CN111126799B - Shared network driver and crew matching method based on bipartite graph - Google Patents
Shared network driver and crew matching method based on bipartite graph Download PDFInfo
- Publication number
- CN111126799B CN111126799B CN201911257853.2A CN201911257853A CN111126799B CN 111126799 B CN111126799 B CN 111126799B CN 201911257853 A CN201911257853 A CN 201911257853A CN 111126799 B CN111126799 B CN 111126799B
- Authority
- CN
- China
- Prior art keywords
- request
- taxi
- passenger
- matching
- requests
- 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
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000029305 taxis Effects 0.000 claims abstract description 47
- 238000004422 calculation algorithm Methods 0.000 claims description 35
- 238000013507 mapping Methods 0.000 claims description 13
- 238000012216 screening Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 description 7
- 238000004088 simulation Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 6
- 238000002474 experimental method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000035755 proliferation Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06311—Scheduling, planning or task assignment for a person or group
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- 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 application discloses a shared network driver and crew matching method based on bipartite graph, which comprises the following steps: the passenger submits a request for sharing and uploads the request to the cloud server, and the cloud server screens taxis meeting the maximum taxi carrying number constraint and meeting the time constraints of the passengers on the taxi and the passengers sending the request aiming at the request of each passenger according to the request of each passenger and the positions and passenger carrying information of all taxis in the taxi set, so as to obtain a candidate taxi set corresponding to the request of each passenger; performing many-to-many matching on all requests and all candidate taxis to obtain the optimal matching between the taxis and the requests, and generating an order according to a matching result; for requests that are not accepted, the cloud server matches its cycles until all requests are accepted or no taxis can meet the request constraints. The method can be suitable for the matching problem of many-to-many ride sharing of multiple passengers and multiple vehicles, the ride sharing rate can be substantially improved, and traffic pollution is reduced.
Description
Technical Field
The application relates to the technical field of computers and transportation, in particular to a shared network driver and passenger matching method based on bipartite graph.
Background
Urban population proliferation and increasing travel demand are among the causes of urban traffic congestion. Transportation travelers have been encouraged to use public transportation services throughout recent years. With the rapid development of artificial intelligence and automatic driving technology, the future transportation mode will be changed into a shared transportation mode mainly based on the automatic driving taxi service. Unlike traditional taxis, an autonomous taxi system must rely on an efficient ride-sharing matching algorithm to determine the match between a vehicle that is still available and a new passenger.
The existing taxi dynamic sharing matching method comprises the following steps:
a 'dynamic ride combination matching multistage screening method' provides a ride combination matching method with three-stage matching processing, wherein three stages are time and detour constraint, a near principle and a user preference principle. And screening out a candidate vehicle set in a first level and a second level, and matching in a third level. The disadvantages are that: the patent carries out the ride-sharing matching according to the preference of the user, but the method for screening the candidate set by using the principle of proximity only provides local options for the user.
An intelligent scheduling method and system for taxi dynamic ride sharing provides a ride sharing candidate set screening method based on a utility function, wherein the utility function comprises information of two aspects of path topology and riding cost. The patent uses a utility function to estimate a ride-through probability to screen a candidate set, and performs ride-through matching according to the selection of a user. The disadvantage is that the overall optimum of the ride-share cannot be further achieved.
A taxi co-taking matching and paying method provides a multi-person co-taking idea in a mode of rewarding by reducing price, and has the defect that a practical matching method and a practical paying method are not provided.
A complex road network-based taxi carpooling cluster optimization system and an optimization method thereof provide a complex road network-based multi-target carpooling matching method. The solution comprises three objective functions, namely the minimum sum of the bypassing proportions of the co-passenger, the shortest objective function of the traveling distance of the co-passenger vehicle and the maximization of the co-passenger matching solution, and the solving method is a genetic algorithm. The method has the disadvantages that the solving algorithm belongs to a searching algorithm, the global optimum can not be strictly ensured, and the variable is difficult to be expressed by a simple data structure.
The patent of passenger and passenger matching based sharing network sharing model is provided, and is suitable for sharing between two persons and three persons. Compared with the scheme, the method has the defects of more classification conditions, more formulas and large calculation amount, and the calculation time for co-multiplication matching among three people cannot meet the practical requirement.
The scheme mainly provides different methods for screening the co-multiplication candidate set, and user selection is used as a matching basis, but the co-multiplication optimization matching method is not further researched. The method for matching according to the selection of the user is only suitable for solving the optimal matching condition of a single passenger or a single vehicle, often results in that other multiple passengers cannot obtain an ideal ride-sharing scheme, cannot realize cluster optimization of ride-sharing, cannot achieve integral optimization, and cannot effectively improve the ride-sharing rate.
Disclosure of Invention
The application aims to provide a bipartite graph-based shared network driver and passenger matching method, which can be suitable for the many-to-many ride matching problem of multiple passengers and multiple vehicles, substantially improves the ride rate of the vehicles, and reduces traffic pollution.
In order to realize the task, the following technical scheme is adopted in the application:
a shared network driver and multiplier matching method based on bipartite graph comprises the following steps:
a passenger inputs a request for carpooling in a taxi carpooling reservation system through a handheld terminal device and uploads the request to a cloud server;
the method comprises the steps that a cloud server collects requests for carpooling initiated by all passengers in a set area within a period of time, and searches all taxi sets with vacant positions in the set area; screening taxis which meet the maximum passenger carrying number constraint of the taxis and simultaneously meet the time constraints of passengers on the taxis and passengers who send the requests according to the request of each passenger and the positions and passenger carrying information of all the taxis in the taxi set, and obtaining a candidate taxi set corresponding to the request of each passenger;
performing many-to-many matching on all requests and all candidate taxis to obtain the optimal matching between the taxis and the requests, and generating an order according to a matching result;
for requests that are not accepted, add to the unmatched set; and the cloud server matches all the request loops of the unmatched set again until all the requests are accepted or no taxi can meet the request constraint.
Further, the request for the ride combination comprises a departure point, a destination point, a departure time, a latest arrival time and the number of passengers.
Further, the algorithm adopted when all requests and all candidate taxis are subjected to many-to-many matching and all requests in an unmatched set are circularly matched is a bipartite graph maximum matching algorithm.
Further, the time constraints of the on-board passenger and the requesting passenger are expressed as:
for passengers aboard the vehicle to disembark before a new passenger gets on the vehicle, i.e. e w ≤t r The time constraint is:
for passengers on board the vehicle after a new passenger disembarks, i.e. e w ≥e r The time constraint is:
for passengers on board the vehicle to disembark new passengers between boarding and disembarking, i.e. t r ≤e w ≤e r The time constraint is:
wherein e is w Indicating the latest arrival time, t, of the taxi r ,e r Respectively representing departure time and latest arrival time, o, set in the passenger's request r ,o w Respectively indicating the requested departure place and the current position of the taxi, d r ,d w Respectively indicating the destination point of the request, the taxi destination, c uv Representing travel time between locations u, v, u representing o r 、o w 、d r Or d w 。
Further, if the tolerance value of the passenger is recorded as θ, the maximum delay time that the passenger can accept at the starting point and the destination is represented as t' r =t r +θ,e′ w =e w +θ,e′ r =e r And + theta, substituting the time constraint condition formula with the time constraint condition formula, wherein the formula form is unchanged.
Further, the performing many-to-many matching on all requests and all candidate taxis includes:
all the requests R in the request set R are corresponding toSet of candidate taxis W C Converting the mapping relation M of the inner taxi w into a bipartite graph B; the two subsets of the bipartite graph are respectively a request set R and a taxi set W with an empty parking space, the sides of the bipartite graph B consist of mapping relations M between all requests R in the request set R and taxis W in the corresponding candidate set W, and the bipartite graph B consists of a left-column request set R, a right-column taxi set W and a mapping M connecting left and right vertexes;
and solving the optimal matching set by using a bipartite graph maximum matching algorithm of graph theory.
Further, all the requests R in the request set R and the corresponding candidate taxi sets W C Converting the mapping relation M of the inner taxi w into a bipartite graph B, comprising the following steps:
firstly, all request R pairs in a request set R are arranged on one side in a column mode, all taxis W in a taxi set W are arranged on the other side in a column mode, and then the request R and a candidate taxi set W are utilized c And (4) connecting the request r with the corresponding candidate taxi, and eliminating unconnected objects to obtain a bipartite graph B.
Further, the solving of the optimal matching set by using a bipartite graph maximum matching algorithm of graph theory includes:
the optimal matching set consists of edges, each match in the set represents that a new request r is accepted by a corresponding taxi w, a co-riding relationship can be formed, and a new order is generated; and updating the corresponding order and the taxi state according to each match in the optimal matching set.
The application has the following technical characteristics:
1. the vehicle ride-share rate is improved, and the vehicle utilization rate is further improved.
According to the method, the demand information is uploaded through the handheld terminal device, the demand information and the supply information are collected and processed through the cloud server, and the optimization model is established by taking the maximum matching as a target, so that the real-time supply and demand information can be utilized, the method based on the graph theory is used for optimizing and matching, the ride share rate of the vehicle is improved, and the utilization rate of the vehicle is greatly improved.
2. The calculation amount is less, the time complexity is lower, and the practical requirements can be met.
Firstly, the problem of the co-riding matching between people is converted into the problem of the co-riding matching between people and vehicles, and the calculation amount is reduced. For the ride-sharing matching between people, no passenger is supposed to be in the vehicle, three times of comparison is needed for judging the relative position situation of the starting point and the destination of two requests in time, four times of comparison is needed for judging whether two passengers can ride together or not after the relative position is determined, and twelve times of comparison is needed. For the carpooling matching of people and the vehicle, because passengers exist on the vehicle, three times of comparison are needed for judging the relative positions of the starting point and the destination of the two requests, and nine times of comparison are needed for judging whether two passengers can carpool or not after the relative positions are determined. The number of comparison times is reduced, i.e. the amount of calculation is reduced.
And secondly, the advantages of the bipartite graph maximum matching algorithm are inherited. The algorithm is a combined optimization algorithm for solving task allocation problems in polynomial time, and the specific expression of the scheme is to solve the optimal matching of taxi co-taking.
3. Has stronger compatibility and expansibility
It is proposed to use a bipartite graph to describe the relation of requests to vehicle objects. On one hand, the data structures such as an array (adjacency matrix) or a dictionary (adjacency list) of the computer can be efficiently used for storage and calculation, on the other hand, a plurality of mature graph theory algorithms are convenient to solve, and the expansibility is strong.
Drawings
FIG. 1 is a schematic flow chart of a matching method of the present application;
FIG. 2 is a schematic bipartite graph of a request and a vehicle;
FIG. 3 is a schematic diagram of an optimal match set;
FIG. 4 is a schematic diagram of a solution process of the Hungarian algorithm;
FIG. 5 is an exemplary graph of simulation data;
FIG. 6 is an exemplary diagram of the solution result of the primary driver-multiplier matching algorithm;
fig. 7 is a graph of the variation of the multiplying power.
Detailed Description
The method converts the co-taking matching problem among passengers into the co-taking matching problem of the passengers and a taxi, provides a bipartite graph for describing the relation between a request and a taxi object, and finally solves the optimal matching of the co-taking by using a bipartite graph maximum matching algorithm of the graph theory; the technical solution of the present application is further described in detail below.
Three objects of a road network, a request and a taxi are defined in the scheme.
The road network uses a directed graph G = (V, E) table, where V is a set of vertices (Vertex) in the road, E is a set of edges (Edge), and each Edge is marked as (u, V). The value of the edge represents the travel time, using c uv Where u, v are vertices.
Request r = < o r ,d r ,t r ,e r >。o r ,d r Respectively representing a request starting point (getting-on point) and a request destination (getting-off point), which are both a certain vertex on the road network G; t is t r ,e r Respectively representing the departure time and the latest arrival time set by the passenger.
Taxi w = < o w ,d w ,e w >。o w ,d w Respectively representing the current position of the taxi and the destination of the taxi (the departure point of a passenger on the taxi), which are all a certain vertex on the road network G; e.g. of the type w Indicating the latest arrival time of the taxi (the latest time at which the passenger on the arrival car gets off the taxi). When there is no passenger in the vehicle, then o w =d w 。
The following is an explanation of the basic variables:
a bipartite graph-based shared network driver and passenger matching method can be used for a dynamic co-passenger matching process of two passengers, and as shown in FIG. 1, the method comprises the following steps:
s1, when a passenger needs taxi sharing service, the passenger inputs a request for sharing in a taxi sharing reservation system through a handheld terminal device and uploads the request to a cloud server; wherein the request for the ride combination comprises a departure location, a destination location, a departure time, a latest arrival time, and a number of passengers.
S2, collecting requests for carpooling initiated by all passengers in a set area within a period of time, such as one minute, by the cloud server, and searching all taxi sets with vacant positions in the set area; screening taxis which meet the maximum passenger carrying number constraint of the taxis and simultaneously meet the time constraints of passengers on the taxis and passengers who send the requests according to the request of each passenger and the positions and passenger carrying information of all the taxis in the taxi set, and obtaining a candidate taxi set corresponding to the request of each passenger; i.e. a collection of taxis that will deliver the passenger to the destination before the latest arrival in between without overloading.
In this step, the size of the setting region may be set according to actual requirements; in the set area, the pool requests R of all passengers form a request set R, and in the area, all available taxis, i.e., taxis with empty spaces, form a set W. For each request r, screening out a corresponding candidate taxi set W from taxis with empty parking spaces by using a time constraint condition c Thereby obtaining the request r and the corresponding candidate taxi set W c Mapping M of interior taxis.
Wherein, the time constraint is as follows:
for a new request r and a taxi with a vacancy, the condition for being able to ride the same car is that a time constraint is met. The time constraint can be divided into three situations that a passenger on the vehicle gets off before a new passenger gets on the vehicle, a passenger on the vehicle gets off after the new passenger gets off the vehicle, and a passenger on the vehicle gets on or off between the new passenger and the vehicle, wherein the last two situations belong to the riding-in situation. Let c r ,c w Respectively indicating the arrival of taxis from the current positionThe travel time required for the passenger destination in the vehicle and the travel time from the newly requested origin to the destination, the time constraint may be expressed as follows:
for passengers aboard the vehicle to disembark before a new passenger gets on the vehicle, i.e. e w ≤t r The time constraint needs to be satisfied:
for passengers on board the vehicle after a new passenger disembarks, i.e. e w ≥e r The time constraint needs to be satisfied:
for passengers on board the vehicle to disembark new passengers between boarding and disembarking, i.e. t r ≤e w ≤e r The time constraint needs to be satisfied:
if the tolerance value of the passenger is recorded as theta, the maximum delay time that the passenger can accept at the starting point and the destination is represented. Let t' r =t r +θ,e′ w =e w +θ,e′ r =e r And + θ, substituting them into the above three time constraint condition formulas (formula 1, formula 2, formula 3), and keeping the formula form unchanged.
S3, after all the requests in the time interval are screened and selected to obtain a corresponding candidate taxi set, all the requests and all the candidate taxis are subjected to many-to-many matching by using a bipartite graph maximum matching algorithm in a graph theory to obtain the optimal matching of the taxis and the requests, and an order is generated according to a matching result; a successful match means that the request is accepted, generating a new order.
S3.1, all the requests R in the request set R and the corresponding candidate taxi sets W C Internal taxiThe mapping relation M of (a) is converted into a bipartite graph B, and the bipartite graph of the request and the taxi is shown in fig. 2.
Let G = (V, E) be an undirected graph, and if the vertex V can be divided into two mutually disjoint subsets (a, B), and the two vertices i and j associated with each edge (i, j) in the graph belong to the two different vertex sets (iinA, jinB), respectively, then the graph G is called a bipartite graph.
In the scheme, two subsets of the bipartite graph are a request set R and a taxi set W with empty parking spaces respectively, and the edges of the bipartite graph B are formed by all requests R in the request set R and corresponding candidate sets W C And the mapping relation M of the internal taxis. As shown in fig. 2, the bipartite graph B is composed of a request set R in the left column, a taxi set W in the right column, and a map M connecting the left and right vertices. Thus request r and set of candidate taxis W c The steps of bipartite graph conversion are described as follows:
according to the S2 of the scheme, all the requests r and corresponding candidate taxi sets W are obtained c And the mapping relation M formed by the inner taxis w. Firstly, all request R pairs in a request set R are arranged on one side in a column mode, all taxis W in a taxi set W are arranged on the other side in a column mode, and then the request R and a candidate taxi set W are utilized c And (4) connecting the request r with the corresponding candidate taxi, and eliminating unconnected objects to obtain a bipartite graph B.
And S3.2, solving an optimal matching set by using a bipartite graph maximum matching algorithm of graph theory. As shown in fig. 3, the optimal matching set is composed of edges, and each match in the set represents that a new request r is accepted by a corresponding taxi w, so that a co-taking relationship can be formed, and a new order is generated. And updating the corresponding order and the taxi state according to each match in the optimal matching set.
In graph theory, an edge points to a point as a match. The matched points cannot be matched again. The maximum match refers to the match of the maximum cardinality. Given a bipartite graph, the algorithm for solving the maximum matching of the bipartite graph has the time complexity of o (n) 2 ) Hungarian Algorithm of (Hungarian Algorithm), complexity ofHopkelov-kappak algorithm (Hopcroft-karp algorithm). The hungarian algorithm is exemplified below.
The hungarian algorithm has the following steps:
(1) Establishing a directed graph G which is divided into the left side and the right side of a bipartite graph
(2) The sequence number on the left side is preferably selected to be smaller for matching.
(3) If the target points of the two points on the left side conflict, the point with the small sequence number selects another possible target point for matching. If there are no other points that can be matched, the match fails.
(4) And (5) continuously repeating the steps (2) and (3) until the vertex on the left side is traversed, and returning to the matching set.
In the scheme, as shown in fig. 2, in the first step, a bipartite graph B composed of a request set R and a taxi set W is obtained by performing step (1) and step (2). And the second step is to match the requests 1, 2 and 3 in sequence according to the sequence number on the left side, as shown in steps (1) (2) (3) (4) of fig. 4, wherein step (3) shows the condition that two target nodes conflict, the processing method is to select another possible target point for matching by the point with the small sequence number, and the processed result is shown in step (4). And finally, obtaining an optimal matching set, namely, the request 1 is matched with the taxi B, the request 2 is matched with the taxi C, and the request 3 is matched with the taxi A.
S4, adding the requests which are not accepted into the unmatched set; the cloud server then recycles the bipartite graph maximum matching algorithm for all requests of the unmatched set until all requests are accepted or no taxis can meet the request constraints.
Requests that are not matched are added to the unmatched set. If the unmatched set in the time period is an empty set or all the requested candidate taxi sets in the unmatched set are empty, ending the algorithm, otherwise repeating S3 and S4.
Data simulation example:
the example uses a simulation experiment to test the algorithm, wherein the objects to be simulated include road network, taxi and request. The path planning algorithm adopted in the experiment is an A-x algorithm, and the selected evaluation index is the combined multiplication rate in a period of time. The specific simulation method is described as follows:
1. data simulation method
As shown in the simulation data example of fig. 5, the road network is represented in the form of a matrix, and the number of rows and columns is 10, which is 100 nodes and 180 edges.
For the request, the generation time of the request is randomly simulated by using a Poisson process, wherein the average time interval between two selected requests is 1 minute, and the starting point and the key point of the request are randomly selected from nodes in the whole road network without repetition. As shown in FIG. 5, the dots on the graph represent the start of the request. The requested cutoff time is selected from the minimum distance and 1,5 times the minimum distance.
As shown in fig. 5, the asterisks indicate taxis, and for taxis, 7 nodes are randomly selected from the road network as taxi positions.
2. Experimental procedure
In the first step, a 10 × 10 matrix road network is generated according to the first part of data simulation method, and 7 vehicles are initialized with request data for one day with an average time interval of 1 minute.
And step two, collecting request data by taking 10 minutes as a batch, dispatching the request by using the matching method of the application, updating the request information and the taxi information in real time, and iterating until no data exists or the day is finished.
And thirdly, collecting the total number of the combined multiplier and the total number of successful orders in every 10 minutes, calculating the combined multiplying rate, and drawing a combined multiplying rate change chart.
3. Results and analysis of the experiments
Fig. 6 shows the result of the solution of the one-time driver-multiplier matching algorithm. The figure has 8 vehicles and 15 requests, and the requests comprise 8 matched requests and 1 unmatched request. It should be noted that the reason that only 5 vehicles are shown in the figure is that the remaining two vehicles are located at the same point as the start of the request, the asterisk being covered by the matched request. The two requests shown in (7, 4) and (9, 4) are successively picked up by taxis at (9, 8), and the carpool is realized.
The evaluation index used in this experiment was the average of the two hour-by-hour multiplications during the day. The scale of the road network is 10, the average time interval for the generation of requests is 1 minute, and the total time length is one day. As shown in fig. 7, the average total multiplication rate in one day is 53.4%, the average total multiplication rate in every two hours is 47.8% at the lowest and 64.2% at the highest, and the minor fluctuation range with the variance of 0.00193 fluctuates around the average value, which shows that the total multiplication matching algorithm has the characteristic of strong robustness.
Claims (2)
1. A shared network driver and multiplier matching method based on bipartite graph is characterized by comprising the following steps:
a passenger inputs a request for carpooling in a taxi carpooling reservation system through a handheld terminal device and uploads the request to a cloud server;
the method comprises the steps that a cloud server collects requests for carpooling initiated by all passengers in a set area within a period of time, and searches all taxi sets with vacant positions in the set area; screening taxis which meet the maximum passenger carrying number constraint of the taxis and simultaneously meet the time constraints of passengers on the taxis and passengers sending the requests according to the request of each passenger and the positions and passenger carrying information of all the taxis in the taxi set, and obtaining a candidate taxi set corresponding to the request of each passenger;
performing many-to-many matching on all requests and all candidate taxis to obtain the optimal matching between the taxis and the requests, and generating an order according to a matching result;
for requests that are not accepted, add to the unmatched set; the cloud server matches all the request loops of the unmatched set until all the requests are accepted or no taxi can meet the request constraint;
the request for carpooling comprises a departure point, a destination point, a departure time, a latest arrival time and the number of passengers;
the algorithm adopted when performing many-to-many matching on all requests and all candidate taxis and performing matching on all request cycles of the unmatched sets is a bipartite graph maximum matching algorithm;
the time constraints for the on-board passenger and the requesting passenger are expressed as:
for passengers aboard the vehicle to disembark before a new passenger gets on the vehicle, i.e. e w ≤t r The time constraint is:
for passengers on board the vehicle after a new passenger disembarks, i.e. e w ≥e r The time constraint is:
for passengers on board the vehicle to disembark new passengers between boarding and disembarking, i.e. t r ≤e w ≤e r The time constraint is:
wherein e is w Indicating the latest arrival time, t, of the taxi r ,e r Respectively representing departure time and latest arrival time, o, set in the passenger's request r ,o w Respectively indicating the requested departure place and the current position of the taxi, d r ,d w Respectively indicating the destination point of the request, the taxi destination, c uv Representing travel time between locations u, v, u representing o r 、o w 、d r Or d w ;
The many-to-many matching of all requests and all candidate taxis comprises:
all the requests R in the request set R and the corresponding candidate taxi sets W C Converting the mapping relation M of the inner taxi w into a bipartite graph B; wherein the two subsets of the bipartite graph are request sets respectivelyR and a taxi set W with empty parking spaces, wherein the edges of the bipartite graph B are formed by all requests R in the request set R and corresponding candidate sets W C The mapping relation M of the inner taxis W is formed, and the bipartite graph B is formed by a request set R of a left column, a taxi set W of a right column and a mapping M connecting a left vertex and a right vertex;
solving an optimal matching set by using a bipartite graph maximum matching algorithm of graph theory;
all requests R in the request set R and the corresponding candidate taxi set W C Converting the mapping relation M of the inner taxi w into a bipartite graph B, which comprises the following steps:
firstly, all request R pairs in a request set R are arranged on one side in a column mode, all taxis W in a taxi set W are arranged on the other side in a column mode, and then the request R and a candidate taxi set W are utilized c The mapping relation M is obtained by connecting the request r with the corresponding candidate taxi and eliminating the unconnected object to obtain a bipartite graph B;
the solving of the optimal matching set by using the bipartite graph maximum matching algorithm of the graph theory comprises the following steps:
the optimal matching set consists of edges, each match in the set represents that a new request r is accepted by a corresponding taxi w, a co-riding relationship can be formed, and a new order is generated; and updating the corresponding order and the taxi state according to each match in the optimal matching set.
2. The bipartite graph-based shared network driver-ride matching method according to claim 1, wherein if the tolerance value of the passenger is recorded as θ, the maximum delay time that the passenger can accept at the start point and the destination is represented as t' r =t r +θ,e′ w =e w +θ,e′ r =e r And + theta, substituting the two into a time constraint condition formula, wherein the formula form is unchanged.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911257853.2A CN111126799B (en) | 2019-12-10 | 2019-12-10 | Shared network driver and crew matching method based on bipartite graph |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911257853.2A CN111126799B (en) | 2019-12-10 | 2019-12-10 | Shared network driver and crew matching method based on bipartite graph |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111126799A CN111126799A (en) | 2020-05-08 |
CN111126799B true CN111126799B (en) | 2023-04-11 |
Family
ID=70498033
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911257853.2A Active CN111126799B (en) | 2019-12-10 | 2019-12-10 | Shared network driver and crew matching method based on bipartite graph |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111126799B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11436556B2 (en) * | 2019-01-31 | 2022-09-06 | Roadie, Inc. | Method and system for delivery assignment and routing |
CN111738550B (en) * | 2020-05-21 | 2023-07-25 | 深圳赛安特技术服务有限公司 | Travel guest-building method, device, equipment and storage medium based on dynamic programming |
CN112949686B (en) * | 2021-01-29 | 2022-12-09 | 中国重汽集团济南动力有限公司 | Matching method based on optimal local distance |
CN113393026B (en) * | 2021-06-09 | 2022-07-26 | 广东工业大学 | Unmanned taxi transfer and path matching method |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107101643A (en) * | 2017-04-10 | 2017-08-29 | 浙江工业大学 | A kind of share-car matching process |
CN109657820A (en) * | 2018-10-23 | 2019-04-19 | 厦门大学 | A kind of taxi matching process reserved |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170167882A1 (en) * | 2014-08-04 | 2017-06-15 | Xerox Corporation | System and method for generating available ride-share paths in a transportation network |
-
2019
- 2019-12-10 CN CN201911257853.2A patent/CN111126799B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107101643A (en) * | 2017-04-10 | 2017-08-29 | 浙江工业大学 | A kind of share-car matching process |
CN109657820A (en) * | 2018-10-23 | 2019-04-19 | 厦门大学 | A kind of taxi matching process reserved |
Also Published As
Publication number | Publication date |
---|---|
CN111126799A (en) | 2020-05-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111126799B (en) | Shared network driver and crew matching method based on bipartite graph | |
CN106557829B (en) | Method and device for acquiring region with unmatched demand and transport capacity in vehicle service | |
CN107194128B (en) | Multi-mode public transport network design method based on central radiation type network framework | |
WO2017028333A1 (en) | Planning method for highway electric vehicle fast charging stations | |
TW201741993A (en) | System and method for determining routes of transportation service | |
CN110222786B (en) | Dynamic car pooling method and system based on travel information | |
CN110909434B (en) | Urban public transport trunk line network design method under low-carbon guidance | |
CN108256969B (en) | Public bicycle leasing point dispatching area dividing method | |
CN112230675B (en) | Unmanned aerial vehicle task allocation method considering operation environment and performance in collaborative search and rescue | |
CN112381472B (en) | Subway connection bus route optimization method and device and storage medium | |
CN105825303A (en) | Drop and pull transport task allocation method | |
CN113763700A (en) | Information processing method, information processing device, computer equipment and storage medium | |
CN115409388A (en) | Multi-vehicle type customized bus operation optimization method | |
Lowalekar et al. | Zone path construction (zac) based approaches for effective real-time ridesharing | |
CN115713207A (en) | Hybrid bus service combination optimization method | |
CN107316094A (en) | One kind commuting circuit method for digging and device | |
CN114358386A (en) | Double-trip-mode ride-sharing site generation method based on reserved trip demand | |
CN113240265B (en) | Urban space division method based on multi-mode traffic data | |
CN108564810B (en) | Parking space sharing system and method | |
CN114372830A (en) | Network taxi booking demand prediction method based on space-time multi-graph neural network | |
Wan et al. | Incremental recommendation system for large-scale taxi fleet in smart cities | |
CN112949987A (en) | Taxi dispatching and matching method, system, equipment and medium based on prediction | |
Huang et al. | Utility-aware batch-processing algorithms for dynamic carpooling based on double auction | |
CN113592246B (en) | PCA-TOPSIS model-based road construction period bus network evaluation method | |
CN115759660A (en) | Scheduling method, device, equipment and medium for unmanned vehicle |
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 |