CN113987730A - Large-scale bus trunk line automatic selection method based on land utilization - Google Patents

Abstract

The invention discloses a large-scale bus trunk line automatic selection method based on land utilization, which mainly comprises the following steps: s1 public transportation network simplification; s2, constructing a mathematical model; s3 solving the single-target planning; s4 dual-target planning solution; the automatic selection method quantitatively selects the route based on the existing public transport network scheme and the forecast origin-destination passenger flow matrix, so as to solve the problems of difficult scheme optimization and the like caused by high difficulty of a subjective design planning scheme, various station and line network structure forms, more influence factors and the like, and improve the scientificity and rationality of public transport route planning.

Description

Large-scale bus trunk line automatic selection method based on land utilization

Technical Field

The invention belongs to the technical field of public transport network planning, and particularly relates to a large-scale automatic bus trunk line selection method based on land utilization.

Background

Overview of public transportation route planning

The public transport line planning is a precondition for efficiently and reliably constructing an urban public transport system, and aims to guide reasonable setting and arrangement of urban public transport lines according to a judicious principle and guarantee functional applicability and operation efficiency of the lines. At present, a plurality of domestic cities carry out large-scale public transportation planning or construction, a scientific and quantitative public transportation line planning method is provided for planning workers, the workload of the planning workers is reduced, and the standard and the scientificity of the line planning work are improved.

Public transportation network planning contains two main research contents: the method comprises the steps of overall network design and route selection, wherein the route selection refers to the determination of public transportation routes on a network and the length and the trend of the public transportation routes based on an existing network.

Brief introduction to the traditional line selection theory research and practice

As shown in fig. 1, the theoretical research of the conventional line selection mainly includes three main directions, namely, based on the line topology structure, based on the line operation cost, and based on the line operation efficiency.

(1) Based on circuit topological structure

Under the given target (such as total line length, line coverage area, maximum flow and the like), a mathematical model (such as an integer programming model) is established to solve the optimal network scheme.

(2) Based on line operation cost and efficiency

New public transportation networks and lines are designed over existing networks or directly to minimize or maximize the efficiency of public transportation service operating over line solutions.

At present, the domestic public transport line planning practice adopts a qualitative and quantitative combined method, namely a series of alternative schemes for planning the public transport line are compiled by the qualitative method, then a quantitative index evaluation system is adopted to select a preferred scheme from the alternative schemes, and the preferred scheme is analyzed and adjusted, and the specific flow can be summarized as shown in fig. 2.

Problems existing in the traditional public transport line planning theory research and practice method

For the traditional public transport line planning theory research, the simultaneous selection of a large number of lines is rarely considered in line design, the setting of special forms such as transfer, complex lines and the like among the lines lacks scientific standards, the traditional research in China mainly adopts a subjective method, and the reasonability of a line scheme is questioned.

Similar to the existing problems of theoretical research, the existing public transportation route planning practice method has the following problems:

(1) the planning scale at the initial stage is small, and the prospect is lacked

The specific problems include:

1. the urban land utilization changes frequently, the urban updating speed is high, and the old route planning scheme has insufficient support for urban development;

2. the preset scale of the planned route lacks analysis in traffic quantification and does not consider urban expansion;

3. in the aspect of rail transit, the layout of the rails in the center area of the city is not deeply considered, the lines are only evenly distributed on each development main shaft of the city, and the coverage rate of the rails in the center urban area is low;

4. in the aspect of rail transit, due to the lack of stable rail network guidance, transfer conditions are not reserved basically, so that subsequent stations and first-stage projects mainly adopt channel transfer and the transfer is inconvenient;

5. the planning is not reserved with elasticity, and the selection of a wire network planning index is improper, so that the passenger flow congestion is serious in the peak period of the current situation;

6. the line network has single level, is lack of urban fast lines, and has no obvious difference from the skeleton line and the common line.

(2) Planning one stage, construction one stage, operation one stage, lack of integrity

The specific problems include:

1. the later-stage adjustment planning is optimized and adjusted on the basis of the original wire network, the defect compensation difficulty of the original wire network is extremely high, particularly in the aspect of conventional public transport, a new line is basically and directly added on the basis of the original urban wire network, and the adjustment in a certain degree scale is rarely carried out;

2. transfer conditions are not reserved, so that the transfer and connection project construction of a subsequent station and a previous project is difficult and transfer redundancy is insufficient;

3. the problems of line overlapping and compound line of an early line and a later line are not considered in the early line planning;

4. under the existing domestic conventional public transport and track construction operation mechanism and system, the possible problems of co-linearity, especially the problem of co-linearity of lines at different levels, are not comprehensively considered overall;

5. the stable public transport network is the basis and basis for construction planning and compilation, and the selection of second-stage engineering construction projects and the stability of a scheme are influenced due to the lag of the track network planning and compilation;

6. the mode of connecting the new city line and the central city area line is single, the mode of connecting the city new city line or the peripheral line and the central city line or the current main line network is single, and most of the connection modes are single line connection. The biggest problem brought by the connection transfer mode is that as the transfer points are positioned at the edge of a city, and a large number of employment posts are lacked at the periphery of the transfer station, most passengers need to select to transfer to other lines to enter a central city, which brings larger passenger flow and operation pressure to the transfer lines and the transfer stations;

7. the capacity of transfer stations is insufficient, so that network bottleneck is caused, and as part of the transfer stations are used as transportation hubs, sectional planning can cause a plurality of lines to be continuously added to a single hub transfer station, so that the transfer capacity of the transfer station is insufficient, and the transfer station becomes the network bottleneck.

Disclosure of Invention

In order to solve the problems of the traditional public transport network planning theory research and practice method, the technical difficulties are as follows:

1. the network scale is increased, the design difficulty of a circuit planning scheme is increased, and particularly, the early-stage network planning is mainly based on subjective design;

2. with the enlargement of the station scale, the more line combinations need to be considered, and the existing line planning scheme evaluation system is difficult to support;

3. the planning difficulty of the dense part of the station is high, the line selection difficulty is increased, and the line trend is difficult to determine;

4. when a plurality of lines are planned simultaneously, transfer stations and transfer forms are difficult to determine;

5. urban traffic track route planning schemes have many influence factors, so that junction schemes are difficult to stabilize, and related track schemes are also difficult to stabilize.

Aiming at the problems, the invention provides a large-scale automatic bus trunk line selection method based on land utilization, which quantitatively selects lines on the basis of the existing network based on the existing bus network scheme and the forecast origin-destination passenger flow matrix, so as to overcome the problems of difficult scheme optimization and the like caused by high difficulty of a subjective design planning scheme, various station and line network structure forms, more influence factors and the like, and improve the scientificity and rationality of public traffic line planning.

The technical scheme of the invention is as follows:

a large-scale bus trunk line automatic selection method based on land utilization comprises the following specific steps:

s1 public transportation network simplification: the public transport network comprises a rail transit network and a conventional public transport network, and for the rail transit network, the conventional rail transit network is simplified into a simplified network only comprising a starting and ending station and a transfer point station; for a conventional public transportation network, the conventional public transportation network is simplified into a simplified network only comprising the shortest path between a public transportation station and a connection station;

s2 constructs a mathematical model:

s2.1 is provided withThe collection of sites of the existing public transport network is

The set of edges of the existing public transport network is represented by a matrix X, and one element in the matrix is

The practical meaning is as follows:

the line number set which enables all the stations to be accessed is set as

Selecting decision variables of the line selection model as a matrix

A certain element in the matrix is

The practical meaning is as follows:

s2.2, constraining the model, wherein the constraining comprises: the method comprises the following steps of line existence constraint, line continuity constraint, site access constraint, line length constraint, line overlapping quantity constraint, line overlapping occurrence frequency constraint, line section operation energy constraint, angle constraint and loop line constraint;

s2.3, constructing an objective function: i.e. to maximize the direct passenger flow ratior _d And a one-time passenger transfer ratior _h And the value range is 0-100%:

in the formula (I), the compound is shown in the specification,

-a linekDirect passenger flow volume;

-a linek ₁ Tok ₂ The primary transfer passenger flow;

-all line pairs (k ₁ ,k ₂ ) The arrangement of (a);

T-total volume of passenger flow

，

As a stationiTo the sitejThe passenger flow volume of the room;

s3 single-objective planning solution: firstly, generating an initial feasible line scheme based on a greedy strategy to obtain an initial feasible line, and then combining a tabu algorithm with a special neighborhood construction and correction method to obtain a single target optimal line;

s3.1 initial line generation: generating an initial line with a better specified operation efficiency index based on a greedy strategy;

s3.2, taking the initial line scheme as the current line scheme and the optimal line scheme, calculating an operation efficiency index, and taking the result as a front target value and an optimal target value;

s3.3, constructing a neighborhood, repairing the neighborhood, and selecting an optimal line scheme from the constructed neighborhood to be used as a current line scheme;

s3.3.1 constructing a neighborhood;

the following fields are constructed:

1) circuit switching neighborhood: the two mutually crossed circuits are respectively switched with one part of the circuit of the two mutually crossed circuits, and the switching operation is only used for switching the components of the circuits, so that new circuits cannot be generated, the existing circuits cannot be deleted, and the total length of all the circuits cannot be changed;

2) line disassembly neighborhood: part of the longer line is disassembled and reconnected with other shorter lines;

3) line direct neighborhood: selecting two starting and ending point stations, taking the shortest path between the two points as a new route, and if the new route conflicts with part of the original route, generating an overlapped part which does not meet the constraint; for the original lines with conflicts, all overlapped parts are removed, and the overlapped parts of the original lines are replaced by the generated new lines;

4) replacing the neighborhood by multiple lines: trying to replace the position and the mode of the line overlapping to disassemble part of the longer line and reconnect the longer line with other shorter lines;

5) line merging neighborhood: the shorter lines in the four neighborhoods in the aforementioned 1) to 4) are merged into a longer line as much as possible under the condition that the constraint is met;

s3.3.2 neighborhood repair: if the adjacent domains contain lines which do not meet partial constraints, the same method as the method for constructing the line switching adjacent domains and the line merging domains is used for correction;

s3.4, calculating the operation efficiency index of the current line scheme, updating the current target value, updating the optimal target value and the optimal line scheme if the current target value is larger than the optimal target value, and updating a tabu table;

s3.5, if the iteration frequency does not reach the upper limit, returning to the step S3.3, otherwise, carrying out the next step;

s3.6 removing the ultra-short and redundant lines of the optimal line scheme, ending the process to obtain the line with the optimal target, and recording the optimal direct passenger flow ratio corresponding to the optimal liner _d And a one-time passenger transfer ratior _h ；

S4 dual target planning solution:

obtaining an optimal direct passenger flow ratio and a primary passenger flow ratio after the single-target planning is carried out, converting a double-target planning problem into a single-target planning problem by adopting an ideal point method, and carrying out primary single-target planning solution to obtain an optimal line;

the objective function corresponding to the single-target planning problem is converted into:

in the formula (I), the compound is shown in the specification,

-weight of direct passenger flow ratio;

-once weighting of passenger flow ratios;

-ignoringr _h After the target, the optimal obtained by solving as a single-target planning problemr _d A value;

-ignoringr _d After the target, the optimal obtained by solving as a single-target planning problemr _h The value is obtained.

Preferably, in the step S1 of simplifying the public transportation network, for the conventional public transportation network, the conventional public transportation network is simplified into a simplified network only including the shortest paths between the bus stops and the connection stops, the conventional bus stop only connected with one shortest path is regarded as the start stop, and the conventional bus stop connected with three or more shortest paths is regarded as the transfer stop.

Preferably, step S3.1 initial route generation: the specific steps of generating an initial line with a better specified operation efficiency index based on a greedy strategy are as follows;

s3.1.1 Generation of a Single line: starting from any starting and ending point station or transfer point station which is not passed by any line in the simplified network, continuously selecting the station which is not passed by the line for prolonging until the constraint is not met or the station is prolonged to another starting and ending point station; when the line is extended to a certain station, selecting other stations adjacent to the station, trying to extend the current line to the selected point, and continuing the line to a point with the maximum passenger flow increment under the condition of meeting the constraint;

s3.1.2 Generation of initial line: repeating the step S3.1.1 until no new lines can be generated;

s3.1.3 not processed by visited site: after a greedy strategy is used for generating feasible lines, for the possible sites which are not accessed yet, extra extremely short lines which do not meet the line length constraint are generated to cover the sites;

s3.1.4 line merge: any two lines in the initial line are merged into a longer line until no lines can be merged if the constraints are met.

Preferably, when the model is constrained in step S2.2, the line existence constraint, the line continuity constraint, the site access constraint, the line length constraint, the line overlapping number constraint, the line overlapping occurrence number constraint, the line section performance constraint, the angle constraint, and the loop constraint are defined as follows:

there is a constraint on the line: the layout and selection of the line are based on the existing network, and the premise of the existing line between any two sites is that a network path exists between the two sites;

line continuity constraint: the line must be continuous and cannot be broken into several parts;

site access constraints: all sites must be covered by at least one line;

and (3) line length constraint: a limit is made to the total length of the line, a first total length constraint being the actual total length of the line, expressed as total path length or total time; a second length constraint is that the number of sites covered by the line is kept within a certain range;

line overlap length constraint: if overlap occurs between the lines, the total length of these overlapping portions should be kept within a certain range;

constraint of the number of line overlaps: the number of lines in the overlapping portion between the lines is kept within a certain range;

constraint of the number of times of line overlapping: in the process of line selection, setting an upper limit for the times of line overlapping of a single line scheme;

and (3) line section operation energy constraint: when calculating the line passenger flow, the one-way maximum transportation capacity of the line section needs to be considered;

angle constraint: larger angles are not suitable to exist among three continuous stations on the line;

and (3) loop line constraint: except for the loop wire, the loop and the ringlet are avoided.

Preferably, in step S3.4:

the tabu table is set in two types: (1) a T-shaped cross exchange taboo table and (2) a cross-shaped cross exchange taboo table; when the T-shaped or cross-shaped cross-exchange neighborhood is selected as the current optimal line scheme, the corresponding position of the corresponding taboo table is updated and a taboo algebra is added;

for a neighborhood within a tabu algebra, if the target value of the neighborhood is better than the target value of the currently feasible line plan, the neighborhood directly disregards the tabu table.

Compared with the prior art, the invention adopting the technical scheme has the following beneficial effects:

the automatic selection method of the public transport line can be directly applied to line selection of rail transit or public electric vehicle traffic, guides and reduces construction cost, increases operation efficiency and improves scientificity of public transport line planning.

Taking a certain actual rail transit network as an example, corresponding to the target planning and solving step of the bus route selection model, the optimal route scheme has the following results:

1. solving single-target planning: the direct passenger flow proportion of the initial route scheme is 0.3673, and the one-time transfer passenger flow proportion is 0.3519, which have reached a better level. The optimal direct passenger flow proportion and the optimal one-time transfer passenger flow proportion obtained by the algorithm are 0.4315 and 0.3972 respectively, which are far higher than 0.4284 and 0.3268 of the original route scheme.

2. Solving by multi-target planning: after the multi-target planning is converted into the single-target planning problem by using an ideal point method, the optimal direct passenger flow proportion and the optimal one-time transfer passenger flow proportion are respectively 0.4300 and 0.3406 and are still higher than the operation efficiency index of the original route scheme.

Compared with the original route scheme, the automatic generation method of the wire network can increase the direct passenger flow proportion of about 0.4 percent and the transfer passenger flow proportion of 4.2 percent.

Drawings

Fig. 1 is a schematic diagram of the main contents of a conventional line selection theory.

Fig. 2 is a flow chart of a conventional public transportation line planning.

Fig. 3 is a schematic view of a public transportation line selection model.

Fig. 4 is a schematic diagram of a public transport network architecture extraction.

FIG. 5 is a schematic diagram of a possible solution expression method.

Fig. 6 is a schematic diagram of T-type cross-line switching.

Fig. 7 is a schematic diagram of X-cross circuit switching.

Fig. 8 is a schematic diagram of line disassembly (in station number sequence).

FIG. 9 is a schematic diagram of elimination of overlapping portions when constructing a direct neighborhood.

Fig. 10 is a schematic diagram of the exchange of two complex line parts of the network.

Fig. 11 is a schematic diagram of six neighborhoods constructed by the tabu algorithm.

FIG. 12 is a schematic diagram of a single-objective planning solution flow.

Fig. 13 is a flow chart of a dual target planning solution algorithm.

Fig. 14 is a schematic illustration of a mass transit network scheme used in an example embodiment.

FIG. 15 is a schematic diagram of an initial wiring scheme constructed based on a greedy strategy.

Fig. 16 is a schematic diagram of the tabu algorithm convergence process.

Fig. 17 is a schematic diagram of a tabu algorithm search route scheme process.

Fig. 18 is a schematic diagram of an optimal route plan obtained by solving the dual-target planning.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples.

The invention discloses a large-scale bus trunk line automatic selection method based on land utilization, which can automatically select and arrange lines based on a public transport network and an origin-destination passenger flow matrix, wherein the origin-destination passenger flow matrix refers to a matrix for storing passenger flow in unit time among different stations.

Automatic bus trunk line selection method

General idea

The invention relates to a large-scale bus trunk line automatic selection method based on land utilization, which realizes the automatic selection of a large-scale bus trunk line through a model consisting of three main steps, as shown in figure 3, the method specifically comprises the following steps:

1. and (3) network simplification: for rail transit, the existing network scheme is simplified into a simplified network only comprising a starting and ending station and a transfer station; for conventional public transportation, a conventional public transportation network is simplified into a simplified network only comprising the shortest path between a public transportation station and a connection station, so that the calculation difficulty under the condition of a large-scale network is reduced;

2. generating an initial line scheme: generating a line scheme with a better specified operation efficiency index based on a Greedy Strategy (Greedy Strategy);

3. solving single-target planning: on the basis of an initial line scheme, a line scheme with optimal operation efficiency is obtained by using a Tabu Search algorithm, and the algorithm can automatically determine reasonable line forms such as line trend, length, transfer station position and the like according to an origin-destination passenger flow matrix and designated constraints;

4. and (3) solving through dual-objective planning: after the single-target planning solution obtains the optimal values which can be reached by two different operation efficiency indexes, the double-target planning problem is converted into the single-target planning problem by adopting an ideal point method, and the line scheme when the two indexes are both in the optimal values is solved.

The method aims to obtain a line scheme that the direct passenger flow of the whole line network and the one-time transfer passenger flow are minimized.

The automatic selection method of the large-scale public transport trunk line based on land use of the invention is described in detail below.

Network structure extraction (network simplification)

In the aspect of rail transit, the non-transfer stations and the non-start and stop stations in the existing network do not affect the trend of the line, so that the stations can be removed under the condition of not changing the main network characteristics of the generated network, a simplified network only containing the transfer stations and the start and stop stations is obtained, the calculation complexity during line selection is reduced, and the schematic diagram of structure extraction is shown in fig. 4.

In the aspect of conventional public transportation, because conventional public transportation directly uses a conventional public transportation network, road nodes of all unconventional public transportation stations can be directly removed, the conventional public transportation network is simplified into a network only containing shortest path roads between the conventional public transportation stations and stations, the conventional public transportation stations only connected with one shortest path are taken as starting and ending stations, and the conventional public transportation stations connected with three or more shortest paths are taken as transfer stations.

In a simplified network (the set of sites and edges is

) The common stations are removed and used as a part of the edge, so when the line selection is performed on the simplified network, the initially obtained line only includes the starting point station and the transfer point station, and the line passes through the common stations on the edge and is not included temporarily, so that the common stations need to be added to the station sequence corresponding to each line again after the line selection is completed.

Mathematical model representation

The set of the stations with the existing public transport network is

The set of edges of the existing public transport network is represented by a matrix X, one element of the matrixIs prepared from

The practical meaning is as follows:

the line number set which enables all the stations to be accessed is set as

The decision variable of the line selection model is a matrix

A certain element in the matrix is

The practical meaning is as follows:

as shown in fig. 5, since representing the feasible solution by a matrix may result in an ultra-large search space, when solving by using a tabu algorithm, the feasible solution needs to be equivalently represented by a site number sequence directly instead of the matrix to reduce the search space. Circuit scheme

The line in (1) is represented as

At the same time

。

The model requires a plurality of constraints for constraints (such as line continuity constraints, line length constraints, line overlap constraints, section operation energy constraints, loop constraints, etc.), which will be satisfied by the neighborhood construction process described below.

1. There is a constraint on the line: the routing and selection of routes is based on an existing network, and if a route exists between any two sites, the constraint can be expressed as:

2. line continuity constraint: the line must be continuous and cannot be broken into several parts, so the line must satisfy:

3. site access constraints: the individual sites must be covered by at least one line, and for each site it should satisfy:

4. and (3) line length constraint: setting the distance matrix between the stations as D if the stationsiAnd sitejThere is an adjacency between them, the elements of whichd _ij Representing the adjacency distance between the stations; if siteiAnd sitejThere is no adjacent relation between them,d _ij representing the linear distance between the stations; if it isi=jThen, thend _ij =0. A limit should be made to the total length of the line, and a first total length constraint is the actual total length of the line, which can be expressed as total path length or total time; a second length constraint is that the number of stations covered by a line should be kept within a certain range, namely:

in the formula (I), the compound is shown in the specification,

most total length of wireA small value;

-minimum number of arbitrary line sites;

-any total line length maximum;

-maximum number of arbitrary line sites.

5. Line overlap length constraint: line-to-line overlaps may occur, i.e. a plurality of consecutive edges in the network may be simultaneously occupied by a plurality of lines, these consecutive edges beingl’Should be kept within a certain range, it is necessary to satisfy:

in the formula (I), the compound is shown in the specification,

-obtaining a function of a set of consecutive edges;

the maximum number of sites allowed on any one continuous edge.

6. Constraint of the number of line overlaps: any edge of a line network can be provided with a plurality of lines, the number of the lines occupying the edge should be kept within a certain range, and the maximum number of the lines on any continuous edge occupied by the plurality of lines should meet the following requirements:

in the formula (I), the compound is shown in the specification,

-the maximum number of lines on any edge of the net.

7. Constraint of the number of times of line overlapping: in the process of line selection, the number of times of line overlapping occurrence should set an upper limit, so that the process of line selection should satisfy:

in the formula (I), the compound is shown in the specification,

-maximum number of line overlaps in the net.

8. And (3) line section operation energy constraint: when calculating the line passenger flow, the one-way maximum transportation capacity of the line section needs to be considered, namely:

in the formula:

-on-line

Upper adjacent stationiTojThe cross section passenger flow volume;

-a line

The section of the pipe has single-direction maximum transportation energy.

9. Angle constraint: for the line

Three consecutive stations

The following relationship should be satisfied:

in the formula (I), the compound is shown in the specification,ω-a line

The maximum steering angle of.

10. And (3) loop line constraint: in the process of line selection, the line may form a small loop, a U-shaped loop, and other undesirable forms, and in order to avoid such undesirable forms, i.e., the generation of sub-loops, the sub-loop constraint proposed by Miller C E is introduced:

according to the current operation conditions of various public transportation in China, the goal of public transportation line selection is to maximize the direct passenger flow ratior _d And a one-time passenger transfer ratior _h (the value range is 0-100%):

in the formula (I), the compound is shown in the specification,

-a linekDirect passenger flow volume;

-a linek ₁ Tok ₂ The primary transfer passenger flow;

-all line pairs (k ₁ ,k ₂ ) The arrangement of (a);

T-total volume of passenger flow

，

As a stationiTo the sitejThe passenger flow volume of the room.

To obtain

，

And

need to be according to the line scheme

And predicting the traffic of the origin-destination passenger flow matrix on the network, wherein the distribution method is based on the routes of the origin-destination points and adopts a capacity-limited multi-path distribution method (the predicted origin-destination matrix is equally divided into a plurality of parts, when each matrix is distributed, the proportion of the output quantity of each origin-destination point distributed to the corresponding route is calculated by using a Logit model based on the route length among the origin-destination points, and then the traffic is loaded on the corresponding route, if the traffic of a certain path does not meet the operation energy constraint of a line section, the path is moved out of the effective route until the distribution of a plurality of origin-destination point matrixes is completed).

Single-target planning solution

The premise of simultaneously maximizing the direct passenger flow and the one-time transfer passenger flow is to respectively solve a line scheme when one target is minimized. The specific operation is that firstly, an initial feasible line scheme is generated based on a greedy strategy to obtain an initial feasible line scheme with higher quality, and then a Tabu algorithm (Tabu Search) is used in combination with a special neighborhood construction and correction method to complete single-target planning solution to obtain an optimal line scheme of a single target.

(1) Greedy policy based initial line plan generation

The specific operation of generating a higher quality initial line plan is as follows

1. Generation using a single line

Starting from any starting and ending station or transfer station which is not passed by any line in the simplified network, continuously selecting the station which is not passed by the line for prolonging until the constraint is not met or the station is prolonged to another starting and ending station. When the line is extended to a certain station, selecting other stations adjacent to the station, trying to extend the current line to the selected point, and continuing the line to a point with the maximum passenger flow increment under the condition of meeting the constraint; in the case where no station satisfying the constraint is available for extension, the line ends up extending at the current station.

2. Generation of an initial routing plan

Repeating the generation of the single line until no new line satisfying the constraint can be generated;

3. non-visited site handling

After a greedy strategy is used for generating a feasible line scheme, for the possible sites which are not accessed yet, extra extremely short lines which do not meet the line length constraint are generated to cover the sites, and the generation method is the same as the step 1;

4. line merging

To reduce the number of shorter routes, an attempt is made to merge any two routes in the original route scheme, extending them into longer routes until no routes can be merged if the constraints are met.

(2) Neighborhood structure

The tabu algorithm will be derived from all current line solutions in each iteration

The available neighborhoods of (a) select the most optimal route scheme for the target, and thus it is necessary to provide sufficient diversity of neighborhoods to obtain the best route schemeA more targeted solution, for which the following neighborhoods are constructed:

1. circuit switched neighborhood

The two lines crossing each other are each made to exchange a part of their own line, and the exchange operation exchanges only the composition of the line without generating a new line, deleting an existing line, and changing the total length of all lines, as shown in fig. 6 and 7. The specific operation steps are as follows:

1) is provided with

；

2) Random slave

Selecting any two lines

；

3) According to

The line cross pattern of (2) is schematically switched according to fig. 6 and 7. Two crossed lines are disassembled by taking crossed sites as breakpoints, and then the crossed lines are subjected to other crossing modes except the existing crossing mode

Reconnecting to complete the switching operation, and if all the new lines generated by the switching operation do not meet the constraint, giving up the switching;

4) repeating the steps 2-3 for a specified number of times, if

Then, then

As one possible routing scheme for the neighborhood.

2. Neighborhood for line disassembly

A portion of the longer lines are disassembled and reconnected to other shorter lines as shown in fig. 8. The specific operation steps are as follows:

1) is provided with

；

2) Random slave

Selecting arbitrary line

；

3) Except for the first site and the last site, in

In randomly selecting a site

；

4) Will be provided with

At site

The disconnection mode is as shown in fig. 8, if two newly generated lines both satisfy the constraint, the two lines are used for replacing

Original line in

Otherwise, abandoning the disassembly;

5) repeating the steps 2-4 for a specified number of times, if

Then, then

As one possible routing scheme for the neighborhood.

3. Direct line neighborhood

In order to improve the quality of the line scheme, it is necessary to increase the number of long-distance lines in the scheme as much as possible, as shown in fig. 9. The specific operation steps are as follows:

1) is provided with

；

2) Randomly selecting two starting and ending point sites, taking the two points and the shortest path site sequence between the two points as a new line, and if the line meets the constraint, adding the line to the new line

；

3) The new line will and

part of the original lines collide to generate an overlapping part which does not meet the constraint. And for the original lines with conflicts, removing all overlapped parts, and replacing the original lines with the generated new lines. If the removing part can break the original line into two parts, two new lines are generated to replace the original line; and if the removed part is completely the same as the original line, deleting the original line. A schematic diagram of the elimination of overlapping parts is shown in fig. 9. If new line and

if the old lines are the same, no operation is performed.

4) If it is

Then, then

As one possible routing scheme for the neighborhood.

4. Multiple wire replacement neighborhood

Since the position of the line overlapping and the line where the line overlapping occurs are different from each other, which affects the objective function value, it is necessary to try to replace the position and manner of the line overlapping to find the optimal complex line form, as shown in fig. 10. The specific operation steps are as follows:

1) is provided with

；

2) Randomly selecting a complex line part according to the circuit scheme

Randomly selecting one of multiple overlapped lines on the complex line part

Will be

Sequence of sites on, obviously

Can be broken down into two shorter lines or reduced in length, adding shorter lines satisfying constraints to

And is removed therefrom

. If it is

Then the line is removed in between without adding any new line to

；

3) Randomly selecting two sites in the simplified network, wherein the two sites are adjacent in the simplified network, the shortest path site sequence between the two sites is used as a new line, if the new line meets three line overlapping constraints (length constraint, quantity constraint and occurrence number constraint),add the new line to

；

4) If it is

Then, then

As one possible routing scheme for the neighborhood.

5. The field of line merging

The main purpose of constructing the line merging neighborhood is to further improve the direct passenger flow ratio of the neighborhood, and shorter lines in the four neighborhoods are merged into longer lines as much as possible under the condition that the constraint is met. The specific operation steps are as follows:

1) is provided with

；

2) Random slave

Selecting any two connected lines with the same start and end station, and if the new line after the two lines are connected meets the constraint, then connecting the original two shorter lines

Is removed and a new line is added to

；

3) Repeat step 2 until no lines can be merged, if

Then, then

As one possible routing scheme for the neighborhood.

(3) Through the above five neighborhood construction methods, six neighborhoods in fig. 11 can be obtained.

Neighborhood repairing: the line solutions that may not satisfy partial constraints (e.g., too short a line length) in each neighborhood are modified by the same method as used to construct the cs neighborhood and the line merge neighborhood. The specific operation steps are as follows:

1) finding out an over-short route which does not meet the constraint of the route length from the route schemes in all neighborhoods;

2) if the too short route crosses any line meeting the constraint, exchanging two routes by using a method same as a line exchange neighborhood structure to generate two lines meeting the constraint for substitution, and otherwise, turning to the next step;

3) if the too short line is connected with any line which meets the constraint, combining the two lines by using the method which is the same as the line combination neighborhood structure to generate a line which meets the constraint for substitution, and if not, turning to the next step;

4) if the too short line can not be corrected by the two steps, the line scheme is abandoned.

5) And repeating the step 2-4 until the line schemes of all the neighborhoods are checked or corrected.

(4) Tabu list setup and scofflaw rules

The purpose of tabu table setting is to reduce the amount of calculation, prevent the algorithm from searching repeatedly, and set two types of tabu tables: (1) t-shape cross exchange tabu list

And (2) a cross-shaped cross-exchange tabu chart

The two tabu tables respectively correspond to two switching modes of the cross line switching neighborhood. For a site as a switching cross pointiWhen using a certain switching pattern

ObtainedWhen the circuit switched neighborhood is selected as the optimal solution,

can be added with contraindication algebra

. At the following stage

In the second iteration, scofflaw rules cannot be satisfied andfform is crossed withiThe two lines can no longer be switched.

According to

Is determined as

。

Scofflaw criteria is a common criteria in heuristic algorithms, i.e. for a neighborhood within a taboo algebra, if the target value of the neighborhood is better than the target value of the current feasible route scheme, the neighborhood can directly disregard the taboo table.

(5) Objective function (target value)

When single-target planning solution is carried out, the targets are respectively direct passenger flow ratior _d And a one-time passenger transfer ratior _h Specifically, the direct passenger flow ratio and the primary passenger flow ratio are respectively taken as targets, a taboo algorithm is used for solving a route scheme with the optimal target, and the direct passenger flow ratio and the primary passenger flow ratio corresponding to the optimal route scheme are recorded.

(6) Algorithm flow

The flow of solving the single-target programming is shown in fig. 12, and the flow is an improved flow of a tabu algorithm, and comprises the following steps:

1. generating an initial line scheme based on a greedy strategy;

2. the initial line scheme is used as the current line scheme and the optimal line scheme, the operation efficiency index is calculated, and the result is used as a previous target value and an optimal target value;

3. constructing a neighborhood, repairing the neighborhood, and selecting an optimal line scheme from the constructed neighborhood to be used as a current line scheme;

4. calculating the operation efficiency index of the current line scheme, updating the current target value, updating the optimal target value and the optimal line scheme if the current target value is larger than the optimal target value, and updating a tabu table;

5. if the iteration times of the algorithm do not reach the upper limit, returning to the step 3, otherwise, performing the next step;

6. and removing the ultra-short and redundant lines of the optimal line scheme and ending the process.

Dual target planning solution

After the single-target rule is solved, the optimal direct passenger flow ratio and the optimal one-time passenger flow ratio can be obtained, because the line selection model needs to maximize the direct passenger flow ratio and the one-time passenger flow ratio, the two-target planning problem belongs to the two-target planning problem, and the improvement of the two target values conflicts with each other, the two-target planning problem is converted into the single-target planning problem by adopting an ideal point method, and the corresponding target function is as follows:

in the formula (I), the compound is shown in the specification,

-weight of direct passenger flow ratio;

-once weighting of passenger flow ratios;

-ignoringr _h After the target, the optimal obtained by solving as a single-target planning problemr _d A value;

-ignoringr _d After the target, the optimal obtained by solving as a single-target planning problemr _h The value is obtained.

The objective function is such that the search path of the tabu algorithm always remains in the direction in which both objectives increase simultaneously.

As shown in fig. 13, a specific flow of the dual-target planning solution is to first perform single-target planning solution on two operation efficiency indexes serving as targets to obtain independent optimal values of the two targets. And then converting the two targets into a single target by using an ideal point method, and performing single-target planning solution once again to obtain an optimal line scheme. The invention will now be described in further detail with reference to the following examples:

example (b):

referring to fig. 14, considering a railway traffic network in a certain area, the number of stations is 228, which contains 15 lines, the minimum line length is 9 stations, and the maximum line length is 32 stations. The direct passenger flow ratio of the line network of the scheme is 0.4284, the one-time passenger flow ratio is 0.3268, and the total length of the line is 475.765 km. And selecting and comparing lines on the basis of the network of the current network.

Firstly, an initial feasible solution is constructed based on a greedy strategy, an obtained initial line scheme is shown in fig. 15, and a part of lines in the initial line scheme are too short and do not meet the line length constraint. During the iteration of the tabu algorithm, these ultra-short lines will be repaired to meet the line length constraint. The direct passenger flow ratio of the initial line plan is 0.3673, and the one-time passenger flow ratio is 0.3519, which shows that the quality of the initial feasible solution reaches a better level.

The optimal direct passenger flow ratio 0.4315 is obtained through a single-target solving algorithm, and the optimal one-time passenger flow ratio is 0.3972. Then in the known

Foundation of (2)And performing double-target planning solution.

FIG. 16 shows the convergence process of the tabu algorithm in the solution process of the dual-target planning, which has strong local search capability and can make the tabu algorithm have strong local search capability

Quickly converging to a local optimum. Fig. 17 shows the procedure of the tabu algorithm searching for the line solutions, and the quality of the feasible line solutions gradually approaches the ideal point (the thick point at the upper right corner in fig. 16).

Fig. 18 shows the optimal route scheme obtained by the tabu algorithm, with a direct passenger flow ratio of 0.4300, a one-time passenger change ratio of 0.3406, and a transfer coefficient of 1.7992. The optimal line plan has 14 lines, the total number of lines is basically the same as that of the present line network, and the sum of the lengths of the lines is 469.124km, which is slightly higher than the total length of the network, because part of the lines share the same part of the network, and a complex line is generated.

Compared with the prior art, the invention adopting the technical scheme has the following beneficial effects:

the automatic selection method of the public transport line can be directly applied to line selection of rail transit or public electric vehicle traffic, guides and reduces construction cost, increases operation efficiency and improves scientificity of public transport line planning.

Taking a certain actual rail transit network as an example, corresponding to the target planning and solving step of the bus route selection model, the optimal route scheme has the following results:

1) solving single-target planning: the direct passenger flow proportion of the initial route scheme is 0.3673, and the one-time transfer passenger flow proportion is 0.3519, which have reached a better level. The optimal direct passenger flow proportion and the optimal one-time transfer passenger flow proportion obtained by the algorithm are 0.4315 and 0.3972 respectively, which are far higher than 0.4284 and 0.3268 of the original route scheme.

2) Solving by multi-target planning: after the multi-target planning is converted into the single-target planning problem by using an ideal point method, the optimal direct passenger flow proportion and the optimal one-time transfer passenger flow proportion are respectively 0.4300 and 0.3406 and are still higher than the operation efficiency index of the original route scheme.

Compared with the original route scheme, the automatic generation method of the wire network can increase the direct passenger flow proportion of about 0.4 percent and the transfer passenger flow proportion of 4.2 percent.

It will be apparent to those skilled in the art that various modifications and improvements can be made to the embodiments of the present invention without departing from the inventive concept of the present application, which falls within the scope of the present application.

Claims (5)

1. A large-scale bus trunk line automatic selection method based on land utilization is characterized by comprising the following specific steps:

s1 public transportation network simplification: the public transport network comprises a rail transit network and a conventional public transport network, and for the rail transit network, the conventional rail transit network is simplified into a simplified network only comprising a starting and ending station and a transfer point station; for a conventional public transportation network, the conventional public transportation network is simplified into a simplified network only comprising the shortest path between a public transportation station and a connection station;

s2 constructs a mathematical model:

s2.1 set the set of stations of the existing public transport network as

The set of edges of the existing public transport network is represented by a matrix X, and one element in the matrix is

The practical meaning is as follows:

the line number set which enables all the stations to be accessed is set as

Selecting decision variables of the line selection model as a matrix

A certain element in the matrix is

The practical meaning is as follows:

s2.2, constraining the model, wherein the constraining comprises: the method comprises the following steps of line existence constraint, line continuity constraint, site access constraint, line length constraint, line overlapping quantity constraint, line overlapping occurrence frequency constraint, line section operation energy constraint, angle constraint and loop line constraint;

s2.3, constructing an objective function: i.e. to maximize the direct passenger flow ratior _d And a one-time passenger transfer ratior _h And the value range is 0-100%:

in the formula (I), the compound is shown in the specification,

-a linekDirect passenger flow volume;

-a linek ₁ Tok ₂ The primary transfer passenger flow;

-all line pairs (k ₁ ,k ₂ ) The arrangement of (a);

T-total volume of passenger flow

，

As a stationiTo the sitejThe passenger flow volume of the room;

s3 single-objective planning solution: firstly, generating an initial feasible line scheme based on a greedy strategy to obtain an initial feasible line, and then combining a tabu algorithm with a special neighborhood construction and correction method to obtain a single target optimal line;

s3.1 initial line generation: generating an initial line with a better specified operation efficiency index based on a greedy strategy;

s3.2, taking the initial line scheme as the current line scheme and the optimal line scheme, calculating an operation efficiency index, and taking the result as a front target value and an optimal target value;

s3.3, constructing a neighborhood, repairing the neighborhood, and selecting an optimal line scheme from the constructed neighborhood to be used as a current line scheme;

s3.3.1 constructing a neighborhood;

the following fields are constructed:

1) circuit switching neighborhood: the two mutually crossed circuits are respectively switched with one part of the circuit of the two mutually crossed circuits, and the switching operation is only used for switching the components of the circuits, so that new circuits cannot be generated, the existing circuits cannot be deleted, and the total length of all the circuits cannot be changed;

2) line disassembly neighborhood: part of the longer line is disassembled and reconnected with other shorter lines;

3) line direct neighborhood: selecting two starting and ending point stations, taking the shortest path between the two points as a new route, and if the new route conflicts with part of the original route, generating an overlapped part which does not meet the constraint; for the original lines with conflicts, all overlapped parts are removed, and the overlapped parts of the original lines are replaced by the generated new lines;

4) replacing the neighborhood by multiple lines: trying to replace the position and the mode of the line overlapping to disassemble part of the longer line and reconnect the longer line with other shorter lines;

5) line merging neighborhood: the shorter lines in the four neighborhoods in the aforementioned 1) to 4) are merged into a longer line as much as possible under the condition that the constraint is met;

s3.3.2 neighborhood repair: if the adjacent domains contain lines which do not meet partial constraints, the same method as the method for constructing the line switching adjacent domains and the line merging domains is used for correction;

s3.4, calculating the operation efficiency index of the current line scheme, updating the current target value, updating the optimal target value and the optimal line scheme if the current target value is larger than the optimal target value, and updating a tabu table;

s3.5, if the iteration frequency does not reach the upper limit, returning to the step S3.3, otherwise, carrying out the next step;

s3.6 removing the ultra-short and redundant lines of the optimal line scheme, ending the process to obtain the line with the optimal target, and recording the optimal direct passenger flow ratio corresponding to the optimal liner _d And a one-time passenger transfer ratior _h ；

S4 dual target planning solution:

obtaining an optimal direct passenger flow ratio and a primary passenger flow ratio after the single-target planning is carried out, converting a double-target planning problem into a single-target planning problem by adopting an ideal point method, and carrying out primary single-target planning solution to obtain an optimal line;

the objective function corresponding to the single-target planning problem is converted into:

in the formula (I), the compound is shown in the specification,

-weight of direct passenger flow ratio;

-once weighting of passenger flow ratios;

-ignoringr _h After the target, the optimal obtained by solving as a single-target planning problemr _d A value;

-ignoringr _d After the target, the optimal obtained by solving as a single-target planning problemr _h The value is obtained.

2. The land-utilization-based automatic selection method for large-scale bus trunk lines according to claim 1, wherein in the step S1 of public transportation network simplification, for a conventional public transportation network, the conventional public transportation network is simplified into a simplified network only including shortest paths between bus stops and connection stops, the conventional bus stop only connected with one shortest path is taken as a starting stop and a finishing stop, and the conventional bus stops connected with three or more shortest paths are taken as transfer stops.

3. The land-utilization-based automatic selection method for large-scale bus trunk lines according to claim 1, characterized in that step S3.1 is an initial line generation: the specific steps of generating an initial line with a better specified operation efficiency index based on a greedy strategy are as follows;

s3.1.1 Generation of a Single line: starting from any starting and ending point station or transfer point station which is not passed by any line in the simplified network, continuously selecting the station which is not passed by the line for prolonging until the constraint is not met or the station is prolonged to another starting and ending point station; when the line is extended to a certain station, selecting other stations adjacent to the station, trying to extend the current line to the selected point, and continuing the line to a point with the maximum passenger flow increment under the condition of meeting the constraint;

s3.1.2 Generation of initial line: repeating the step S3.1.1 until no new lines can be generated;

s3.1.3 not processed by visited site: after a greedy strategy is used for generating feasible lines, for the possible sites which are not accessed yet, extra extremely short lines which do not meet the line length constraint are generated to cover the sites;

s3.1.4 line merge: any two lines in the initial line are merged into a longer line until no lines can be merged if the constraints are met.

4. The land-utilization-based large-scale bus trunk line automatic selection method according to claim 3, wherein when the model is constrained in the step S2.2, line existence constraint, line continuity constraint, station access constraint, line length constraint, line overlapping number constraint, line overlapping occurrence frequency constraint, line section operation energy constraint, angle constraint and circular line constraint are defined as follows:

there is a constraint on the line: the layout and selection of the line are based on the existing network, and the premise of the existing line between any two sites is that a network path exists between the two sites;

line continuity constraint: the line must be continuous and cannot be broken into several parts;

site access constraints: all sites must be covered by at least one line;

and (3) line length constraint: a limit is made to the total length of the line, a first total length constraint being the actual total length of the line, expressed as total path length or total time; a second length constraint is that the number of sites covered by the line is kept within a certain range;

line overlap length constraint: if overlap occurs between the lines, the total length of these overlapping portions should be kept within a certain range;

constraint of the number of line overlaps: the number of lines in the overlapping portion between the lines is kept within a certain range;

constraint of the number of times of line overlapping: in the process of line selection, setting an upper limit for the times of line overlapping of a single line scheme;

and (3) line section operation energy constraint: when calculating the line passenger flow, the one-way maximum transportation capacity of the line section needs to be considered;

angle constraint: larger angles are not suitable to exist among three continuous stations on the line;

and (3) loop line constraint: except for the loop wire, the loop and the ringlet are avoided.

5. The land-utilization-based automatic selection method for large-scale public transportation trunk lines according to claim 4, characterized in that in step S3.4:

the tabu table is set in two types: (1) a T-shaped cross exchange taboo table and (2) a cross-shaped cross exchange taboo table; when the T-shaped or cross-shaped cross-exchange neighborhood is selected as the current optimal line scheme, the corresponding position of the corresponding taboo table is updated and a taboo algebra is added;

for a neighborhood within a tabu algebra, if the target value of the neighborhood is better than the target value of the currently feasible line plan, the neighborhood directly disregards the tabu table.

Images