CN109657820A - A kind of taxi matching process reserved - Google Patents
A kind of taxi matching process reserved Download PDFInfo
- Publication number
- CN109657820A CN109657820A CN201811281523.2A CN201811281523A CN109657820A CN 109657820 A CN109657820 A CN 109657820A CN 201811281523 A CN201811281523 A CN 201811281523A CN 109657820 A CN109657820 A CN 109657820A
- Authority
- CN
- China
- Prior art keywords
- taxi
- request
- matching
- time
- passenger
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000008569 process Effects 0.000 title claims abstract description 16
- 230000004907 flux Effects 0.000 claims abstract description 9
- 230000036962 time dependent Effects 0.000 claims abstract description 4
- 238000004364 calculation method Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000029305 taxis Effects 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 2
- 230000000977 initiatory effect Effects 0.000 abstract 1
- 238000004422 calculation algorithm Methods 0.000 description 12
- 230000008859 change Effects 0.000 description 6
- 238000002474 experimental method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000009826 distribution Methods 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000012217 deletion Methods 0.000 description 2
- 230000037430 deletion Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 101000869717 Homo sapiens Probable mitochondrial glutathione transporter SLC25A40 Proteins 0.000 description 1
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 102100032418 Probable mitochondrial glutathione transporter SLC25A40 Human genes 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000009885 systemic effect Effects 0.000 description 1
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
- G06Q10/02—Reservations, e.g. for tickets, services or events
-
- 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
- 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
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0645—Rental transactions; Leasing transactions
-
- G06Q50/40—
Abstract
The present invention is a kind of taxi matching process that can be reserved, including the following steps: 1) road net data is initialized, time dependent speed networks are constructed；2) vehicle-state real-time update is sent to central server and handles to system, passenger's initiation trip request Concurrency；3) ready request and unloaded taxi are abstracted into node and establish matching bipartite graph by server, and calculate the efficiency of passenger's request and taxi pairing；4) the taxi matching problem in system is converted to minimum price flux problem, generates optimal " taxi-passenger " matching scheme；5) according to matching scheme, system sends notification to the successful passenger of request, and taxi then goes to designated place to meet passenger；6) step 2) is jumped to, periodically incremental update request and matching.This method and system can solve the allotment problem of taxi reserve requests and Real time request, reduce passenger's Waiting time, improve the running efficiency of taxi, improve benefit.
Description
Technical field
The present invention relates to response type field of traffic, especially a kind of taxi matching process that can be reserved.
Background technique
Demand response formula traffic system (Demand Responsive Transit System) is public transport, traditional taxi
Etc. trip modes important alternative solution.System is made of passenger, vehicle and dispatch service quotient tripartite.Dispatch service quotient,
If Uber and drop drop call a taxi, a large amount of vehicle and passenger are docked.Passenger issues request in real time, and by system matches, searching is most closed
Suitable vehicle is provided the service of " one-to-one ".
However, current most of dispatching service system, is all only capable of processing and requests at random with real-time customer, it is pre- for Gu
It about requests, supports also not enough.This is mainly due to difficulties and challenge that Dispatch by appointment has some inherences.Firstly, pre-
It is important that about trip seeks to pick punctual.People are to have some important activities using the usual reason of reservation vehicle
(as by train or aircraft) need to arrange.If vehicle is late so that passenger waits, the service experience for reserving passenger can mutually be on duty.Its
It is secondary, the complexity of urban road, so that being still difficult accurately to estimate the running time of vehicle on the way at present.Therefore, it reserves
If vehicle want reaching on the time, it is necessary to arrange when reserved the looser time ahead of time reach.But this also band simultaneously
Carry out other some problems.For example the reservation vehicle reached in advance needs to find parking facility, but parking facility is limited；Moreover,
It reserves vehicle service and needs more time and efforts, cost also will increase, this can reduce its market share.It is all these because
Element, so that current vehicle dispatch system or reservation vehicle service or the service is not supported to be independently of current scheduling system
Come carry out.
Reserve requests and Real time request are uniformly accounted for range, " passenger-vehicle " that the overall situation solves by the method for the present invention
Matching problem, reasonable distribution taxi resource can reduce the waiting time of passenger, improve the running efficiency of taxi.
Summary of the invention
It is a primary object of the present invention to overcome drawbacks described above in the prior art, a kind of taxi that can be reserved is proposed
Method of completing the square can satisfy the demand of passenger's reservation by bus, globally deploy taxi resource, arranged rational travel route.
The present invention adopts the following technical scheme:
A kind of taxi matching process reserved, which is characterized in that include the following steps；
1) initialization road net data constructs time dependent speed networks according to taxi historical trajectory data；
2) location information where taxi real-time update, the status information of unloaded/heavy duty, when passenger initiates trip request
To central server, trip request is divided into Real time request and reserve requests according to the time window of trip request by central server,
The reserve requests that wherein Real time request and part meet condition are ready request, can be deployed；
3) ready request and unloaded taxi are abstracted into node and establish matching bipartite graph by central server, and are calculated
The efficiency of ready request and unloaded taxi pairing；If the taxi of the zero load can be in the corresponding time window of ready request
Interior arrival is got on the bus a little, then by the taxi of the zero load, the corresponding node of ready request has connected a line with this, and with the pairing
Efficiency weight of the negative value as the side；
4) two point Source and Sink are added into matching bipartite graph, wherein all taxi nodes of Source connection,
All ready requesting nodes of Sink connection calculate the minimum cost stream scheme of matching bipartite graph, to generate optimal taxi
Vehicle-passenger's matching scheme；
5) it according to taxi-passenger's matching scheme, sends notification to trip and requests successful passenger, taxi then goes to finger
Fixed getting on the bus a little meets passenger, and is sent to destination；
6) step (2) are jumped to.
In step 2), the data format of the trip request includes the following:
q(olon, olat, dlon, dlat, t1, t2, t0)
Wherein, olonExpression is got on the bus a longitude, olatExpression is got on the bus a latitude, dlonIndicate terminal longitude, dlatIndicate terminal
Latitude, time window [t1, t2] indicate passenger's best time for wishing taxi to pick, t0 indicate that trip request generates when
Between；
According to the time window [t1, t2] in trip request Q, trip request is divided into two classes: Real time request Q1It is invited with pre-
Seek Q3
Wherein T1 is preset value, poor for the minimum time of making reserve requests；
By another preset value T2 > T1, by reserve requests Q2Carry out further division:
The current system time of t ' expression defines ready request according to above classification:
To wait request, the ready request can enter matching system and find suitable taxi and be matched clothes
Business waits request to be waited for, until it is converted to ready requesting party and can match.
In step 3), the operational effectiveness formula is as follows:
U (c, q, t)=α * serv (c, q, t)+(1- α) * trac (c, q, t)
Wherein, c expression taxi, u (c, q, t) expression matching efficiency, serv (c, q, t) expression service level, trac (c,
Q, t) indicate track efficiency, α indicates a fixed proportionality coefficient, and value range is [0,1], and t indicates described unloaded and goes out
It hires a car at the time of start matching and request Q to a ready trip.
The calculation of the service level serv (c, q, t) of the Real time request is as follows:
Wherein, pt (c, q, t) is at the time of estimating taxi to be connected to passenger:
Pt (c, q, t)=t+w (c, q, t)+δ
Wherein, w (c, q, t) be estimate taxi sail for it is described get on the bus a little time used, δ >=0 is the taxi waiting time,
It is defined as follows
It gets on the bus a little if waiting time δ indicates that taxi is reached before the q.t1 moment, taxi needs the time waited.
The calculation formula of the service level serv (c, q, t) of the reserve requests is as follows:
Wherein, K is a parameter more than or equal to 1.
The calculation formula of the track efficiency is as follows:
Wherein, U is preset preset parameter, is set as taxi and is connected to trip request to being connected to putting down for empty driving before passenger
The equal time.
In step 4), taxi matching is converted to minimum price flux problem and is solved, specifically, taxi matching is asked
It is entitled to seek a matching schemeSo that the matched overall efficiency of all taxis and passenger is maximum:
Wherein,For feasible matched set all in current matching system, t ' is present system time, GrFor road
Net, Q be ready request set, C be can matched taxi set will be hired out by the matching bipartite graph of foundation
Vehicle matching problem is converted into minimum price flux problem:
Wherein, EbFor the set on side in matching bipartite graph, cw (u, v) is in matching bipartite graph with u, and v is the side on vertex
Cost weight, f (u, v) are in matching bipartite graph with u, and v is the flow on the side on vertex, the matching bipartite graph to meet it is following about
Beam:
Capacity-constrained: f (u, v)≤capacity (u, v), i.e., while upper flow no more than while capacity
Symmetry: f (u, v)=- f (v, u)
Flow conservation:To all u ≠ Source, Sink
Meet target flow:And
Wherein, VbFor the set put on matching bipartite graph, w is a vertex in set Vb, and d is target flow, and m, n are
It is taxis quantity and trip number of requests respectively, solves the minimum price flux problem, obtained matching result is described best
Taxi-passenger's matching scheme.
Further include that gridding operation is carried out to road net data in step 6), by taxi and occurs requesting to navigate to grid
In, using the minimal weight matched between matrix record grid, incrementally updating is carried out to matching bipartite graph.
By the above-mentioned description of this invention it is found that compared with prior art, the invention has the following beneficial effects:
The historical track and call data of this system combination taxi, the overall situation solve taxi reserve requests and Real time request
Allotment problem, reasonable distribution taxi resource can reduce the Waiting time of passenger, improve the running efficiency of taxi, improve
The income of taxi driver.
Detailed description of the invention
Fig. 1 is the flow chart of present system；
Fig. 2 is the architecture diagram of present system；
Fig. 3 is that " vehicle-passenger " matches bipartite graph exemplary diagram；
Fig. 4 is the schematic diagram that " vehicle-passenger " matches that bipartite graph updates；
Traffic network grid schematic diagram in the Xiamen City Fig. 5 island；
The comparison for requesting Service Efficiency of tri- kinds of Fig. 6 MCF, MUF, RRO methods；
The comparison for requesting Service Efficiency of tri- kinds of Fig. 7 MCF, MUF, RRO methods；
The comparison of tri- kinds of Fig. 8 MCF, MUF, RRO method driver's total incomes.
Specific embodiment
Below by way of specific embodiment, the invention will be further described.
The taxi matching process that one kind of the invention can be reserved comprising have: initialization city road network matches efficiency
It calculates, the foundation of matching bipartite graph, the calculating of Optimum Matching, the incrementally updating for matching bipartite graph.Referring to Fig. 1, Fig. 2, below
It introduces main each step and realizes details:
1) initialization road net data constructs time dependent speed networks according to taxi historical trajectory data.
The most basic data of road network are map data files, can be derived from openstreet.First by ground picture and text
The element of non-rice habitats on part filters out, and then imports the file into database (such as neo4j database).Meanwhile it reading
Taxi positioning and traveling record, are calculated speed networks.Wherein, taxi positioning and traveling record may be from hiring out
All valid data, format in some period of Che company are as follows:
D(lon,lat,dir,speed,day,hour,minute,date)
Data from left to right, be followed successively by longitude lon, latitude lat, direction dir, speed speed, date day (n days July),
Hour hour, minute minute, date date.There are 144 values in every section in speed networks, respectively indicates each 10 in one day
The travel speed (144=24*60/10) in the section in minute.The calculation method of each speed is taxi note more than calculating
Record the corresponding period, corresponding road section speed weighted average.
2) calculating of efficiency is matched.Location information where taxi real-time update, the status information of unloaded/heavy duty；When
Passenger initiates trip request to central server, and trip request is divided into reality according to the time window of trip request by central server
When request and reserve requests, wherein Real time request and part meet condition reserve requests be ready request, can be adjusted
Match.
The data format of the trip request of passenger is as follows:
q(olon, olat, dlon, dlat, t1, t2, t0)
Wherein, olonIndicate starting point longitude, olatIndicate starting point latitude, dlonIndicate terminal longitude, dlatIndicate terminal latitude,
Time window [t1, t2] indicates the best time that passenger wishes taxi to pick, and t0 indicates the time that trip request generates.
According to the time window [t1, t2] of trip request, trip request is divided into two classes: Real time request Q1With reserve requests Q2
Wherein T1 is preset value, poor for the minimum time of making reserve requests；
By another preset value T2 > T1, by reserve requests Q2Carry out further division:
According to above classification, ready request is defined:
To wait request, the ready request can enter matching system and find suitable taxi and be matched clothes
Business waits request to be waited for, until it is converted to ready requesting party and can match.
When considering whether a taxi will go to pick certain passenger, can be assessed with a score value, this score value is again
Referred to as match efficiency.It is higher to match efficiency, then it represents that this taxi goes to pick this passenger more reasonable, economical and quick.
At the time of indicating that a taxi starts matching to a trip request q with t.A matching is calculated with following formula
Efficiency:
U (c, q, t)=α * serv (c, q, t)+(1- α) * trac (c, q, t)
Wherein, c indicates that taxi, u (c, q, t) indicate matching efficiency, and serv (c, q, t) indicates service level efficiency,
Trac (c, q, t) indicates track efficiency, and α indicates a fixed proportionality coefficient, and value range is [0,1].
The calculation of Real time request and the serv (c, q, t) of reserve requests are different；
For Real time request, the calculation formula of serv (c, q, t) is as follows:
Wherein, pt (c, q, t) is at the time of estimating taxi to be connected to passenger
Pt (c, q, t)=t+w (c, q, t)+δ
Wherein, w (c, q, t) is to estimate taxi to sail for the time used in starting point by bus, and δ >=0 is the waiting time of taxi,
It is defined as follows
If waiting time δ indicates that taxi reaches starting point taxi by bus before the q.t1 moment and needs the time waited.
For reserve requests, the calculation formula of serv (c, q, t) is as follows:
Wherein, K is a parameter more than or equal to 1, and K value is bigger, and system is more inclined to the allotment of reserve requests.
The calculation formula of track efficiency is as follows:
Wherein, U is one preset parameter of systemic presupposition, is traditionally arranged to be taxi and is connected to request to before being connected to passenger,
The average time of empty driving.
For entire matching system, it is assumed that its road network is Gr, it is requested by n in current time t, request set Q,
There is m taxi in taxi set C, then, the matching between Q and C can be indicated with the matrix M of a n × m.In matrix
Element aI, jDefault value is 0, as a taxi ciIt is matched and requests q to a passengerjWhen, aI, jIt is 1.This matching is can
Capable matching will meet following two condition:
Condition 1: if aI, j=1, then pt (c, q, t) ∈ [q.t1, q.t2]
Condition 2:
Condition 1 indicates that taxi needs to be connected to passenger at the appointed time, and condition 2 indicates that a passenger can only be by one
Taxi is connected to.This is feasible matched to be defined as follows in t moment overall efficiency:
3) match the foundation of bipartite graph: ready request and unloaded taxi are abstracted into node and established by central server
Bipartite graph is matched, and calculates the efficiency of ready request and unloaded taxi pairing；If the taxi of the zero load can be
Thread is requested to reach in corresponding time window and be got on the bus a little, then by the taxi of the zero load, the corresponding node of ready request has been connected with this
A line, and using the negative value of the efficiency of the pairing as the weight on the side.
The present invention solves the problems, such as taxi and the matched global optimum of passenger with matching bipartite graph.By idle taxi
Vehicle is abstracted into node and is placed on left side, does not determine that matched request is abstracted into node and puts right side.Each passenger is requested, if
Taxi can reach starting point and be picked in its request time window [t1, t2], then by the taxi node and the request section
Point connects line.In addition, adding two special point Source and Sink into bipartite graph, Source node is connected all with line
Taxi node, Sink node connect all requesting nodes with line.Use EbIndicate the set on all sides on bipartite graph, each edge
There are a capacity cap and a cost weight cw, be defined as follows:
cap(vi, vj)=1, (vi, vj)∈Eb
It can be found in Fig. 2
4) two point Source and Sink, wherein Source connection the calculating of Optimum Matching: are added into matching bipartite graph
All taxi nodes, all ready requesting nodes of Sink connection calculate the minimum cost stream scheme of matching bipartite graph, thus
Generate optimal taxi-passenger's matching scheme
Optimum Matching is the maximum feasible matching of overall efficiency.Assuming thatFor in all feasible matched set of t moment,
The then efficiency of Optimum Matching are as follows:
For feasible matched set all in current matching system, t ' is present system time, seeks a certain moment
The Optimum Matching of t 'The problem of be known as static maximum matching Efficacy Problem.It, can be by static state most by the matching bipartite graph of foundation
Big matching Efficacy Problem is converted into minimum price flux problem:
EbFor the set on side in matching bipartite graph, cw (u, v) is to be weighed in matching bipartite graph with the cost that u, v are the side on vertex
Weight, f (u, v) are with u in matching bipartite graph, and when v solves minimum cost Liu problem for the flow on the side on vertex, bipartite graph will expire
It is enough lower constraint:
Capacity-constrained: f (u, v)≤capacity (u, v), i.e., while upper flow no more than while capacity
Symmetry: f (u, v)=- f (v, u)
Flow conservation:To all u ≠ Source, Sink
Meet target flow:And
Wherein, VbFor the set put on bipartite graph, d is target flow, m, n be vehicle number and number of request respectively.
The step of solving minimum cost stream is as follows:
(1) d=min (| C- Δ C |, | Q- Δ Q |), wherein Δ C be matching bipartite graph in requesting node without line
The quantity of taxi node, Δ Q are the requesting node quantity in bipartite graph with taxi node without line；
(2) if d > 0, minimum cost stream is solved with minimum-cost flow algorithm；Otherwise, algorithm terminates, and the bipartite graph is without most
Small cost stream；
(3) if the minimum cost stream for meeting flow d can be acquired, retain the matching result, algorithm terminates；If nothing
Method acquires the minimum cost stream for meeting flow d, then it represents that flow d is excessive, enables d=d-1, continues (2).
Minimum price flux problem (Minimum-cost flow problem, MCFP) in step (2) is operational research and figure
A classical problem in.Due to target and constraint condition be all it is linear, linear programming (linear can be passed through
Programming method) solves.Typical algorithm has Cycle Cancelling algorithm (Morton Klein, " A
primal method for minimal cost flows with applications to the assignment and
transportation problems".Management Science.14:205–220.doi:10.1287/
Mnsc.14.3.205,1967), Network Simplex algorithm (Orlin, James B. " A polynomial time
primal network simplexalgorithm for minimum cost flows".Mathematical
Programming.78 (2): 109-129.doi:10.1007/BF02614365.ISSN 0025-5610,1997) etc..
5) it according to taxi-passenger's matching scheme, sends notification to trip and requests successful passenger, taxi then goes to finger
Fixed getting on the bus a little meets passenger, and is sent to destination.
6) step (2) are jumped to, the incrementally updating of bipartite graph is matched.
The state of the trip request of taxi and passenger can change with the operating of system, in each period
Matching maximum overall efficiency be possible to cannot, matching scheme is also different.Matching system goes to change match party with time change
For case to reach bigger overall efficiency, this is particularly important for the request of reservation type.When a request still has ample time
Etc. it is to be serviced when, matched taxi can be replaced with and reach bigger overall efficiency originally.But it continually counts
It calculates, change matching scheme can make system burden aggravate.Therefore, by the way of incrementally updating.
Referring to fig. 4, time t '=t+ut moment request set and taxi set are illustrated respectively in Q ' and C '.Ut is
System makes the time interval of matching scheme adjustment.In ut this period, some taxis may be become from passenger carrying status
Complete vehicle curb condition, some new requests it is possible that, some passengers may be connect away by the taxi reached.
Q ' consists of two parts: Q '=Q-∪Q+, wherein Q-It is derivative by the request set Q in t moment and goes out, and Q+
Then represent the request being newly added.Likewise, C '=C_∪C+, wherein C-It is to be gone out by the taxi set C derivative in t moment
, and C+Then represent the taxi being newly added.Q-Q-And C-C-Node corresponding to interior request and taxi will be from bipartite graph
Middle deletion.Need to delete includes following two situation: (1) taxi c has been matched to request q and w (c, q) < τ.τ is
A preset threshold value.That is when a taxi will reach the starting point for the passenger that its preparation is picked, the group
Matching can not be revoked.(2) request q does not have matched taxi and q (t+ δ) < t '.δ is a threshold value of setting, and t ' is to work as
Preceding system time, that is to say, that when request is serviced without suitable taxi for a long time, be then from matching by the request
It is deleted in system.
Road network is divided into k row k column, then road network is divided into k2A grid, uses giIt indicates, wherein 1≤i≤k2.Herein
On the basis of, define cost matrix L (k2, k2), grid g is indicated with dist (i, j)iTo grid gjMinimum cost weight.The matrix
It changes over time.It can be found in attached drawing 3.The update of bipartite graph is matched shown in steps are as follows:
A) passenger being connected to for being hired out vehicle deletes the group from bipartite graph and matches corresponding taxi node
And requesting node；
B) for the request that services without suitable taxi of long-time, request is rejected, by its node from bipartite graph
Middle deletion；
C) for the taxi of starting point of passenger that its preparation is picked will be reached, the taxi is deleted from bipartite graph
The line of node and other requesting nodes；Equally, the line side of the requesting node Yu other taxi nodes is also deleted；
D) belong to C for all+Taxi, calculate separately its cw and tax with reachable requests all in bipartite graph and arrive
On the side that they are connected；
E) belong to C for all-Taxi c, it and all Q+In q, if meeting dist (gc, gq) < max
{ q.t2-t ': q ∈ Q+, then recalculate the cw of it and all requesting nodes in Q '；Otherwise, see that this is connected with taxi node
All requesting nodes, if dist (gc, gq) > q.t2-t ', then by the side being connected of the taxi node and the requesting node
It deletes, otherwise, cw is updated according to the current traveling-position of taxi.
Experimental verification
Experimental verification uses actual road net data and taxi driving trace data.Road net data is from open source map
OpenStreetMap (abbreviation OSM), body of a map or chart are [118.0660E, 118.1980E] × [24.4240N, 24.5600N],
OSM road net data is stored in chart database Neo4j to use for experiment.It realizes by the library Neo4j Spatial to road network
Read inquiry operation.Neo4j Spatial is importing, storage and the inquiry of spatial data etc. for being directed to Neo4j database
The routine call library of operation.Taxi track data uses the driving trace data in Xiamen City's on July 1st, 2014 to July 10.
The track data is mainly used for two aspects, is that data positioned, counted, calculated etc. with operation on one side, establishes speed
Network.On the other hand, obtain track data in taxi starting point and anchor point, to simulate passenger request departure place and
Destination.
Experiment is write based on the road network being stored in Neo4j database with JAVA language.Detailed process is as follows for experiment:
It the operation such as is read out, positions, counting to taxi track data in advance, establish speed networks and storing into road network, wherein
Every section is according to one speed of every 10 minutes statistics, so there are 144 speed (144=24*60/10) in every section.Then,
Gridding is carried out to Xiamen map, is equally divided into 100 small rectangles.Fig. 5 gives in experimentation to Xiamen traffic network grid
Schematic diagram, there is the rectangle of road net data to share 72, to this 72 small rectangle find closest to its geographic center road section
Point is used as anchor point, and timesharing (the every 10 minutes) shortest distances calculated between each rectangle anchor point set up cost metrix, there are
To 144 cost metrixs, and store into database.Simulator reads the message file of vehicle and request, generates vehicle and asks
It asks, carries out the simulation of vehicle and request.First in the current time vehicle and request establish two shunting figures, and calculate it
Between effectiveness then calculate minimum cost flow scheme using minimum cost flow algorithm, change road further according to this scheme
The matching of vehicle and request in net determines matching relationship, updates bipartite graph, more new vehicle, request.When simulated time exhausts
When, simulation terminates.
Experiment simulation generates 50 taxis, generates passenger's request according to every five seconds one frequency, simulated time 90 is divided
Clock.Wherein, the 10% of reserve requests Zhan total request amount.Experiment first just matches plan using minimum cost flow in conjunction with reserve requests
The matching process of slightly (Minimum-cost Flow, MCF) is tested, and request Service Efficiency, passenger waiting time and driver are obtained
Three results of total income.Hereafter we refer to present aspect method with MCF.In addition, two comparison algorithms of experimental setup, wherein one
A comparison algorithm combines reservation type request, using Utility preference strategy (Maximum Utility First, MUF), i.e.,
The maximum matching of Utility preferentially selects and determines, rather than pursues whole maximum strategy；Another comparison algorithm is not examined
The request for considering reservation type, using matching strategy (the Realtime Request for handling reserve requests as Real time request
Only,RRO)。
The comparing result of the Service Efficiency of request is as shown in Figure 6.It can be seen that on total request Service Efficiency, since MCF is adopted
With the strategy of total optimization, so MCF ratio MUF is slightly higher, compared with the RRO for equally pursuing total optimization, total request of MCF meets
Rate is not much different with RRO.On the Service Efficiency of reserve requests, MCF ratio MUF high, this should update matching each due to MUF
When, total matching for selecting maximum Utility preferentially determines, causes the reserve requests occurred below that can not be matched to vehicle.RRO by
In handling reserve requests as Real time request, so its reserve requests Service Efficiency is 0.On Real time request Service Efficiency, MCF
It is not much different with MUF, RRO is apparently higher than other two algorithms, the reason is that RRO treats as reserve requests at Real time request
Reason.
The waiting time comparing result of passenger is as shown in Figure 7.In conjunction with Fig. 6 Service Efficiency comparison it can be found that three kinds of sides
In method, MCF reserve requests Service Efficiency is maximum, and reserve requests would generally reserve the more time to allow vehicle driving to go out to what is requested
Hair point.So the MCF waiting time can be small compared with what other two methods were come.And RRO handles reserve requests as Real time request,
So that its average latency is apparently higher than other two algorithms.It is finally the Comparative result of driver's income, as shown in Figure 8.
MUF is whole optimal due to not accounting for, so that its income is lower than income brought by MCF.Although and whole without reservation matching
The request Service Efficiency of body is not low, but it handles reserve requests as Real time request, in income, not no reserve requests
It additionally increases income, so that income is lower than other two methods.
From simulated experiment as can be seen that global taxi reserve requests and the Real time request of solving that the method for the present invention proposes
Allotment problem can be reduced the Waiting time of passenger, be improved the running efficiency of taxi, mentioned with reasonable distribution taxi resource
The income of high taxi driver.
The above is only a specific embodiment of the present invention, but the design concept of the present invention is not limited to this, all to utilize this
Design makes a non-material change to the present invention, and should all belong to behavior that violates the scope of protection of the present invention.
Claims (8)
1. the taxi matching process that one kind can be reserved, which is characterized in that include the following steps；
1) initialization road net data constructs time dependent speed networks according to taxi historical trajectory data；
2) location information where taxi real-time update, the status information of unloaded/heavy duty, when passenger initiates trip request into
Trip request is divided into Real time request and reserve requests according to the time window of trip request by central server, central server, wherein
The reserve requests that Real time request and part meet condition are ready request, can be deployed；
3) ready request and unloaded taxi are abstracted into node and establish matching bipartite graph by central server, and are calculated just
The efficiency of thread request and unloaded taxi pairing；If the taxi of the zero load can arrive in the corresponding time window of ready request
Up to getting on the bus a little, then by the taxi of the zero load, the corresponding node of ready request has connected a line with this, and with the effect of the pairing
Weight of the negative value of energy as the side；
4) two point Source and Sink are added into matching bipartite graph, wherein all taxi nodes of Source connection, Sink
All ready requesting nodes are connected, the minimum cost stream scheme of matching bipartite graph is calculated, multiplies to generate optimal taxi-
Objective matching scheme；
5) it according to taxi-passenger's matching scheme, sends notification to trip and requests successful passenger, taxi then goes to specified
It gets on the bus and a little meets passenger, and be sent to destination；
6) step (2) are jumped to.
2. the taxi matching process that one kind as described in claim 1 can be reserved, it is characterised in that: in step 2), it is described go out
Row request data format include the following:
q(olon, olat, dlon, dlat, t1, t2, t0)
Wherein, olonExpression is got on the bus a longitude, olatExpression is got on the bus a latitude, dlonIndicate terminal longitude, dlatIndicate terminal latitude,
Time window [t1, t2] indicates the best time that passenger wishes taxi to pick, and t0 indicates the time that trip request generates；
According to the time window [t1, t2] in trip request Q, trip request is divided into two classes: Real time request Q1With reserve requests Q3
Wherein T1 is preset value, poor for the minimum time of making reserve requests；
By another preset value T2 > T1, by reserve requests Q2Carry out further division:
The current system time of t ' expression defines ready request according to above classification:
To wait request, the ready request can enter matching system and find suitable taxi and be matched service, etc.
Wait request to be waited for, until it is converted to ready requesting party and can match.
3. the taxi matching process that one kind as claimed in claim 2 can be reserved, it is characterised in that: described in step 3)
Operational effectiveness formula is as follows:
U (c, q, t)=α * serv (c, q, t)+(1- α) * trac (c, q, t)
Wherein, c indicates that taxi, u (c, q, t) indicate matching efficiency, and serv (c, q, t) indicates service level, trac (c, q, t)
Indicate track efficiency, α indicates a fixed proportionality coefficient, and value range is [0,1], and t indicates the taxi of the zero load
At the time of starting matching and request Q to a ready trip.
4. the taxi matching process that one kind as claimed in claim 3 can be reserved, it is characterised in that: the clothes of the Real time request
Be engaged in horizontal serv (c, q, t) calculation it is as follows:
Wherein, pt (c, q, t) is at the time of estimating taxi to be connected to passenger:
Pt (c, q, t)=t+w (c, q, t)+δ
Wherein, w (c, q, t) be estimate taxi sail for it is described get on the bus a little time used, δ >=0 is the taxi waiting time, fixed
Justice is as follows
It gets on the bus a little if waiting time δ indicates that taxi is reached before the q.t1 moment, taxi needs the time waited.
5. the taxi matching process that one kind as claimed in claim 3 can be reserved, it is characterised in that: the clothes of the reserve requests
Be engaged in horizontal serv (c, q, t) calculation formula it is as follows:
Wherein, K is a parameter more than or equal to 1.
6. the taxi matching process that one kind as claimed in claim 3 can be reserved, it is characterised in that: the meter of the track efficiency
It is as follows to calculate formula:
Wherein, U is preset preset parameter, is set as taxi and is connected to trip request to the mean time for being connected to empty driving before passenger
Between.
7. the taxi matching process that one kind as claimed in claim 3 can be reserved, it is characterised in that: in step 4), will hire out
Vehicle matching is converted to minimum price flux problem and is solved, specifically, taxi matching problem is to seek a matching schemeSo that the matched overall efficiency of all taxis and passenger is maximum:
Wherein,For feasible matched set all in current matching system, t ' is present system time, GrFor road network, Q is
The set of ready request, C be can matched taxi set taxi is matched by the matching bipartite graph of foundation
Problem is converted into minimum price flux problem:
Wherein, EbFor the set on side in matching bipartite graph, cw (u, v) is to be weighed in matching bipartite graph with the cost that u, v are the side on vertex
Weight, f (u, v) are in matching bipartite graph with u, and v is the flow on the side on vertex, which will meet following constraint:
Capacity-constrained: f (u, v)≤capcity (u, v) i.e. while upper flow no more than while capacity
Symmetry: f (u, v)=- f (v, u)
Flow conservation:To all u ≠ Source, Sink
Meet target flow:And
Wherein, VbFor the set put on matching bipartite graph, w is a vertex in set Vb, and d is target flow, and m, n are difference
Be taxis quantity and trip number of requests, solve the minimum price flux problem, obtained matching result be it is described it is optimal go out
It hires a car-passenger's matching scheme.
8. the taxi matching process that one kind as described in claim 1 can be reserved, it is characterised in that: in step 6), further include
Gridding operation is carried out to road net data, by taxi and occurs requesting to navigate in grid, using between matrix record grid
The minimal weight of pairing carries out incrementally updating to matching bipartite graph.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811281523.2A CN109657820B (en) | 2018-10-23 | 2018-10-23 | Taxi matching method capable of making reservation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811281523.2A CN109657820B (en) | 2018-10-23 | 2018-10-23 | Taxi matching method capable of making reservation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109657820A true CN109657820A (en) | 2019-04-19 |
| CN109657820B CN109657820B (en) | 2020-10-20 |
Family
ID=66110543
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811281523.2A Active CN109657820B (en) | 2018-10-23 | 2018-10-23 | Taxi matching method capable of making reservation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109657820B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110264133A (en) * | 2019-06-12 | 2019-09-20 | 浙江大搜车软件技术有限公司 | Vehicle deploying method, apparatus, computer equipment and storage medium |
| CN111126799A (en) * | 2019-12-10 | 2020-05-08 | 广东工业大学 | Shared network driver and crew matching method based on bipartite graph |
| CN111178716A (en) * | 2019-12-19 | 2020-05-19 | 广东工业大学 | Auction type automatic driving taxi empty dispatching method |
| CN111191899A (en) * | 2019-12-23 | 2020-05-22 | 华南理工大学 | Vehicle scheduling method based on region division parallel genetic algorithm |
| CN113642752A (en) * | 2021-08-24 | 2021-11-12 | 长沙理工大学 | Method and device for determining simultaneous operation number of urban taxies |
| US20230385723A1 (en) * | 2021-02-18 | 2023-11-30 | Grabtaxi Holdings Pte. Ltd. | System and method for bipartite matching |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103680128A (en) * | 2013-11-26 | 2014-03-26 | 上海交通大学 | Intelligent taxi scheduling system |
| CN105761482A (en) * | 2016-05-10 | 2016-07-13 | 北京交通大学 | Taxi real-time appointing method and system based on fairness |
| CN106781447A (en) * | 2017-02-28 | 2017-05-31 | 广州交通信息化建设投资营运有限公司 | A kind of intelligent taxi scheduling communication method |
| US20170167882A1 (en) * | 2014-08-04 | 2017-06-15 | Xerox Corporation | System and method for generating available ride-share paths in a transportation network |
-
2018
- 2018-10-23 CN CN201811281523.2A patent/CN109657820B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103680128A (en) * | 2013-11-26 | 2014-03-26 | 上海交通大学 | Intelligent taxi scheduling system |
| US20170167882A1 (en) * | 2014-08-04 | 2017-06-15 | Xerox Corporation | System and method for generating available ride-share paths in a transportation network |
| CN105761482A (en) * | 2016-05-10 | 2016-07-13 | 北京交通大学 | Taxi real-time appointing method and system based on fairness |
| CN106781447A (en) * | 2017-02-28 | 2017-05-31 | 广州交通信息化建设投资营运有限公司 | A kind of intelligent taxi scheduling communication method |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110264133A (en) * | 2019-06-12 | 2019-09-20 | 浙江大搜车软件技术有限公司 | Vehicle deploying method, apparatus, computer equipment and storage medium |
| CN110264133B (en) * | 2019-06-12 | 2022-04-19 | 浙江大搜车软件技术有限公司 | Vehicle allocation method and device, computer equipment and storage medium |
| CN111126799A (en) * | 2019-12-10 | 2020-05-08 | 广东工业大学 | Shared network driver and crew matching method based on bipartite graph |
| CN111126799B (en) * | 2019-12-10 | 2023-04-11 | 广东工业大学 | Shared network driver and crew matching method based on bipartite graph |
| CN111178716A (en) * | 2019-12-19 | 2020-05-19 | 广东工业大学 | Auction type automatic driving taxi empty dispatching method |
| CN111178716B (en) * | 2019-12-19 | 2023-05-23 | 广东工业大学 | Automatic driving taxi empty scheduling method for auction type |
| CN111191899A (en) * | 2019-12-23 | 2020-05-22 | 华南理工大学 | Vehicle scheduling method based on region division parallel genetic algorithm |
| CN111191899B (en) * | 2019-12-23 | 2022-04-12 | 华南理工大学 | Vehicle scheduling method based on region division parallel genetic algorithm |
| US20230385723A1 (en) * | 2021-02-18 | 2023-11-30 | Grabtaxi Holdings Pte. Ltd. | System and method for bipartite matching |
| CN113642752A (en) * | 2021-08-24 | 2021-11-12 | 长沙理工大学 | Method and device for determining simultaneous operation number of urban taxies |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109657820B (en) | 2020-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109657820A (en) | A kind of taxi matching process reserved | |
| US11614751B2 (en) | System for on-demand high-capacity ride-sharing via dynamic trip-vehicle assignment and related techniques | |
| CN110110871B (en) | Order allocation method and system | |
| US11619951B2 (en) | On-demand high-capacity ride-sharing via dynamic trip-vehicle assignment with future requests | |
| Amirgholy et al. | Optimal design of sustainable transit systems in congested urban networks: A macroscopic approach | |
| Zhu et al. | Public vehicles for future urban transportation | |
| Agatz et al. | Optimization for dynamic ride-sharing: A review | |
| Taniguchi et al. | Intelligent transportation system based dynamic vehicle routing and scheduling with variable travel times | |
| CN107101643B (en) | A kind of share-car matching process | |
| CN105279955B (en) | A kind of share-car method and apparatus | |
| CN107103383B (en) | Dynamic taxi sharing scheduling method based on taxi-taking hotspot | |
| US20190339087A1 (en) | Deep reinforcement learning for optimizing carpooling policies | |
| CN105917376A (en) | Optimizing selection of drivers for transport requests | |
| CN109948854B (en) | Intercity network taxi booking order distribution method based on multi-objective optimization | |
| Li et al. | Optimizing flexible one-to-two matching in ride-hailing systems with boundedly rational users | |
| CN108288101B (en) | Auction mechanism-based vehicle resource allocation and pricing method for online taxi appointment reservation service | |
| Zwick et al. | Impact of service design on urban ridepooling systems | |
| Araldo et al. | On the Importance of demand Consolidation in Mobility on Demand | |
| Liang et al. | An optimization model for vehicle routing of automated taxi trips with dynamic travel times | |
| JP2002024659A (en) | Taxi dispatch reserving system | |
| CN112562309B (en) | Network taxi appointment scheduling method based on improved Dijkstra algorithm | |
| WO2020207430A1 (en) | Order distribution method, device, electronic apparatus, and computer readable storage medium | |
| Foti et al. | The nash equilibrium among taxi ridesharing partners | |
| CN114819819B (en) | Path planning implementation method under instant logistics scene | |
| CN115879658A (en) | Customized bus route optimization method, system, storage medium and equipment |
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 |