Method and system for evaluating and optimizing urban traffic network based on different stakeholders

Abstract

The application discloses a city traffic network evaluation and optimization method and system based on different interests, and relates to the technical field of traffic network analysis. Firstly, based on road network information, traffic flow and passenger flow data, three directed and weighted complex networks representing different interests are constructed: high-speed traffic and convenience network, urban motor vehicle road traffic and convenience network, and public transportation network. Then, the weighted edge community detection algorithm is used to divide each network into communities, and the traffic flow, passenger flow, vehicle speed and average travel time of each community area are evaluated. Through network connectivity analysis, the traffic accessibility of each area is evaluated. Combined with population density, enterprise distribution and other data, the weighted vertex importance algorithm is used to identify important nodes and evaluate their rationality and role. Finally, according to the evaluation results, optimization schemes such as road repair, widening, new road construction and additional public transportation lines are proposed.

Description

Technical Field

[0001] This invention relates to the field of traffic network analysis technology, and in particular to a method and system for evaluating and optimizing urban traffic networks based on different stakeholders. Background Technology

[0002] The rational distribution of urban public transportation routes is crucial for improving service quality, enhancing residents' quality of life, and promoting regional economic development. As the foundation of regional spatial connections, the transportation network directly reflects the rationality of urban planning and the level of regional development, determining the smoothness of material and information exchange between cities and regions. Therefore, a comprehensive analysis of transportation network layout is an urgent need for urban planning, design, and management.

[0003] Currently, transportation network evaluation methods mainly target road networks and public transportation networks. They use various methods such as optimization methods, statistical and machine learning methods, GIS technology, complex network analysis, and spatiotemporal data analysis to analyze information such as network topology, connectivity, shortest path, traffic flow distribution, path optimization, and resource scheduling to evaluate the performance of transportation networks.

[0004] However, many of these methods fail to consider that transportation networks are networks with both direction and weight. Furthermore, many methods cannot fully utilize all traffic data to analyze the traffic situation of a region or city. More importantly, existing research methods primarily analyze the overall transportation network, with less focus on the needs of different transportation stakeholders. Analyzing traffic based on different stakeholders can more comprehensively reflect service coverage, fairness of travel opportunities, and fairness of resource allocation, thus providing a scientific basis for optimizing resource allocation and promoting social equity.

[0005] Therefore, we propose a method and system for evaluating and optimizing urban transportation networks based on different stakeholders, and propose scientific optimization schemes based on the evaluation results to comprehensively improve the fairness, efficiency and service quality of urban transportation networks. Summary of the Invention

[0006] The purpose of this invention is to solve the above-mentioned problems by providing a method and system for evaluating and optimizing urban transportation networks based on different stakeholders.

[0007] To achieve the above objectives, the technical solution adopted by the present invention includes the following steps:

[0008] S1: The data is preprocessed using a five-logistic regression filling method, and then a density-based method combined with the Pauta criterion is used to detect outliers in the dataset. According to the calculated density of each data point, if the data point x satisfies the formula (1), then the point is identified as an outlier.

[0009] (1)

[0010] in, Let x be the density of the data point. The mean of the inverse of the density within the dataset. The standard deviation of the inverse density within the dataset;

[0011] S2: Construct three types of transportation networks representing different traffic interest groups, including a high-speed accessibility and convenience network for those who benefit from high-speed travel, an urban motor vehicle accessibility and convenience network for residents who drive, and a public transportation network for those who benefit from public transportation. The specific construction is as follows:

[0012] Construction of a high-speed accessibility and convenience network representing the interests of high-speed travelers

[0013] (1) Network nodes and edges: The entrances and exits, toll stations, service areas, intersections and monitoring points of each highway in and around the city are taken as nodes of the traffic network, and the actual path between two nodes is taken as the edge between the two points. All the nodes and edges form a complex network.

[0014] (2) The direction of the edge in the network: The direction of the vehicle between two points is taken as the direction between two nodes in the complex network. The complex network is an asymmetric network, and the direction between two nodes is also bidirectional.

[0015] (3) Road data: Obtain the average daily traffic flow, maximum speed limit and number of lanes for each direction of each edge in the network;

[0016] (4) Weight between two points

[0017] (2)

[0018] in It represents the average daily traffic flow of the edges between node i and node j. It is the highest daily traffic volume on the network. Let $i$ be the maximum speed limit for the edge between node $i$ and node $j$. The number of lanes between node i and node j. and This is to scale the data to the same range.

[0019] (5) The physical significance of cyberspace and the stakeholders it represents

[0020] because It is the average daily traffic flow between the two points, therefore The larger the number of lanes, the more people use the road, and the better the convenience of the road segment. Therefore, the first term of formula (2) can reflect the convenience of this expressway network. For the second term of formula (2), the number of lanes... More, speed limit The larger the volume, the higher the average daily traffic volume. The smaller the value, the more vehicles pass through this section of road per unit of time, so this value represents the convenience of this section of road.

[0021] A network representing the interests of urban drivers regarding urban road accessibility and convenience.

[0022] (1) Network nodes and edges: All traffic light intersections in this city are taken as nodes of the traffic network, and the actual path between two nodes is taken as the edge between the two points. All points and edges form a complex network.

[0023] (2) The direction of the edge in the network: The direction of the vehicle between two points is taken as the direction between two nodes in the complex network. The complex network is an asymmetric network, and the direction between two nodes is also bidirectional.

[0024] (3) Road data: Obtain the traffic flow, maximum speed limit, and number of lanes for each direction of each edge in the network.

[0025] (4) Weight between two points

[0026] (3)

[0027] in It represents the average daily traffic flow of the edges between node i and node j. It is the highest daily traffic volume on the network. Let $i$ be the maximum speed limit for the edge between node $i$ and node $j$. The number of lanes between node i and node j. and This is to scale the data to the same range;

[0028] (5) The physical significance of cyberspace and the stakeholders it represents

[0029] because It is the average daily traffic flow between the two points, therefore The larger the number of lanes, the more people use the road, and the better the convenience of the road segment. Therefore, the first term of formula (2) can reflect the convenience of this expressway network. For the second term of formula (2), the number of lanes... The more, the faster the speed limit. The larger, The lower the traffic volume, the more vehicles pass through this section of road per unit time, so this item represents the convenience of this section of road.

[0030] Public transport network construction representing the interests of public transport users

[0031] (1) Network nodes and edges: Each bus, subway, and light rail station in this city and its suburbs is taken as a node of the transportation system, and the actual transportation route between two points is taken as an edge between the two points. All the points and edges constitute a complex network.

[0032] (2) The direction of the edges in the network: The direction of traffic flow of the public transportation line between two points is taken as the direction between two nodes in the complex network. Since the vehicles on the public transportation line travel in both directions, this complex network is an asymmetric network, and the direction between two nodes is also bidirectional.

[0033] (3) Road data: Obtain the average daily traffic flow in each direction of each edge in the network, the total number of passengers getting on and off at transfer stations, and the maximum average traffic flow of a single bus stop in this city.

[0034] (4) Weight between two points:

[0035] (4)

[0036] in Let $\mathbf{i}$ be the average daily traffic flow between nodes $i$ and $j$, and $max(ADT)$ be the maximum average daily traffic flow in the public transport network. and They represent two adjacent stations respectively. i, j The number of passengers getting on and off the bus is divided by 2 to separate the two nodes. i and j The number of passengers between them was averaged. The maximum number of passengers boarding and alighting at a single node in a public transportation system;

[0037] (5) The physical significance of networks and the stakeholders they represent:

[0038] because It is the average daily traffic flow between the two points, therefore The larger the value, the more people use the public transportation line, and the better the convenience of the line. Therefore, the first term of formula (4) can reflect the convenience of this public transportation network. As for the second term of formula (4), it is the total number of passengers getting on and off at the two stations. The larger the area, the greater the traffic flow between the two points. The smaller the value, the more likely it is to represent the two nodes of this transportation route. i, j The more passengers a public transport network has, the more rational its public transport network becomes.

[0039] S3: Conduct community detection for each transportation network, divide it into communities, and conduct a transportation convenience assessment of the corresponding area of ​​each community, including traffic flow, pedestrian flow, vehicle speed, and average travel time.

[0040] A weighted edge-based LPA algorithm considering weights and directions is used for community detection. Interrelated traffic nodes eventually form several communities. The process consists of the following two steps:

[0041] The first step is to initialize the labels of all nodes in the transportation network so that each node initially belongs to an independent community;

[0042] The second step is to perform an iterative label propagation process. First, the nodes in the network are arranged in a random order to obtain a random node sequence, denoted as X. Then, the nodes in X are traversed in an asynchronous manner, and each node is assigned to the community with the largest cumulative weight of all directed edges linked to it. The algorithm terminates when each node has been assigned to the community with the largest cumulative weight of all directed edges linked to it. Finally, several communities can be obtained. Nodes within the same community have the same label, while nodes between different communities have different labels. Each community corresponds to an associated traffic node region.

[0043] S4: Traffic Accessibility Assessment: Treat each community detected in each network as an independent graph, and use graph connectivity analysis to assess the traffic accessibility of the area corresponding to each community, while also analyzing the rationality and reasons for insufficient accessibility.

[0044] S5: Evaluate key nodes in each community by combining data such as population density, enterprise distribution, infrastructure service distribution, and commercial location;

[0045] The PageRank algorithm, improved with weighted edge addition, redefines the importance of each node and detects important nodes, as shown in the following formula:

[0046] (6)

[0047] in, Represents a node The set of nodes it points to. Indicates a path from other nodes to a node. The cumulative sum of edges, Representing an edge Weights;

[0048] The process of the weighted directed PageRank algorithm is as follows:

[0049] 1. Initialize weights: Assign an initial W-PR value to each node. For fairness, the initial W-PR value of each node is usually set to an equal probability.

[0050] 2. Iterative calculation: The W-PR value of each node is calculated iteratively according to formula (6), and the difference between the W-PR value of the current node and the W-PR value of the previous round is calculated;

[0051] 3. Termination condition: Repeat step 2 until the difference is less than the preset convergence threshold or the preset number of convergences is reached;

[0052] S6: Based on the evaluation results, propose corresponding urban transportation network optimization plans for each community:

[0053] (1) For the high-speed traffic and convenience network, based on the evaluation results, optimization schemes such as building new roads, widening local roads and setting local speed limits are proposed according to the important nodes in the community;

[0054] (2) For the urban motor vehicle road traffic and convenience network, based on the evaluation results and according to the important nodes in the community, optimization schemes such as building new roads, widening local roads, and changing local speeds are proposed;

[0055] (3) For the public transport network, based on the evaluation results, propose optimization schemes to increase or decrease bus routes and bus stops according to the important nodes in the community.

[0056] Furthermore, including:

[0057] The transportation network construction module is used to build three types of transportation networks that represent different groups of transportation stakeholders.

[0058] The evaluation module is used to detect communities in each transportation network and divide them into communities. Combining data such as population density, enterprise distribution, infrastructure service distribution, and commercial areas, it is used to evaluate the rationality of important nodes within each community, the accessibility of the corresponding area of ​​each community, and the convenience of transportation in the corresponding area of ​​each community.

[0059] The optimization scheme generation module is used to propose corresponding urban transportation network optimization schemes for each community based on the evaluation results.

[0060] Furthermore, the transportation network construction module includes:

[0061] The high-speed accessibility and convenience network construction unit is used to determine the nodes, edges, weights, and directions of the high-speed accessibility and convenience network.

[0062] The urban motor vehicle road traffic and convenience network construction unit is used to determine the nodes, edges, weights, and directions of the urban motor vehicle road traffic and convenience network.

[0063] The public transport network construction unit is used to determine the nodes, edges, weights, and directions of the public transport network.

[0064] Furthermore, the evaluation module employs a weighted edge-based community detection algorithm and a weighted edge-based vertex importance algorithm, including a community detection unit, an important node evaluation unit within the community, and a traffic convenience evaluation unit within the community.

[0065] Furthermore, the optimization scheme generation module includes a high-speed trafficability and convenience network optimization scheme generation unit, an urban motor vehicle road trafficability and convenience network optimization scheme generation unit, and a public transport network optimization scheme generation unit.

[0066] According to one aspect of the present invention, a method and system for evaluating and optimizing urban transportation networks based on different stakeholders are provided. Compared with the prior art, the present invention has the following advantages:

[0067] 1. By constructing three traffic networks representing highway users, urban drivers, and public transport passengers respectively, the needs of different traffic interest groups are comprehensively covered, making up for the shortcomings of existing research on the analysis of single traffic networks or overall networks.

[0068] 2. By employing a weighted edge-based community detection algorithm and a vertex importance algorithm, combined with multi-dimensional data such as population density, enterprise distribution, and infrastructure distribution, the system scientifically assesses the trafficability, convenience, and accessibility of the transportation network. Furthermore, it tailors optimization solutions for each region, significantly improving the traffic capacity and convenience service level of the transportation network.

[0069] 3. Fully consider the needs of different transportation stakeholders, optimize resource allocation, promote fairness in transportation services, and drive coordinated regional economic and social development.

[0070] 4. Based on rich road network information, traffic flow, pedestrian flow and other data, realize data-driven traffic network evaluation and optimization, and improve the scientificity and reliability of decision-making. Attached Figure Description

[0071] Figure 1 This is a schematic flowchart of the method of the present invention;

[0072] Figure 2 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0073] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0074] like Figure 1 As shown, this invention provides a method and system for evaluating and optimizing urban traffic networks based on different stakeholders. It makes full use of data such as road network information, traffic flow, and pedestrian flow to construct three directed and empowered complex networks representing different stakeholders: a high-speed accessibility and convenience network representing high-speed access stakeholders; an urban motor vehicle road accessibility and convenience network representing residents' driving interests; and a public transportation network representing those who take public transportation interests.

[0075] Then, based on these three different networks, this invention uses an edge-weighted community detection algorithm and a node importance algorithm, combined with data such as population density, enterprise distribution, infrastructure service distribution, and commercial areas, to evaluate the accessibility, convenience, and availability of each transportation network for different traffic stakeholders. Specifically, the weighted edge-weighted community detection algorithm is used to divide each network into communities, and the traffic flow, pedestrian flow, vehicle speed, and average travel time of each community's corresponding area are evaluated. Network connectivity analysis is used to evaluate the traffic accessibility of each community's corresponding area. Furthermore, community detection is combined with a vertex prediction algorithm. First, the weighted edge-weighted vertex importance algorithm is used to find important nodes in the network, and then the rationality, function, and shortcomings of the important traffic nodes corresponding to each community are evaluated. Finally, based on data such as population density, enterprise distribution, infrastructure service distribution, and commercial areas, combined with accessibility, convenience, and availability, this project proposes urban transportation network optimization schemes such as road repair and widening, road construction, and the addition of bus routes in the areas corresponding to each community, significantly improving the city's traffic capacity and service level.

[0076] The specific process is as follows:

[0077] 1. Data standardization

[0078] Missing value supplementation

[0079] To more effectively imput missing values ​​in the dataset, we selected the First Five Last Three Logistic Regression Imputation (FTLRI) method. This method uses a sliding window-based "first five, last three" model to handle missing data, where "first five (F)" represents the five data points with complete values ​​immediately preceding the data point with a missing value, and "last three (T)" represents the three data points with complete values ​​immediately following the data point with a missing value. The basic steps of the First Five Last Three (FT) model are as follows:

[0080] Step 1 If an attribute p With an attribute that has missing values q Pearson correlation between r pq If the value is ≥0.6, then this attribute is the target attribute.

[0081] Step 2: If and only if F (first Five) and T (last Three) are data points in the dataset with complete values ​​consisting of the target attribute and the attribute with missing values, find the first five and last three data points that are temporally related to the data point with missing values ​​by using a sliding window in Step 1.

[0082] After selecting eight data points with complete attribute values ​​using the FT algorithm, FTLRI will train an imputed model suitable for these eight data points using a logistic regression algorithm. Finally, the missing values ​​will be obtained using this imputed model.

[0083] Anomaly handling

[0084] We employ a density-based approach combined with the Pauta criterion to detect outliers in the dataset. Based on the calculated density of each data point, if the data point... x If the condition is met, the point is identified as an outlier.

[0085] (1)

[0086] in, For data points x density, The average of the inverse of the density within the dataset. The standard deviation of the inverse density within the dataset.

[0087] Data Integration

[0088] To obtain more comprehensive and accurate data to support subsequent analysis, we systematically integrated multiple datasets, including nodes_all, edges_all, nodes_drive, edges_drive, Bus_Stops, and Bus_Routes. Based on the relationships between datasets (such as data point IDs, temporal relationships, and spatial relationships), we integrated important feature attributes from different datasets to construct multiple unified datasets. This process not only preserved the key information of the original data but also enhanced the multidimensionality and completeness of the data through feature extraction and attribute association.

[0089] 2. Transportation Network Construction

[0090] To ensure our network reflects the demands of different interest groups, we constructed three traffic networks based on three different weighting methods and datasets, each representing a different interest group: a high-speed accessibility and convenience network for those who benefit from high-speed access; an urban motor vehicle accessibility and convenience network for those who benefit from driving; and a public transportation network for those who benefit from public transportation.

[0091] Construction of a high-speed accessibility and convenience network representing the interests of high-speed travelers

[0092] (1) Network nodes and edges: The entrances and exits, toll stations, service areas, intersections and monitoring points of each highway in and around the city are taken as nodes of the traffic network, and the actual path between two nodes is taken as the edge between the two points. All the nodes and edges form a complex network.

[0093] (2) The direction of the edges in the network: The direction of the vehicle's travel between two points is taken as the direction between two nodes in the complex network. Since the vehicle travels in both directions at high speeds, this complex network is an asymmetric network, and the direction between two nodes is also bidirectional.

[0094] (3) Road data: Obtain the average daily traffic flow, maximum speed limit and number of lanes for each direction of each edge in the network.

[0095] (4) Weight between two points

[0096] (2)

[0097] in It represents the average daily traffic flow of the edges between node i and node j. It is the highest daily traffic volume on the network. Let $i$ be the maximum speed limit for the edge between node $i$ and node $j$. The number of lanes between node i and node j. and This is to scale the data to the same range.

[0098] (5) The physical significance of cyberspace and the stakeholders it represents

[0099] because It is the average daily traffic flow between the two points, therefore The larger the value, the more people use the road, and the better the convenience of that section. Therefore, the first term of formula (2) reflects the convenience of this highway network. For the second term of formula (2), the number of lanes... The more, the faster the speed limit. The larger the volume, the higher the average daily traffic volume. The smaller the value, the more vehicles pass through this section of road per unit of time, so this value represents the convenience of this section of road.

[0100] Because this network uses highway data, it can be viewed as a network representing the high-speed accessibility and convenience for those who benefit from high-speed travel.

[0101] A network representing the interests of urban drivers regarding urban road accessibility and convenience.

[0102] (1) Network nodes and edges: All traffic light intersections in this city are taken as nodes of the traffic network, and the actual path between two nodes is taken as the edge between the two points. All points and edges form a complex network.

[0103] (2) The direction of the edges in the network: The direction of vehicle travel between two points is taken as the direction between two nodes in the complex network. Since many roads in the city are bidirectional, this complex network is an asymmetric network, and the direction between two nodes is also bidirectional.

[0104] (3) Road data: Obtain the traffic flow, maximum speed limit and number of lanes for each direction of each edge in the network.

[0105] (4) Weight between two points

[0106] (3)

[0107] in It represents the average daily traffic flow of the edges between node i and node j. It is the highest daily traffic volume on the network. Let $i$ be the maximum speed limit for the edge between node $i$ and node $j$. The number of lanes between node i and node j. and This is to scale the data to the same range.

[0108] (5) The physical significance of cyberspace and the stakeholders it represents

[0109] because It is the average daily traffic flow between the two points, therefore The larger the value, the more people use the road, and the better the convenience of that section. Therefore, the first term of formula (2) reflects the convenience of this highway network. For the second term of formula (2), the number of lanes... The more, the faster the speed limit. The larger, The lower the traffic volume, the more vehicles pass through this section of road per unit time, so this item represents the convenience of this section of road.

[0110] Because this network uses urban motor vehicle road data, it can be viewed as a network representing the interests of urban drivers and improving the accessibility and convenience of urban motor vehicle roads.

[0111] Public transport network construction representing the interests of public transport users

[0112] (1) Network nodes and edges: Each bus, subway, and light rail station in this city and its suburbs is taken as a node of the transportation system, and the actual transportation route between two points is taken as an edge between the two points. All the points and edges constitute a complex network.

[0113] (2) The direction of edges in the network: The direction of traffic flow on the public transportation route between two points is taken as the direction between two nodes in the complex network. Since the vehicles on the public transportation route travel in both directions, this complex network is an asymmetric network, and the direction between two nodes is also bidirectional.

[0114] (3) Road data: Obtain the average daily traffic flow in each direction of each edge in the network, the total number of passengers getting on and off at transfer stations, and the maximum average traffic flow of a single bus stop in this city.

[0115] (4) Weight between two points:

[0116] (4)

[0117] in Let $\mathbf{i}$ be the average daily traffic flow between nodes $i$ and $j$, and $max(ADT)$ be the maximum average daily traffic flow in the public transport network. and They represent two adjacent stations respectively. i, j The number of passengers getting on and off the bus is divided by 2 to separate the two nodes. i and j The number of passengers between them was averaged. This refers to the maximum number of passengers boarding and alighting at a single node in a public transportation system.

[0118] (5) The physical significance of networks and the stakeholders they represent:

[0119] because It is the average daily traffic flow between the two points, therefore The larger the value, the more people use the public transportation line, and the better the convenience of the line. Therefore, the first term of formula (4) reflects the convenience of this public transportation network. For the second term of formula (4), the total number of passengers getting on and off at the two stations... The larger the area, the greater the traffic flow between the two points. The smaller the value, the more likely it is to represent the two nodes of this transportation route. i, j The more passengers a public transport network has, the more rational its operation becomes.

[0120] Since this network is a public transportation network, it can be viewed as a public transportation network representing the interests of those who ride public buses.

[0121] 3. Assessment

[0122] Weighted edge-based community detection algorithm

[0123] (1) Improve the Label Propagation Community Detection (LPA) algorithm to enable it to handle weighted edge networks.

[0124] a. Label propagation algorithm

[0125] The Label Propagation Algorithm (LPA) is the earliest proposed community detection algorithm based on label propagation. Due to its simplicity and near-linear time complexity, it has been widely adopted by researchers. The LPA algorithm is based on the assumption that each node in the network chooses to join the community with the largest number of its neighbors. It initializes each node with a unique label and allows the label to propagate through the undirected, unweighted graph. As the label propagates, densely connected groups of nodes quickly reach a consensus on a unique label. When many such dense consensus groups are created throughout the network, they continue to expand outwards as much as possible. When the propagation process ends, nodes with the same label are assigned to the same community.

[0126] The LPA algorithm employs an information propagation mechanism. It detects communities by propagating the label information of nodes in the network. The specific execution process of the algorithm is as follows:

[0127] (1) Initialize the labels of all nodes in the network and assign a unique label to each node.

[0128] (2) Arrange the nodes in the network in a random order to obtain a random node sequence, denoted as X.

[0129] (3) Traverse the nodes in X in order and update the label of each node with the label that appears most frequently among its neighboring nodes.

[0130] (4) If the label of each node in the network is consistent with the label that appears most frequently in its neighboring nodes, the algorithm terminates; otherwise, repeat steps (2) and (3).

[0131] Following the steps described above, closely connected nodes in the network are quickly assigned the same label, and eventually, nodes with the same label form a community.

[0132] b. Weighted edge label propagation algorithm capable of handling traffic networks

[0133] The LPA algorithm studies undirected, unweighted networks. This means that during the algorithm's iterations, label propagation is undirected, and each node has equal importance. A node's label update depends on the label that appears most frequently among its neighbors. However, in a transportation network, a transportation node's contribution to the transportation of its neighbors is not necessarily equal. Therefore, in a directed, weighted transportation network, a node's label depends on the label of the community with the largest cumulative weight among all directed edges pointing to it.

[0134] To identify associated communities from each constructed traffic network, we propose a weighted edge-based LPA algorithm that considers weights and directions. The algorithm is executed in two steps: First, it initializes the labels of all nodes in the traffic network, ensuring each node initially belongs to an independent community. Second, it performs an iterative label propagation process. First, the nodes in the network are arranged in a random order, resulting in a random node sequence, denoted as X. Then, the algorithm iterates through the nodes in X, asynchronously updating each node to be assigned to the community with the largest cumulative weight of all directed edges linking to it. The algorithm terminates once every node has been assigned to the community with the largest cumulative weight of all directed edges linking to it. Ultimately, several communities are obtained, where nodes within the same community have the same label, and nodes between different communities have different labels. Each community corresponds to an associated traffic node region.

[0135] By using an improved LPA algorithm to detect communities in a traffic network, interconnected traffic nodes eventually form several communities. It's worth noting that because the label update order of traffic nodes is randomized in the new algorithm, the detected community structure may differ each time. To obtain stable and high-quality community detection results, we run the algorithm multiple times to find the community structure with optimal modularity as the final community partitioning result.

[0136] (2) Use the weighted edge-adding LPA algorithm to detect the communities in the above five networks respectively.

[0137] Weighted edge vertex importance algorithm

[0138] (1) Improve the PageRank algorithm by adding weighted edges so that it can handle weighted edge networks.

[0139] a. PageRank algorithm

[0140] PageRank is an algorithm used by the Google search engine. It treats the internet as a giant directed graph, where web pages are represented as nodes and links between pages are represented as directed edges. The PageRank algorithm assumes that visitors on each web page have an equal probability of jumping to the next page based on hyperlinks on that page, and that this random jumping continues across the network. This process forms a random walk model, also known as a first-order Markov chain. The algorithm assigns an initial PageRank (PR) value to each web page and then iteratively updates these PR values ​​until the change in the PR value is sufficiently small or a predetermined number of iterations is reached. The formula for calculating the PR value of node vi is shown below.

[0141] (5)

[0142] in, The probability that a user will continue browsing after visiting a certain page is called the damping factor. It points to a node The set, It is a node The number of edges that extend from the chain.

[0143] b. A weighted edge-based PageRank algorithm capable of handling traffic networks.

[0144] In transportation networks, the importance of different nodes varies significantly, with key nodes typically having a greater impact on traffic. To analyze transportation systems more accurately, we employ an improved PageRank algorithm to rank the importance of nodes in the network. This algorithm comprehensively considers the topological characteristics of nodes and traffic flow attributes, enabling us to accurately identify key nodes in the network. Based on the ranking results, we further select nodes with higher importance for in-depth analysis to reveal the characteristics of the transportation system.

[0145] For key traffic data points in urban transportation networks, we made two major improvements to the PageRank algorithm. The original PageRank algorithm only works for directed, unweighted networks. During iteration, a page's PR value is evenly distributed among all the pages it points to by dividing it by the total number of its forward links. However, in a transportation network, a node's traffic contribution to its surrounding nodes is not necessarily equal. Its traffic contribution should be distributed to its pointed-to neighboring nodes according to the weight of its directed edges. Therefore, in our improved algorithm, a node's PR value is proportionally distributed to all neighboring nodes according to the weight of its forward links. Based on this, we redefined the importance of each node, called the Weighted PageRank (W-PR).

[0146] (6)

[0147] in, Represents a node The set of nodes it points to. Indicates a path from other nodes to a node. The cumulative sum of edges, Representing an edge Weights.

[0148] The process of the weighted directed PageRank algorithm is as follows:

[0149] 1. Initialize weights: Assign an initial W-PR value to each node. For fairness, the initial W-PR value for each node is usually set to an equal probability.

[0150] 2. Iterative calculation: The W-PR value of each node is calculated iteratively according to formula (6), and the difference between the W-PR value of the current node and the W-PR value of the previous round is calculated.

[0151] 3. Termination condition: Repeat step 2 until the difference is less than the preset convergence threshold or the preset number of convergences is reached.

[0152] (2) Use the weighted edge-adding PageRank algorithm to detect important nodes in each network.

[0153] An assessment will be conducted for each community group's corresponding region.

[0154] 1) Assessment of traffic capacity and convenience for residents

[0155] a) For the high-speed accessibility and convenience network, combine data such as suburban population density, urban population density, out-of-city population density, enterprise distribution, and infrastructure service distribution to analyze the accessibility and convenience of the highways in the corresponding areas of each community in the high-speed transportation network for highway users.

[0156] b) For the urban motor vehicle road traffic and convenience network, based on data such as urban population density, commercial areas, enterprise distribution and infrastructure service distribution, analyze the traffic capacity and convenience of urban roads in the corresponding area of ​​each community in the public transport network for urban drivers.

[0157] c) For public transport networks, analyze the convenience of bus routes in each community area for public transport users by combining data such as population density, number of passengers, enterprise distribution, infrastructure service distribution, and commercial areas.

[0158] 2) Transportation accessibility assessment

[0159] Each community detected in each network is treated as an independent graph. Graph connectivity analysis is used to evaluate the traffic accessibility of the area corresponding to each community, and the rationality and reasons for insufficient accessibility are analyzed at the same time.

[0160] 3) Evaluation of key nodes within a club for club-wide testing and key node testing.

[0161] Based on the key nodes in each community, the actual key traffic nodes in the corresponding areas are identified. Combining data such as population density, enterprise distribution, infrastructure service distribution, and commercial locations, the traffic flow, pedestrian flow, vehicle speed, and average travel time of the area corresponding to each actual key traffic node are evaluated.

[0162] 4. Based on the evaluation results, propose optimization plans for each community group, focusing on key milestones.

[0163] (1) For the high-speed accessibility and convenience network, based on the above evaluation results, optimization schemes such as building new roads, widening local roads and setting local speed limits are proposed according to the important nodes in the community.

[0164] (2) Based on the above assessment results, optimization schemes such as building new roads, widening local roads, and changing local speed limits are proposed for the traffic and convenience network of urban motor vehicle roads, according to the important nodes in the community.

[0165] (3) For the public transport network, based on the above evaluation results, and according to the important nodes in the community, propose optimization schemes to increase or decrease bus routes and bus stops.

[0166] Systems used to perform the above methods include:

[0167] The transportation network construction module is used to build three types of transportation networks that represent different groups of transportation stakeholders.

[0168] The assessment module is used to detect communities in each transportation network and divide them into communities. Combining data such as population density, enterprise distribution, infrastructure service distribution, and commercial areas, it is used to assess the rationality of important nodes within each community, the accessibility of the corresponding area of ​​each community, and the convenience of transportation in the corresponding area of ​​each community.

[0169] The optimization scheme generation module is used to propose corresponding urban transportation network optimization schemes for each community based on the evaluation results.

[0170] The transportation network construction module includes:

[0171] The high-speed accessibility and convenience network construction unit is used to determine the nodes, edges, weights, and directions of the high-speed accessibility and convenience network.

[0172] The urban motor vehicle road traffic and convenience network construction unit is used to determine the nodes, edges, weights, and directions of the urban motor vehicle road traffic and convenience network.

[0173] The public transport network construction unit is used to determine the nodes, edges, weights, and directions of the public transport network.

[0174] The evaluation module employs a weighted edge-based community detection algorithm and a weighted edge-based vertex importance algorithm. The evaluation module includes a community detection unit, an evaluation unit for important nodes within a community, and an evaluation unit for traffic convenience within a community.

[0175] The optimization scheme generation module includes a high-speed traffic flow and public convenience network optimization scheme generation unit, an urban motor vehicle road traffic flow and public convenience network optimization scheme generation unit, and a public transport network optimization scheme generation unit.

[0176] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0177] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for evaluating and optimizing urban transportation networks based on different stakeholders, characterized in that: Includes the following steps: S1: Construct three types of transportation networks that represent different traffic interest groups, including a high-speed accessibility and convenience network representing those who benefit from high-speed access, an urban motor vehicle accessibility and convenience network representing those who benefit from driving, and a public transportation network representing those who benefit from taking public transportation. The steps for constructing the three types of transportation networks include: Determine the nodes and edges of the high-speed accessibility and convenience network, with each highway entrance / exit, toll station, service area, intersection, and monitoring point in and around the city as nodes, and the actual path between two nodes as edges. Determine the nodes and edges of the urban motor vehicle road network for traffic flow and convenience, using all traffic light intersections in the city as nodes and the actual paths between two nodes as edges; Determine the nodes and edges of the public transport network, using each bus, subway, and light rail station in the city and suburbs as nodes and the actual transportation routes between two points as edges. S2: Perform community detection on each traffic network, divide it into communities, and evaluate the traffic flow, pedestrian flow, vehicle speed, and average travel time in the corresponding area of ​​each community; The step of performing community detection for each traffic network adopts a weighted edge-based community detection algorithm, which includes the following sub-steps: Initialize the labels of all nodes in the network so that each node initially belongs to an independent community; Arrange the nodes in the network in a random order to obtain a random node sequence; Traverse the nodes in the node sequence in order, and assign each node to the community with the largest cumulative weight of all directed edges linked to it; Repeat the above steps until each node has been assigned to the community with the largest cumulative weight of all directed edges linked to it, at which point the algorithm terminates. S3: Assess the accessibility of the area corresponding to each community; S4: Evaluate key nodes in each community by combining data on population density, business distribution, infrastructure service distribution, and commercial locations; The step of evaluating important nodes in each community uses a weighted vertex importance algorithm with added edges, and includes the following sub-steps: Initialize weights: Assign an initial WR value to each node; Iterative calculation: The PageRank algorithm with weighted edge addition is used to iteratively calculate the WR value of each node, and calculate the difference between the WR value of the current node and the WR value of the previous round; Repeat the above steps until the difference is less than the preset convergence threshold or the preset number of convergences is reached. S5: Based on the evaluation results, propose corresponding urban transportation network optimization plans for each community.

2. The urban transportation network evaluation and optimization method based on different stakeholders as described in claim 1, characterized in that, The steps for proposing corresponding urban transportation network optimization schemes for each community include: For high-speed traffic and convenient network, propose solutions for building new roads, widening existing roads, and optimizing speed limits in certain areas based on key nodes within the community. Regarding the traffic flow and convenience network of urban motor vehicle roads, optimization schemes are proposed based on important nodes in the community, including the construction of new roads, widening of existing roads, and local speed adjustments. For the public transportation network, based on the important nodes in the community, propose optimization schemes to increase or decrease bus routes and bus stops.

3. A system for evaluating and optimizing urban traffic networks based on different traffic stakeholders, used to execute the method described in any one of claims 1 to 2, characterized in that, include: The transportation network construction module is used to build three types of transportation networks that represent different groups of transportation stakeholders. The evaluation module is used to detect communities in each transportation network and divide them into communities; combined with data on population density, enterprise distribution, infrastructure service distribution, and commercial areas, it is used to evaluate the rationality of important nodes within each community, the accessibility of the corresponding area of ​​each community, and the convenience of transportation in the corresponding area of ​​each community. The optimization scheme generation module is used to propose corresponding urban transportation network optimization schemes for each community based on the evaluation results.

4. The urban traffic network evaluation and optimization system based on different traffic stakeholders as described in claim 3, characterized in that, The transportation network construction module includes: The high-speed accessibility and convenience network construction unit is used to determine the nodes, edges, weights, and directions of the high-speed accessibility and convenience network. The urban motor vehicle road traffic and convenience network construction unit is used to determine the nodes, edges, weights, and directions of the urban motor vehicle road traffic and convenience network. The public transport network construction unit is used to determine the nodes, edges, weights, and directions of the public transport network.

5. The urban traffic network evaluation and optimization system based on different traffic stakeholders as described in claim 3, characterized in that, The evaluation module employs a weighted edge-based community detection algorithm and a weighted edge-based vertex importance algorithm, including a community detection unit, an important node evaluation unit within a community, and a traffic convenience evaluation unit within a community.

6. The urban traffic network evaluation and optimization system based on different traffic stakeholders as described in claim 3, characterized in that, The optimization scheme generation module includes a high-speed trafficability and convenience network optimization scheme generation unit, an urban motor vehicle road trafficability and convenience network optimization scheme generation unit, and a public transport network optimization scheme generation unit.

Images