CN111985710B - Bus passenger travel station prediction method, storage medium and server - Google Patents
Bus passenger travel station prediction method, storage medium and server Download PDFInfo
- Publication number
- CN111985710B CN111985710B CN202010833582.7A CN202010833582A CN111985710B CN 111985710 B CN111985710 B CN 111985710B CN 202010833582 A CN202010833582 A CN 202010833582A CN 111985710 B CN111985710 B CN 111985710B
- Authority
- CN
- China
- Prior art keywords
- bus
- passenger
- cluster
- station
- boarding
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/01—Satellite radio beacon positioning systems transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/13—Receivers
- G01S19/14—Receivers specially adapted for specific applications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/123—Traffic control systems for road vehicles indicating the position of vehicles, e.g. scheduled vehicles; Managing passenger vehicles circulating according to a fixed timetable, e.g. buses, trains, trams
- G08G1/127—Traffic control systems for road vehicles indicating the position of vehicles, e.g. scheduled vehicles; Managing passenger vehicles circulating according to a fixed timetable, e.g. buses, trains, trams to a central station ; Indicators in a central station
- G08G1/13—Traffic control systems for road vehicles indicating the position of vehicles, e.g. scheduled vehicles; Managing passenger vehicles circulating according to a fixed timetable, e.g. buses, trains, trams to a central station ; Indicators in a central station the indicator being in the form of a map
-
- 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
- H04W4/30—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
-
- 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
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A30/00—Adapting or protecting infrastructure or their operation
- Y02A30/60—Planning or developing urban green infrastructure
Abstract
The invention discloses a bus passenger travel station prediction method, a storage medium and a server, wherein the method comprises the following steps: acquiring GPS track data of buses, IC card swiping record data of passengers and bus stops and line vector data in a certain city area; based on the GPS track data of the bus, the card swiping record data of the passenger bus IC card and the bus stop and line vector data, matching the historical bus card swiping boarding stations of the passenger and constructing a historical boarding station outgoing link of the passenger; spatial clustering is carried out on bus stops, and a passenger history boarding clustering travel link is constructed; predicting bus clusters based on a variable order Markov model and a passenger history bus cluster travel link, and predicting specific bus stops based on a probability model; based on analysis of the mixed travel chains of the buses and subways, bus departure clustering and departure station points are predicted. The method can be used for predicting the bus stops at the bus stops, and the accuracy of predicting the bus stops is obviously improved.
Description
Technical Field
The invention relates to the field of spatial information application, in particular to a bus passenger travel station prediction method, a storage medium and a server.
Background
Along with the popularization of the intelligent public transport system, the intelligent transport IC card is widely applied to a public transport fare payment system, and massive travel track information of users is recorded. The method can be used for accurately predicting the information of the bus stops of urban residents when the buses go out and get off, greatly reducing the financial expenditure of government traffic investigation, maximizing the profit of related enterprises, being beneficial to grasping the change of bus passenger flow, assisting the bus operation decision and optimizing bus routes, and providing a new view angle for researching various fields such as individual/group public traffic behavior characteristics, urban traffic flow estimation, urban space structure and functional adaptability analysis and the like.
Today, a large number of public buses in many mass cities in China adopt a ticket system, which is considered as an 'incomplete transaction mode' in the industry, and is mainly expressed in the following aspects: 1) The passengers only swipe the card when getting on or off the car, and do not need to swipe the card when getting off; 2) The card swiping data only records the information of the card swiping passenger ID, card swiping time, boarding vehicles and the like, and does not record the accurate boarding site information of the passengers; 3) The user does not have to swipe the card when the vehicle stops at a stop, and a large number of station-crossing card-swiping phenomena exist in the peak period. It can be seen that how to accurately infer the historical boarding points of passengers based on historical data and further predict future boarding and disembarking points is a very challenging task.
In recent years, due to the development of information communication technology and location-based services, the research on the space characteristics of human travel behaviors, such as travel distance distribution characteristics, movement track turning radius, access area probability density distribution and the like, is greatly improved based on a large amount of high-precision individual travel track data acquired by media such as intelligent traffic systems, mobile phone communication, social networks and the like. Important progress has also been made in theoretical studies of predictability of human spatial movement patterns, which experiments show that human spatial movement behavior is highly regular and predictable. The track prediction is divided into position prediction under continuous tracks and position prediction under discontinuous tracks, and mainly comprises a set meter model and a personal selection model. Because bus swiping data generally has no real data of the departure station, the estimation of the departure station is also more fuzzy, the real travel link is difficult to construct, and the cost of large-scale demonstration experiments is higher. In general, the academia is still in an exploration stage for predicting the passenger travel station based on the public transportation IC card swiping record data, and the related research is not deep enough.
Accordingly, the prior art is still in need of improvement and development.
Disclosure of Invention
In view of the shortcomings of the prior art, the invention aims to provide a bus passenger travel stop prediction method, a storage medium and a server, and aims to solve the problem that the prior art cannot accurately predict a passenger travel stop based on bus IC card swiping record data.
The technical scheme of the invention is as follows:
a bus passenger travel station prediction method comprises the following steps:
acquiring GPS track data of buses, IC card swiping record data of passengers and bus stops and line vector data in a certain city area;
matching the position of the passenger using the IC card to the nearest bus stop of the bus route according to the GPS track data of the bus, the IC card swiping record data of the passenger and the bus stop and route vector data, acquiring the station point of the passenger on the bus by swiping, and constructing a historical station point out-of-the-way link of the passenger;
spatial clustering is carried out on bus stops, and a passenger historical boarding station outgoing link is converted into a passenger historical boarding clustering outgoing link;
based on a partial matching prediction variable order Markov model, the boarding cluster of the passenger is predicted, and a specific boarding station is deduced through the predicted boarding cluster based on a probability model.
And predicting the get-off clusters and the get-off stops of the bus passengers based on the analysis result of the bus subway mixed travel chain.
With reference to the first aspect, an embodiment of the present invention provides a first possible implementation manner of the first aspect, where the passenger bus IC card swipe record data includes a passenger anonymous ID, a swipe time, a trip mode, a subway boarding and disembarking station, a bus license plate, and a consumption price.
With reference to the first aspect, the embodiment of the present invention provides a second possible implementation manner of the first aspect, wherein the bus GPS track data includes a bus license plate, track data acquisition time, an instantaneous speed, an instantaneous direction angle and an instantaneous position of the bus when the bus is running.
With reference to the first aspect, an embodiment of the present invention provides a third possible implementation manner of the first aspect, where the bus stop and line vector data includes a bus stop number, a bus stop location, a bus stop name, and a bus route to which the stop belongs.
With reference to the first aspect, an embodiment of the present invention provides a fourth possible implementation manner of the first aspect, where according to the GPS track data of the bus, the IC card swipe record data of the passenger bus, and the bus stop and line vector data, matching a position of the passenger using the IC card swipe to a bus stop nearest to the bus line, obtaining a bus stop of the passenger, and constructing a historical bus stop exit link of the passenger, the steps specifically include:
carrying out space-time track restoration on the GPS track data of the buses by adopting a low-frequency floating car map matching algorithm;
based on an average speed interpolation algorithm, matching the instantaneous position of the passenger when swiping the card to a bus track;
based on the nearest neighbor rule, taking the bus stop with the nearest instant card swiping position as the boarding station of the passenger according to the position information of the bus stop;
and connecting the travel stations of the same passenger on a plurality of dates into a station outgoing link on the history of the passenger according to the time sequence.
With reference to the first aspect, an embodiment of the present invention provides a fifth possible implementation manner of the first aspect, where, according to the bus stop and line vector data, converting a historical passenger boarding uplink to a historical passenger boarding cluster travel link, the steps specifically include:
based on density connectivity among bus stops, adopting a density clustering algorithm to distribute bus stops meeting density threshold conditions around a core point (current stop) into the same cluster;
traversing all bus stops, and judging whether the same-name stops opposite to the current stop and the closest stops to the current stop are in the same cluster with the current stop;
if the same-name site and the site closest to the current site are not in the same cluster with the current site, merging the cluster where the same-name site is located and the cluster where the site closest to the current site is located into the cluster of the current site, and further obtaining the historical boarding cluster travel link of the passenger.
With reference to the first aspect, an embodiment of the present invention provides a sixth possible implementation manner of the first aspect, wherein, according to the variable-order markov model based on the partial matching prediction algorithm and the historical boarding cluster travel link of the passenger, the boarding cluster of the passenger is predicted, and based on the probability model, a specific boarding point of the passenger is predicted, and the steps include:
taking a card swiping track formed by bus card swiping records of a single passenger in one month as a text string sequence, wherein each bus card swiping cluster is taken as a character, so that a bus history card swiping cluster travel link sequence of each passenger is formed;
using the formulaProcessing the condition that a certain cluster does not appear after a certain cluster link string in the training set, wherein s represents a cluster link string with the length of k, c represents any cluster in the cluster set, Σs represents all clusters in the training set which appear after the cluster link string s, and s' represents the longest suffix of the cluster link string s;
defining the algorithm based on the partial matching prediction as:
for any one of the cluster link strings s and the cluster c, the number of times the cluster link string sc appears in the training set is represented by N (sc), Σs represents all clusters in the training set that appear after the cluster link string s, and the following formula is given:and->
Since a cluster center typically contains multiple sites, the cluster in which the sites are located can characterize the predictive upper limit of the upper stops. For the prediction of boarding stops, the probabilistic model will select the historical boarding stops with the highest probability of passengers in the cluster center as the predicted boarding stops.
Based on the partial matching prediction algorithm, the historical bus boarding cluster travel link sequence of each passenger is trained to predict, and then the bus boarding station point travel link sequence of the passenger is predicted according to the probability model, so that the bus boarding cluster and the bus boarding station point prediction of each individual are realized.
With reference to the first aspect, an embodiment of the present invention provides a seventh possible implementation manner of the first aspect, where, according to an analysis result of the bus subway mixed travel chain, a get-off cluster and a get-off station point of a passenger are predicted, and the steps include:
constructing a bus subway mixed travel chain, carrying out blind measurement on the model by utilizing subway departure records with definite departure stations, counting the repetition degree of next subway departure clusters and current subway departure clusters, and deducing the descending amplitude of the prediction accuracy of the subway departure clusters relative to the prediction of the subway departure clusters;
and further deducing the prediction accuracy of the bus departure cluster/station based on the prediction accuracy decline amplitude of the subway departure cluster and the prediction result of the bus departure cluster/station.
In a second aspect, an embodiment of the present invention further provides a computer readable storage medium, where one or more programs are stored, where the one or more programs are executed by one or more processors, so as to implement the steps of the above-mentioned method for predicting a bus passenger trip station.
In a third aspect, an embodiment of the present invention further provides an application server, where the application server includes at least one processor, a display screen, a memory, a communication interface and a bus, where the processor, the display screen, the memory and the communication interface complete communication between each other through the bus, and the processor invokes a logic instruction in the memory to execute a step of the bus passenger trip station prediction method according to the above claim.
The embodiment of the invention has the following beneficial effects:
the embodiment of the invention matches the bus swiping and getting-on station of the passenger based on the bus GPS track data, the passenger bus IC card swiping record data and the bus station and line vector data. And carrying out spatial clustering on bus stops, constructing a passenger history bus trip link, predicting the boarding clusters based on a variable-order Markov model of a partial matching prediction algorithm, and predicting specific boarding stops based on a probability model. Based on the analysis result of the bus and subway mixed travel chain, bus departure clustering and departure station points are predicted, and verification analysis is carried out according to a specific embodiment of the invention. The analysis result shows that the model has good prediction performance of the station points on and off the public traffic, the prediction accuracy of the model is gradually improved along with the increase of the training data, and the prediction accuracy of the station points on/off the traffic clustering center based on the daily data of the month is obviously improved.
Drawings
Fig. 1 is a flowchart of a preferred embodiment of a method for predicting a bus passenger trip stop according to an embodiment of the present invention.
Fig. 2 is a flowchart showing the step S20 of fig. 1 according to the present invention.
Fig. 3 is a flowchart illustrating the step S30 of fig. 1 according to the present invention.
Fig. 4 is a block diagram of a preferred embodiment of an application server according to an embodiment of the present invention.
Detailed Description
The invention provides a bus passenger travel station prediction method, a storage medium and a server, and the technical scheme of the invention is further described below with reference to the accompanying drawings in order to make the purposes, the technical schemes and the effects of the embodiment of the invention clearer and more definite. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, fig. 1 is a flowchart of a preferred embodiment of a bus passenger trip station prediction method provided by the present invention, where, as shown in the figure, the method includes the steps of:
step S10, acquiring GPS track data of buses in a certain city area, card swiping record data of passenger bus IC cards, bus stops and line vector data;
step S20, matching the position of the passenger using the IC card to the nearest bus stop of the bus route according to the GPS track data of the bus, the IC card swiping record data of the passenger and the bus stop and route vector data, acquiring the station point of the passenger on the bus, and constructing a historical station point out-of-service link of the passenger;
step S30, spatial clustering is carried out on bus stops, and a passenger historical boarding station outgoing link is converted into a passenger historical boarding clustering outgoing link;
step S40, predicting boarding clusters of passengers based on a partial matching prediction variable order Markov model and a passenger history boarding cluster travel link, and predicting specific boarding stations of the passengers based on a probability model;
and S50, predicting the get-off clusters and the get-off stops of passengers according to the analysis result of the bus subway mixed travel chain.
As one specific embodiment, the time coverage is based on data within one month, and the bus GPS track data comprises information such as bus license plates, data acquisition time, instantaneous speed, instantaneous direction angle, instantaneous position and the like when the bus runs; bus track sampling frequency is unequal: in some embodiments, a portion of the bus is sampled at a low frequency time with a sampling period of about 60 seconds.
In the embodiment of the invention, the passenger public transportation IC card swiping record data comprises passenger public transportation swiping consumption records (including buses and subways) and comprises information such as card swiping passenger anonymous ID, card swiping time, travel mode, subway boarding and disembarking stations or bus license plates, consumption prices and the like.
The bus stop and line vector data, wherein the bus stop data information comprises a stop number, a stop position, a stop name and a route to which the stop belongs; the bus route vector data contains all shape point information of the route. In some embodiments, the bus stop and route vector data further includes subway data, the subway data including subway route and subway stop information.
In some embodiments, the longitude and latitude data may adopt different coordinate systems, and the longitude and latitude data in the different coordinate systems (such as a GCJ-02 longitude and latitude coordinate system and the like) may be statistically converted into the WGS84 longitude and latitude coordinate system through an open source code.
Aiming at the characteristics of the passenger bus IC card swiping record data and the passenger travel characteristics, the embodiment of the invention makes the following limitation on the prediction of the passenger travel station based on incomplete IC card transaction records: public transportation route selection and travel station selection of passengers have high repeatability and regularity in a relatively short time domain; the station selected by the same passenger at the last time is usually the same-name station of the last getting-off station or other stations in the relative adjacent space domain, and the problem of predicting the getting-off station of the passenger is converted into the problem of predicting the getting-on station.
The passenger bus IC card swiping record is a low-frequency track sample, and the greatest problem caused by sample sparseness is that the position transfer rule used for training and learning cannot truly describe the movement behavior of the passenger. The phenomenon of 'getting on the bus before swiping cards' widely exists, particularly the position of the bus at the moment of swiping cards of a large number of passengers at the peak of business hours can not coincide with the boarding station, and the boarding and alighting stations of the passengers can be selected in various ways, so that the probability of transferring between the frequent boarding and alighting OD (Origin Destination trip traffic) is further diluted. The embodiment of the invention fully considers the conditions and designs a clustering center-site two-stage prediction model. Firstly, spatial clustering and expansion are carried out on stations (including subway stations), and stations with the same name and close distance are combined into a cluster to the greatest extent; converting a historical train station departure chain into a historical train station departure cluster travel chain, training the cluster travel chain by using a variable order Markov model, predicting train departure clusters of passengers, and deducing the train departure clusters to specific train station according to a probability model; based on analysis results of the bus subway mixed travel chain, the getting-off clusters and the getting-off station points of passengers are predicted.
In an embodiment of the present invention, as shown in fig. 2, the step S20 specifically further includes:
s21, acquiring GPS track data of buses in a certain city area, card swiping record data of passenger bus IC cards, bus stops and line vector data;
s22, carrying out space-time track restoration on the GPS track data of the bus by adopting a low-frequency floating car map matching algorithm;
step S23, based on an average speed interpolation algorithm, matching the instantaneous position of the passenger when swiping a card to a bus track;
and step S24, based on the nearest neighbor rule, taking the bus stop with the nearest instant card swiping position as the boarding station of the passenger according to the position information of the bus stop.
Specifically, one premise of constructing a bus travel link of a passenger is to match the position of the passenger at the time of swiping a card to a bus stop according to the running track and the stop position information of the bus. The premise of realizing the work is to match the map of the card swiping position of the passenger into the track of the bus. Because the GPS original track point of the bus is a low-frequency sampling point, the embodiment of the invention refers to a low-frequency floating vehicle map matching method MDP-MM proposed by Chen and the like, and space-time track reduction is carried out on the track of the bus. Based on an average speed interpolation algorithm, the instantaneous position of the passenger card swiping is matched to the bus track. For the matching of bus stops, the simplest mode is to adopt a nearest matching mode, namely, according to the position information of the bus stops, the bus stop with the nearest instantaneous card swiping position is used as the passenger stop.
The bus position at the moment of swiping the card of a large number of users is inconsistent with the station position of the passenger at the moment of swiping the card of the bus at first, so that the bus is particularly obvious in rush hours and even the phenomenon of swiping the card at multiple stations during going-off is caused, and the result of the mode of nearest neighbor matching has larger error. Aiming at the distance distribution characteristic of the passenger 'off-station card swiping', the 'maximum probability matching' mode is adopted to carry out secondary correction on the directly matched bus stops. And selecting the passengers with travel records in 7:00-9:00 hours in the morning more than half of the working days in one month as study objects, and counting the frequency of the passengers most closely matched with each boarding station. Assuming that the highest-frequency station is the passenger normal boarding station, calculating the route distance from the GPS position of each passenger to the respective station at the time of card swiping in the time period, and counting the probability distribution and the cumulative probability distribution of the distance of the passenger swiping out from the station. The improved 'maximum probability matching' algorithm for each passenger builds a boarding station candidate set, wherein the candidate set comprises 5 stations, namely, the station with the nearest card swiping position, the next station with the nearest card swiping position, the first 3 stations with the nearest card swiping position, and the station with the highest frequency in the candidate set is used as the boarding station for correcting the passenger.
In the embodiment of the present invention, as shown in fig. 3, the step S30 specifically further includes:
step S31, based on density connectivity among bus stops, adopting a density clustering algorithm to distribute bus stops meeting density threshold conditions around a core point (current stop) into the same cluster;
step S32, traversing all bus stops, and judging whether the same-name stops opposite to the current stop and the closest stops to the current stop are in the same cluster with the current stop;
and step S33, if the same-name site and the site closest to the current site are not in the same cluster with the current site, merging the cluster where the same-name site is located and the cluster where the site closest to the current site is located into the cluster of the current site, and further obtaining the historical boarding cluster travel link of the passenger.
Specifically, the public transport travel link constructed by the embodiment of the invention is a mixed travel link of buses and subways. In the public transportation travel process, besides subway card swiping records, the truth value of a passenger getting-off station is difficult to obtain on a large scale. Therefore, the travel link of "get-on station-get-off station- …" can only be degraded into the travel link of "get-on station- …". Assuming that the spatial domain where the passenger gets off the bus stop at the previous moment is the spatial domain where the next get-on bus stop is located, the embodiment of the invention approximately converts the problem of predicting the passenger gets off the bus stop into the problem of predicting the next get-on bus stop. In order to evaluate the accuracy of the prediction algorithm in the embodiment of the invention, a dark down link is maintained for subway records with clear down stations while an up station point link is constructed.
After a passenger gets off at station a, if the travel link is not broken, there are two situations in which the passenger gets off the station: 1) Selecting a site A with the same name or a site B with a relatively close distance from the site A to return; 2) And continuing riding along the original direction, and selecting the current station A or the station C which is closer to the current station A for transfer. Because of sparsity of card swiping records and diversity of site selection, in order to increase stability of passenger site transfer probability, the embodiment of the invention needs to spatially cluster the sites, and all the sites in adjacent spatial domains are clustered into a meaningful position. The principle of spatial clustering is to ensure that uplink and downlink sites with the same name on the same line and sites very close to each other in space (such as a plurality of sites at an intersection) of different lines are merged together, and the adjacent sites in each direction of the same line are effectively segmented into different clusters to the greatest extent, so that the smaller the cluster radius is, the better the cluster radius is. The embodiment of the invention introduces a DBSCAN density clustering algorithm, clusters according to the density connectivity relation among the sites, and distributes the sites meeting the density threshold condition around the core points into the same cluster. After the clustering is completed, it cannot be guaranteed that all sites a with the same name have the same cluster with the closest site B from the site a, and a cluster expansion check needs to be performed again. This expansion traverses all sites, ensuring that a and B are already in the same cluster as a, and if not, expanding and merging the clusters in which a and B are located with the clusters of a. According to the embodiment of the invention, first, travel stations of the same passenger on a plurality of dates are connected into a station outgoing link [ S1-S2, … -Sn ] on the history of the passenger according to the time sequence; through spatial clustering and expansion operation, the boarding station point outgoing links [ S1-S2, … -Sn ] of passengers are converted into boarding cluster outgoing links [ C1-C2, … -Cm ].
To verify the previous assumptions, the present invention introduces Fano inequality in the theory of information to theoretically calculate the upper limit of the passenger travel link prediction probability. For a given bus swiping data set, if the value of the upper limit max of the prediction accuracy of the departure station point is to be estimated, the lower limit of the prediction error calculated by using the Fano inequality, namely (1-max) can be equivalently converted. The idea of Fano inequality is to relate the error probability of the speculative random variable X to its conditional entropy H (x|y) assuming that the random variable Y is known, the value of the random variable X to which Y is related is speculative. Assuming that the estimated value X ' of X is calculated by Y, the value space of X is χ, and a lower bound is made on the probability of X ' +.X, where X-Y-X ' constitutes a Markov chain. The error probability can be defined as
Pe＝P{X′≠X} (3.1)
According to the Fano inequality, there is an inequality
H(Pe)+Pe·log|χ|≥H(X|Y) (3.2)
Because the value of Pe only has two cases, H (Pe) is a binary entropy, the maximum value of H (Pe) is 1, and therefore the formula (3.2) can be weakened to be
1+Pe·log|χ|≥H(X|Y) (3.3)
I.e.
This is the prediction error probability, which characterizes an unequal lower probability bound, i.e. the lower bound of prediction errors, from which the upper bound of prediction correctness can be derived.
From max=1-Pe (3.5)
Has the following components
Equation 3.6, the correct upper bound for prediction, whichThe method is used for measuring the performance of a prediction model and judging whether the model has reached higher prediction accuracy. According to the embodiment of the invention, passenger intelligent IC card swiping records of 1 month working days are selected, and shannon entropy distribution and highest predictability pi max distribution of travel link nodes after group clustering are respectively calculated. The results show that the peak of entropy is located near 0.65 and pi max is generally higher than 0.65. The lower entropy shows that the passenger's travel station selection has very low uncertainty (2 0.65 Approximately 1.57), while a high n max demonstrates at the population level that the selection of passenger boarding and disembarking stops can be predicted, as well as that the passenger routing and stopping selections are highly repeatable and regular in a relatively short horizon.
In the embodiment of the present invention, the step S40 specifically further includes:
step S41, predicting boarding clusters of passengers by using a partial matching prediction variable order Markov model and a passenger history boarding cluster link;
step S42, predicting the specific boarding station of the passenger based on the predicted boarding cluster of the passenger by using the probability model.
Specifically, the markov model can effectively solve the problem of position prediction, and the traditional N-order markov model and the hidden markov model predict the next position of the passenger through the state transition matrix when calculating the transition probability. The limitation of the method is that the space complexity is too high, and the variable-order Markov model can well solve the problem of low space utilization rate. The prediction performance of the partial match predicted (Prediction by Partial Match, PPM) variable order Markov model is superior to other models by comparison of the effects of the various variable order Markov model algorithms on predicting discrete sequences under a common dataset.
The PPM algorithm utilizes a dictionary tree mechanism to address the problem of low space utilization of the traditional Markov model. Constructing a dictionary tree of a passenger travel clustering link requires specifying a model order D, and the model uses an escape mechanism to treat the zero frequency problem existing in the traditional conditional probability model. For the location prediction problem, the escape mechanism is used for processing the situation that a certain cluster does not appear after a certain cluster link in the training set. For each clustered link string s of length k (k.ltoreq.D), the escape mechanism assigns a conditional probability P (escape|s) and a remaining probability 1-P (escape|s) to clusters in the training set that do not occur and that occur after clustered link string s, respectively. The escape mechanism can be expressed by the following formula:
for an empty cluster link string ε, there is P (c|ε) =1/|Σ|, where Σ represents all clusters that have appeared in the training set and c represents any one of the clusters in the cluster set. Σs represents all clusters in the training set that occur after the cluster link string s, and s' represents the longest suffix of the cluster link string s.
For any one of the cluster link strings s and the cluster c, the number of times that the cluster link string sc appears in the training set is represented by N (sc), Σs represents all clusters in the training set that appear after the cluster link string s, i.e., Σs= { c: N (sc) >0}, the following formula is given:
in the training stage, the PPM algorithm constructs a dictionary tree from the training set to realize prediction. Each node in the tree represents a cluster and a counter, and if the specified order is D, the maximum depth of the tree is d+1. The algorithm firstly constructs a cluster link string epsilon with a root node representing the null, then incrementally analyzes the training set sequence, analyzes one cluster c and a sub-link string with the length D corresponding to the cluster each time, and forms a path in the dictionary tree. After the first D clusters of the training set sequence are parsed, the length of each newly constructed path is d+1, and the value of the counter along this path is incremented by 1. Thus, the value of the counter for each node indicates the number of times the cluster c for that node occurs after the cluster link string sequence s represented by the path, i.e., N (sc). After constructing the dictionary tree, the conditional probability P (c|s) can be calculated through the dictionary tree, wherein c is any one cluster in the cluster set, and s is a cluster link string with the length less than or equal to D. At the time of calculation, the dictionary tree is traversed from the root node, and the traversing rule is whether the longest suffix s' of s is matched. Then s' c represents a complete path from the root node to the leaf and the conditional probability is calculated using equations (3.7-3.9).
When the next boarding station of the passengers is predicted, for each passenger, a historical boarding station outgoing link can be formed by the bus card swiping record in one month, and the boarding station outgoing link is converted into a historical boarding cluster outgoing link after clustering. According to the embodiment of the invention, the PPM model is trained by utilizing the historical bus-on-bus clustering travel link sequence of each passenger, the probability of the passenger in each possible space domain in the future is calculated according to the current position of the passenger by combining the formulas, and then the space domain with the highest probability is used as a prediction result, so that the bus-on-bus clustering prediction of the passenger is realized.
Because the route selection and the station selection of the passengers have high repeatability and regularity in a relatively short time domain, the embodiment of the invention uses a traditional probability model in the second stage of the clustering center-station model, and further predicts the specific boarding stations of the passengers by combining the passenger boarding clusters predicted by the partial matching prediction model.
In the embodiment of the present invention, the step S50 specifically further includes:
step S51, constructing a mixed travel chain of the subway of the bus, carrying out blind measurement on the model by utilizing a subway getting-off record with a definite getting-off station, counting the repetition degree of a next subway getting-on cluster and a current subway getting-off cluster, and deducing the descending amplitude of the prediction accuracy of the subway getting-off cluster relative to the prediction of the subway getting-on cluster;
step S52, further deducing the prediction accuracy of the bus departure clusters/stations based on the prediction accuracy reduction amplitude of the subway departure clusters and the prediction results of the bus departure clusters/stations.
Specifically, the embodiment of the invention provides a setting that a station selected by the same passenger at the last time is usually the same name station of the last get-off station or other stations in a relative adjacent space domain, and in order to verify the setting, the embodiment of the invention performs blind test on a model by utilizing a subway get-off station hidden chain with a clear get-off station based on a built subway mixed travel chain of a bus. The result shows that if the last subway getting-on station point cluster is used as the current subway getting-off station point cluster, the repeatability of the current subway getting-off cluster and the actual subway getting-off cluster is up to 86.91 percent, that is, if the last subway getting-on station point cluster is directly used as the current getting-off station point cluster, the accuracy of the prediction method of the current getting-off cluster/station can be considered to be reduced by about 13.09 percent compared with the prediction accuracy of the actual getting-off cluster/station, namely the previous setting is verified. Based on the prediction result of the bus-in clusters/stations and the downlink amplitude, the embodiment of the invention can further infer the prediction accuracy of the bus-out clusters/stations.
The embodiment of the invention solves the prediction problem of the resident bus travel boarding station, and converts the prediction problem of the resident bus travel alighting station into the boarding station prediction problem for analysis. Firstly, the embodiment of the invention needs to cluster and expand the sites in a certain space domain. Secondly, for the problem of predicting bus stops, the location prediction technique requires a large amount of data to train the model. Empirically, the more data is trained, the more the model is likely to reflect the travel rules of the passengers. According to the embodiment of the invention, the get-on clustering travel links are constructed according to the two modes of 'nearest neighbor matching' and 'maximum probability matching', and the get-on clustering prediction accuracy based on PPM with different orders when sample data are increased is calculated through the two modes respectively. Finally, since a cluster center typically contains multiple sites, predictions of the cluster in which the sites are located can characterize the upper prediction limit of the upper stops. For the prediction of the station, the embodiment of the invention can select the station with the highest historical boarding probability of passengers in the clustering center as the boarding station point.
For the problem of predicting bus stops, the embodiment of the invention deduces the prediction result of the bus stops based on the prediction result of the bus stops. In some embodiments, when predicting the get-off station of the current passenger, the prediction method of the current get-off station is slightly different from the prediction method of the next get-on station, and the space domain with the largest probability of occurrence of the passenger in the history link in the current passenger travel route in the cluster C where the next get-on station is used as the get-off station (if the cluster C does not include any station of the current route, the station with the largest probability of occurrence of the space domain except the current get-on station is used as the get-off station).
Based on the above bus passenger trip station prediction method, the embodiment of the present invention further provides a computer readable storage medium, where one or more programs are stored, and the one or more programs may be executed by one or more processors to implement the steps in the method in any of the above embodiments.
Based on the above bus passenger trip station prediction method, the embodiment of the present invention further provides an application server, as shown in fig. 4, where the application server includes at least one processor 20 (processor); a display screen 21; and a memory 22 (memory), which may also include a communication interface (Communications Interface) 23 and a bus 24. Wherein the processor 20, the display 21, the memory 22 and the communication interface 23 may communicate with each other via a bus 24. The display screen 21 is configured to display a user guidance interface preset in the initial setting mode. The communication interface 23 may transmit information. The processor 20 may invoke logic instructions in the memory 22 to perform the methods of the embodiments described above.
Further, the logic instructions in the memory 22 may be implemented in the form of software functional units and stored in a computer readable storage medium when sold or used as a stand alone product.
The memory 22 is provided as a computer readable storage medium and may be configured to store a software program, a computer executable program, and program instructions or modules corresponding to the methods in the embodiments of the present invention. The processor 30 performs the functional applications and data processing, i.e. implements the methods of the embodiments described above, by running software programs, instructions or modules stored in the memory 22.
The memory 22 may include a storage program area that may store an operating system, at least one application program required for functions, and a storage data area; the storage data area may store data created according to the use of the terminal device, etc. In addition, the memory 22 may include high-speed random access memory, and may also include nonvolatile memory. The storage medium may be any of various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or may be a transitory storage medium.
In addition, the specific processes that the storage medium and the plurality of instruction processors in the mobile terminal load and execute are described in detail in the above method, and are not stated here.
In summary, the invention designs a two-stage passenger travel station prediction model of a clustering center-station based on bus GPS track data, passenger bus IC card swiping record data and bus station and line vector data, matches historical upper station points of passengers, constructs historical upper station point outgoing links, spatially clusters bus stations and constructs historical upper car clustering travel links, predicts bus upper station clusters and upper station points based on a variable order Markov model and a probability model of a partial matching prediction algorithm, predicts bus lower station clusters and lower station points based on analysis results of a bus subway mixed travel chain, and performs evidence analysis according to a specific embodiment of the invention. The analysis result shows that the model has good recognition rate of the station points on the bus and the station points off the bus, the prediction accuracy of the model is gradually improved along with the increase of the training data, and the prediction accuracy of the station points on the bus/station clustering center/station is obviously improved based on the data of the working days of the month.
It is to be understood that the invention is not limited in its application to the examples described above, but is capable of modification and variation in light of the above teachings by those skilled in the art, and that all such modifications and variations are intended to be included within the scope of the appended claims.
Claims (8)
1. A bus passenger trip station prediction method, the method comprising:
acquiring GPS track data of buses, IC card swiping record data of passengers and bus stops and line vector data in a certain city area;
matching the card swiping position of the passenger to the nearest bus stop of a bus route according to the GPS track data of the bus, the card swiping record data of the IC card of the passenger and the bus stop and route vector data, acquiring the card swiping stop point of the passenger, and constructing a historical bus stop point outgoing link of the passenger;
spatial clustering is carried out on bus stops, and a passenger historical boarding station outgoing link is converted into a passenger historical boarding clustering outgoing link;
predicting the boarding clusters of the passengers based on the historical boarding cluster travel links of the passengers and a variable-order Markov model of a partial matching prediction algorithm, and deducing the predicted boarding clusters to specific boarding sites based on a probability model;
based on analysis results of a bus subway mixed travel chain, predicting a bus passenger getting-off cluster and a bus passenger getting-off station point;
the method comprises the steps of matching the position of the passenger IC card swiping to the nearest bus stop of a bus route according to the GPS track data of the bus, the passenger IC card swiping record data and the bus stop and route vector data, obtaining the bus stop of the passenger, and constructing a historical bus stop outgoing link of the passenger, wherein the method comprises the following steps:
carrying out space-time track restoration on the GPS track data of the buses by adopting a low-frequency floating car map matching algorithm;
based on an average speed interpolation algorithm, matching the instantaneous position of the passenger when swiping the card to a bus track;
based on the nearest neighbor rule, taking the bus stop with the nearest instant card swiping position as the boarding station of the passenger according to the position information of the bus stop;
connecting travel stations of the same passenger on a plurality of dates into a historical travel link of the passenger at the station point according to the time sequence;
predicting a get-off cluster and a get-off station point of passengers according to the analysis result of the bus subway mixed travel chain, wherein the method comprises the following steps:
constructing a bus subway mixed travel chain, carrying out blind measurement on the model by utilizing subway departure records with definite departure stations, counting the repetition degree of next subway departure clusters and current subway departure clusters, and deducing the descending amplitude of the prediction accuracy of the subway departure clusters relative to the prediction of the subway departure clusters;
and further deducing the prediction accuracy of the bus departure cluster/station based on the prediction accuracy decline amplitude of the subway departure cluster and the prediction result of the bus departure cluster/station.
2. The bus passenger trip stop prediction method according to claim 1, wherein the passenger bus IC card swiping record data comprises a passenger anonymous ID, a swiping time, a trip mode, a subway boarding and disembarking stop, a bus license plate and a consumption price.
3. The bus passenger trip stop prediction method according to claim 1, wherein the bus GPS track data comprises bus license plates, track data acquisition time, instantaneous speed, instantaneous direction angle and instantaneous position of a bus in operation.
4. The bus passenger trip stop prediction method according to claim 1, wherein the bus stop and route vector data includes, but is not limited to, bus stop numbers, bus stop positions, bus stop names, and bus routes to which the stops belong.
5. The bus passenger trip station prediction method according to claim 1, wherein the step of converting the passenger historically standing point outgoing link into the passenger historically boarding cluster outgoing link according to the bus station and line vector data comprises the steps of:
based on density connectivity among bus stops, adopting a density clustering algorithm to distribute bus stops meeting density threshold conditions around the current stops into the same cluster;
traversing all bus stops, and judging whether the same-name stops opposite to the current stop and the closest stops to the current stop are in the same cluster with the current stop or not: if the same-name site and the site closest to the current site are not in the same cluster with the current site, merging the cluster where the same-name site is located and the cluster where the site closest to the current site is located into the cluster of the current site, and further obtaining the historical boarding cluster travel link of the passenger.
6. The bus passenger travel station prediction method according to claim 1, wherein the step of predicting the passenger boarding cluster based on the passenger history boarding cluster travel link and a variable-order markov model of a partial matching prediction algorithm, and predicting the passenger specific boarding station based on the probability model further comprises:
taking a card swiping track formed by bus card swiping records of single passengers in one month as a text string sequence to form a bus history boarding cluster trip link sequence of each passenger;
based on the partial matching prediction algorithm, the bus history boarding cluster travel link sequence of each passenger is trained to predict, and then the boarding station point uplink sequence of the passenger is predicted according to the probability model, so that boarding cluster and boarding station point prediction of each individual are realized.
7. A computer readable storage medium storing one or more programs for execution by one or more processors to implement the steps of the bus passenger trip station prediction method of any one of claims 1-6.
8. An application server comprising at least one processor, a display, a memory and a communication interface and a bus, wherein the processor, the display, the memory and the communication interface are in communication with each other via the bus, and the processor invokes logic instructions in the memory to perform the steps of the bus passenger trip station prediction method of any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010833582.7A CN111985710B (en) | 2020-08-18 | 2020-08-18 | Bus passenger travel station prediction method, storage medium and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010833582.7A CN111985710B (en) | 2020-08-18 | 2020-08-18 | Bus passenger travel station prediction method, storage medium and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111985710A CN111985710A (en) | 2020-11-24 |
| CN111985710B true CN111985710B (en) | 2023-08-25 |
Family
ID=73434037
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010833582.7A Active CN111985710B (en) | 2020-08-18 | 2020-08-18 | Bus passenger travel station prediction method, storage medium and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111985710B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112562378B (en) * | 2020-12-01 | 2023-04-18 | 平安科技（深圳）有限公司 | Bus scheduling method and device, computer equipment and medium |
| CN112733891B (en) * | 2020-12-29 | 2023-08-01 | 华侨大学 | Method for identifying bus IC card passengers to get off station points during travel chain breakage |
| CN112765284A (en) * | 2021-01-21 | 2021-05-07 | 广州羊城通有限公司 | Method and device for determining relevant location of user |
| CN113010507A (en) * | 2021-03-23 | 2021-06-22 | 浩鲸云计算科技股份有限公司 | Public transport OD (origin-destination) calculation method and system based on time calibration and genetic algorithm |
| CN113658433B (en) * | 2021-08-18 | 2022-08-30 | 苏州工业园区测绘地理信息有限公司 | Method for extracting passenger flow characteristics based on bus card swiping and code scanning data |
| CN113821738A (en) * | 2021-11-22 | 2021-12-21 | 北京交研智慧科技有限公司 | Travel chain acquisition method and device, electronic equipment and storage medium |
| CN114363842B (en) * | 2021-12-31 | 2023-04-28 | 中山大学 | Bus passenger departure station prediction method and device based on mobile phone signaling data |
| CN114331234B (en) * | 2022-03-16 | 2022-07-12 | 北京交通大学 | Rail transit passenger flow prediction method and system based on passenger travel information |
| CN114743401B (en) * | 2022-03-23 | 2023-05-09 | 华录智达科技股份有限公司 | Data visualization bus dispatching management platform based on bus digital transformation |
| CN114882693B (en) * | 2022-03-23 | 2022-11-18 | 昆明理工大学 | Bus passenger getting-off station prediction method based on card swiping data deep mining |
| CN114580787B (en) * | 2022-04-02 | 2023-07-14 | 佳都科技集团股份有限公司 | Passenger flow prediction method and device based on bus subway connection station |
| CN115081762B (en) * | 2022-08-24 | 2022-11-01 | 北京交通大学 | Passenger integrated intelligent travel method and system based on urban rail transit |
| CN116187585B (en) * | 2023-04-19 | 2023-07-14 | 杭州数知梦科技有限公司 | Method, device and application for predicting BRT bus route of passenger |
| CN116778739A (en) * | 2023-06-20 | 2023-09-19 | 深圳市中车智联科技有限公司 | Public transportation scheduling method and system based on demand response |
| CN117035232A (en) * | 2023-08-08 | 2023-11-10 | 广东家祥智能科技有限公司 | Smart city data processing method and system based on big data platform |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007145625A1 (en) * | 2006-06-14 | 2007-12-21 | Lucent Technologies, Inc. | Dynamic route prediction based on travel patterns of mobile units |
| JP2008084256A (en) * | 2006-09-29 | 2008-04-10 | Japan Tobacco Inc | Information providing system |
| CN103149576A (en) * | 2013-01-29 | 2013-06-12 | 武汉大学 | Map matching method of floating car data |
| CN105229931A (en) * | 2014-02-27 | 2016-01-06 | 华为技术有限公司 | Line synchronization method in OSD system, system and vector quantization controlled entity |
| CN105825297A (en) * | 2016-03-11 | 2016-08-03 | 山东大学 | Markov-model-based position prediction method |
| CN106570182A (en) * | 2016-11-12 | 2017-04-19 | 中兴软创科技股份有限公司 | Getting-off station recognition method and system for bus |
| CN110264710A (en) * | 2019-05-21 | 2019-09-20 | 天津大学 | It is swiped the card the bus passenger flow estimating method with public transport GPS data based on IC card |
| CN110276977A (en) * | 2019-07-29 | 2019-09-24 | 广东工业大学 | A kind of bus station matching process, device, equipment and readable storage medium storing program for executing |
| CN110390349A (en) * | 2019-06-20 | 2019-10-29 | 浙江大学 | Bus passenger flow volume based on XGBoost model predicts modeling method |
| CN111275241A (en) * | 2020-01-08 | 2020-06-12 | 天津大学 | Bus passenger getting-off station inference method based on machine learning decision tree |
-
2020
- 2020-08-18 CN CN202010833582.7A patent/CN111985710B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007145625A1 (en) * | 2006-06-14 | 2007-12-21 | Lucent Technologies, Inc. | Dynamic route prediction based on travel patterns of mobile units |
| JP2008084256A (en) * | 2006-09-29 | 2008-04-10 | Japan Tobacco Inc | Information providing system |
| CN103149576A (en) * | 2013-01-29 | 2013-06-12 | 武汉大学 | Map matching method of floating car data |
| CN105229931A (en) * | 2014-02-27 | 2016-01-06 | 华为技术有限公司 | Line synchronization method in OSD system, system and vector quantization controlled entity |
| CN105825297A (en) * | 2016-03-11 | 2016-08-03 | 山东大学 | Markov-model-based position prediction method |
| CN106570182A (en) * | 2016-11-12 | 2017-04-19 | 中兴软创科技股份有限公司 | Getting-off station recognition method and system for bus |
| CN110264710A (en) * | 2019-05-21 | 2019-09-20 | 天津大学 | It is swiped the card the bus passenger flow estimating method with public transport GPS data based on IC card |
| CN110390349A (en) * | 2019-06-20 | 2019-10-29 | 浙江大学 | Bus passenger flow volume based on XGBoost model predicts modeling method |
| CN110276977A (en) * | 2019-07-29 | 2019-09-24 | 广东工业大学 | A kind of bus station matching process, device, equipment and readable storage medium storing program for executing |
| CN111275241A (en) * | 2020-01-08 | 2020-06-12 | 天津大学 | Bus passenger getting-off station inference method based on machine learning decision tree |
Non-Patent Citations (1)
| Title |
|---|
| 基于隐马尔可夫模型的公交乘客出行链识别方法;崔洪军;张晓阳;朱敏清;;科学技术与工程(第19期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111985710A (en) | 2020-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111985710B (en) | Bus passenger travel station prediction method, storage medium and server | |
| Diao et al. | A hybrid model for short-term traffic volume prediction in massive transportation systems | |
| Kato et al. | Comparative analysis of transit assignment: evidence from urban railway system in the Tokyo Metropolitan Area | |
| Perugia et al. | Designing a home-to-work bus service in a metropolitan area | |
| Huang et al. | A novel bus-dispatching model based on passenger flow and arrival time prediction | |
| Si et al. | A multi‐class transit assignment model for estimating transit passenger flows—a case study of Beijing subway network | |
| He et al. | Evaluation of city-scale built environment policies in New York City with an emerging-mobility-accessible synthetic population | |
| CN110598917B (en) | Destination prediction method, system and storage medium based on path track | |
| CN110348601A (en) | A kind of short-term passenger flow forecast method of subway based on two-way shot and long term memory network | |
| CN111160622A (en) | Scenic spot passenger flow prediction method and device based on hybrid neural network model | |
| Mo et al. | Calibrating path choices and train capacities for urban rail transit simulation models using smart card and train movement data | |
| Zhang et al. | A method of emission and traveller behavior analysis under multimodal traffic condition | |
| Mei et al. | Identifying commuters based on random forest of smartcard data | |
| Tay et al. | Bayesian optimization techniques for high-dimensional simulation-based transportation problems | |
| Hu et al. | Vehicle travel destination prediction method based on multi-source data | |
| Xu et al. | A taxi dispatch system based on prediction of demand and destination | |
| Hu et al. | Passenger congestion alleviation in large hub airport ground-access system based on queueing theory | |
| CN114202120A (en) | Urban traffic travel time prediction method aiming at multi-source heterogeneous data | |
| Faial et al. | A methodology for taxi demand prediction using stream learning | |
| Munir et al. | Route optimization for the cement industry under the CPEC initiative | |
| Chabchoub et al. | Bike sharing systems: a new incentive rebalancing method based on spatial outliers detection | |
| Feng et al. | Optimising departure intervals for multiple bus lines with a multi‐objective model | |
| Volkovska | Methods and models for passenger transportations forecasting on air routes | |
| Lin | Recurrent neural network models of human mobility | |
| CN112561128B (en) | Method for predicting daily passenger capacity of conventional buses for future urban rail transit transfer |
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 |