CN115713206A - Bus individual trip decision model - Google Patents
Bus individual trip decision model Download PDFInfo
- Publication number
- CN115713206A CN115713206A CN202211424041.4A CN202211424041A CN115713206A CN 115713206 A CN115713206 A CN 115713206A CN 202211424041 A CN202211424041 A CN 202211424041A CN 115713206 A CN115713206 A CN 115713206A
- Authority
- CN
- China
- Prior art keywords
- station
- bus
- individual
- trip
- time
- 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.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Traffic Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a bus individual trip decision model, which is used for obtaining a trip rule of a bus individual based on the analysis of bus individual trip data and inputting the rule as initial basic data of the model; identifying and calculating the information of the station points of the individual bus trip through a time matching method, a trip chain theory and bus station characteristics, and finishing the identification and division of the individual trip chain by combining the connotation and the function level of the multi-mode bus, defining a virtual starting and ending point for the departure or arrival station cluster of the individual bus, and finishing the analysis and processing of the individual trip data; then based on individual historical travel information obtained by the method, an individual travel decision model related to departure time selection, getting-on stop selection and bus route selection is established by taking an individual as a unit and based on a Markov decision process theory; the individual can be enabled to reach the travel terminal by itself with the least total cost.
Description
Technical Field
The invention relates to the technical field of public transportation, in particular to a bus individual trip decision model.
Background
In order to relieve the adverse effect of social vehicles which are increased sharply in road traffic on cities, meet the demand of urban residents on traffic and travel as much as possible and ensure the fairness of the urban residents on the travel, the multi-mode public traffic system quickly becomes a key component of a large-city comprehensive traffic transportation system, and plays a role in bearing urban large-traffic volume traffic transportation service and relieving urban traffic pressure;
the bus travel OD calculation method research is basic work of bus passenger flow prediction, the occurrence and attraction amount of each bus stop can be further determined by calculating the OD amount of bus passengers, and the method plays an important role in judging whether the current urban bus network setting is reasonable and further optimizing and adjusting the network. The calculation of the early urban bus passenger flow OD is mainly based on a manual investigation method by using an questionnaire; with the popularization and application of the IC card technology in the public transportation field, students begin to research the travel information of passengers hidden in the IC card for calculating the getting-on and getting-off station of the bus passengers, but the number of the passengers getting-off at the station is calculated mainly based on the attraction of the bus station and by combining factors such as land utilization properties around the station, but the method does not consider the travel characteristic difference of the individual passengers, and cannot obtain the specific getting-off station of the passengers, so that a bus individual travel decision model is urgently needed to solve the above problems.
Disclosure of Invention
The invention provides a method for establishing a bus individual travel decision model by calculating the travel chains of the bus upper and lower station points of an individual to realize the multi-mode bus recent passenger flow prediction driven by individual travel data, so as to solve the problems in the prior art.
In order to achieve the purpose, the invention provides the following technical scheme: a bus individual trip decision model, comprising:
obtaining travel rules of the bus individuals based on analysis of the travel data of the bus individuals, and inputting the travel rules as initial basic data of a model;
based on the meaning and the definition of the function level of the multi-mode bus, optimizing initial basic data: dividing individual trip chains into multi-mode bus trips and single-mode bus trips according to a transfer time threshold, performing density clustering according to spatial position attributes of individual departure or arrival stations in a road network, establishing corresponding virtual OD pairs, and identifying information of the individual trip chains;
constructing state spaces and action space sets of different trips according to the information of the individual trip chains, and continuously updating the action sets according to the incomplete information environment of the public transport network; and establishing an individual trip decision model comprising departure time selection, getting-on station selection and trip path selection by taking an individual as a unit by combining historical individual trip time cost, trip habit definition and direct return of a calibration state action pair, future return and state transition probability.
Preferably, the bus individual trip data comprises IC card swiping data, bus-mounted GPS data and bus route stop position information, wherein the bus route stop position information is combined, time is used as a main connecting field and is supplemented with the GPS data of the bus to obtain a bus arrival time table, sparse GPS data is supplemented through an interpolation method, the bus arrival time is obtained by matching the sparse GPS data with the station longitude and latitude, the IC card swiping time of the passenger is obtained by using a time matching method, and the boarding stop identification process of the passenger with the IC card is completed.
Preferably, the GPS data interpolation process: in the selected acquired GPS data, the longitude and latitude of any two adjacent GPS records are respectively (x) b,w,j ,y b,w,j )、(x b,w,j+1 ,y b,w,j+1 ) The recording time is respectively RT b,w,j 、RT b,w,j+1 And the interval time is Λ, inserting Λ data between two GPS records by taking 1s as an interpolation interval, and calculating the longitude and latitude of the inserted λ -th GPS data:
therein correspond toRecording time as RT b,w,j +λ。
Preferably, the trips of the individual buses are divided into a closed bus trip chain and a non-closed bus trip chain according to a trip chain theory, and the getting-off stops of the different types of bus trip chains are calculated by combining the identification result of the getting-on stops and the characteristics of the bus stops.
Preferably, the getting-off station estimation specifically includes:
step 2, judging the getting-on station BS of the current riding train number p,k And the boarding station BS of the next ride train p,k+1 Whether the vehicles belong to the same line or not, if so, taking the getting-on station of the next ride train AS the getting-off station of the current ride train, namely AS p,k ＝BS p,k Otherwise, executing step 3;
step 3, determining to use BS p,k+1 In a circular area with the circle center and the maximum walking distance D as the radius, the CAS is set from the upstream to the downstream stations where the current riding train passes p,k ＝{cas m M is more than or equal to 0 and is more than or equal to M }; if the set is an empty set, the trip chain belongs to a non-public traffic trip chain; if the number of stations in the set is M =1, the only station of the set is a get-off station AS of the current riding train number p,k (ii) a If M > 1, respectively calculating the most upstream station cas 1 As a starting point, via CAS p,k Any station cas m And with BS p,k+1 Trip time t as terminal point m Selecting the station cas with the shortest travel time * ＝argmin(t m ) Get-off station AS for current riding number p,k The calculation formula is as follows:
wherein v is w Is the walking speed; d m For station cas m And BS p,k+1 The distance between them; l is m-1,m For station cas m-1 And cas m The length of the bus route between; v. of b The bus operating speed;
step 4, judging the getting-on station BS of the last train taking p,k And a boarding station BS for taking a first train p,1 Whether belonging to the same line or not, if so, taking the getting-on station of the first-time train AS the getting-off station of the last-time train, namely AS p,k ＝BS p,1 Otherwise, executing step 5;
step 5, determining the getting-on station BS of the first taking train number p,1 And (3) in a circular area with the circle center and the maximum walking distance D as the radius, collecting stations from upstream to downstream of the last ride pass, if the collection is an empty collection, determining that the passenger trip chain belongs to a non-closed bus trip chain, and otherwise, determining a get-off station of the last ride according to the step 3.
Preferably, the individual outgoing behavior further comprises a non-public transport outgoing chain, and the station of getting off is calculated as follows:
Step 2, calculating that the passenger is at any station das n Probability of getting off the vehicle PA p,n The calculation formula is as follows:
PA p,n ＝f n,1 ·f n,2 /∑(f η,1 ·f η,2 )；
wherein f n,1 、f n,2 Respectively representing the development strength around the station and the bus accessibility characteristic parameters; f. of n,1 ＝BF n /∑BF η ，f n,2 ＝R n /∑R η ，BF n For site das n The average passenger capacity in a certain period indirectly reflects the development strength around the site; r n For passing through site das n The number of bus lines.
Step 3, according to the situation that passengers are at any station das n Probability of getting off the vehicle PA p,n Adopting a roulette method to calculate the get-off station AS of the current riding car p,k 。
Preferably, the individual trip chains are divided according to the transfer time threshold, specifically:
step 1: judging the boarding station BS of two times before and after the passenger according to the recognition result of the boarding station p,k And BS p,k+1 Whether the passengers belong to the same line or not, if the passengers belong to the same line, the daily trip of the passengers belongs to single-mode bus trip, if the passengers do not belong to the same line, whether the two bus lines belong to the same function level or not is judged, if the passengers belong to the same function level, the daily trip of the passengers still belongs to single-mode bus trip, and if the passengers do not belong to the same function level, the daily trip of the passengers belongs to multi-mode bus trip;
step 2: according to the get-off station AS p,k Determining the site AS p,k Corresponding bus operation shift w b Obtaining the arrival schedule of the vehicles in the shift, and connecting AS p,k The station name of the vehicle is matched with the station in the arrival schedule to obtain the AS of the vehicle at the station p,k AT arrival time AT b,w,j (ii) a Taking 0.25 of the current travel time of the bus as the getting-off time delta t of the passenger p Then the alighting time AT of the passenger b,w,j,l The calculation formula is as follows:
AT b,w,j,l ＝AT b,w,j +Δt p ；
and step 3: calculating a certain get-off time AT b,w,j,l And the next station BS of getting on the bus p,k+1 Corresponding boarding time TT p,k+1 Time interval mu of p,l ，μ p,l ＝TT p,k+1 -AT b,w,p,l ；
And 4, step 4: when the passenger transfer belongs to the same-station transfer, namely the boarding stations of two times before and after the passenger belong to the same line, determining the maximum waiting time WT of the passenger at the station according to the departure intervals of the lines of different function levels of the multi-mode bus p,k Taking the maximum waiting time as the maximum transfer time threshold eta of the transfer at the same station p,k I.e. eta p,k ＝WT p,k (ii) a When the passenger transfer belongs to the different station transfer, namely the two boarding stations before and after the passenger do not belong to the same line, the passenger walking speed v and the maximum walking distance D acceptable by the passenger are used as the basis w Calculating a maximum transfer time threshold;
and 5: will go out the time interval mu p,l With a maximum transfer time threshold η p,k Make a comparison as if mu p,k ≤η p,k Then the bus individual completes single-time multi-mode or single-mode bus trip on the same day, and mu p,l Namely the transfer time; on the contrary, the bus individual finishes multiple multi-mode or single-mode bus trips on the same day, namely mu p,l Is the active time.
Preferably, the establishment of the virtual OD pair specifically includes:
step 2, setting neighborhood parameters epsilon according to the radiation range of different functional level bus stops, and correspondingly adjusting and setting MinPts according to specific research objects; from the set D of neighborhood parameter (ε, minPts) pairs p,BS Performing core object search to obtain a core object set omega p,BS ；
Step 3, from the set omega p,BS Randomly selecting a core object as a seed to execute a cluster generation algorithm, and searching all sites which can be reached by the density of the core object, thereby forming a first cluster
Then will be
The core object contained in (2) is from Ω p,BS Screening, then respectively randomly selecting one seed from the updated set to generate a next cluster, and repeating until the set is empty;
step 4, according to each cluster
Screening out the stop get-off stations of the trip chains corresponding to the numbers by the trip chain numbers corresponding to each object, and correspondingly forming each set
Each object in the set also comprises the number of each trip chain and the longitude and latitude coordinates of the corresponding get-off station;
step 5, according to the neighborhood parameter (epsilon, minPts) pair set
Searching core objects, removing abnormal points and obtaining a core object set
Step 6, calculating the set
And
a center point of (a); the central points of each group of corresponding sets jointly form a plurality of groups of virtual OD point pairs of the passenger p; the calculation formula is as follows:
wherein the content of the first and second substances,
an X coordinate representing a center point of each set;
y coordinates representing the center point of each set; x is the number of b,w,n X coordinates representing the objects contained in each set; y is b,w,n Y coordinates representing the objects contained in each set; n represents an object in each set; n represents the value contained in each setAll of the objects of (a).
Preferably, the state space S:
S＝{Destination/Origin,Departure-time,Boarding-stion,En-route,Alighting-station}；
wherein, destination/Origin is the longitude and latitude coordinates of a pair of virtual OD points in the individual history; the department-time is the specific Departure time information selected by the individual; the Boarding-sting is the longitude and latitude of the getting-on station selected by the individual and the collinear route information contained in the station; en-route is information such as vehicles, operation train numbers and the like of a corresponding line of an individual selected station point; the adjusting-station is information such as routes, vehicles, operating times, longitude and latitude of the get-off station selected by the individual.
An action space A:
a = { Select-future-time, select-routing-station, select-bus-route, select-addressing-station, to-the-destination } each action space is a data set To be selected, wherein the data set To be selected comprises all departure time information of individual historical trips; the Select-pairing-station to-be-selected set comprises all getting-on site information of individual historical trips; the Select-bus-route to-be-selected set comprises collinear bus route information corresponding to all boarding stations of individual historical trips; the Select-aligning-station to-be-selected set comprises all historical get-off station information corresponding to individual 'bus route selection'; the To-the-destination candidate set includes longitude and latitude coordinates of all the established virtual OD points.
Preferably, the action selection of each state in the individual trip decision model determines the state transition probability according to the future reward, so as to select the action with the maximum state transition probability as the currently selected action and transfer to the next state, and when the action is transferred to the next state, the direct reward of the action selection becomes known and updates the historical experience value set.
Compared with the prior art, the invention has the beneficial effects that: according to the method, the information of the station points of the individual bus trip is identified and calculated through a time matching method, a trip chain theory and bus station characteristics, the identification and division of the individual trip chain are completed by combining the connotation and the function level of the multi-mode bus, a virtual starting and ending point is defined for the starting or arriving station cluster of the individual, and the analysis and processing of the individual trip data are completed; then based on individual historical travel information obtained by the method, an individual travel decision model related to departure time selection, getting-on stop selection and bus route selection is established by taking an individual as a unit and based on a Markov decision process theory; the individual can be enabled to reach the travel terminal by itself with the least total cost.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.
In the drawings:
FIG. 1 is a flow chart of a bus individual travel decision model of the invention;
FIG. 2 is a flow chart of passenger boarding station identification according to the present invention;
FIG. 3 is a flowchart of passenger disembarkation site deduction according to the present invention;
FIG. 4 is a graph illustrating the number of passengers getting on the vehicle at 108 stations according to an embodiment of the present invention;
fig. 5 is a transition diagram of the individual state of travel in the multi-mode public transportation network according to the invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.
The embodiment is as follows: as shown in fig. 1, a decision model for bus individual trip includes:
the identification and calculation of the station of getting on or off bus for individual bus trip, the identification of the individual multi-mode bus trip chain, the establishment of the virtual OD pair and the establishment of the individual multi-mode bus trip decision model are carried out, wherein:
obtaining travel rules of the bus individuals based on analysis of the travel data of the bus individuals, and inputting the travel rules as initial basic data of a model;
based on the meaning and the definition of the function level of the multi-mode bus, optimizing initial basic data: dividing individual trip chains into multi-mode bus trips and single-mode bus trips according to a transfer time threshold, performing density clustering according to spatial position attributes of individual departure or arrival stations in a road network, establishing corresponding virtual OD pairs, and identifying information of the individual trip chains;
constructing state spaces and action space sets of different trips according to the information of the individual trip chains, and continuously updating the action sets according to the incomplete information environment of the public transport network; and establishing an individual travel decision model comprising departure time selection, boarding station selection and travel route selection by taking an individual as a unit by combining historical individual travel time cost, travel habit definition and direct return, future return and state transition probability of a calibration state action pair.
The bus individual travel data are basic data of bus passenger flow prediction, and various element information including individual getting-on/off stops, getting-on/off time and the like can be obtained through analysis and mining of the travel data, so that the travel rule of the bus individual is mastered, and initial basic data input is provided for building a passenger flow prediction simulation model; the bus individual trip data comprises IC card swiping data, bus-mounted GPS data and bus line stop position information;
referring to fig. 2, the boarding station identification is specifically:
step 1: and selecting research data.
Selecting a bus line r to be researched, and acquiring any passing station s i Latitude and longitude (x) i ,y i ) Selecting a certain operating vehicle b on the line and a certain operating vehicle number w b And IC card data of any passenger p riding in the vehicle.
Step 2: clearing redundant data: calculating the interval time IT between each card swiping record and the two card swiping records p,k If 0 < IT p,k ＜L S,S+1 /v b (L S,S+1 Calculating the distance between the current boarding station and the downstream adjacent station by using a hemiversine formula; v. of b Speed of bus operation), sayThe time interval of card swiping is short, so that a passenger p can be considered to have a follower and share one IC card with the follower, the two card swiping times are regarded as one trip, and the time recorded in the time before the card swiping time is taken as the card swiping time of the time;
and step 3: GPS data interpolation processing: if the bus GPS data belongs to the sparse sample returned by the low frequency and is not related to the arrival and departure of the bus, the GPS data is supplemented by an interpolation method in the selected GPS data, and the longitude and latitude of any two adjacent GPS records are respectively (x) b,w,j ,y b,w,j )、(x b,w,j+1 ,y b,w,j+1 ) Recording time is respectively RT b,w,j 、RT b,w,j+1 And the interval time is Λ, inserting Λ data between two GPS records by taking 1s as an interpolation interval, and calculating the longitude and latitude of the inserted λ -th GPS data:
wherein the corresponding recording time is RT b,w,j +λ；
And 4, step 4: determining the arrival position, arrival time and departure time of the vehicle: for any site s on line r i Calculating the number of vehicles w supplemented by the step 2 by adopting a semi-versine formula b Selecting the longitude and latitude of the GPS data closest to the station according to the distance between the GPS data and the station
As the arrival position of the vehicle, the corresponding time of recording
As arrival time AT b,w,j (ii) a Identifying moments AT in GPS data b,w,j The time when the rear speed is changed from 0 to non-0 state is the corresponding vehicle departure time DT b,w,j ；
And 5: identifying passenger boarding stations: will and the train number w b The associated IC card records are sorted according to the card swiping time, and the kth card swiping time of the passenger p is known as TT p,k (K is more than or equal to 1 and less than or equal to K, wherein K is the card swiping times of the passenger on the same day), comparing the card swiping time recorded by any card swiping time with the arrival time of the train number AT any station, and when AT is detected b,w,i ≤TT p,k ＜AT b,w,i+1 Then AT b,w,i Corresponding site s i Getting-on station BS for swiping the card for the passenger p,k 。
Referring to fig. 4, the trip of the individual bus is divided into a closed bus trip chain, a non-closed bus trip chain and a non-bus trip chain according to a trip chain theory, wherein the calculation of the get-off station of the taking times except the last time in the non-closed bus trip chain is the same as the calculation method of the get-off station of the closed bus trip chain; the getting-off station of the last trip in the current day needs to be calculated by means of historical trip data and judgment of characteristics of downstream stations, and the method is the same as a non-bus trip chain processing method
Wherein, the station calculation of the non-closed bus trip chain is specifically as follows:
step 2, judging the getting-on station BS of the current riding train number p,k And the boarding station BS of the next ride train p,k+1 Whether the vehicles belong to the same line or not, if so, taking the getting-on station of the next ride train AS the getting-off station of the current ride train, namely AS p,k ＝BS p,k Otherwise, executing step 3;
step 3, determining to use BS p,k+1 In a circular area with the circle center and the maximum walking distance D as the radius, the CAS is set from the upstream to the downstream stations where the current riding train passes p,k ＝{cas m M is more than or equal to M and less than or equal to 0; if the set is an empty set, the trip chain belongs to a non-public traffic trip chain; if the number of stations in the set is M =1, the only station of the set is a get-off station AS of the current riding train number p,k (ii) a If M > 1, then divideRespectively calculating with the most upstream station cas 1 As a starting point, via CAS p,k Any station cas m And with BS p,k+1 Time of trip t as terminal point m Selecting the station cas with the shortest travel time * ＝argmin(t m ) Get-off station AS for current ride number p,k The calculation formula is as follows:
wherein v is w Is the walking speed; d is a radical of m For station cas m And BS p,k+1 The distance between them; l is m-1,m For station cas m-1 And cas m The length of the bus route between; v. of b The bus operation speed is set;
step 4, judging the getting-on station BS of the last train taking p,k And pick-up station BS of the first ride p,1 Whether belonging to the same line or not, if so, taking the getting-on station of the first-time train AS the getting-off station of the last-time train, namely AS p,k ＝BS p,1 Otherwise, executing step 5;
step 5, determining the getting-on station BS of the first taking train number p,1 And (3) in a circular area with the circle center and the maximum walking distance D as the radius, collecting stations from upstream to downstream of the last ride pass, if the collection is an empty collection, determining that the passenger trip chain belongs to a non-closed bus trip chain, and otherwise, determining a get-off station of the last ride according to the step 3.
The calculation of the getting-off station of the non-public transport travel chain is as follows:
Step 2, calculating das of passengers at any station n Probability of getting off the vehicle PA p,n The calculation formula is as follows:
PA p,n ＝f n,1 ·f n,2 /∑(f η,1 ·f η,2 )；
wherein f n,1 、f n,2 Respectively representing the development intensity around the station and the bus accessibility characteristic parameters; f. of n,1 ＝BF n /∑BF η ，f n,2 ＝R n /∑R η ，BF n For site das n The average passenger loading in a certain period indirectly reflects the development intensity around the site; r n For passing through site das n The number of bus lines.
Step 3, according to the situation that passengers are at any station das n Probability of getting off the vehicle PA p,n Calculating the getting-off station AS of the current riding number by roulette method p,k 。
In a specific embodiment, taking 108 buses in Dongxian county as research objects, the basic data source adopts bus IC card data and vehicle GPS data acquired from a local bus system; the 108 east-west lines respectively pass through 30 bus stops and 28 bus stops, and the lines are mainly residential land and industrial land. Considering that the public transportation demand of residents in Dongxian county is smaller than that of a large city and the traveling times of passengers in one day may not meet the analysis needs of the city, data of a complete week from 6 months to 4 months to 6 months and 10 days in 2018 are selected for research; using a MySQL program to build a database to complete preprocessing work such as key field screening, missing and redundant data cleaning, and obtaining a total of 1405 IC card swiping records to be analyzed in one week; meanwhile, python language programming is used for completing calculation of a subsequent calculation process of upper and lower station points;
and (3) identification of boarding stations: if the time interval of GPS data in the Dongxian public transportation system is 1-3min, the system belongs to sparse GPS data, and the arrival time of each train in the research period is determined according to the method; the following table shows arrival time of a certain train at a partial station under the vehicle number 110286; identifying the boarding stations of all passengers on the current day according to the arrival time of each train number on the 108-way road from 6/4/6/10/2018, and counting to obtain the distribution of the number of boarding persons on each station, as shown in FIG. 4; the original GPS data and the supplemented GPS data are respectively matched with the IC card data, so that 11.3% of boarding station matching results of 108 paths are different, the original GPS data matching results have errors, and the supplemented GPS data can correct the errors and effectively improve the time matching precision;
and (4) calculating the get-off station: according to the method, a Python language is adopted to write codes for calculation; and 1405 IC card swiping records to be analyzed are arranged in 108 paths, 1183 get-off stations are successfully calculated, and the recognition success rate is 84.2%.
The individual transfer behavior of the urban public transport trip can be divided according to two modes, namely a transfer traffic mode and a transfer space distance, and the individual trip chain is divided according to a transfer time threshold value, specifically:
step 1: judging the boarding station BS of two times before and after the passenger according to the recognition result of the boarding station p,k And BS p,k+1 Whether the passengers belong to the same line or not, if the passengers belong to the same line, the daily trip of the passengers belongs to single-mode bus trip, if the passengers do not belong to the same line, whether the two bus lines belong to the same function level or not is judged, if the passengers belong to the same function level, the daily trip of the passengers still belongs to single-mode bus trip, and if the passengers do not belong to the same function level, the daily trip of the passengers belongs to multi-mode bus trip;
and 2, step: according to the get-off station AS p,k Determining the site AS p,k Corresponding bus operation shift w b Obtaining the arrival schedule of the vehicles in the shift, and connecting AS p,k The station name of the vehicle is matched with the station in the arrival time table to obtain the AS of the vehicle at the station p,k AT arrival time AT b,w,j (ii) a Taking 0.25 of the current travel time of the bus as the getting-off time delta t of the passenger p Then the time AT of getting-off of the passenger b,w,j,l The calculation formula is as follows:
AT b,w,j,l ＝AT b,w,j +Δt p ；
and step 3: calculating a certain get-off time AT b,w,j,l And the next station BS of getting on the bus p,k+1 Corresponding boarding time TT p,k+1 Time interval mu of p,l ，μ p,l ＝TT p,k+1 -AT b,w,p,l ；
And 4, step 4: when the passenger transfer belongs to the same-station transfer, namely the boarding stations of two times before and after the passenger belong to the same line, determining the maximum waiting time WT of the passenger at the station according to the departure intervals of the lines of different function levels of the multi-mode bus p,k Taking the maximum waiting time as the maximum transfer time threshold eta of the transfer of the same station p,k I.e. eta p,k ＝WT p,k (ii) a When the passenger transfer belongs to the different station transfer, namely the boarding stations of two times before and after the passenger do not belong to the same line, according to the maximum walking distance D acceptable by the passenger and the walking speed v of the passenger w Calculating a maximum transfer time threshold;
and 5: interval of travel time mu p,l With a maximum transfer time threshold η p,k Make a comparison as if mu p,k ≤η p,k Then the bus individual completes single-time multi-mode or single-mode bus trip on the same day, and mu p,l Namely the transfer time; on the contrary, the bus individual finishes multiple multi-mode or single-mode bus trips on the same day, namely mu p,l Is the active time.
After the identification and division of individual trip chains are completed, virtual O points and D points of the individuals going out in the multi-mode public transport network are defined for each individual according to historical trip habits, the Manhattan distance between different departure stations (arrival stations) is calculated according to the spatial position attribute of the individual departure stations (arrival stations) in a road network, the departure stations (arrival stations) are clustered, and a virtual O point (D point) is defined for each departure station cluster (arrival station cluster); DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is the most classical Density Clustering algorithm, the Clustering structure of a research object can be determined according to the tightness of sample data distribution, one or more core objects in a group of samples are obtained by setting a group of neighborhood parameters, partial abnormal points can be found while Clustering, the abnormal points in the data set are insensitive, and the virtual OD pairs of the finally determined core objects (which can determine individual rows) are obtained;
the establishment of the virtual OD pairs specifically comprises the following steps:
step 2, setting neighborhood parameters epsilon according to the radiation range of different functional level bus stops, and correspondingly adjusting and setting MinPts according to specific research objects; set D of pairs according to neighborhood parameters (ε, minPts) p,BS Performing core object search to obtain a core object set omega p,BS ；
Step 3, from the set omega p,BS Randomly selecting a core object as a seed to execute a cluster generation algorithm, and searching all sites which can be reached by the density of the core object, thereby forming a first cluster
Then will be
The core object contained in (2) is from Ω p,BS Screening, then respectively randomly selecting one seed from the updated set to generate a next cluster, and repeating until the set is empty;
step 4, according to each cluster
Screening out the stop get-off stations of the trip chains corresponding to the numbers by the trip chain numbers corresponding to each object, and correspondingly forming each set
Each object in the set also comprises the number of each trip chain and the longitude and latitude coordinates of the corresponding get-off station;
step 5, according to the neighborhood parameter (epsilon, minPts) pair set
Searching core objects, removing abnormal points and obtaining a core object set
Step 6, calculating a set
And
a center point of (a); the central points of each group of corresponding sets jointly form a plurality of groups of virtual OD point pairs of the passenger p; the calculation formula is as follows:
wherein the content of the first and second substances,
an X coordinate representing the center point of each set;
y coordinates representing the center point of each set; x is a radical of a fluorine atom b,w,n X coordinates representing the objects contained in each set; y is b,w,n Y coordinates representing the objects contained in each set; n represents an object in each set; n represents all the objects contained in each set.
After the identification and division of the individual trip chains and the establishment of the virtual OD pairs are completed, the individuals form a complete trip chain. Considering the continuity of travel habits, an individual has multiple travel records on the same travel chain, and the travel record data form a basic data source of the multi-mode bus recent passenger flow prediction model. In order to enable the individual trip decision model to have accurate data input, key information of an individual trip chain needs to be identified, and the information comprises an individual trip starting and ending point, time for going to and getting on a bus, time for getting on the bus, station waiting time, vehicle stopping time, time for getting on the bus, station for getting off the bus, time for getting off the bus, transfer time and time for going to an ending point.
As shown in fig. 5, in the multi-mode public transportation network, a trip individual is used as a decision maker to make a departure time selection decision in a "trip end point (i.e., a starting point of a next trip)" state, and after determining a departure time, the trip individual transitions to a "departure time selected" state. And then selecting the boarding station according to the self travel principle and transferring to the selected boarding station state. The individual selects the bus route between the starting and ending points of the trip at the station, and reaches the 'in-vehicle trip' state after the selection is finished; in the bus route selection state, an individual determines whether the individual takes the bus to directly reach a terminal or gets off the bus at some stations to transfer to other routes to reach the terminal, if the individual does not need to transfer, the individual is transferred from the in-bus journey state to the off-bus station selected state, goes to the terminal and returns to the trip terminal state; if the transfer is needed, the getting-on station selection is carried out in the state that the getting-off station is selected, and a new round of action selection decision and state transfer are started; the state of the individual in the trip process in the public transport network is transferred;
wherein, the state space S:
S＝{Destination/Origin,Departure-time,Boarding-stion,En-route,Alighting-station}；
wherein, destination/Origin is the longitude and latitude coordinates of a pair of virtual OD points in the individual history; the department-time is the specific Departure time information selected by the individual; the Boarding-sting is the longitude and latitude of the getting-on station selected by the individual and the collinear route information contained in the station; en-route is information such as vehicles, operation train numbers and the like of a corresponding line of an individual selected station point; the adjusting-station is information such as routes, vehicles, operating times, longitude and latitude of the get-off station selected by the individual.
An action space A:
a = { Select-future-time, select-routing-station, select-bus-route, select-addressing-station, to-the-destination } each action space is a data set To be selected, wherein the data set To be selected comprises all departure time information of individual historical trips; the Select-pairing-station to-be-selected set comprises all getting-on site information of individual historical trips; the Select-bus-route set to be selected comprises collinear bus route information corresponding to all boarding stations of individual historical trips; the Select-aligning-station to-be-selected set comprises all historical get-off station information corresponding to individual 'bus route selection'; the To-the-destination candidate set comprises longitude and latitude coordinates of all the established virtual OD points.
The body row decision model comprises direct return R, future return F and state transition probability P, the action selection of each state is to determine the state transition probability P according to the future return F, so that the action with the maximum P is selected as the currently selected action and is transferred to the next state, and when the action is transferred to the next state, the direct return R selected by the action becomes known and updates the historical empirical value set; first, the corresponding states and motion spaces R, F and P are defined as shown in the following table:
the direct return value is represented by the trip time cost, but the time cost of an individual in the trip process is actually a penalty value, which is contrary to the definition of the return value, so that the direct return corresponding to each action is represented by the negative number of the time cost;
(1)R SDT : the direct return of the selection of the departure time is determined by two factors of the expected arrival time and the historical total trip time of the individual; when the individual selects the departure time, the individual firstly determines the expected arrival time, and determines whether the trip belongs to early arrival (negative value), punctuality (0) or delay (positive value) according to the difference value delta t between the actual arrival time and the expected arrival time, wherein the actual arrival time of the individual is determined by the sum of the departure time and the actual total trip time; the calculation formula of delta t is：Δt＝t depart +t trip -t desire Wherein, t depart Indicates the departure time, t trip Represents the actual travel time, t desire Indicating the desired arrival time.
As the individual is taken as the trip target of the individual on time in the actual trip process, the early arrival and the delay are both positive numbers when defining the direct return so as to ensure the minimum on-time return in order to accord with the actual life; r SDT Can be expressed as:
wherein, Δ t early Indicating the time difference of early arrival, Δ t late Indicating a delayed time difference.
(2)R SBS : the trip individual can determine the getting-on station of the trip at the trip end point, and R is SBS Can be defined as the travel time t of the individual from the travel terminal to the boarding station D/O,BS Negative number of (d): r SBS ＝-t D/O,BS ；
The mode of the individual arriving at the boarding station includes walking, bicycle, automobile, etc., generally walking; when an individual needs to transfer to another station for waiting at a certain getting-off station, R SBS Is defined as the travel time t of an individual from a get-off station to another get-on station BS,BS' Negative number of (d): r SBS ＝-t BS,BS' ；
(3)R SBR : the individual selects the bus route at the getting-on stop, the direct report corresponding to each bus route reflects the total time consumed by the route, including the waiting time t of the trip individual at the current stop wait The stop time t of the bus at the station stop And the travel time t between the boarding station and the final arrival station travel When the individual needs to take the intermediate transfer, R SBR Also includes individual transfer travel time t BS,BS' 。t wait Specifically, the difference between the time when the individual arrives at the station and the time when the vehicle arrives at the current station; t is t stop Specifically, the difference between the time when the vehicle arrives at a certain station and the time when the vehicle leaves the station is referred to; t is t travel Particularly, the total time of the individual in the in-vehicle transit state; when the individual directly reaches the travel terminal, R SBR Expressed as: r SBR ＝-(t wait +t stop +t travel )；
When the individual needs to transfer to the terminal station in the middle, R SBR Expressed as: r is SBR ＝-(t wait +t stop +t travel +t BS,BS' )；
(4)R TD : the get-off site and the travel destination of the individual trip are determined in the action selection, so that the direct return to the destination, namely the travel time t between the individual get-off site and the travel destination TD,D/O Negative number of (d): r TD ＝-t TD,D/O ；
2. Future return F
The future return F is the basis of the calculation of the state transition probability; the direct return value of the individual when the individual selects the action in the dynamic public transportation information environment is unknown, and the direct return value can be accurately obtained only after the individual actually completes the action; therefore, according to the continuity of individual trip habits, the future return can be estimated through the historical experience values; each action selection of the individual in the trip has a corresponding action selection set, each action in the set has a plurality of records in the historical trip, the record values are corresponding direct return values R of the individual at that time after the action is finished, and the future return F is calculated according to the average number of the direct return values R of each action historical record.
(1)F SDT : selecting future return of departure time according to the average value of the difference between the actual arrival time and the expected arrival time of a certain departure time selected from individual historical trips
Determining; f SDT Expressed as:
wherein the content of the first and second substances,
representing the average of the time differences of early arrivals in the historical trip,
representing the average of the time differences delayed in the historical trip.
(2)F SBS ：F SBS Namely the average of the historical travel time of the individual from the travel terminal to the boarding station
Negative number of (d); if the individual is transferred from the getting-off station to another getting-on station, F SBS Average historical travel time for an individual to transfer from a disembarking station to another boarding station
Negative number of (d); f SBS Respectively expressed as:
(3)F SBR : the individual needs to select the bus route in the selected state of the boarding station; considering that an individual is always in a dynamic trip environment, the trip route which can minimize the trip time cost of the individual can be timely adjusted and reselected according to the vehicle arrival condition of a station, so that the individual waits for a bus at the station based on the station waiting time average value of each route in historical trip data
Negative number, average value of stop time of bus at the station
And the average of the travel times between the boarding station and the final arrival station
The sum of the negative numbers of the trip determines the preferred path of the trip; when the individual needs to be transferred midway, the average value of the transfer trip time of the individual needs to be considered
Negative number of (1), then F SBR Can be expressed as:
meanwhile, according to the principle that travel time cost is minimum, the individual needs to consider that the vehicle arriving for the first time in the waiting process is not the pre-boarding route vehicle, and at the moment, the future return F needs to be recalculated SBR '，F SBR ' different in that the individual does not consider the current site
F SBR ' is represented as:
then, after the individual waits for a period of time at the boarding station, when the first arriving bus is the individual pre-boarding route vehicle, the individual gets on the bus and finishes the selection of the bus route action; if the bus is not an individual pre-boarding route vehicle, the individual needs to be according to the current route of the arriving vehicle and the restF of the path SBR ' comparing the state transition probability, and then selecting whether to get on the bus to finish the current action selection.
It should be noted that, because the individual has considered the whole travel process when selecting the bus route, and the transition probability of each action is determined here, the selection of the getting-off station and the arrival at the destination do not need to define the future return to calculate the state transition probability, and the action decision involved in the process from the selection of the bus route to the arrival at the destination can be regarded as a whole consideration.
3. Probability of state transition P
When each action selection is carried out by an individual, the size of the state transition probability needs to be calculated so as to determine the action selection to be finally executed. The calculation of the state transition probability P adopts a logistic regression expression, the Pmax is the optimal selection of the decision of the current action, and the individual gives priority to the action; the respective state transition probabilities P are respectively expressed as:
the travel individuals have 5 states in the multi-mode public transport network, and then, the state transition probability matrix M (5) Is defined as:
finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. The utility model provides a bus individual trip decision-making model which characterized in that includes:
the method comprises the steps of obtaining travel rules of bus individuals based on analysis of travel data of the bus individuals, and inputting the travel rules as initial basic data of a model;
based on the meaning and the definition of function level of the multi-mode bus, the initial basic data are optimized: dividing individual trip chains into multi-mode bus trips and single-mode bus trips according to a transfer time threshold, performing density clustering according to spatial position attributes of individual departure or arrival stations in a road network, establishing corresponding virtual OD pairs, and identifying information of the individual trip chains;
constructing state spaces and action space sets of different trips according to the information of the individual trip chains, and continuously updating the action sets according to the incomplete information environment of the public transport network; and establishing an individual trip decision model comprising departure time selection, getting-on station selection and trip path selection by taking an individual as a unit by combining historical individual trip time cost, trip habit definition and direct return of a calibration state action pair, future return and state transition probability.
2. The transit individual trip decision model of claim 1, characterized in that: the bus individual trip data comprises IC card swiping data, bus-mounted GPS data and bus route stop position information, wherein the bus route stop position information is combined, time is used as a main connecting field and is supplemented with the GPS data of the bus to obtain a vehicle arrival time table, sparse GPS data is supplemented through an interpolation method, the vehicle arrival time is obtained by matching the sparse GPS data with the station longitude and latitude, the IC card swiping time of a passenger is obtained by using a time matching method, and the getting-on stop identification process of the passenger with the IC card is completed.
3. The transit individual trip decision model of claim 2, characterized in that: GPS data interpolation processing: in the selected acquired GPS data, the longitude and latitude of any two adjacent GPS records are respectively (x) b,w,j ,y b,w,j )、(x b,w,j+1 ,y b,w,j+1 ) Recording time is respectively RT b,w,j 、RT b,w,j+1 And the interval time is Λ, inserting Λ data between two GPS records by taking 1s as an interpolation interval, and calculating the longitude and latitude of the inserted λ -th GPS data:
wherein the corresponding recording time is RT b,w,j +λ。
4. The transit individual trip decision model of claim 3, characterized in that: according to the trip chain theory, the trip of the individual bus is divided into a closed bus trip chain and a non-closed bus trip chain, and the getting-off station of the different types of bus trip chains is calculated by combining the identification result of the getting-on station and the characteristics of the bus stations.
5. The bus individual trip decision model of claim 4, wherein: the calculation of the get-off station is as follows:
step 1, according to the getting-on station BS p,k Judging whether the current card swiping record is the last time of the current day, if K is less than K, executing the step 2, otherwise, executing the step 4;
step 2, judging the getting-on station BS of the current riding train number p,k And pick-up station BS of next riding train number p,k+1 Whether the line belongs to the same line, if so, thenTaking the getting-on station of the next ride train AS the getting-off station of the current ride train, namely AS p,k ＝BS p,k Otherwise, executing step 3;
step 3, determining to use BS p,k+1 In a circular area with the circle center and the maximum walking distance D as the radius, the CAS is set from the upstream to the downstream stations where the current riding train passes p,k ＝{cas m M is more than or equal to 0 and is more than or equal to M }; if the set is an empty set, the trip chain belongs to a non-public traffic trip chain; if the number of stations in the set is M =1, the only station of the set is a get-off station AS of the current riding train number p,k (ii) a If M > 1, respectively calculating the most upstream station cas 1 As a starting point, via CAS p,k Any station cas m And with BS p,k+1 Time of trip t as terminal point m Selecting the station cas with the shortest travel time * ＝argmin(t m ) Get-off station AS for current ride number p,k The calculation formula is as follows:
wherein v is w Is the walking speed; d m For station cas m And BS p,k+1 The distance between them; l is m-1,m For station cas m-1 With cas m The length of the bus line between; v. of b The bus operating speed;
step 4, judging the getting-on station BS of the last taking train p,k And pick-up station BS of the first ride p,1 Whether belonging to the same line or not, if so, taking the getting-on station of the first-time train AS the getting-off station of the last-time train, namely AS p,k ＝BS p,1 Otherwise, executing step 5;
step 5, determining the getting-on station BS of the first taking train number p,1 In a circular area with the circle center and the maximum walking distance D as the radius, the last time of taking a bus passes through a station set from upstream to downstream, if the set is an empty set, the passenger trip chain belongs to a non-closed bus trip chain, otherwise, the last time of taking the bus is determined according to the step 3And a get-off station of the secondary taking train.
6. The transit individual trip decision model of claim 4, characterized in that: the individual outgoing behavior also comprises a non-public transport outgoing chain, and the station calculation of getting off is as follows:
step 1, determining boarding station BS of current riding train number p,k Downstream site aggregation DAS p,k ＝{das n 1≤n≤N}；
Step 2, calculating that the passenger is at any station das n Probability of getting off the vehicle PA p,n The calculation formula is as follows:
PA p,n ＝f n,1 ·f n,2 /∑(f η,1 ·f η,2 )；
wherein f n,1 、f n,2 Respectively representing the development intensity around the station and the bus accessibility characteristic parameters; f. of n,1 ＝BF n /∑BF η ，f n,2 ＝R n /∑R η ，BF n For site das n The average passenger loading in a certain period indirectly reflects the development intensity around the site; r n For passing through site das n The number of bus lines.
Step 3, according to the situation that passengers are at any station das n Probability of getting off the vehicle PA p,n Adopting a roulette method to calculate the get-off station AS of the current riding car p,k 。
7. The transit individual trip decision model according to claim 5 or 6, characterized in that: dividing individual trip chains according to a transfer time threshold, specifically:
step 1: judging boarding station BS of two times before and after the passenger according to the identification result of the boarding station p,k And BS p,k+1 Whether the passenger belongs to the same line or not, if the passenger belongs to the same line, the current trip of the passenger belongs to the single-mode bus trip, if the passenger does not belong to the same line, whether the two bus lines belong to the same function level or not is judged, if the passenger belongs to the same function level, the current trip of the passenger still belongs to the single-mode bus trip, and if the passenger does not belong to the same function levelThe energy level belongs to multi-mode bus travel;
step 2: according to the get-off station AS p,k Determining the site AS p,k Corresponding bus operation shift w b Obtaining the arrival schedule of the vehicles of the shift, and comparing AS p,k The station name of the vehicle is matched with the station in the arrival schedule to obtain the AS of the vehicle at the station p,k AT arrival time AT b,w,j (ii) a Taking 0.25 of the current travel time of the bus as the getting-off time delta t of the passenger p Then the time AT of getting-off of the passenger b,w,j,l The calculation formula is as follows:
AT b,w,j,l ＝AT b,w,j +Δt p ；
and step 3: calculating a certain get-off time AT b,w,j,l And the next station BS of getting on the bus p,k+1 Corresponding boarding time TT p,k+1 Time interval mu of p,l ，μ p,l ＝TT p,k+1 -AT b,w,p,l ；
And 4, step 4: when the passenger transfer belongs to the same-station transfer, namely the boarding stations of two times before and after the passenger belong to the same line, determining the maximum waiting time WT of the passenger at the station according to the departure intervals of the lines of different function levels of the multi-mode bus p,k Taking the maximum waiting time as the maximum transfer time threshold eta of the transfer of the same station p,k I.e. eta p,k ＝WT p,k (ii) a When the passenger transfer belongs to the different station transfer, namely the boarding stations of two times before and after the passenger do not belong to the same line, according to the maximum walking distance D acceptable by the passenger and the walking speed v of the passenger w Calculating a maximum transfer time threshold;
and 5: will go out the time interval mu p,l With a maximum transfer time threshold η p,k Make a comparison as if mu p,k ≤η p,k Then the bus individual completes single-time multi-mode or single-mode bus trip on the same day, and mu p,l Namely the transfer time; on the contrary, the bus individual finishes multiple multi-mode or single-mode bus trips on the same day, namely mu p,l Is the active time.
8. The transit individual trip decision model of claim 7, characterized in that: the establishment of the virtual OD pairs specifically comprises the following steps:
step 1, identifying all complete trip chains of the public transportation individuals, numbering the complete trip chains, extracting an initial boarding station of each trip chain, and forming a set D p,BS Each object in the set comprises the number of each trip chain and longitude and latitude coordinates of the corresponding boarding station;
step 2, setting neighborhood parameters epsilon according to the radiation range of different functional level bus stops, and correspondingly adjusting and setting MinPts according to specific research objects; set D of pairs according to neighborhood parameters (ε, minPts) p,BS Performing core object search to obtain a core object set omega p,BS ；
Step 3, from the set omega p,BS Randomly selecting a core object as a seed to execute a cluster generation algorithm, and searching all sites which can be reached by the density of the core object, thereby forming a first cluster
Then will be
The core object contained in (2) is from Ω p,BS Screening, then respectively randomly selecting one seed from the updated set to generate a next cluster, and repeating until the set is empty;
step 4, according to each cluster
Screening out the stop get-off stations of the trip chains corresponding to the numbers by the trip chain numbers corresponding to each object, and correspondingly forming each set
Each object in the set also comprises the number of each trip chain and the longitude and latitude coordinates of the corresponding get-off station;
step 5, according to the neighborhood parameter (epsilon, minPts) pair set
Searching core objects, removing abnormal points and obtaining a core object set
Step 6, calculating a set
And
a center point of (a); the central points of each group of corresponding sets jointly form a plurality of groups of virtual OD point pairs of the passenger p; the calculation formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,
an X coordinate representing a center point of each set;
y coordinates representing the center point of each set; x is the number of b,w,n X coordinates representing the objects contained in each set; y is b,w,n Y coordinates representing the objects contained in each set; n represents an object in each set; n represents all the objects contained in each set.
9. The transit individual trip decision model of claim 1, characterized in that: the state space S:
S＝{Destination/Origin,Departure-time,Boarding-stion,En-route,Alighting-station}；
wherein, destination/Origin is the longitude and latitude coordinates of a pair of virtual OD points in individual history; the department-time is the specific Departure time information selected by the individual; the Boarding-sting is the longitude and latitude of the getting-on station selected by the individual and the collinear route information contained in the station; en-route is information such as vehicles, operation train numbers and the like of a corresponding line of an individual selected station point; the height-station is information such as a route, a vehicle, an operation number, longitude and latitude of a get-off station selected by an individual.
An action space A:
a = { Select-future-time, select-routing-station, select-bus-route, select-addressing-station, to-the-destination } each action space is a data set To be selected, wherein the data set To be selected comprises all departure time information of individual historical trips; the Select-pairing-station to-be-selected set comprises all getting-on site information of individual historical trips; the Select-bus-route to-be-selected set comprises collinear bus route information corresponding to all boarding stations of individual historical trips; the Select-weighting-station to-be-selected set comprises all historical getting-off station information corresponding to individual 'bus route selection'; the To-the-destination candidate set includes longitude and latitude coordinates of all the established virtual OD points.
10. The transit individual trip decision model of claim 9, characterized in that: and determining the state transition probability according to the future return for the action selection of each state obtained by the individual trip decision model, so as to select the action with the maximum state transition probability as the currently selected action and transfer to the next state, wherein when the action is transferred to the next state, the direct return of the action selection becomes known and the historical empirical value set is updated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211424041.4A CN115713206A (en) | 2022-11-14 | 2022-11-14 | Bus individual trip decision model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211424041.4A CN115713206A (en) | 2022-11-14 | 2022-11-14 | Bus individual trip decision model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115713206A true CN115713206A (en) | 2023-02-24 |
Family
ID=85233193
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211424041.4A Pending CN115713206A (en) | 2022-11-14 | 2022-11-14 | Bus individual trip decision model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115713206A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116709215A (en) * | 2023-06-09 | 2023-09-05 | 武汉江汉城市科技发展有限公司 | Method, equipment and storage medium for reminding messages of public transportation |
-
2022
- 2022-11-14 CN CN202211424041.4A patent/CN115713206A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116709215A (en) * | 2023-06-09 | 2023-09-05 | 武汉江汉城市科技发展有限公司 | Method, equipment and storage medium for reminding messages of public transportation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110428117B (en) | Passenger flow accurate induction method and system under urban rail transit multi-scene | |
| Guo et al. | A deep reinforcement learning approach to ride-sharing vehicle dispatching in autonomous mobility-on-demand systems | |
| CN102867408B (en) | Method for selecting bus trip route | |
| CN110390349A (en) | Bus passenger flow volume based on XGBoost model predicts modeling method | |
| CN112700072A (en) | Traffic condition prediction method, electronic device, and storage medium | |
| CN110836675B (en) | Decision tree-based automatic driving search decision method | |
| CN105702035B (en) | A kind of method for complexity of being ridden using history public transport data assessment | |
| CN107194491A (en) | A kind of dynamic dispatching method based on Forecasting of Travel Time between bus passenger flow and station | |
| Xie et al. | A hybrid method combining Markov prediction and fuzzy classification for driving condition recognition | |
| Shafahi et al. | SpeedRoute: Fast, efficient solutions for school bus routing problems | |
| CN111753910A (en) | Method and device for predicting drip order demand based on LSTM | |
| CN107978148A (en) | A kind of traffic status prediction method based on multi-source traffic data dynamic reliability | |
| CN115713206A (en) | Bus individual trip decision model | |
| CN111899511A (en) | Bus arrival time prediction method for AVL data of collinear line | |
| CN116663811A (en) | Scheduling matching method and device for reciprocating dynamic carpooling of inter-city passenger transport | |
| CN113379159B (en) | Taxi driver passenger searching route recommendation method based on gray model and Markov decision process | |
| Chen et al. | Online eco-routing for electric vehicles using combinatorial multi-armed bandit with estimated covariance | |
| CN111723871B (en) | Estimation method for real-time carriage full load rate of bus | |
| Liu et al. | Optimization approach to improve the ridesharing success rate in the bus ridesharing service | |
| CN114066503A (en) | Method for forecasting taxi demand based on virtual service area construction | |
| Hariz et al. | Mobility traffic model based on combination of multiple transportation forms in the smart city | |
| Yu et al. | Optimization of urban bus operation frequency under common route condition with rail transit | |
| CN112949939B (en) | Taxi passenger carrying hotspot prediction method based on random forest model | |
| Li et al. | Route restoration method for sparse taxi GPS trajectory based on Bayesian network | |
| CN111815017B (en) | Electric vehicle charging load prediction method based on travel data |
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 |