A bus network modeling method for station demand-responsive service

Abstract

The present invention proposes a bus network modeling method for site demand responsive service. The method is aimed at introducing a bus network with site demand responsive service, and expresses the uncertain topological characteristics of the bus network in an abstract and modeled manner, including multi-layer network modeling of the bus network, setting length thresholds and site sets for each bus line; establishing a constraint programming model according to the line length thresholds and site sets, solving all possible path sets for each line, and constructing a random network model of the bus line and the bus network; designing edge weight parameters based on the service probability of the edges of the bus network, and constructing an aggregate network model; by defining the weighted adjacency matrix of the aggregate network of the random network, abstracting and modeling it, thereby laying a theoretical foundation for the planning, evaluation and analysis of the introduction of site demand responsive service into the urban bus system.

Description

Technical Field

[0001] The present invention relates to the technical field of network modeling, and in particular to a bus network modeling method for station demand responsive services. Background Art

[0002] For a long time, my country's conventional public bus transportation (hereinafter referred to as conventional public transportation) has provided operating services according to fixed routes, fixed stops, and fixed timetables. With the increase in citizens' demand for convenient and flexible public transportation services, demand-responsive public transportation services have flourished in various places. In 2020, the Ministry of Transport clearly put forward the overall requirements for improving the development system of new transportation formats and building a demand-responsive travel service system in the "Opinions on Several Issues Concerning the Promotion of Modernization of Transportation Governance System and Governance Capabilities".

[0003] Demand-responsive public transportation originated in Europe and the United States. With its flexible and changeable service model, it has unique service advantages under specific traffic needs. Station demand-responsive public transportation service is a type of demand-responsive public transportation service. It takes into account the regularity of conventional public transportation services and the flexibility of demand-responsive public transportation services. It has little impact on the existing operation and management system and is one of the main innovation directions of urban public transportation services in my country. Under this service model, public transportation vehicles travel along a baseline route with several fixed stations. Demand-responsive stations are arranged within a certain service radius around the baseline route. The stop demand depends on whether there are passengers getting on or off at these stations, that is, whether they are activated. The vehicle's driving route from the baseline route to the demand-responsive station and back to the baseline route is fixed.

[0004] Station demand-responsive bus services have been put into use in some foreign cities and have achieved good results. A large number of operational cases have shown that this service is an effective way to solve the problem of bus services in low travel density areas. The conclusion of the US Transportation Research Board's research report "A Guide for Planning and Operating Flexible Public Transportation Services" is that station demand-responsive bus services are suitable for use in towns and urban suburbs with low road network density and low-demand communities planned around the main roads. It can effectively reduce the waste of transportation resources caused by laying multiple lines, and achieve the purpose of reducing vehicle idle mileage, improving vehicle turnover efficiency, and maintaining high service accessibility. In my country, in addition to the above-mentioned scope of use, the last-mile bus service connecting to subway stations in the suburbs of large cities, the inter-city adjacent area bus service in the metropolitan area, and some urban tourist bus services all have station demand-responsive service needs. Therefore, the introduction of station demand-responsive bus services in the existing urban public transportation system is an inevitable trend for the development of refined and high-quality urban public transportation, and it also has broad application prospects that meet the actual needs of urban public transportation.

[0005] Establishing a bus network model that introduces station demand-responsive services and calculating and analyzing network performance based on it are the primary issues in designing and evaluating bus networks that introduce this service model. Unlike traditional conventional bus networks, bus networks that provide station demand-responsive services have uncertain route directions and stop settings. Reflected in the network structure, the former has a certain network topology, while the latter has an uncertain network topology, which poses a great challenge to the network modeling and analysis of buses. However, in the current network analysis research involving demand-responsive bus services, almost all the focus is on network planning, such as the layout of demand-responsive stations and the optimization of key service parameters of bus routes that provide demand-responsive services. Few studies have been conducted on the modeling of bus networks with uncertain topologies for station demand-responsive services. This technical gap further restricts the planning, design, and operation organization of public transportation systems that introduce station demand-responsive services in the future. Summary of the invention

[0006] The purpose of the present invention is to propose a method for abstracting and modeling a bus network with uncertain topology that can be oriented to station demand-responsive services, thereby laying a theoretical foundation for the planning, evaluation and analysis of introducing station demand-responsive services into urban bus systems.

[0007] To achieve the above object, the present invention proposes a bus network modeling method for station demand responsive service, which abstracts and models the uncertain topological characteristics of the bus network introduced with station demand responsive service, and includes the following steps:

[0008] Step 1: Conduct multi-layer network modeling for the bus network and set the length threshold and station set of each bus line;

[0009] Step 2: According to the line length threshold and the station set, a constraint programming model is established to solve all possible path sets and construct a random network model of bus lines and bus networks;

[0010] Step 3: Design edge weight parameters based on the service probability of the edges of the bus network and construct an aggregate network model.

[0011] Furthermore, the step 1 specifically includes:

[0012] 1.1: Establish an undirected multi-layer bus network, where each layer represents a bus route. The nodes in the layer represent bus stops, and the edges represent the intervals between stops on the route. The edges between layers represent the replication relationship of the same stop on different routes, which can be understood as passengers transferring at the same stop between different routes.

[0013] 1.2: Set a station set for each bus line, wherein the station set includes a fixed station set and a demand response station set. The fixed station defines the stations that must be served by the line, while the demand response station defines the stations that can be optionally served;

[0014] 1.3: Set a path length threshold for each bus line to constrain the maximum detour distance that the bus line can take when detouring to serve a demand response station.

[0015] Furthermore, the step 2 specifically includes:

[0016] 2.1: Establish decision variables of the constraint programming model, including variables indicating whether to pass through a certain station interval and whether to serve a certain demand response station;

[0017] 2.2: Establishing constraints for the constraint programming model, including constraints for the fixed first and last stations of the bus line, constraints for the fixed intermediate stations of the bus line, constraints for demand response stations, and constraints for path length restrictions;

[0018] 2.3: Solve the constraint programming model to obtain all solutions that satisfy the constraints. The solution space is the path set of the bus routes. The Cartesian product of all route solution spaces is the solution space of the entire bus network, that is, the random network model of the bus network.

[0019] Furthermore, the step 3 specifically includes:

[0020] 3.1: Taking all bus stops as point sets, construct a clustered random bus network;

[0021] 3.2: According to the possible path set of each bus line contained in the solution space of the bus random network model, the edge weight of the aggregate random bus network is set to the probability that the station interval is served by at least one line;

[0022] 3.3: Obtain the aggregate network of the undirected weighted random bus network, which is used to model the bus network that provides demand-responsive services.

[0023] Compared with the prior art, the advantages of the present invention are:

[0024] The present invention is oriented to site demand-responsive services, filling the gap in the existing technology in this field. By comprehensively using multi-layer network and random network models and defining the weighted adjacency matrix of the aggregate network of the random network, the uncertain topological bus network when oriented to site demand-responsive services is abstracted and modeled, and a generalized modeling method is provided for the bus network that introduces site demand-responsive services, thereby laying a theoretical foundation for the planning, evaluation and analysis of the introduction of site demand-responsive services into urban bus systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of a method flow chart of an embodiment of the present invention;

[0026] Figure 2 It is a schematic diagram of constructing an aggregate random bus network model based on a random network model in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be further described below.

[0028] This embodiment proposes a bus network modeling method for station demand responsive service. The method introduces a bus network with station demand responsive service and abstracts and models the uncertain topological characteristics of the bus network. Figure 1 As shown, it mainly includes the following steps:

[0029] S101: Perform multi-layer network modeling on the bus network, set the length threshold and station set of each bus line; in this embodiment, the bus network is an undirected multi-layer network, represented as Where Y = {α|α∈{1,2,…,||L||}} represents a set of layers, where each layer α represents a bus line and L is the set of bus lines; G = (G1,…,G α ,…,G ||L|| ) is the set of networks within the layer, where G α =(V α ,E α ) represents a node set, V α represents a bus stop, and the edge set E α represents the set of intervals between stations, and the bus stops at each layer are given by the same node set S, which represents all the bus stops in the network. That is, for any α∈Y, there is V α =S, which means there is a one-to-one mapping between nodes in different layers, called replicated nodes. It is a collection of bipartite networks that represent the interconnections between different layers, representing the same-stop transfer relationship between passengers between bus lines; in addition, these inter-layer connections only occur between replicated nodes, so the multi-layer bus network can be regarded as a multiple network.

[0030] Consider the general case of station demand-responsive bus service, where the bus stops associated with bus line l are Classified as a fixed site and demand-responsive sites Right now Fixed stations are stations that bus routes must visit under any circumstances, while demand-responsive stations provide services selectively only when there is a temporary demand. Without loss of generality, assume that each bus route has two fixed terminal stations, namely, The demand-responsive sites are all intermediate sites, that is, Further set the path length threshold of the bus line to h l , which constrains the maximum distance that a bus line can detour when it detours to serve demand response stations. In this case, the topological structure of each bus line is uncertain, that is, each random bus line has a set of possible paths, which leads to the randomness of the entire multi-layer network.

[0031] S102: According to the line length threshold and the station set, a constraint programming model is established for each bus line to solve all possible path sets of each bus line, and a random network model of the bus line and bus network is constructed based on the constraint programming model; in this embodiment, the random network defines the set of all possible networks, and the bus network set that provides station demand response services is composed of the path set of each bus line. The path set refers to the set of all possible paths of a bus line, which is mainly determined by two factors: the distribution of demand response stations and the length threshold of the bus line. Based on the constraint programming model established for each bus line, all possible path sets are solved.

[0032] Establishing decision variables of constraint programming model: defining 0-1 variables Indicates whether the bus line l∈L passes through the bus stop i∈V l and j∈V l The station interval between them; define 0-1 variables Indicates whether the bus line l∈L passes through a demand response station

[0033] Establish the constraint conditions of the constraint programming model: Based on the above decision variables, establish the constraint conditions defined by formulas (1)-(7):

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041] Among them, formulas (1)-(3) require that the path is a single path connecting two terminal stations, and set independent constraints on the degrees of nodes representing the first and last stations, fixed intermediate stations, and demand response stations: formula (1) requires that the node degree of the first and last stations is 1; formula (2) requires that the node degree of the fixed intermediate station is 2; formula (3) requires that if it passes through a demand response station, its node degree is 2, otherwise it is 0. Formula (4) ensures that the total path length of line l∈L does not exceed the given threshold h l Finally, formula (5) defines an undirected network, formula (6) prohibits the existence of self-loops, and formula (7) prohibits the existence of edges connecting to stations outside the set of stations on the bus line.

[0042] Solve the solution space specified by the above constraints, that is, define the path set of bus line l in is the set of all edges that make up all possible paths. Therefore, every possible path Corresponding to a solution G in the constraint programming model l,k (S,E l,k ),Right now in is the kth solution The solution space of the random network of the bus network is defined by the Cartesian product of the possible path sets of all bus lines. That is, each possible network is obtained by selecting a path combination from the path set of each bus line. Therefore, there is There are possible ways to form a bus network, such as Figure 2 shown.

[0043] S103: Design edge weight parameters based on the service probability of the edges of the bus network and construct an aggregate network model; for bus line l∈L, set its path set Each possible path in has an equal probability of occurrence, that is, when randomly observing bus route l, path E is observedl,k The probability is set to in yes Based on this, the aggregate network of the random bus network is defined as G A =(S,E A ), which is used to model the bus network that provides demand-responsive services. Its weighted adjacency matrix is Define the edge weight parameter based on the edge service probability: the probability that the station interval between nodes i∈S and j∈S is served by at least one line As shown in formula (8):

[0044]

[0045] in represents the probability of the bus line l∈L serving the station interval between nodes i∈S and j∈S, as shown in formula (9):

[0046]

[0047] Based on this, the aggregate network of the random network of buses for station demand responsive service is obtained. This network model takes into account the topological uncertainty of the bus routes that provide station demand responsive service. By defining the weighted adjacency matrix of the aggregate network of the random network, it is abstracted and modeled, thus providing a generalized modeling method for the bus network that introduces station demand responsive service.

[0048] The above is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any technician in the relevant technical field, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, which does not depart from the content of the technical solution of the present invention and still falls within the protection scope of the present invention.

Claims

1. A bus network modeling method for station demand responsive service, characterized in that, for introducing a bus network with station demand responsive service, the characteristics of the uncertain topology of the bus network are abstracted and modeled, and the bus network modeling method comprises the following steps: Step 1: Perform multi-layer network modeling on the bus network and set the length threshold and station set of each bus line; Step 1 specifically includes: 1.1: Establish an undirected multi-layer bus network, where each layer represents a bus route, the nodes in the layer represent bus stops, and the edges represent the intervals between stops on the route; the edges between layers represent the replication relationship of the same stop on different routes; 1.2: Setting a site set for each bus line, wherein the site set includes a fixed site set and a demand response site set; 1.3: Setting a path length threshold for each bus line to constrain the maximum detour distance that the bus line can take when detouring to serve a demand response station; Step 2: According to the line length threshold and the station set, a constraint programming model is established to solve all possible path sets and construct a random network model of bus lines and bus networks; Described step 2 specifically comprises: 2.1: Establish decision variables of constraint programming model; 2.2: Establishing constraints for the constraint programming model, including constraints for the fixed first and last stations of the bus line, constraints for the fixed intermediate stations of the bus line, constraints for demand response stations, and constraints for path length restrictions; 2.3: Solve the constraint programming model to obtain all solutions that satisfy the constraints. The solution space is the path set of the bus routes. The Cartesian product of all route solution spaces is the solution space of the entire bus network, that is, the random network model of the bus network. Step 3: Design edge weight parameters based on the service probability of the edges of the bus network and construct an aggregate network model; the service probability of the edge is the probability that the station interval between two stations in the bus random network is served by at least one bus line; the aggregate network model is a model that abstracts the bus random network into a probability weighted network.

2. The bus network modeling method for station demand responsive service according to claim 1 is characterized in that: The step 3 specifically includes: 3.1: Taking all bus stops as point sets, construct a clustered random bus network; 3.2: According to the possible path set of each bus line contained in the solution space of the bus random network model, the edge weight of the aggregate random bus network is set to the probability that the station interval is served by at least one line; 3.3: Obtain the aggregate network of the undirected weighted random bus network, which is used to model the bus network that provides demand-responsive services.

Images