Serviceability-based public transportation network wiring method and device

Abstract

This invention relates to the field of traffic data processing technology. To address the problem of prioritizing efficiency over service experience in existing public transportation network optimization technologies, this invention proposes a public transportation network layout method and apparatus based on serviceability. The method includes: acquiring station information; data modeling using a Space-L and Space-P dual network model; normalizing and designing serviceability evaluation indicators including direct access rate, average direct in-vehicle time, average direct waiting time, non-accessibility penalty, average number of transfers, and public transportation network redundancy, and establishing a multi-objective optimization model; and designing a public transportation network genetic algorithm based on serviceability to solve for the optimized layout scheme. This invention, by comprehensively considering multiple serviceability indicators to construct an optimization model, efficiently achieves multi-objective public transportation network layout planning, improving the serviceability and sustainable development of the public transportation system.

Description

Technical Field

[0001] This invention relates to the field of traffic data processing technology, and specifically to a public transportation network cabling method and apparatus based on serviceability. Background Technology

[0002] With the continuous aggravation of urban traffic congestion and the rapid development of subway networks in large and medium-sized cities, optimizing public transportation networks has become a key measure to improve the efficiency of public transportation systems. The public transportation network routing and optimization scheme in this invention relates to a public transportation system (a broad public transportation system) that includes at least bus and subway modes. It mainly consists of public transportation lines (or simply lines) and public transportation stops (or simply stops) located on these lines. The public transportation stops include at least bus stops and subway stations, and the public transportation lines include at least bus lines and subway lines.

[0003] Traditional public transportation network planning methods have many limitations when dealing with large-scale, multi-objective optimization. Heuristic methods relying on human experience are highly subjective and struggle to guarantee global optimality; mathematical programming methods, while capable of yielding exact solutions, often fail to solve complex, large-scale problems due to computational complexity; graph theory-based methods focus primarily on network topology but neglect actual passenger flow and transfer behavior; metaheuristic optimization algorithms, while improving solution quality, suffer from insufficient stability and reliance on initial solutions; comprehensive simulations, while providing realistic evaluation of solutions, are computationally expensive, inefficient, and lack flexibility. Overall, traditional methods are deficient in efficiency, solution capability, and robustness.

[0004] On the other hand, the serviceability of a public transportation system has a dual meaning: on the one hand, it is reflected in the ability of the public transportation network to meet passengers' needs for convenience, accessibility, and comfort; on the other hand, it represents the standardization and robustness of the network system. These factors together constitute the core elements influencing public travel willingness and the attractiveness of public transportation.

[0005] Genetic Algorithms (GAs), with their powerful global search capabilities, are suitable for optimizing public transportation network layouts. By simulating natural selection, these algorithms can find optimal route layout schemes in a complex solution space. However, despite the application of GAs in public transportation network optimization, existing research largely focuses on traditional optimization objectives, such as minimizing operating costs, maximizing route coverage, and minimizing average travel time. These objectives tend to be biased towards the operator's perspective, emphasizing resource allocation and operational efficiency while neglecting a serviceability evaluation system based on the actual travel experience of passengers. This leads to the following problems: while the optimized public transportation network improves coverage, the proportion of direct trips is low; some areas have route coverage, but too many transfers result in insufficient travel convenience; in areas with dense subway lines, public transportation routes have significant duplication and redundancy, leading to resource waste; and the overall optimization results tend to focus on improving operational efficiency rather than fully reflecting the service quality perceived by passengers.

[0006] In summary, to address the problem of prioritizing efficiency over service experience in existing public transportation network optimization technologies, it is necessary to propose a public transportation network layout method based on serviceability. This method can efficiently achieve large-scale, multi-objective, and multi-constraint public transportation network planning and optimization, while also comprehensively considering factors such as coverage, direct access ratio, convenience, resources, and operational efficiency, especially serviceability indicators. Summary of the Invention

[0007] In view of this, in order to solve the problems existing in the prior art, the present invention provides a public transportation network layout method and apparatus based on serviceability. The method takes the serviceability-related indicators of the public transportation network map as the comprehensive optimization objective, and comprehensively considers the redundancy of the public transportation network, the directness rate, the average direct in-vehicle time, the average direct waiting time, the non-accessibility penalty, and the average number of transfers. By proposing a public transportation network genetic algorithm based on serviceability and designing a global search and evolution mechanism, the method performs multi-objective optimization of the public transportation network, realizes the principle of public transportation network map optimization based on serviceability, and promotes the efficient and sustainable development of the transportation system while improving the serviceability of the public transportation system.

[0008] This invention provides a public transportation network cabling method based on serviceability, comprising:

[0009] Step 110: Obtain station information of the public transportation network; use Space-L and Space-P dual network models for data modeling to construct a multi-modal integrated public transportation network, which is used to characterize the physical structure of the public transportation network and reflect the transfer accessibility between stations;

[0010] Step 120: Based on the multi-modal integrated public transportation network, design normalized multi-dimensional serviceability evaluation indicators; the multi-dimensional serviceability evaluation indicators include at least the direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy.

[0011] Step 130: Using the multi-dimensional serviceability evaluation indicators, establish a multi-objective optimization model for public transportation network layout based on serviceability; obtain the objective function of the multi-objective optimization model for public transportation network layout by weighted summation of the multi-dimensional serviceability evaluation indicators;

[0012] Step 140: Design a public transport network genetic algorithm based on serviceability to solve the public transport network routing multi-objective optimization model and obtain an optimized public transport network routing scheme based on serviceability; use the objective function of the public transport network routing multi-objective optimization model as the fitness function of the public transport network genetic algorithm based on serviceability.

[0013] On the other hand, the present invention also provides a public transportation network cabling device based on serviceability, comprising:

[0014] The first module is used to obtain station information of the public transportation network; it uses Space-L and Space-P dual network models for data modeling to construct a multi-modal integrated public transportation network, which is used to characterize the physical structure of the public transportation network and reflect the transfer accessibility between stations;

[0015] The second module is used to design normalized multi-dimensional serviceability evaluation indicators based on a multi-modal integrated public transportation network. The multi-dimensional serviceability evaluation indicators include at least the direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy.

[0016] The third module is used to establish a multi-objective optimization model for public transportation network layout based on serviceability by utilizing the multi-dimensional serviceability evaluation indicators; and to obtain the objective function of the multi-objective optimization model for public transportation network layout by weighted summation of the multi-dimensional serviceability evaluation indicators.

[0017] The fourth module is used to design a public transport network genetic algorithm based on serviceability, solve the multi-objective optimization model of the public transport network layout, and obtain an optimized public transport network layout scheme based on serviceability; the objective function of the public transport network layout multi-objective optimization model is used as the fitness function of the public transport network genetic algorithm based on serviceability.

[0018] In summary, the present invention provides a public transportation network cabling method and apparatus based on serviceability. Compared with the prior art, the method and apparatus of the present invention have the following beneficial effects:

[0019] (1) Data modeling based on Space-L and Space-P dual network models is adopted to construct a multi-mode integrated transportation network, accurately depict the physical connection and transfer relationship between stations, and provide a reliable data foundation for public transportation network planning.

[0020] (2) By introducing indicators such as direct access rate, average direct in-vehicle time, average direct waiting time, average number of transfers, non-access penalty, and public transportation network redundancy, a target function is constructed by normalizing and weighting the summation. This allows for flexible balancing of different optimization objectives and adaptability to diverse urban transportation needs and policy orientations. The introduction of public transportation network redundancy indicators improves the system's fault tolerance and reliability in the event of partial line failures or surges in passenger flow, thereby enhancing the resilience of the public transportation system. The optimized wiring scheme reduces transfers, shortens waiting time, enhances directness and convenience, reduces travel complexity, improves the efficiency of the public transportation system, and also enhances the travel experience and accessibility service perception of passengers.

[0021] (3) By designing a public transport network genetic algorithm based on serviceability, the design of public transport network planning is applicable to the genetic algorithm framework, adapting to the optimization needs of large-scale and complex public transport networks, and having the ability to handle large-scale and multi-constraint network planning problems, ensuring the feasibility and scalability of the scheme. The model supports the integrated optimization of multiple modes of transportation such as buses, subways, and light rails, adapting to the development trend of modern urban multi-mode public transport systems. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the steps of a public transportation network cabling method based on serviceability in one embodiment of the present invention.

[0024] Figure 2This is a schematic diagram of a Space-L spatial modeling topology example in one embodiment of the present invention, wherein nodes 1, 3, 6, and 9 represent stations on line 1 of the public transportation network, nodes 2, 3, 4, and 5 represent stations on line 2 of the public transportation network, and nodes 7, 8, 9, and 10 represent stations on line 3 of the public transportation network.

[0025] Figure 3 This is a schematic diagram illustrating an example of the Space-L topology structure of a public transportation network in one embodiment of the present invention;

[0026] Figure 4 This is a schematic diagram of a Space-P spatial modeling topology example in one embodiment of the present invention, wherein nodes 1, 3, 6, and 9 represent stations on line 1 of the public transportation network, nodes 2, 3, 4, and 5 represent stations on line 2 of the public transportation network, and nodes 7, 8, 9, and 10 represent stations on line 3 of the public transportation network.

[0027] Figure 5 This is a schematic diagram comparing service areas based on public transportation routes and public transportation stops in one embodiment of the present invention, wherein... Figure 5 (a) is the service area based on public transportation lines in the prior art, specifically defined as the parallel dashed strip area of ​​the public transportation lines. Figure 5 (b) is the service area based on public transportation stations in this invention, specifically set as a disc-shaped area with each public transportation station as the center, and the six-star marked position is the non-service area;

[0028] Figure 6 This is a schematic diagram of the genetic algorithm framework for public transportation networks based on serviceability in one embodiment of the present invention;

[0029] Figure 7 This is a schematic diagram illustrating the change of the total target value during the public transportation network layout optimization process in the experiment of this invention;

[0030] Figure 8 This is a schematic diagram illustrating the changes in target values ​​of different indicators during the public transportation network layout optimization process in the experiment of this invention. Figure 8 (a) is the optimized curve of the direct access rate. Figure 8 (b) is the optimized curve of average direct travel time to the train interior. Figure 8 (c) is the optimized curve of average direct arrival waiting time. Figure 8 (d) is the optimization curve for the unreachability penalty. Figure 8 (e) is the optimized curve of average number of transfers. Figure 8 (f) is the optimization curve of public transportation network redundancy;

[0031] Figure 9 This is a schematic diagram of the optimized public transportation network layout used in the experiment of this invention, wherein... Figure 9 (a) shows the optimized public transportation network in a part of City C. Figure 9 (b) is Figure 9 (a) shows the optimized public transport network with map geographic information removed, retaining only public transport stops and routes;

[0032] Figure 10 This is a schematic diagram comparing the public transportation network before and after the optimization of the public transportation network layout in the experiment of this invention. Figure 10 (a) is the initially generated public transport network. Figure 10 (b) The optimized public transport network;

[0033] Figure 11 This is a schematic diagram of the public transportation network optimization curve using a traditional genetic algorithm in the experiment of this invention.

[0034] Figure 12 The diagram shows the target curves of the serviceability index using the public transportation network genetic algorithm based on serviceability and the traditional genetic algorithm, respectively, in the experiments of this invention.

[0035] Figure 13 This is a schematic diagram comparing the public transportation network optimization results in the experiment of this invention, wherein, Figure 13 (a) A visual schematic diagram of the public transport network layout for the initial population with a map background. Figure 13 (b) is a schematic diagram of the public transportation network layout with a map background, obtained by optimization using a traditional genetic algorithm. Figure 13 (c) is a schematic diagram of the public transportation network layout with a map background, obtained by optimizing the public transportation network based on the serviceability genetic algorithm of the present invention. Figure 13 (d) is Figure 13 (a) A schematic diagram of the public transport network layout, with only public transport stops and routes retained after removing map geographic information. Figure 13 (e) is Figure 13 (b) A public transport network layout diagram with map geographic information removed, retaining only public transport stops and routes. Figure 13 (f) is Figure 13 (c) A schematic diagram of the public transportation network layout with the map geographic information removed and only public transportation stations and routes retained. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0037] In one embodiment, such as Figure 1 As shown, the present invention provides a public transportation network cabling method based on serviceability, comprising:

[0038] Step 110: Obtain station information of the public transportation network; use Space-L and Space-P dual network models for data modeling to construct a multi-modal integrated public transportation network, which is used to characterize the physical structure of the public transportation network and reflect the transfer accessibility between stations;

[0039] Step 120: Based on the multi-modal integrated public transportation network, design normalized multi-dimensional serviceability evaluation indicators; the multi-dimensional serviceability evaluation indicators include at least the direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy.

[0040] Step 130: Using the multi-dimensional serviceability evaluation indicators, establish a multi-objective optimization model for public transportation network layout based on serviceability; obtain the objective function of the multi-objective optimization model for public transportation network layout by weighted summation of the multi-dimensional serviceability evaluation indicators;

[0041] Step 140: Design a public transport network genetic algorithm based on serviceability to solve the public transport network routing multi-objective optimization model and obtain an optimized public transport network routing scheme based on serviceability; use the objective function of the public transport network routing multi-objective optimization model as the fitness function of the public transport network genetic algorithm based on serviceability.

[0042] Specifically, step 110 includes:

[0043] Step 111: Obtain station information for the public transportation network and determine the stations and their latitude and longitude within the public transportation network to be optimized within the cabling area. The public transportation network stations include at least bus stops and subway stations; specific bus and subway station information can be obtained using station information provided by the public transportation service platform in conjunction with map application software, or provided by transportation companies. The data format of the station information is shown in Table 1 below, including station number, station name, station type, and the station's geographical longitude and latitude; the station types include at least bus and subway.

[0044] Table 1: Example of station information for public transportation networks

[0045]

[0046] Step 112: Using the station information of the public transportation network, construct a basic transportation network and obtain the travel time matrix of the public transportation network, including:

[0047] First, the shortest travel time between public transportation stations is obtained by using map application software interfaces or navigation data provided by transportation companies. This shortest travel time between stations is then used as the navigation time between stations. The map application software interface can be either Gaode Map API or Baidu Map PAI.

[0048] Using connectivity thresholds Perform edge constraint judgment on the distance between stations: if two stations The distance between them is greater than If two stations cannot be directly connected, then they are not suitable as adjacent public transportation stops; if two stations... The distance between them is less than or equal to Then, based on traffic navigation time Used as an edge weight to connect two sites;

[0049] Using stations as network nodes and edge weights, a basic transportation network is formed. , ; ; yes The node scale, that is, the station scale of the public transportation network; based on Construct a travel time matrix based on edge weights. , Each element in the array is a triplet. The travel time matrix The data is stored in the form of an edge_list for easy retrieval and updating during genetic algorithm calculations. The connectivity threshold... The settings and adjustments are based on the distance standards between adjacent public transportation stations.

[0050] In one embodiment, the connectivity threshold is taken as .

[0051] In one embodiment, connectivity thresholds for different areas are dynamically set based on the city density determined by population, building, and commercial facility density. .

[0052] Step 113: The basic transportation network is topologically modeled using both Space-L and Space-P dual-network models to obtain the necessary data, and a multi-modal integrated public transportation network is constructed, including:

[0053] A Space-L model is used to model the basic transportation network, constructing a public transportation Space-L network. Specifically, all stations (bus stops and subway stations) of the public transportation network are used as nodes, and adjacent stations on each existing public transportation line of the basic transportation network are connected. Stations on public transportation lines include bus stops and / or subway stations.

[0054] The Space-L model is used to accurately depict the physical structure (actual operational structure) of the public transportation network, providing a physical topological basis for subsequent route optimization. Therefore, the Space-L model is also known as the physical operational network, and can be used to calculate indicators such as average direct train travel time and out-of-reach penalties.

[0055] In the Space-L network of public transportation, a node represents a station in the public transportation network. If two stations are consecutive stops on a public transportation line, an undirected edge is established between the two stations. The edge can be unweighted (only indicating the connection relationship) or weighted, such as the number of public transportation lines passing through the station, the frequency of vehicle passage within a certain period, the transfer distance, or the passenger flow. Figure 2 The image shows an example of a three-line public transport Space-L network; Figure 3 The image shows a real-world example of a city's public transportation Space-L network.

[0056] The Space-P (Spatial Accessibility Path) model is used to model public transportation transfer relationships in the basic transportation network, constructing a public transportation Space-P network. For example... Figure 4 As shown, the Space-P model uses public transportation stations as nodes. As long as any two stations are on the same public transportation line, even if they are not adjacent stops, an edge is established between the two stations, thereby reflecting the possibility of reaching each other through the line.

[0057] Unlike the Space-L model, which only reflects physical adjacency, the Space-P model can characterize the path structure of potential transfers and reachable routes, making it more suitable for analyzing the convenience of transfers between routes and the accessibility of passenger travel.

[0058] Combining the two processes described above, the Space-L model and the Space-P model are applied to the basic transportation network to construct a multimodal integrated public transportation network, including a public transportation Space-L network. and the Space-P public transportation network In this invention, the Space-L model is used to calculate indicators related to physical connectivity, such as direct in-vehicle time and inaccessibility penalty, while the Space-P model is used to calculate indicators related to transfer convenience, such as directness rate and average number of transfers. The two models complement each other, jointly providing network modeling support for serviceability evaluation.

[0059] In summary, this invention, in the data modeling stage, combines public transportation station data with map navigation information to construct a travel time matrix and employs a dual-network model (Space-L and Space-P) to perform topological modeling of the basic transportation network. Unlike existing methods that rely on only a single model, this invention's data modeling approach can depict both the physical structure of the public transportation network and reflect the accessibility of transfers between stations, thus providing complete and efficient data support for subsequent serviceability evaluation and multi-objective optimization.

[0060] In step 120, based on the Space-L and Space-P network models, a multimodal integrated public transportation network is constructed, and normalized multi-dimensional serviceability evaluation indicators are designed, including:

[0061] (1) Direct access rate

[0062] In public transportation planning, the directness rate (or non-transfer rate) is a core indicator for measuring passenger travel convenience, representing the proportion of trips where passengers can reach their destination from their origin without transferring. By increasing the directness rate, this invention can significantly improve the passenger travel experience. In the Space-P public transportation network, any two public transportation stations... The shortest path reachability index of the route is It is given by the following formula:

[0063]

[0064] in, A value greater than 1 indicates the number of transfers when using public transportation. (The number of stops is...) Space-P public transportation network In this context, the total number of direct trips only considers the number of stations. to station Direct access, given by the following formula:

[0065] ;

[0066] Considering the direct access rate as one of the algorithm optimization objectives, the direct access rate is expressed as:

[0067] ;

[0068] This invention optimizes direct access rates, thereby reducing the average number of transfers for passengers and improving travel convenience.

[0069] (2) Average time to reach the vehicle

[0070] Average direct in-vehicle time (Avg.) refers to the weighted average of the travel time of direct passengers within the vehicle, reflecting network operational efficiency. It directly impacts passenger travel experience and system competitiveness, and is a key indicator for evaluating operational efficiency. In the Space-L network model of public transport networks, for each direct OD (Origin, Destination) station... Based on the current routes, search for all feasible direct routes. The direct origin-destination (OD) pair represents a pair of origin stations on the route that can be directly reached. and the final stop Calculate the travel time for each route (cumulate the travel time of adjacent stations along the route). Travel time , Sourced from travel time matrix (Calculated from actual traffic navigation travel time), then the number of stations is... Space-L public transportation network In the middle, the site to station Direct route between The shortest travel time is The shortest travel time is taken as the representative value for that OD station pair. For all direct OD station pairs... The average of the summation of the shortest travel times is:

[0071]

[0072] in, This indicates the total number of direct OD (Original Destination) station pairs. This invention, by optimizing the average direct travel time within the vehicle, can shorten the actual travel time for passengers and enhance the competitiveness of public transportation systems compared to efficient modes of transportation such as subways.

[0073] (3) Average waiting time for direct routes

[0074] Average direct wait time (Avg.) refers to the average waiting time for passengers at the originating station, reflecting the quality of service frequency. Waiting time is a crucial factor in passengers' perceived service quality and directly influences their travel choices. The waiting time at any station is calculated as the departure interval, which is determined by the reciprocal of the departure frequency. Assuming passengers and buses arrive randomly, and passengers prioritize routes with lower travel time costs, when multiple alternative public transportation routes meet the requirements, the passenger waiting time... The calculation formula is:

[0075]

[0076] In the above formula, Indicates direct route The frequency of departures, Indicates a set of direct routes. To characterize the constant of uniform passenger arrival at the station, when The statement assumes that passengers arrive at the bus stop at a uniform rate. For all direct OD (Original Direction) stops... The average of the summed waiting times is:

[0077]

[0078] in, This indicates the number of direct OD (Original Departure Point) stations. This invention effectively reduces waiting time uncertainty and improves passenger satisfaction by optimizing waiting times.

[0079] (4) Unattainable penalty

[0080] Considering residents' travel needs, a threshold of 2 transfers is defined. In the Space-L model of the public transport network composed of lines, any two stations... If the destination is unreachable or requires more than 2 transfers, then the two stations... Unreachable penalty If the number of arrivals or transfers is less than or equal to 2, then... The value is 0. The number of stations is... Space-L public transportation network In this context, the objective function corresponding to the unreachable penalty is:

[0081]

[0082] This invention avoids large-scale travel blind spots and ensures fair accessibility of the network by introducing an inaccessibility penalty.

[0083] (5) Average number of transfers

[0084] Average number of transfers (Avg. Transfers) is the number of transfers between any two stations in a public transport network. The average minimum number of transfers required to complete a trip reflects network connectivity efficiency. The number of transfers is the gold standard for measuring travel convenience and directly impacts passenger willingness to travel. In the Space-P network model of a public transport network composed of lines, the average minimum number of transfers required to complete a trip reflects network connectivity efficiency. The shortest path reachability index of the route is , >1 indicates that a transfer is required, and the number of transfers is [number missing]. -1, then the average number of transfers is:

[0085]

[0086] in, This indicates the number of direct OD (Original Destination) station pairs and the number of OD station pairs accessible by transfer. This invention reduces passenger travel complexity by limiting the average number of transfers, and is particularly effective for the elderly and commuters.

[0087] (6) Redundancy of public transportation network

[0088] Public transportation network redundancy measures the degree of redundant investment in network resources. Reasonable redundancy can increase system robustness, but excessive redundancy means resource waste and low operational efficiency. Especially with the increase in subway lines, the redundancy of public transportation networks increases dramatically. This invention is the first in the field of public transportation network optimization to use redundancy as an optimization objective.

[0089] Existing technologies consider the line density of public transport networks to quantify the redundancy of public transport lines. However, based on the service area of ​​a public transport line, it cannot accurately represent the serviceability of that line to residents. In fact, the integral amount based on the coverage area of ​​public transport stations provides a more accurate characterization of the redundancy of the public transport network. For example... Figure 5 As shown, Figure 5 The strip-shaped area in (a) is the service area based on public transportation lines in the prior art. Figure 5 (b) The circular area is the service area based on public transportation stations in this invention. As can be seen from the hexagonal star mark in the icon, the location between two stations of a public transportation line (the non-service area) is not actually served by that line. Service redundancy based on network density is not accurate in actual transportation network evaluation. Therefore, the public transportation network redundancy in this invention is developed based on the coverage rate of public transportation stations.

[0090] The process of defining redundancy in public transportation networks includes:

[0091] For each bus stop The number of public transportation lines passing through this bus stop is Covering radius around it Within its coverage area, the site adds value to the surrounding services. for: For each subway station The number of public transportation lines passing through this subway station is Covering radius around it Within its coverage area, this site adds value to the surrounding services. for: (Weight 2 reflects the higher transport and service capacity of the subway.) After calculating the service added value for each station, all service added values ​​are summed up according to spatial location to obtain the service value of the entire public transport network. The distribution of .

[0092] In one embodiment, for bus stops Take the coverage radius 800 meters; for subway station Take the coverage radius 1500 meters.

[0093] In one embodiment, the service value of a portion of a city's public transportation coverage area is calculated based on the service area of ​​public transportation stations and then visualized. The service values ​​of different areas are different, thereby enabling the determination of redundancy.

[0094] Based on the travel needs of urban residents and the standards of urban rail network planning, a reasonable service saturation threshold should be set. Assuming the urban public transport network is a connected graph, geographical locations exceeding this threshold are considered redundant service points. Integrating the areas exceeding the threshold, the total redundancy is calculated as follows:

[0095]

[0096] in, This is a pre-set service saturation threshold.

[0097] Service Area Internal redundancy: .

[0098] The redundancy of the public transportation network is defined by the following formula:

[0099]

[0100] This invention proposes for the first time an algorithm for calculating the redundancy of public transportation networks based on the integral redundancy quantification method in the above formula, as shown in Algorithm 1 in Table 2. This algorithm can effectively reduce the waste of resources caused by duplicate routes while ensuring system robustness.

[0101] Table 2: Public Transportation Network Redundancy Algorithm

[0102]

[0103] In summary, this invention proposes a multi-dimensional evaluation system centered on serviceability for public transportation network optimization. For the first time, it systematically incorporates six indicators—normalized direct access rate, average direct onboard time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy—into the optimization objective. Direct access rate and average number of transfers measure passenger travel convenience; average direct onboard time and average waiting time reflect passenger time efficiency; non-access penalty ensures network connectivity and fairness; and public transportation network redundancy innovatively suppresses resource waste and improves system robustness. Unlike existing optimization methods that only focus on operator-oriented indicators such as operating costs and route coverage, this invention achieves comprehensive optimization of convenience, efficiency, fairness, and resource utilization through six passenger-experience-oriented indicators. This enables the generation of public transportation network schemes that better meet actual needs, providing a scientific, feasible, and widely applicable technical solution for urban public transportation planning and operation.

[0104] Compared with the optimization target indicators of the prior art, the above-mentioned multi-dimensional serviceability evaluation indicators in this invention have the following improvements:

[0105] (1) Shifting from cost / coverage / time to serviceability-driven. Most existing public transport network optimizations focus on operating costs, route coverage, or average travel time, emphasizing the operator's perspective and failing to comprehensively reflect passenger experience. This invention proposes a serviceability evaluation system centered on passenger travel experience and network accessibility. It systematically introduces six indicators into the optimization objectives: direct access rate, average direct onboard time, average direct waiting time, non-accessibility penalty, average number of transfers, and public transport network redundancy. This ensures the objective function directly corresponds to "convenience, fairness, and accessibility." It comprehensively depicts the travel experience from the passenger's perspective, improving the practicality and public acceptance of the optimization results.

[0106] (2) Introducing the quantification of "public transportation network redundancy" and incorporating it into optimization. Existing technologies usually use "line density" or "overlapping coverage" to roughly measure redundancy, which cannot truly reflect the service contribution of the lines. This invention proposes for the first time a redundancy modeling method based on the station coverage contribution integral: within the service area near the station, for example, within a bus coverage radius of about 800m and a subway coverage radius of about 1500m, the service value of each line is accumulated. Redundant areas are divided based on service saturation thresholds, and the redundancy is obtained by integration. The service areas are then normalized to form a comparable and optimizable redundancy index, which is then incorporated into the objective function with weights. This approach can improve accessibility while effectively suppressing redundant investment and resource waste, ensuring a more streamlined and efficient network structure.

[0107] (3) A multi-objective overall objective function and a consistency mechanism based on serviceability are proposed. Existing works often simply add multiple objectives linearly without handling the order-of-magnitude differences between different objectives, leading to optimization bias. This invention proposes a two-level mechanism of "sub-objective normalization + adaptive weights": First, each sub-objective is normalized to make it dimensionless, eliminating the order-of-magnitude differences between different objectives; then, corresponding weights are applied, and by dynamically adjusting the weight parameters, it is ensured that each indicator is effectively considered in the optimization evolution process, which can avoid a single indicator dominating the optimization result, achieve multi-objective balanced optimization, and improve the convergence stability and scalability of the algorithm.

[0108] In step 130, using the aforementioned multi-dimensional serviceability evaluation indicators, a multi-objective optimization model for public transportation network layout based on serviceability is established, including:

[0109] Step 131: In the process of optimizing the public transportation network, the six serviceability indicators included in the multi-dimensional serviceability evaluation indicators—direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy—correspond to multiple dimensions such as travel convenience, travel efficiency, waiting experience, network connectivity, and resource utilization. To achieve comprehensive optimization, the multi-dimensional serviceability evaluation indicators are weighted and summed, and the objective function in the multi-objective optimization model is specifically designed as follows:

[0110]

[0111] in, This is the weighting coefficient of direct access rate in multi-objective optimization; the negative sign indicates that the smaller the objective value, the better during optimization. This represents the weighting coefficient of the average direct travel time to the vehicle interior in multi-objective optimization. The average direct arrival time is the weighting factor in multi-objective optimization. The weighting coefficient of the unreachability penalty in multi-objective optimization; This represents the weighting coefficient of the average number of transfers in multi-objective optimization. This represents the weighting coefficient of redundancy in multi-objective optimization.

[0112] Step 132: Find the minimum value of the objective function and construct a multi-objective optimization model for public transportation network layout based on serviceability:

[0113] .

[0114] Step 133: Weight coefficients of each index in the objective function of the multi-objective optimization model. As an adaptive adjustment parameter of the model, it is used to adjust the contribution of different optimization objectives to the overall optimization.

[0115] Specifically, when setting weight parameters, an adaptive adjustment mechanism is provided based on different urban traffic demands: when urban planning prioritizes the needs of special groups such as the elderly and children who are sensitive to transfers, the direct access rate is increased. Average number of transfers The weighting; when cities prioritize overall travel efficiency, the average direct travel time to the vehicle is increased. Compared to average direct waiting time Weighting; in the early stages of network planning, to avoid service blind spots, penalties should be imposed for non-compliance. It needs to be given greater weight; in cities with existing high-capacity transportation systems such as subways, the redundancy of the public transportation network should be considered. Improvements are needed to control resource waste.

[0116] By designing the above-mentioned multi-objective optimization model, this invention can ensure passenger travel experience while taking into account operational efficiency and resource utilization, and achieve global, balanced and scalable optimization of public transportation networks.

[0117] Step 140: Design a public transport network genetic algorithm based on serviceability to solve the multi-objective optimization model of the public transport network layout and obtain an optimized public transport network layout scheme based on serviceability.

[0118] This invention proposes a public transportation network genetic algorithm based on serviceability. The algorithm is based on the natural selection and genetic mechanism of the Genetic Algorithm (GA). It encodes public transportation route schemes into station sequences and iteratively optimizes them through evolutionary operations such as selection, crossover, and mutation to gradually generate a public transportation network structure that better meets urban passenger flow needs and operational constraints. Compared with traditional methods, this invention has significant advantages in the following aspects: First, in terms of global search capability, this invention utilizes a population parallel evolution mechanism to effectively avoid single local search methods from getting trapped in local optima, enabling it to explore high-quality solutions in a vast, non-convex transportation network space; Second, in terms of multi-objective optimization capability, this invention incorporates multi-dimensional serviceability indicators such as direct access rate, average number of transfers, direct in-vehicle time, waiting time, inaccessibility penalty, and public transportation network redundancy into the fitness function, and combines weighted combination and normalization mechanisms to achieve comprehensive optimization of passenger travel experience and network resource utilization; Third, in terms of adaptability and robustness, this invention can effectively handle road capacity constraints, transfer hub distribution, and passenger flow uncertainty, and can still maintain stable convergence performance even under conditions of incomplete data or irregular objective functions.

[0119] The optimization process of the algorithm framework designed in this invention is as follows: Figure 6 As shown: First, the travel time matrix, public transportation station information, and existing public transportation network are read as input; then, the adjacent station edges in the travel time matrix and the shortest path algorithm (such as Dijkstra's algorithm) are used to generate... A public transportation network was constructed by identifying several routes; subsequently, a genetic iteration process was initiated: the computer-generated and input public transportation network formed the initial population. and its departure frequency The algorithm selects the best-performing individuals through an elite retention strategy, and then generates a new generation of population through crossover and multi-strategy mutation operators. The objective function of the public transportation network layout multi-objective optimization model based on serviceability is used as the fitness function of the genetic algorithm to calculate the fitness of each individual in the population. The fitness is re-evaluated based on the travel time matrix in each generation of the population. When the iteration reaches the preset number of generations, the public transportation network scheme with the highest fitness (maximum fitness function value) is output as the final optimization result.

[0120] Through this serviceability-based genetic optimization mechanism, the present invention can generate public transportation network schemes that take into account travel convenience, time efficiency, network fairness and resource utilization within a limited computation time, avoiding excessive duplication and inefficient routes, significantly improving the overall service level and feasibility of the public transportation system, and providing a brand-new intelligent solution for urban transportation planning and decision-making.

[0121] Step 140 specifically includes:

[0122] Step 141: Input the objective function of the multi-mode integrated public transport network and public transport network layout multi-objective optimization model, and use the objective function as the fitness function of the public transport network genetic algorithm based on serviceability;

[0123] Step 142, Construct the initial population Each individual represents an initial public transportation network scheme. This invention proposes two methods to obtain the initial population of individuals:

[0124] (1) External import: If the input is an existing public transportation network, its routes are statistically analyzed and directly encoded as individuals in the initial population.

[0125] In one embodiment, the line is encoded, specifically in the form of: Individual The numbers represent the public transportation station numbers.

[0126] (2) Computer generation: Set the number of lines within the area of ​​the public transport network to be optimized, for example, control the number of stops per line to a certain level. Between. First, the origin and destination stations (OD, origin station and destination station) are selected based on Dijkstra's algorithm. The algorithm generates shortest path reachability metrics for routes and performs validity checks (e.g., acyclicity, length constraints). If routes are insufficient, new, simpler routes are generated from the underlying transportation network stored in `edge_list` through random walks. The generated routes undergo constraint filtering and path signature mechanisms. The path signature mechanism prevents duplicate individuals by hashing the sequence of nodes on the route, avoiding the generation of identical route schemes and ensuring legality and non-repetition. Different individuals can be generated by changing the random seed (individuals in the population) until the population size reaches a preset number.

[0127] Using edge_list to store graph structure data is an on-demand storage strategy. By sacrificing the constant-time query advantage of the adjacency matrix, it achieves a significant improvement in memory usage and core computing efficiency, enabling it to handle larger-scale urban traffic networks and complete iterative evolution more quickly.

[0128] In one embodiment, an initial population may be set. The public transport network in China as an individual entity includes Lines, initial population The number of individuals .

[0129] For each route in the initial public transport network plan, randomly generate an initial departure frequency. .

[0130] Initial population It is a group of 0th generation (first generation) wire mesh schemes in the algorithm, each containing a wire mesh structure and its corresponding departure frequency.

[0131] Step 143: Perform genetic operations based on iterative cycles. In each iteration, the current population is processed... The public transportation network plan uses genetic manipulation to obtain the next generation of population. The public transportation network scheme, wherein the genetic operation for each cycle includes:

[0132] (1) For the current population The public transportation network scheme uses the shortest path algorithm to recalculate the travel time matrix, and then uses the obtained travel time matrix... The fitness of the current population's public transportation network scheme is evaluated by calculating the fitness (fitness function value) of each individual; at the same time, a departure frequency is generated for each route in the current population's public transportation network scheme. .

[0133] (2) Elite Retention and Selection

[0134] To prevent the optimal solution from being lost during evolution, this invention employs an elite retention strategy, directly retaining the top-fittest solutions in each generation. One individual is carried over to the next generation; the remaining individuals are selected through a tournament selection process, using the best-performing individuals among the top performers. Compare the individual samples and select the best ones to proceed to the crossover and mutation phase.

[0135] In one embodiment, The proportion of the population is .

[0136] (3) Cross

[0137] To achieve the combination and inheritance of superior patterns among different network schemes, this invention adopts a point-crossing strategy based on line sets. First, two parent individuals are randomly selected. and (i.e., two public transportation network schemes), and randomly generate two intersections at the line set level. and The following formula represents the two selected parent nets:

[0138] ;

[0139] ;

[0140] in, Individual The lines in the middle, Individual The lines in;

[0141] Crossover generates new mesh offspring:

[0142] ;

[0143] ;

[0144] Legality verification is still required after crossover. Crossover operations inherit the best line combinations from the parent generation, increasing diversity.

[0145] (4) Variation

[0146] To promote population diversity and drive circuit structure evolution, this invention designs a multi-strategy hybrid mutation operator. This operator uses probability... The core operations that affect each line in the population mainly include the following six types:

[0147] Route Extension: This involves extending the route by one stop at either the beginning or end, aiming to expand the route's coverage and explore new service areas. The original route is:

[0148] ;

[0149] The extended route is as follows:

[0150] ;

[0151] or ;

[0152] Shrink: Removes the first and last stops of a line to shorten excessively long lines and optimize resource allocation efficiency.

[0153] The route after contraction is as follows:

[0154] ;

[0155] or ;

[0156] Inserting a station: Inserts a new station that is adjacent to both consecutive stations, enabling local fine-tuning and optimization of the route. The route after insertion is as follows:

[0157] ;

[0158] Flipping: Reversing the route sequence to change the operating direction and adapt to asymmetrical passenger flow demand. The flipped route is as follows:

[0159] ;

[0160] Splice: This operation includes two modes:

[0161] (a) Remove intermediate nodes to simplify the route as follows:

[0162] ;

[0163] (b) Find a bypass station to replace the original route segment and form a new route.

[0164] ;

[0165] Line Merge: This operation randomly selects two lines with a certain probability. If they are connected end-to-end, they are merged into a new line. If merging reduces the number of lines in the population, a valid line is generated according to the random walk method in the initial population to ensure the number of lines for each individual. The random walk mechanism is used to automatically explore new candidate paths in the network structure. This mutation operation can significantly change the line structure and is one of the key strategies for the algorithm to escape local optima. At the same time, all mutation operations must pass validity verification (e.g., acyclicity, length constraints, etc.) to ensure that the generated new individuals are all feasible solutions.

[0166] (5) If the next generation of the population is... Less than the preset maximum algebra If so, then the next loop will be executed.

[0167] All individuals after crossover and mutation will have their fitness values ​​recalculated to ensure consistency in multi-objective optimization.

[0168] Step 144, if the loop termination condition is met: the next generation of the population... Reaching the preset maximum algebra If the fitness of the individual with the highest fitness in a population does not improve after several consecutive occurrences, the loop terminates and the individual with the highest fitness in the population is output as the optimal public transportation network scheme, including detailed route structure and random departure frequency.

[0169] Finally, this invention was tested in a real-world scenario in a city (City C) to verify the algorithm of the public transportation network layout method described in this invention. Specifically, the following hyperparameter settings were applied to the algorithm: the number of lines in the public transportation network layout area to be optimized was 35, and the number of stops per line was controlled within a certain range. Between; the genetic algorithm part sets the optimization generation. The population size is 1000. After optimization, the objective function value of the public transportation network map converged from 17.49 to 6.63, and the total objective value continued to decrease during the iteration process, such as... Figure 7 As shown, this indicates that the algorithm can converge stably in the direction of optimization.

[0170] Further observation is needed regarding the evolution of various serviceability indicators, such as... Figure 8 As shown, specifically:

[0171] like Figure 8 As shown in (a), the direct rate continues to increase, indicating that the proportion of direct trips among passengers has increased after optimization, and the convenience of travel has been significantly improved.

[0172] like Figure 8 (b) and Figure 8 As shown in (c), the average direct arrival time inside the vehicle and the average direct arrival waiting time remain at a reasonable level. Although there are fluctuations in the local iteration stage, they eventually tend to stabilize, indicating that the present invention ensures time efficiency and waiting experience while improving coverage and directness.

[0173] like Figure 8 As shown in (d), the unreachability penalty is significantly reduced, reflecting the gradual enhancement of the feasibility and rationality of generating solutions;

[0174] like Figure 8 As shown in (e), the average number of transfers (TR) tends to be at a better level during the search process, indicating an improvement in the overall travel convenience for passengers; as Figure 8 As shown in (f), the redundancy of the public transport network remains at a low level, indicating that the route structure effectively avoids duplication and waste while ensuring coverage.

[0175] Geographically visualize the optimization results, such as Figure 9 As shown. The optimized public transport network is compared with the initially generated public transport network, as shown. Figure 10 As shown, where, Figure 10 (a) is the initially generated public transport network. Figure 10(b) The optimized public transport network shows significant improvements in several aspects. First, the spatial distribution of routes is more balanced, avoiding the situation in the initial plan where some areas had excessively dense routes while others were sparse or even had service blind spots. Coverage in peripheral areas is enhanced, and service fairness is significantly improved. Second, the route routes are more rational. The initial plan had problems such as detours and excessively long routes, while the optimized route layout is more direct and efficient, reducing unnecessary redundancy and detours, thereby improving operational efficiency. In terms of transfers, the optimized main line intersection layout is more rational, with both the number and rationality of transfer nodes improved. Passengers can switch between routes at more locations, thereby shortening overall travel time. At the same time, the optimized plan extends routes to the edge of urban areas, significantly improving service coverage, effectively reducing service blind spots and isolated stations, and meeting more travel needs. Finally, the degree of route redundancy is significantly reduced. Compared with the initial network where a large number of routes were concentrated in the same corridor, resulting in resource waste, the optimized public transport network achieves a structure with clear functional division and higher utilization efficiency. In summary, the genetic algorithm based on serviceability index proposed in this invention has achieved systematic improvements in public transportation network balance, operational efficiency, transfer convenience, and coverage. It not only improves the travel experience of passengers but also enhances the overall resource utilization efficiency of the network, fully verifying the practicality and superiority of this invention in urban public transportation network planning.

[0176] This invention, through further experiments, verifies the public transportation network genetic algorithm based on serviceability used in the public transportation network optimization process. Compared with traditional genetic algorithms, it makes several improvements. Traditional genetic algorithms typically generate the initial population completely randomly or construct it through simple shortest paths, lacking control over network coverage and balance. This invention, however, adopts a "two-stage initial population" strategy, combining external import, Dijkstra's algorithm, and road network-based random walks to ensure population diversity and feasibility. Traditional algorithms often use fitness functions with single objectives or simple weighting, generally maximizing route coverage, minimizing travel time, and minimizing operating costs, without normalization between different indicators. This invention introduces six serviceability indicators, including direct access rate, in-vehicle time, waiting time, non-access penalty, average transfer rate, and redundancy, combined with normalization and adaptive weighting mechanisms to ensure balanced multi-objective optimization. Regarding selection strategies, traditional algorithms often use roulette wheel or random tournaments, which can easily lead to the loss of excellent individuals. This invention uses elite retention combined with a tournament selection that restricts selection to the top few excellent individuals, ensuring both convergence speed and population diversity. In terms of mutation operations, traditional methods only randomly replace, insert, or delete nodes. This invention designs a hybrid multi-strategy mutation operator, including line extension, contraction, insertion, flipping, splicing, and line merging, and performs validity verification after each operation. In particular, the line merging operator can significantly change the line structure, helping to escape local optima. Overall, this invention, through elite preservation, adaptive weight adjustment, merging operators, and reasonable mutation, not only ensures solution diversity but also significantly accelerates the convergence speed and improves the superiority of the final solution. Based on these differences, this invention conducts experimental comparisons in examples.

[0177] The objective function of the traditional genetic algorithm is:

[0178] ,

[0179] in, The operating cost of the public transport network is calculated using route duration. Departure frequency indicates other and The average direct arrival time and waiting time are consistent with those described earlier in this invention. This represents the site coverage ratio. In the selection operation of the genetic algorithm, traditional GA uses a roulette wheel selection method.

[0180] In the experiment, all settings except for the improvements mentioned above were kept unchanged, such as the initial population, model input, and other algorithm sub-parts, and a comparative experiment was conducted.

[0181] The optimization curves of the traditional genetic algorithm and the improved genetic algorithm of this invention are as follows: Figure 11 and Figure 12As shown, on the optimization objective curve of the algorithm, traditional GA (such as...) Figure 11 As shown, due to the lack of normalized and adaptive weights, the existence of contradictory sub-objectives within the objective, and the fact that roulette wheel betting in genetic programming does not necessarily preserve the relatively optimal solution inherited to the current generation, the overall objective is not optimized; in contrast, the improved genetic algorithm curve based on serviceability of this invention (as shown) Figure 12 As shown in the lower middle curve, the overall target value gradually decreases with genetics, thus achieving optimization.

[0182] At the same time, from Figure 12 As can be seen from the upper curve, traditional public transport network optimization has not fully optimized the serviceability index of the public transport network. Table 3 shows a comparison of algorithms based on the serviceability index values. The improved genetic algorithm of this invention optimizes the initial population value from 18.75 to 6.63, while the traditional GA searches from 18.75 to 21.03.

[0183] Table 3. Comparison of optimization values ​​between the genetic algorithm and the traditional genetic algorithm for public transportation networks based on serviceability.

[0184]

[0185] from Figure 13 Based on the optimization results, Figure 13 (a) is a visual network of the initial population individuals. Traditional genetic algorithms yield public transportation networks as shown in the image. Figure 13 As shown in (b), there are many isolated nodes, while the public transportation network obtained by the improved genetic algorithm based on serviceability of this invention is as follows: Figure 13 As shown in (c), the optimized public transportation network has the fewest isolated nodes, and overall, it is more reasonable and practical, thus increasing the algorithm's value in real-world applications. Furthermore, Figure 13 (d) is Figure 13 (a) A schematic diagram of the public transport network layout, with only public transport stops and routes retained after removing map geographic information. Figure 13 (e) is Figure 13 (b) A public transport network layout diagram with map geographic information removed, retaining only public transport stops and routes. Figure 13 (f) is Figure 13 (c) A public transport network layout diagram with the map geographic information removed and only public transport stations and routes retained can more clearly show the aforementioned features.

[0186] In one embodiment, the present invention provides a public transportation network cabling method and apparatus based on serviceability, the apparatus comprising:

[0187] The first module is used to obtain station information of the public transportation network; it uses Space-L and Space-P dual network models for data modeling to construct a multi-modal integrated public transportation network, which is used to characterize the physical structure of the public transportation network and reflect the transfer accessibility between stations;

[0188] The second module is used to design normalized multi-dimensional serviceability evaluation indicators based on a multi-modal integrated public transportation network. The multi-dimensional serviceability evaluation indicators include at least the direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy.

[0189] The third module is used to establish a multi-objective optimization model for public transportation network layout based on serviceability by utilizing the multi-dimensional serviceability evaluation indicators; and to obtain the objective function of the multi-objective optimization model for public transportation network layout by weighted summation of the multi-dimensional serviceability evaluation indicators.

[0190] The fourth module is used to design a public transport network genetic algorithm based on serviceability, solve the multi-objective optimization model of the public transport network layout, and obtain an optimized public transport network layout scheme based on serviceability; the objective function of the public transport network layout multi-objective optimization model is used as the fitness function of the public transport network genetic algorithm based on serviceability.

[0191] On the other hand, the present invention provides a computer device including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the public transportation network cabling method based on serviceability provided in any of the above embodiments. The computer device may be a server. The computer device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store sample data. The network interface of the computer device is used for communication with external terminals via a network connection.

[0192] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the public transport network cabling method based on serviceability provided in any of the above embodiments.

[0193] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0194] Matters not covered in this invention are common knowledge. The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered to be within the scope of this specification.

[0195] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A public transportation network cabling method based on serviceability, characterized in that, include: Step 110: Obtain station information for the public transportation network; A dual-network model of Space-L and Space-P is used for data modeling to construct a multi-modal integrated public transportation network, which is used to characterize the physical structure of the public transportation network and reflect the transfer accessibility between stations; including: Step 111: Use the public transportation service platform in conjunction with map application software to obtain station information of the public transportation network, and determine the stations and their latitude and longitude in the public transportation network to be optimized within the wiring area. Step 112: Using the station information of the public transportation network, construct a basic transportation network and obtain the travel time matrix of the public transportation network, including: By utilizing station information from the public transportation network, the shortest travel time between stations can be obtained and used as the navigation time between stations; The distance between stations is determined by using a connectivity threshold: if the distance between two stations is greater than the connectivity threshold, the two stations are considered not directly connected; if the distance between two stations is less than or equal to the connectivity threshold, the two stations are connected by using traffic navigation time as the edge weight; the connectivity threshold is set and adjusted according to the public transportation adjacent station distance standard. Using stations as network nodes, a basic transportation network is formed based on edge weights; Based on the edge weights of the basic transportation network, a travel time matrix is ​​constructed, where each element in the travel time matrix is ​​a triple; the travel time matrix is ​​stored in edge_list format. Step 113: Use Space-L and Space-P dual network models to perform topology modeling of the basic transportation network and construct a multi-modal integrated public transportation network; The Space-L is a spatial route model that physically models the basic transportation network. It constructs a public transportation Space-L network by using all stations of the public transportation network as nodes and establishing connections between adjacent stations in each route of the basic transportation network. The Space-P is an accessibility path model for modeling public transportation transfer relationships in a basic transportation network. It constructs a public transportation Space-P network by using all stations of the public transportation network as nodes and establishing an edge between any two stations in each line of the basic transportation network. By applying the Space-L model and the Space-P model to the basic transportation network, a multimodal integrated public transportation network is constructed, including a public transportation Space-L network and a public transportation Space-P network. Step 120: Based on the multi-modal integrated public transportation network, design normalized multi-dimensional serviceability evaluation indicators; the multi-dimensional serviceability evaluation indicators include at least the direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy. Step 130: Using the multi-dimensional serviceability evaluation indicators, establish a multi-objective optimization model for public transportation network layout based on serviceability; obtain the objective function of the multi-objective optimization model for public transportation network layout by weighted summation of the multi-dimensional serviceability evaluation indicators; Step 140: Design a public transport network genetic algorithm based on serviceability to solve the public transport network routing multi-objective optimization model and obtain an optimized public transport network routing scheme based on serviceability; use the objective function of the public transport network routing multi-objective optimization model as the fitness function of the public transport network genetic algorithm based on serviceability.

2. The public transportation network layout method based on serviceability according to claim 1, characterized in that, The direct access rate is given by the following formula: ； in, It represents the total number of direct routes within the Space-P public transportation network. It refers to the node size of the basic transportation network; This indicates two stations in the Space-P public transportation network. and The shortest path between the two routes is given by the following formula: ； hour, It directly indicates the number of transfers when using public transportation; The average direct travel time to the vehicle is given by the following formula: ； in, It's the time for traffic navigation. This indicates the total number of direct OD station pairs; a direct OD station pair represents a pair of start and end stations that are directly accessible on the route. Indicates site to station Direct routes between them; The average direct arrival waiting time is given by the following formula: ； in, Indicates the passenger's waiting time. Indicates direct route The frequency of departures, Indicates a set of direct routes. This is a constant used to characterize the uniformity of passenger arrivals at stations.

3. The public transportation network layout method based on serviceability according to claim 2, characterized in that, The unreachable penalty is given by the following formula: ； in, Indicates two sites Penalty for unreachability between two sites; If the destination is inaccessible or requires more than 2 transfers, then... If two stations If the number of connections or transfers between the two points is less than or equal to 2, then ; The average number of transfers is given by the following formula: ； in, This represents a set of routes that can be reached directly or by transfer. This indicates the total number of direct OD station pairs and transfer-accessible OD station pairs, where each transfer-accessible OD station pair represents a pair of starting and ending stations accessible by transfer on the line. The redundancy of the public transportation network is given by the following formula: ； in, The service value of the public transportation network is obtained by statistically analyzing the service value of each station and then summing all the service value by spatial location. This is the total redundancy. The pre-set service saturation threshold; Service area Internal redundancy; the service value added for each site, obtained through the following methods: For bus stops ,go through The number of lines is ,exist Surrounding coverage radius Within the coverage area, then the bus stops Service added value for: ; For subway stations ,go through The number of lines is ,exist Surrounding coverage radius Within its coverage area, subway stations Service added value for: .

4. The public transportation network routing method based on serviceability according to claim 3, characterized in that, Step 130 includes: Step 131: Weight the direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transport network redundancy included in the multi-dimensional serviceability evaluation indicators, and design the objective function in the multi-objective optimization model: in, This is the weighting coefficient of direct access rate in multi-objective optimization; the negative sign indicates that the smaller the objective value, the better during optimization. This represents the weighting coefficient of the average direct travel time to the vehicle interior in multi-objective optimization. The average direct arrival time is the weighting factor in multi-objective optimization. The weighting coefficient of the unreachability penalty in multi-objective optimization; This represents the weighting coefficient of the average number of transfers in multi-objective optimization. This represents the weighting coefficient of redundancy in multi-objective optimization. Step 132: Find the minimum value of the objective function and construct a multi-objective optimization model for public transportation network layout based on serviceability: ； Step 133: Adjust the weight coefficients in the objective function of the multi-objective optimization model. As an adaptive adjustment parameter of the model; it provides an adaptive adjustment mechanism based on the traffic needs of different cities, and sets weight coefficient parameters.

5. The public transportation network cabling method based on serviceability according to claim 4, characterized in that, The adaptive adjustment mechanism based on different urban traffic demands includes setting weighting coefficient parameters, including: When prioritizing transfers for special groups and sensitive groups, the weighting factor for increasing the direct access rate should be increased. Weighting coefficient with average number of transfers When prioritizing overall travel efficiency, increase the weighting factor for average direct travel time to the vehicle. Weighting coefficient of average direct waiting time In the initial stages of public transportation network planning, to avoid service blind spots, the weighting factor for inaccessibility penalties is increased. In large urban transportation systems, the weighting coefficient for increasing the redundancy of the public transportation network. Control resource waste.

6. The public transportation network cabling method based on serviceability according to claim 4, characterized in that, Step 140 includes: Step 141: Input the objective function of the multi-mode integrated public transport network and public transport network routing multi-objective optimization model, and use the objective function as the fitness function of the public transport network genetic algorithm based on serviceability to evaluate the applicability of the public transport network scheme; Step 142, Construct the initial population In the initial population, each individual represents an initial public transport network scheme; for each route in the initial public transport network scheme, an initial departure frequency is randomly generated. ; Step 143: Perform genetic operations based on iterative cycles. In each iteration, the current population is processed... Genetic manipulation of public transportation network schemes to obtain the next generation of population. Public transportation network plan; Step 144, if the loop termination condition is met: the next generation of the population... Reach the preset maximum algebra If the fitness of the individual with the highest fitness in the population does not improve after several consecutive occurrences, the loop terminates; the individual with the highest fitness in the population is output as the optimal public transportation network scheme, which includes detailed route structure and departure frequency.

7. The public transport network cabling method based on serviceability according to claim 6, characterized in that, The initial population Obtained through either of the following two methods: Method 1: Directly count the routes of the initially acquired public transportation network and encode them as the initial population. Individuals within; Method 2: Using a computer to generate the initial population ,include: Set the number of lines within the area where the public transportation network layout needs to be optimized; The shortest path algorithm is used to select OD station pairs, generate the shortest path of the route, and verify its validity. If the number of routes is insufficient, a new simple route path is generated by random walk in the basic transportation network stored in the form of edge_list. The new simple route path is prevented from generating duplicate individuals by constraint filtering and path signature mechanism, and the same route is avoided by hashing the node sequence on the route. By changing the random seed, different individuals are generated in the initial population until the size of the initial population reaches the preset number.

8. The public transport network cabling method based on serviceability according to claim 7, characterized in that, Step 143 includes: Perform fitness assessment: for the current population The public transportation network scheme uses the shortest path algorithm to recalculate the travel time matrix, and then uses the obtained travel time matrix... By calculating the fitness of each individual, the fitness of the current population's public transportation network scheme is evaluated; and a departure frequency is generated for each route in the current population's public transportation network scheme. ; An elite retention strategy is adopted to retain and select elites: the current population is directly retained. The most adaptive front From individual to the next generation of population Middle; Current population The remaining individuals that were not directly retained to the next generation were selected through a tournament selection process; the individuals with the highest fitness among the remaining individuals were selected. Compare individual samples and select the best ones to proceed to the crossover and mutation phase; A point-to-point crossing strategy based on a set of lines is used to perform the crossing operation: Randomly select two parent individuals; At the line set level, two intersection points are randomly generated to produce new network offspring: Perform legality verification to confirm that the above cross-operation inherits the excellent route combination of the parent generation and increases diversity; Design a multi-strategy hybrid mutation operator to perform mutation operations; the multi-strategy hybrid mutation operator uses probability... The operation of the multi-strategy hybrid mutation operator, which acts on each path in the population, includes at least the following: Route extension: Extending the original route by one station from its starting / ending point to expand the route coverage and explore new service areas; Line shrinking: Removes the starting / ending stations of a line to shorten excessively long lines and optimize resource allocation efficiency; Inserting a station: Inserting a new station that is adjacent to both consecutive stations allows for local fine-tuning and optimization of the line; Route reversal: Reversing the route sequence and changing the direction of operation to adapt to asymmetrical passenger flow demand; Line splicing: simplifies the route by deleting intermediate nodes; or finds bypass stations to replace the original route segments to form a new route. Route merging: Two routes are randomly selected with a set probability. If the two routes are connected end to end, they are merged into a new route. If merging reduces the number of routes for individuals in the population, a legal route is generated according to the random walk method in the initial population to ensure the number of routes for individuals in the population. If the next generation of the population is... Less than the preset maximum algebra If so, then the next loop will be executed.

9. A public transport network cabling device based on serviceability, characterized in that, The device includes: The first module is used to acquire station information for the public transportation network; it employs Space-L and Space-P dual network models for data modeling to construct a multi-modal integrated public transportation network, which characterizes the physical structure of the public transportation network and reflects the transfer accessibility between stations; it includes: Submodule 1 is used to obtain station information of the public transportation network by combining the public transportation service platform with map application software, and to determine the stations and their latitude and longitude in the public transportation network to be optimized within the wiring area. Submodule two is used to construct a basic transportation network using station information from the public transportation network and obtain the travel time matrix of the public transportation network, including: By utilizing station information from the public transportation network, the shortest travel time between stations can be obtained and used as the navigation time between stations; The distance between stations is determined by using a connectivity threshold: if the distance between two stations is greater than the connectivity threshold, the two stations are considered not directly connected; if the distance between two stations is less than or equal to the connectivity threshold, the two stations are connected by using traffic navigation time as the edge weight; the connectivity threshold is set and adjusted according to the public transportation adjacent station distance standard. Using stations as network nodes, a basic transportation network is formed based on edge weights; Based on the edge weights of the basic transportation network, a travel time matrix is ​​constructed, where each element in the travel time matrix is ​​a triple; the travel time matrix is ​​stored in edge_list format. Submodule 3 is used to perform topology modeling of the basic transportation network using the Space-L and Space-P dual network models, and to construct a multi-modal integrated public transportation network; The Space-L is a spatial route model that physically models the basic transportation network. It constructs a public transportation Space-L network by using all stations of the public transportation network as nodes and establishing connections between adjacent stations in each route of the basic transportation network. The Space-P is an accessibility path model for modeling public transportation transfer relationships in a basic transportation network. It constructs a public transportation Space-P network by using all stations of the public transportation network as nodes and establishing an edge between any two stations in each line of the basic transportation network. By applying the Space-L model and the Space-P model to the basic transportation network, a multimodal integrated public transportation network is constructed, including a public transportation Space-L network and a public transportation Space-P network. The second module is used to design normalized multi-dimensional serviceability evaluation indicators based on a multi-modal integrated public transportation network. The multi-dimensional serviceability evaluation indicators include at least the direct access rate, average direct in-vehicle time, average direct waiting time, non-access penalty, average number of transfers, and public transportation network redundancy. The third module is used to establish a multi-objective optimization model for public transportation network layout based on serviceability by utilizing the multi-dimensional serviceability evaluation indicators; and to obtain the objective function of the multi-objective optimization model for public transportation network layout by weighted summation of the multi-dimensional serviceability evaluation indicators. The fourth module is used to design a public transport network genetic algorithm based on serviceability, solve the multi-objective optimization model of the public transport network layout, and obtain an optimized public transport network layout scheme based on serviceability; the objective function of the public transport network layout multi-objective optimization model is used as the fitness function of the public transport network genetic algorithm based on serviceability.

Images