CN111785025A

CN111785025A - Novel flexible bus dispatching method based on automatic driving module bus

Info

Publication number: CN111785025A
Application number: CN202010712991.1A
Authority: CN
Inventors: 马晓磊; 刘小寒; 李欣
Original assignee: Beihang University
Current assignee: Beihang University
Priority date: 2020-07-22
Filing date: 2020-07-22
Publication date: 2020-10-16
Anticipated expiration: 2040-07-22
Also published as: CN111785025B

Abstract

The invention discloses a novel flexible bus dispatching method based on an automatic driving module bus, which comprises the following steps: calculating a generalized cost matrix of non-checkpoint passenger requirements; constructing an assignment problem based on the generalized cost matrix and solving the assignment problem; constructing an initial solution for solving the scheduling and line scheme based on the solving result; selecting neighbors according to an initial solution of a current vehicle route scheme and by referring to a taboo expression rule and a solution banning strategy; further narrowing the neighbors by referring to the cost change matrix and a preset granularity threshold; solving a vehicle scheduling subproblem, searching an optimal neighbor, and updating a current solution, an optimal solution and a tabu table; it is checked whether the algorithm meets a termination condition. The invention aims to minimize the weighted sum of the passenger cost and the vehicle operation cost, and the flexible bus operation mode based on the scheduling method can reduce the vehicle operation cost and improve the service level of passengers.

Description

Novel flexible bus dispatching method based on automatic driving module bus

Technical Field

The invention relates to the technical field of public transport organization and scheduling, in particular to a novel flexible bus scheduling method based on an automatic driving module bus.

Background

At present, flexible public transport, also known as changeable-line public transport, is a public transport mode combining traditional fixed lines and demand response type public transport. In flexible public transport systems, vehicles depart from stations according to a departure schedule and access predetermined "check points" in sequence, but the vehicles are allowed to deviate from a fixed line in a section formed by any two adjacent "check points" to serve passengers outside the "check points" in a service mode of demand-responsive public transport. The check points are equivalent to stations in a traditional fixed bus line, and are generally selected to be places with intensive passenger flow or transfer centers.

Under ideal conditions, the flexible public transport not only takes the advantage of low cost of the traditional fixed line, but also can exert the characteristic of flexible service like demand response type public transport. As early as the 80 s of the 20 th century, the european and american countries successively proposed the concept of flexible public transportation and started to perform practical operation tests. Subsequently, many scholars developed relevant studies. Currently, flexible public transportation services are available in many areas of the united states. However, flexible public transportation service is difficult to popularize in China, and the reasons are three main points: (1) the space-time complexity of the demands of passengers in China increases the difficulty of implementing the service; (2) lack of support of intelligent transportation technology matched with the method; (3) and a scientific optimal scheduling method suitable for the method is lacked.

In recent years, the technology of automatic driving and mobile interconnection is rapidly developed, and technical conditions are provided for the revolution of urban traffic. Among them is the combination of vehicle communication with automated driving technology that has prompted the introduction of the concept of automated public transportation. One of the features of automated mass transit is the ability to accommodate fluctuations in passenger demand through variable vehicle capacity. This function may be implemented by an electrically driven autopilot module vehicle. The autopilot module car is designed by nextfurturetransfer corporation, located in the united states. The automatic driving module vehicle is mainly characterized by being exquisite and flexible, and capable of automatically splicing and disassembling a plurality of module vehicles. Therefore, the dilemma of popularization of the service in China can be overcome to a certain extent by introducing the automatic driving module vehicle into a flexible public transportation system and designing a scientific dispatching method.

Therefore, how to provide a novel flexible bus scheduling method based on an automatic driving module bus is a problem that needs to be solved urgently by technical personnel in the field.

Disclosure of Invention

In view of the above, the invention provides a novel flexible bus dispatching method based on an automatic driving module vehicle, and the method makes full use of the characteristics of variable capacity and high flexibility of the automatic driving module vehicle. Under the condition of knowing passenger demand data, a departure schedule, the number of module cars carried during departure and the driving routes of different module cars are automatically generated by means of a granularity taboo algorithm. The invention aims to minimize the weighted sum of the passenger cost and the vehicle operation cost, and the flexible bus operation mode based on the scheduling method can reduce the vehicle operation cost and improve the service level of passengers.

In order to achieve the purpose, the invention adopts the following technical scheme:

a novel flexible bus dispatching method based on an automatic driving module bus comprises the following steps:

step S1, calculating a generalized cost matrix of the requirements of the non-checkpoint passengers;

s2, constructing an assignment problem based on the generalized cost matrix and solving the assignment problem;

step S3, constructing an initial solution of the solution scheduling and circuit scheme based on the solution result;

step S4, selecting neighbors according to the initial solution of the current vehicle route scheme and referring to the taboo expression rule and the solution strategy;

step S5, further reducing the neighbors by referring to the cost change matrix and the preset granularity threshold;

s6, solving a vehicle scheduling subproblem, searching an optimal neighbor, and updating a current solution, an optimal solution and a tabu table;

and step S7, checking whether the algorithm meets the termination condition.

Preferably, the step S3 includes:

s3.1, constructing an initial line solution of the flexible bus;

and S3.2, solving the vehicle scheduling subproblem to obtain an initial solution of the scheduling and circuit scheme problem.

Preferably, the vehicle scheduling sub-problem includes solving the total number of the module vehicles carried during each departure from the site and how to perform recombination and decomposition after the module vehicles reach each checkpoint.

Preferably, the step S7 includes: if the result of the checking algorithm meets the termination condition, ending the process and outputting the result; and if the result of the checking algorithm does not meet the termination condition, returning to the step S4.

Preferably, the constructing of the assignment problem in step S2 includes two branches, and one branch constructs an initial route solution by solving the assignment problem, and then executes step S3; the other branch is to solve the corresponding dual problem and construct a cost variation matrix, and then to execute step S5.

Preferably, in the step S6, the vehicle scheduling sub-problem is solved by calculating an optimal scheduling scheme using a nested vehicle scheduling sub-algorithm, so as to obtain an initial solution of the original problem.

Preferably, the assignment problem in step S2 is solved directly by a commercial solver.

Preferably, the granularity threshold is set to 500 in step S5.

Preferably, the taboo step size of the taboo table in the step S4 is 10 steps.

Preferably, the termination condition is set such that the program run time exceeds 4 hours or the number of iterations exceeds 200.

According to the technical scheme, the invention discloses a novel flexible bus dispatching method based on an automatic driving module bus, a generalized cost matrix between non-check point demands is calculated according to the pre-known passenger demand information, and the space distance between two demand points and the distance between the corresponding preset time of each passenger are comprehensively considered for the generalized cost between the non-check point demands. Second, an assignment problem is constructed. The purpose of the build assignment problem is twofold: first, an initial routing solution is constructed inspired from the solution of the assignment problem; second, the dual problem of the assignment problem is solved to construct a cost variation matrix. An initial line solution and a cost variation matrix are constructed respectively according to the solution of the assignment problem and the dual problem thereof. The purpose of constructing the cost variation matrix is to provide for reducing the neighbor space in the tabu algorithm. And (3) calculating an optimal scheduling scheme by adopting a nested vehicle scheduling sub-algorithm to obtain an initial solution of the original problem in the known initial route solution. And selecting the neighbor according to the tabu table and the deblocking strategy, and further reducing the neighbor according to the cost change matrix and a preset granularity threshold. And calculating the optimal neighbor by adopting a nested vehicle scheduling sub-algorithm, updating the current solution, updating the current optimal solution and updating a tabu table. And if the algorithm termination strategy is met, outputting the optimal solution, otherwise, continuously executing the granularity tabu algorithm.

Compared with the prior art, the invention jointly optimizes the vehicle departure schedule, the number of the module vehicles carried during departure and the driving routes of different module vehicles; the granularity tabu algorithm is used for remarkably reducing the calculation time on the basis of ensuring the solving precision; the characteristics of variable capacity and high flexibility of the automatic driving module bus are fully utilized, and a technical method for organizing and scheduling flexible buses in China popularization is provided.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.

FIG. 1 is a flowchart of a process for solving a specific scheduling and routing scheme according to the present invention.

Fig. 2 is a first schematic diagram of an automatic driving module vehicle provided by the invention.

Fig. 3 is a schematic diagram of an automatic driving module vehicle provided by the invention.

Fig. 4 is a schematic diagram of a novel flexible bus operation mode provided by the invention.

Fig. 5 is a geographic information diagram of a novel flexible bus route stop provided by the invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The embodiment of the invention discloses a novel flexible bus scheduling method based on an automatic driving module bus, which comprises the following steps:

and step S7, checking whether the algorithm meets the termination condition.

In order to further optimize the above technical solution, step S3 includes:

s3.1, constructing an initial line solution of the flexible bus;

In order to further optimize the technical scheme, the vehicle scheduling sub-problem comprises the steps of solving the total number of the module vehicles carried during the departure of the site and how to recombine and decompose the module vehicles after the module vehicles reach each check point.

In order to further optimize the above technical solution, step S7 includes: if the result of the algorithm meets the termination condition, ending the process and outputting the result; if the result of the checking algorithm does not satisfy the termination condition, the process returns to step S4.

In order to further optimize the above technical solution, the step S2 of constructing an assignment problem includes two branches, and one branch is used for constructing an initial route solution by solving the assignment problem, and then executing step S3; the other branch is to solve the corresponding dual problem and construct a cost variation matrix, and then to execute step S5.

In order to further optimize the technical scheme, the vehicle dispatching subproblems are solved in the step S6, and the optimal dispatching scheme is calculated by using a nested vehicle dispatching subproblem so as to obtain an initial solution of the original problem.

In order to further optimize the above technical solution, the problem assigned in step S2 is directly solved by a commercial solver.

In order to further optimize the above technical solution, the granularity threshold is set to 500 in step S5.

In order to further optimize the above technical solution, the taboo step size of the taboo table in step S4 is 10 steps.

In order to further optimize the above technical solution, the termination condition is set such that the program run time exceeds 4 hours or the number of iterations exceeds 200.

The invention comprehensively uses the automatic driving module vehicle to replace the traditional large and medium-sized public transport vehicles. As shown in fig. 2 and 3, the capacity of the autopilot module vehicle is generally 5 to 10 people, and vehicle assembly and disassembly can be automatically realized under driving conditions.

The invention is suitable for vehicles which simultaneously meet the following four conditions:

1) an autonomous vehicle;

2) the mutual splicing and decomposition of the vehicles in the running state or the parking state are automatically realized without manual operation;

3) after different vehicles are spliced, passengers in the vehicle can freely transfer carriages in the vehicle;

4) all module vehicle capacity, performance in the motorcade are all unified.

The invention designs a novel flexible bus operation mode based on the automatic driving module vehicle by considering the advantages of the automatic driving module vehicle. The flexible bus operation mode of the present invention is described below with reference to fig. 4. The vehicles are connected with each other by five module vehicles from a station. When driving from checkpoint C1 to checkpoint C2, the three module vehicles are connected to each other and denoted APTV1, APTV1 being allowed to deviate from a fixed route to service passengers not at the checkpoint. At the same time, two modular cars are connected to each other and labeled APTV2, APTV2 serving only passengers at checkpoints along a fixed line. When both the APTV1 and the APTV2 reach the checkpoint C2, the module cars undergo a new round of recombination and decomposition, respectively an APTV3 consisting of three module cars and an APTV4 consisting of two module cars, which travel along a flexible route and a fixed route, respectively. When both the APTV3 and the APTV4 reach the checkpoint C3, the module cars are again recombined and decomposed, respectively an APTV5 consisting of one module car and an APTV6 consisting of four module cars, which travel along the flexible and fixed routes, respectively. Finally, in section S4, all module cars are grouped together to form APTV7 to service passengers at checkpoints.

Fig. 4 is explained below from the passenger perspective. Passenger P1 first rides APTV1 from position O1 to checkpoint C2. At checkpoint C2, passenger P1 is in-vehicle transferred APTV4 and transported from checkpoint C2 to checkpoint C3. When two automatic driving module cars are connected with each other, the glass door between the two automatic driving module cars is automatically opened so that passengers can freely transfer in the cars. At checkpoint C3, passenger P1 rides APTV5 and eventually arrives at destination D1. Passenger P2 rides APTV3 from destination O2 to checkpoint C3 and seamlessly transitions to APTV6 and subsequent APTV7 and eventually to destination D2 (i.e., checkpoint C5). Similarly, passenger P3 rides APTV4 at checkpoint C2 to checkpoint C3 and transfers APTV5 inside checkpoint C3 and is transported by the vehicle to destination D3. Passenger P4 does not need any transfer because the passenger's origin-destination points are checkpoints.

The algorithm flow steps for solving the scheduling and routing scheme are shown in fig. 1, and specifically are as follows:

and step S1, calculating a generalized cost matrix of the requirements of the non-checkpoint passengers.

The generalized costs incurred in the process from satisfying demand i to satisfying demand j are represented for the elements of the matrix. The number of the matrix elements is

The calculation formula is as follows:

wherein R ∪ { -1} represents R departure times and a virtual departure time { -1 }. K involved in the scheduling phase_NPNDPassengers who indicate neither the starting point nor the ending point are at the checkpoint; k_NPDA passenger whose starting point is not at the checkpoint but whose ending point is at the checkpoint; k_PNDRepresenting passengers with a starting point at the checkpoint but an ending point not at the checkpoint.

Indicating the time when the vehicle departed from checkpoint C1 for the r-th departure.

Representing the time elapsed from checkpoint C1 to the boarding point for passenger j.

Representing the time elapsed from the point of departure of passenger i to the return to the vehicle station.

Representing the waiting time required to service the vehicle in front of passenger j, which is determined by the difference between the actual arrival time of the vehicle and the time the passenger can start to receive service. j is a function of⁺Indicating the point where passenger j gets on. i.e. i^-Indicating the point at which passenger i gets off ξ indicates the time it takes for a person to get on or off, here taking the value of 2 seconds.

The number of the collection elements is

The time that the passenger spends from the point k to the point t is p_k,tThe calculation formula is as follows:

wherein FS is a checkpoint set. NFS is a non-checkpoint collection, including boarding and disembarking at non-checkpoints.

Representing the distance from location k to location c_kIs known in advance. In the same way, the method for preparing the composite material,_k,t、_i,i+1、

respectively representing slave points k, i, c_tAverage travel time to locations t, i +1, t. And delta represents the accumulation consumption time of different module vehicles at the check point. If k is a checkpoint, then c_kEqual to the checkpoint number value, otherwise c_kEqual to s (k) + 1. If t is a checkpoint, then c_tEqual to the checkpoint number value, otherwise c_tIs equal to s (t). s (k) and s (t) respectively represent the line section numbers of k and t. Route segment and checkpoint numbering rules referring to fig. 4, C1-C5 are checkpoint numbers and S1-S4 are segment numbers. Finally, the process is carried out in a batch,

values of at most three are possible: (i)⁺,j⁺,i^-,j^-)、(i⁺,i^-,j⁺,j^-)、(i⁺,j⁺,j^-,i^-)。(i⁺,j⁺,i^-,j^-) Indicating that the vehicle departed from the origin of passenger i, arrived first at the origin of passenger j and arrived againThe end of passenger i, and finally passenger j. In the same way, (i) can be obtained⁺,i^-,j⁺,j^-) And (i)⁺,j⁺,j^-,i^-) The meaning of (a). The generalized costs for the three cases are therefore:

wherein, tau_iThis value needs to be known in advance for the time when passenger i arrives at the checkpoint or when the passenger can start to receive service. Finally, the process is carried out in a batch,

the values are the average of the above possible occurrences. Finally, all can be calculated

A generalized cost matrix for the needs of non-checkpoint passengers has thus been found.

Step S2, constructing the following assignment problem based on the generalized cost matrix:

the assignment problem is a 0-1 integer programming problem. Wherein x is_ijIs a 0-1 decision changeAmount of the compound (A). If its value is 1, it indicates that the demands i to j are directly serviced by the vehicle in succession, otherwise it indicates that the demands i to j cannot be directly serviced by the vehicle in succession. Here, the first and second liquid crystal display panels are,

let mu let_iV and v_jA dual variable is divided that assigns a problem first row and second row constraint. Order to

The cost of the arc (i, j) descent. Solving the assignment problem and the dual problem thereof by using a commercial solver to obtain x_ijAnd U_ij. Wherein, is composed of U_ijIs composed of

The matrix is referred to as a cost variation matrix. The implications of this assignment problem are: knowing the generalized cost matrix of passenger demand, how to assign a passenger demand j to a passenger demand i before it is reached

The purpose of solving the assignment problem is to lay down the initial solution for the next step of constructing the solution scheduling and routing scheme problem.

And step S3, constructing an initial solution for solving the scheduling and routing scheme problem.

By using the solution results of the assigned problem in step S2, heuristics may be received to construct an initial solution to the solution scheduling and routing scheme problem. The construction of the initial solution is divided into the following two steps:

1) an initial route solution of the flexible bus is constructed first. X obtained in step S2_ijAnd performing end-to-end connection according to the value of 0-1, and finally obtaining an R +1 main path (corresponding to R +1 departure times) containing the station nodes and a plurality of sub-loops not containing the station nodes. Obviously, the main path means non-checkpoint requirements and requirement sequences for R +1 departure and corresponding service for each departure, while the sub-circle has no practical significance. Therefore, the demands in the sub-circles can be distributed one by one in a random distribution modeInto the main path. After all the sub-circles are distributed, the number of the dispatched cars needs to be preloaded to each shift without considering the requirement of the check points, and the reorganization and the decomposition are carried out after the module cars reach each check point. They are calculated as follows:

the first step is as follows:

the second step is that:

the third step:

wherein the content of the first and second substances,

and

respectively representing the number of vehicles that the shift r needs to allocate to the flexible line and the fixed line at the check point s.

And

defined as the maximum number of passengers in the shift r in the section s flexible route and fixed route, respectively.

Representing the number of shifts r preloaded departure.

2) And secondly, solving the vehicle scheduling subproblem to obtain an initial solution of the scheduling and route scheme problem. The vehicle scheduling sub-problem comprises solving the total number of the module vehicles carried during each time of departure at the site and how to recombine and decompose the module vehicles after the module vehicles reach each check point. The vehicle dispatch sub-problem algorithm pseudo-code is as follows:

input/checkInformation of ordering passengers, vehicle route,

Outputting a weighted sum c of the scheduling scheme, the passenger cost and the vehicle operation cost_min。

1 list of fees

2 list of feasible vehicle scheduling schemes

3 list of timetables

4 checkpoint passenger information list UPD not loaded into vehicle⁰And ← checking point passenger information.

5for r＝1:R+1do

6

7if r＝1,then

8

9if r>1,then

10for i＝1:|L^r-1|do

11

12 from

The excess nodes are subtracted.

13 will

Is inserted into L^r,C^r,Ti^r,UPD^rIn (1).

14c_min←min(C^R+1) And stores the corresponding timetable and scheduling scheme.

Function(s)

The definition is as follows:

15

16for j＝1:2C-1do

17 count the number of passengers on board the vehicle before loading the checkpoint passengers.

18if the above value exceeds

Total passenger capacity of vehicle then

19v _ num ← v _ num +1 and returns 15 lines.

20if the above value does not exceed

Total passenger capacity of vehicle

The checkpoint passengers are loaded 21 into the vehicle and the vehicle arrival time, departure time and the current accumulated passenger costs and vehicle operating costs are calculated.

22 insert vnum, current cumulative cost, vehicle arrival time, departure time, current unloaded passenger information into

In (1).

23

Lines 15-21 are executed 24.

And step S4, selecting the neighbor by referring to the rules of the tabu table and the solution strategy.

And determining the neighbors of the current vehicle route solution according to the current vehicle route solution and the arcs put into the tabu table. The arcs in the tabu table represent the connecting lines between two nodes reached by the vehicle in a continuous sequence, and the neighbors of the current route solution should not include the arcs in the tabu table. The tabu step size is set to 10 steps, i.e. a certain arc is continuously placed in the tabu table in 10 algorithm iterations, then it is removed from the tabu table in the 11 th iteration. In addition, if a certain arc does not meet the 10 iteration requirements, but its constructed neighbors update the historical minimum of the sought target, then it is still removed from the tabu table.

And step S5, further reducing the neighbors by referring to the cost change matrix and the preset granularity threshold value.

The neighbors of the current vehicle route solution are obtained by placing the demand j one step change after the demand i. In order to reduce the search range and improve the calculation efficiency, the neighbors obtained in step S4 need to be screened. Here, the granularity threshold is set to 500. Referring to the cost variation matrix, if U_ijGreater than 500, then the demand j cannot be placed behind the demand i by moving. Finally, the neighbors in step S4 are further reduced.

And step S6, updating.

And respectively calculating the weighted sum of the passenger cost and the vehicle operation cost of each neighbor based on the vehicle scheduling subproblem algorithm in the step S3, and taking the minimum value as the current solution. And updating the historical minimum value and the tabu table at the same time.

And step S7, checking whether the algorithm meets the termination condition.

The termination condition here is set to a program run time of more than 4 hours or a number of iterations of more than 200. If either of the two is satisfied, the calculation is terminated, otherwise, the process returns to step S4.

In the following embodiments of the invention, as shown in fig. 5, the line is a beijing XXX bus XXX line, and the line is located in a beijing XXX area and is a conventional common fixed line. The method is improved into a flexible bus route and a corresponding scheduling and route scheme is formulated through the embodiment. The passenger demand information can be obtained by mining bus card swiping data and mobile internet data, which are known information of the invention and are not further described here. The scheduling period selected in the embodiment is 15: 00-16: 00, this line has 13 checkpoints in total (7 unidirectional). The automatic module adopted here has the passenger capacity of 10 people and the departure interval of 10 minutes.

This step is calculated using the non-checkpoint passenger demand information table shown in table 1. Table 1 records the demand information for 18 non-checkpoint passengers. All the distance information in step S1 can be calculated from the coordinates of the start point and the end point of the non-checkpoint passenger. The coordinate unit in table 1 is meter, so the calculated distance unit needs to be converted into time unit. The average travel speed of the bus is 30 kilometers per hour, so that the average travel speed can be converted into travel time. The required time is the time in minutes that the passenger arrives at the checkpoint or the passenger can begin to receive service. Requirement type 1 represents K_PNDAnd type 2 represents K_NPDAnd type 3 represents K_NPND. The total number of departures was (90/10+1), i.e., 10. And finally, saving the calculated generalized cost matrix of 28 x 28.

Table 1: non-checkpoint passenger demand information

Step S2, the following assignment problem is constructed based on the generalized cost matrix.

And substituting the generalized cost matrix obtained in the step S1 into a coefficient serving as an assignment problem objective function, and directly calculating the original problem and the dual problem by using a commercial solver. A 28 x 28 cost variation matrix is finally obtained. Steps S3-S7 execute a granular tabu algorithm.

The program is written and run using MATALB as described in steps S3-S7 above. Wherein the weighted sum of the passenger cost and the vehicle operation cost is: w1 × C1+ w2 × C2+ w3 × C3+ w4 × N4. C1 represents the total travel time (unit: hour) of all vehicles; c2 represents the sum of all passenger travel times; c3 represents the sum of all passenger waiting times; n4 represents the total number of passengers not served during the scheduled time. Here, w1, w2, w3 and w4 take the values of 20, 15, 30 and 1000, respectively.

And obtaining a vehicle driving route scheme according to the calculation result, wherein the scheme is shown in the table 2. The first class of vehicles in turn arrives at the start and end of passenger 7, the start and end of passenger 10, and the start and end of passenger 13. The route for this shift is therefore (in conjunction with fig. 5 and passenger information table 2): checkpoint 1, checkpoint 2, checkpoint 3, passenger 7 starting point, checkpoint 4, checkpoint 5, checkpoint 6, checkpoint 7, checkpoint 8, checkpoint 9, checkpoint 10, checkpoint 11, passenger 13 starting point, checkpoint 12, passenger 13 ending point, checkpoint 13. The remaining shift routes may be analogized.

Table 2: vehicle driving route scheme

And obtaining a vehicle dispatching scheme according to the calculation result, wherein the scheme is shown in the table 3. Taking shift 1 as an example, the total number of module cars dispatched by the shift station is two. Row 1, column 1 of the table represents the scheduling of 1 shift at checkpoint 1. The first number to the left of the brackets represents the number of module cars assigned to the flexible line and the second number represents the number of module cars assigned to the fixed line.

Table 3: vehicle dispatching scheme

The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A novel flexible bus dispatching method based on an automatic driving module bus is characterized by comprising the following steps:

and step S7, checking whether the algorithm meets the termination condition.

2. The novel flexible bus dispatching method based on automatic driving module vehicle as claimed in claim 1, wherein the step S3 includes:

s3.1, constructing an initial line solution of the flexible bus;

3. The novel flexible bus dispatching method based on automatic driving module cars as claimed in claim 2, characterized in that the sub-problem of vehicle dispatching comprises solving the total number of module cars carried each time when the car is dispatched at the station and how to reorganize and decompose the module cars after reaching each check point.

4. The novel flexible bus dispatching method based on automatic driving module vehicle as claimed in claim 1, wherein the step S7 includes: if the result of the checking algorithm meets the termination condition, ending the process and outputting the result; and if the result of the checking algorithm does not meet the termination condition, returning to the step S4.

5. The novel flexible bus dispatching method based on automatic driving module vehicle as claimed in claim 1, wherein the construction of assignment problem in step S2 includes two branches, one branch is constructed by solving assignment problem, constructing initial route solution, and then executing step S3; the other branch is to solve the corresponding dual problem and construct a cost variation matrix, and then to execute step S5.

6. The novel flexible bus scheduling method based on the automatic driving module bus as claimed in claim 1, wherein the solving of the vehicle scheduling subproblem in the step S6 adopts a nested vehicle scheduling subproblem to calculate an optimal scheduling scheme so as to obtain an initial solution of an original problem.

7. The novel flexible bus dispatching method based on automatic driving module vehicle as claimed in claim 1, wherein the assignment problem in step S2 is solved directly by commercial solver.

8. The novel flexible bus dispatching method based on automatic driving module vehicle as claimed in claim 1, wherein the granularity threshold value in step S5 is set to 500.

9. The novel flexible bus dispatching method based on automatic driving module vehicle as claimed in claim 1, wherein the taboo step size of the taboo table in the step S4 is 10 steps.

10. The novel flexible bus dispatching method based on automatic driving module vehicle as claimed in claim 1, wherein the termination condition is set as program operation time exceeding 4 hours or iteration number exceeding 200.