on floating car technology
CN112784000A (en) Passenger searching method based on taxi track data
CN111222381A (en) User travel mode identification method and device, electronic equipment and storage medium
CN110909434A (en) Urban public transport trunk line network design method under low-carbon guidance
CN116017407A (en) Method for reliably identifying resident trip mode driven by mobile phone signaling data
CN109489679A (en) A kind of arrival time calculation method in guidance path
CN113569977A (en) Travel purpose identification method based on mobile phone signaling data
CN116862325A (en) Urban rail transit passenger travel chain inference method and system
CN113408833A (en) Public traffic key area identification method and device and electronic equipment
ZHAO et al. Big data-driven residents’ travel mode choice: a research overview
CN111931968B (en) Method for optimizing public bicycle station layout by using mobile phone GPS data
CN114882693B (en) Bus passenger getting-off station prediction method based on card swiping data deep mining
CN111275241A (en) Bus passenger getting-off station inference method based on machine learning decision tree

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant