Method for evaluating function resilience of urban multi-modal transportation network under flood disaster

Abstract

The application provides a kind of flood disaster under the evaluation method of urban multi-mode traffic network function resilience. The method comprises: obtaining research city road network data, ground bus network and rail transit network data, and constructing a multi-mode traffic network structure model. POI data is collected to estimate the traffic generation and attraction of the city, and individual daily travel chain is simulated. The multi-mode traffic network structure model and the individual daily travel chain are fused to construct the multi-mode traffic network function model of the city. According to the spatial damage characteristics of flood disaster, the failure model of the multi-mode traffic network function model is constructed based on the structure failure model of the multi-mode traffic network, and the function resilience of the multi-mode traffic network function model under flood disaster is evaluated from the aspects of passable demand and travel delay. The method of the application can be applied to the evaluation of disaster prevention and mitigation strategies before disaster, the evaluation of infrastructure repair strategies after disaster and other aspects, and the urban resilience is enhanced, so as to realize sustainable development.

Description

Technical Field

[0001] This invention relates to the field of traffic management technology, and in particular to a method for assessing the functional resilience of urban multimodal transportation networks under flood disasters. Background Technology

[0002] A well-developed transportation network is a key pillar of economic prosperity, contributing to poverty eradication and sustainable development. Multimodal transportation systems, by coupling multiple modes of transport, can reduce traffic congestion, improve network accessibility, and optimize system efficiency. However, existing research indicates that the interdependence of different modes of transport in multimodal networks can lead to systemic cascading failures: that is, the disruption of one mode of transport may negatively impact other modes, even causing the collapse of the entire transportation system.

[0003] Floods, as one of the most frequent and dangerous extreme events, pose a threat to the safe operation of transportation infrastructure. Floods cause direct physical damage and direct economic losses. It is estimated that natural disasters cause between US$3.1 billion and US$22 billion in physical damage to road and rail assets globally each year, of which about 73% is caused by river floods and monsoon floods. In addition, transportation infrastructure provides opportunities for employment, education, and healthcare; disruption of transportation systems will hinder economic development, social stability, and welfare, resulting in indirect economic losses. Therefore, assessing and understanding the combined impacts of flooding disasters on the structural and functional resilience of urban multimodal transportation networks and improving their resilience to promote sustainable development is of great significance.

[0004] Currently, methods for assessing the functional resilience of transportation systems in response to floods primarily focus on single-modal transportation networks, such as private car networks. These methods employ data-driven or traffic assignment approaches to predict or simulate passenger flow distribution in single transportation networks, such as road networks, after damage, to measure the impact of sudden events like floods on transportation networks. A few scholars have focused on multimodal public transportation networks that couple surface public transport and rail public transport systems, applying cascading failure models to assess the impact of sudden events on the carrying capacity of integrated public transport systems. However, these methods all neglect the complex and dynamic competitive and cooperative relationships between private and public transport systems, and these interactions can amplify or weaken disturbances, thereby affecting the system's carrying capacity and functional resilience.

[0005] Therefore, current urban transportation network functional resilience assessment techniques have the technical problem of being unable to assess the functional resilience of complex multimodal transportation systems coupled with private and public transportation systems under flood disasters. Summary of the Invention

[0006] This invention provides a method for assessing the functional resilience of urban multimodal transportation networks under flood disasters, so as to effectively measure the impact of specific flood disasters on urban integrated transportation systems and the functional resilience of integrated transportation systems.

[0007] To achieve the above objectives, the present invention adopts the following technical solution.

[0008] A method for assessing the functional resilience of urban multimodal transportation networks under flood disasters includes:

[0009] Acquire road network data, surface public transport network data, and rail public transport network data of the city under study, and construct a multimodal transportation network structure model;

[0010] By collecting Points of Interest (POI) data, the traffic generation and attraction of the city are estimated, and the gravity model is improved based on the traffic decay function to simulate the daily travel chain of individuals.

[0011] By integrating the multimodal transportation network structure model and the individual daily travel chain, a multimodal transportation network functional model of the city is constructed, and the multimodal transportation network functional model is calibrated based on real data.

[0012] Based on the spatial damage characteristics of floods, a structural failure model for a multimodal transportation network is constructed.

[0013] Based on the structural failure model of the multimodal transportation network, a failure model of the multimodal transportation network functional model is constructed, and the functional resilience of the multimodal transportation network functional model under flood disaster is evaluated from the aspects of accessibility demand and travel delay.

[0014] Preferably, the step of acquiring road network data, surface public transport network data, and rail public transport network data of the study city, and constructing a multimodal transportation network structure model, includes:

[0015] The road network dataset of the research city was obtained from the developed street map. This road network dataset includes the geospatial location, topological relationship and road segment attributes of the road network nodes. The road network dataset was cropped using the administrative division data of the research city. Intersections were abstracted as nodes and road segments were abstracted as edges.

[0016] Improve road segment attribute information, which includes free-flow speed, number of lanes and traffic capacity; set ground public transport network data, including ground public transport station information, route information and departure timetable information; and map ground public transport stations and routes to the road network.

[0017] The data for setting up the rail transit network includes rail transit station information, route information, and departure timetable information. The station information and route information of the rail transit are obtained, and the rail transit stations are abstracted as nodes, and the topological relationship between two adjacent rail stations of a bus route is abstracted as edges.

[0018] Preferably, the mapping of ground bus stops and routes to the road network includes:

[0019] Based on an open map platform, obtain the latitude and longitude information of ground public transport stations and the sequence information of public transport stations that the routes pass through in sequence;

[0020] Based on the latitude and longitude of the station, the nearest neighbor matching method is used to match each bus stop to the nearest road segment;

[0021] For any two adjacent bus stops in all bus routes, use Dijkstra's algorithm to find the shortest path sequence between the road segments where these two adjacent stops are located, and use this shortest path sequence as the road segments that the bus route passes through in sequence.

[0022] Preferably, the process of collecting POI data to estimate the city's traffic generation and attraction, improving the gravity model based on the traffic decay function, and simulating individual daily travel chains includes:

[0023] The study area was divided into grids with a spatial resolution of 1km × 1km. It was assumed that all origin and destination points were located at the centroid of the grid. POI data for the study area was crawled from an open map platform. Residential communities, dormitories, and mixed-use buildings were identified as origin points. The average area *s* of these three types of origin points was determined. m Traffic occurrence rate per unit area during weekday morning rush hour (a) m Determine the traffic occurrence rate α of the three types of travel origins. m :

[0024] α m =a m s m m∈{residential communities, dormitories, mixed-use commercial and residential buildings} (1)

[0025] Statistically summarize and aggregate the traffic volume of all POIs (Points of Interest) within each grid, and calculate the traffic volume P for grid i. i The calculation formula is:

[0026]

[0027] Where k is the k-th POI of class m, It is a 0-1 variable. If the k-th m-class POI is in the i-th grid, it is 1; otherwise, it is 0.

[0028] Other points of interest (POIs) are considered as travel destinations, including companies, schools, and hospitals. The average area s of the travel destinations of these other POIs is determined. m Traffic attraction rate per unit area during weekday morning rush hour (b) mThe traffic attraction rate β of the other POIs is obtained. m ;

[0029] β m =b m s m m∈{company, school, hospital, etc.} (3)

[0030] The traffic attraction of all travel destinations within each grid is statistically analyzed and summarized, and this is calculated as the attractiveness Q of that grid. i :

[0031]

[0032] The function for the decrease in traffic flow with travel distance is set as follows:

[0033] D(x ij )=(x ij +x0) -β exp(-x ij / K) (5)

[0034] Where, x ij The distance traveled is represented by x0, the minimum distance is K, the cutoff value is β, and the control parameter is set to x0 = 1.5km, K = 80km and β = 1.75.

[0035] The flow T from grid i to grid j ij for:

[0036]

[0037] Based on the daily departure time frequency statistics, a daily travel demand chain is simulated. It is assumed that travelers have only two trips a day, that is, the round trips during the morning and evening rush hours are symmetrical. The departure time required for each commuting trip is estimated from the daily departure time frequency survey results. The minimum threshold of 4 hours between the two trips is used to estimate the departure time required for each commuting return trip.

[0038] Preferably, the process of integrating the multimodal transportation network structure model and the individual daily travel chain to construct the city's multimodal transportation network functional model, and calibrating the multimodal transportation network functional model based on real data, includes:

[0039] Multimodal traffic network structure data and individual daily travel chain data are input into the MATSim multi-agent traffic simulation platform model, and activity utility parameters, car travel cost parameters and public transportation travel cost parameters are initialized in the configuration file respectively.

[0040] The city’s average annual GDP, population and average annual working hours per person are obtained from statistical yearbooks. The average unit time value per person in the city is calculated as the hourly utility parameter of the activity.

[0041] The travel cost per unit mile of a car is determined based on fuel consumption per kilometer and fuel price. The parking cost of a car is improved based on the average parking time and parking fee standard, and the cost parameters per unit travel time are initialized.

[0042] The fare cost of public transportation is determined based on the single-trip fare of ground public transport and rail public transport, and the single-transfer cost and unit travel time cost parameters of public transportation are initialized.

[0043] Multimodal traffic network structure data and daily travel chain data are input into the multi-agent traffic simulation platform MATSim. Activity utility parameters, car travel cost parameters, and public transportation travel cost parameters are initialized in the configuration file. Based on the co-evolutionary algorithm, each individual in the multi-agent traffic simulation platform simultaneously optimizes its own travel chain, including adjusting the departure time, mode of transportation, and travel route of each individual, with the goal of improving the daily travel utility of each individual, until all individuals in the system can no longer benefit from adjusting their own travel plans. Through the continuous convergence and iteration of the multi-agent traffic simulation platform, a user equilibrium state is reached, and the initial traffic network functional model is obtained.

[0044] Based on the average travel time, public transport modal share, and average travel distance output from each simulation, the fixed costs of travel utility for cars and public transport are adjusted until the error is adjusted to a reasonable range. At this point, the total travel volume output by the model is T0, and the average travel time is t0.

[0045] Preferably, the construction of a structural failure model for a multimodal transportation network based on the spatial damage characteristics of floods includes:

[0046] Based on the spatial distribution characteristics of flood events, directly failed nodes and road segments in the multimodal transportation network structure are identified and eliminated. For the road network, a dual method is used, treating edge segments as points and intersections as edges. The inundation depth of midpoint nodes in each road segment is calculated. For the rail network, the inundation depth of initial network nodes is calculated. Inundation thresholds for flood-affected nodes in both the road and rail networks are set. Nodes with flood depths exceeding these thresholds are considered directly failed nodes, and directly failed nodes and their edges are deleted. The node failure criteria for the road network are as follows:

[0047]

[0048] The failure criteria for nodes in the track network are as follows:

[0049]

[0050] Among them, h 0 mThe value represents the step height of the subway station, h represents the submerged water depth of the node, μ represents the submerged threshold of the node, and I is a 0-1 variable, where I = 0 indicates that the node is ineffective and I = 1 indicates that the node is effective.

[0051] Based on network connectivity, identify and eliminate indirect failures in the multimodal transportation network structure. Identify the connected subgraphs and the maximum connected subgraphs of the road network. Treat nodes in non-maximum connected subgraphs as indirect failure nodes and delete the nodes and edges of the nodes. Treat bus stops of bus routes that cannot operate as indirect failures and delete the nodes and edges of the nodes.

[0052] Preferably, the functional failure model for constructing a multimodal transportation network functional model based on the structural failure model of the multimodal transportation network evaluates the functional resilience of the multimodal transportation network functional model under flood disasters from two aspects: accessibility demand and travel delay, including:

[0053] Based on the spatial distribution characteristics of flood disasters, travel demand that is directly inundated and indirectly interrupted by floods is identified. Based on the spatial distribution pattern of floods, the inundation depth of the grid centroid is calculated. Grids with a water depth exceeding the inundation threshold are identified as directly inundated by floods. All travel from grids that serve as the starting or ending point is considered as directly inundated travel demand. Travel from grids whose starting or ending points are surrounded by floods is considered as indirectly interrupted travel demand.

[0054] Using the centroids of the grid as nodes, and connecting any two adjacent centroids of the grid as edges according to the Moore neighbor rule, construct a centroid network. Delete the nodes and edges corresponding to the centroids of directly invalid grids. Calculate the maximum connected subgraph of the remaining network. The centroids of the grids corresponding to nodes not on the maximum connected subgraph are regarded as "islands" surrounded by floods, and the travel demand they generate or attract is deleted.

[0055] By integrating the multimodal transportation network structure model after removing structural failures and the daily travel chain after removing flood-induced demand, the remaining multimodal transportation network structure and remaining travel demand are input into the multi-agent transportation simulation platform MATSim. Based on the co-evolutionary algorithm, each remaining individual aims to maximize its own daily travel utility and continuously adjusts its travel plan on the remaining multimodal transportation network, including departure time, mode of transportation, and travel route, until each individual can no longer benefit from adjusting its travel plan. At this point, the model iteratively converges, reaching a user equilibrium state, and the functional failure model of the multimodal transportation network is obtained.

[0056] The average travel time and accessibility demand are calculated using MATSim. When an individual's daily travel utility converges, the model outputs travel volume T1, average travel time t1, and accessibility demand ratio F. trip for:

[0057]

[0058] The relative travel delay is:

[0059]

[0060] Passable demand ratio F trip Between 0 and 1, F trip The closer the value is to 1, the smaller the impact of flooding on the carrying capacity of the city's multimodal transportation system, and the stronger its functional resilience; conversely, the closer the value is to 1, the weaker the impact of flooding on the system's carrying capacity and the stronger its functional resilience. trip The closer the value is to 0, the greater the impact of the flood on the carrying capacity of the city's multimodal transportation system, and the worse its functional resilience. The relative travel delay τ reflects the change in travel time for the remaining passable demand. A negative relative delay indicates that the remaining demand is more unimpeded on the remaining road network after the disaster; conversely, a positive relative delay indicates that the remaining demand is more congested on the remaining road network.

[0061] As can be seen from the technical solutions provided by the embodiments of the present invention above, the method of the present invention can be applied to pre-disaster road network and rail network planning and design, post-disaster infrastructure recovery strategy optimization and emergency management measures, etc., to enhance urban resilience and thus achieve sustainable development.

[0062] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of the invention. Attached Figure Description

[0063] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments 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 these drawings without creative effort.

[0064] Figure 1 A diagram illustrating the mechanism by which floods disrupt the structure and demand of a multimodal transportation network, provided as an embodiment of the present invention.

[0065] Figure 2 A flowchart illustrating a method for assessing the functional resilience of urban multimodal transportation networks under flood disasters, provided in an embodiment of the present invention.

[0066] Figure 3 This invention provides a modeling result of a multimodal transportation network structure in Nanjing. Detailed Implementation

[0067] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0068] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or couplings. The term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0069] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0070] To facilitate understanding of the embodiments of the present invention, the following will provide further explanation and description with reference to the accompanying drawings and several specific embodiments. These embodiments do not constitute a limitation on the embodiments of the present invention.

[0071] The present invention provides a method for assessing the functional resilience of urban multimodal transportation networks under flood disasters, comprising: constructing a multimodal transportation network structural model using open-source data such as road networks, surface public transport, and rail public transport; simulating daily travel chains of urban individuals using open-source data such as POIs (Points of Interest); integrating the multimodal transportation network structural model and daily travel chains to construct and calibrate a multimodal transportation network functional model; constructing a structural failure model of the multimodal transportation network based on the spatial distribution characteristics of specific flood disasters; constructing a functional failure model of the multimodal transportation network; and assessing the functional resilience of urban multimodal transportation systems under flood disasters from the perspectives of accessibility demand and travel delays.

[0072] Figure 1This invention provides a diagram illustrating the mechanism of flood damage to the structure and demand of a multimodal transportation network. In the diagram, A represents the structural damage of flood to the car network; B represents the structural damage of flood to the public transport network; C represents the demand interruption caused by flood directly inundating the origin or destination; and D represents the demand interruption caused by flood inundating all feasible roads between the origin and destination, even though the origin and destination are not inundated.

[0073] The processing flow of an assessment method for the functional resilience of urban multimodal transportation networks under flood disasters provided in this embodiment of the invention is as follows: Figure 2 As shown, the processing steps include the following:

[0074] Step 1: Collect data on administrative divisions, road networks, surface public transport networks, and rail transit networks to construct a multimodal transportation network structure model.

[0075] Step 1.1: Road network data includes the geospatial location, topological relationships, and road segment attributes of road network nodes. Road network data is obtained from Open Street Map and administrative division data from the Tianditu administrative boundary data interface. The road network dataset is then cropped using the administrative division data of the research city. Intersections are abstracted as nodes, and road segments are abstracted as edges. To ensure the connectivity of the road network, the maximum connected subgraph of the road network is identified, and nodes and edges not on the maximum connected subgraph are deleted.

[0076] Step 1.2: Improve the road segment attribute information according to the "Highway Route Design Specification" (JTG D20—2017). This road segment attribute information includes free-flow speed, number of lanes, and traffic capacity. The surface public transport network data includes station information, route information, and departure timetable information for surface public transport. The surface public transport stations and routes are mapped to the road network, thus integrating the road network data and the surface public transport network data.

[0077] The above process of mapping ground public transport stops and routes to the road network includes:

[0078] (1) First, crawl the ground bus station information (station latitude and longitude) and route information (sequence of bus stops passed through) based on the Gaode Map Open Platform.

[0079] (2) In order to map ground bus stops to the road network, the nearest neighbor matching method is used to match each bus stop to the nearest road segment based on the latitude and longitude of the station.

[0080] (3) In order to map the ground bus routes to the road network, for any two adjacent bus stops in all bus routes, the Dijkstra algorithm is used to find the shortest path sequence between the road segments where the two adjacent stops are located, and the road segment sequence is used as the road segments that the bus routes pass through in sequence.

[0081] Step 1.3: The rail transit network data includes rail transit station information, route information, and departure timetable information. Station information (station latitude and longitude) and route information (passing through rail stations sequentially) are crawled using the Gaode Map Open Platform. Unlike ground public transport networks, the Space L method is applied to construct the rail transit network, which abstracts rail transit stations as nodes and the topological relationship between adjacent rail stations of a bus route as edges.

[0082] Step 2: Collect POI data to estimate the traffic generation and attraction during the morning rush hour, improve the gravity model based on the traffic decay function, and simulate the daily travel chain of individuals.

[0083] Step 2.1: Divide the study area into grids, label the traffic occurrence rate of residential areas and other locations, collect POI data, and estimate the traffic occurrence P of the grid. i Specifically, the study area was divided into grids with a spatial resolution of 1km × 1km. It was assumed that all origin and destination points were located at the centroid of the grid, and short-distance trips within the grid were not considered. POI data for the study area was crawled from the Gaode Maps open platform, with residential communities, dormitories, and mixed-use buildings as the origin points. Referring to the "Transportation Trip Rate Handbook," the average area *s* of these three types of origin points was determined. m Traffic occurrence rate per unit area during weekday morning rush hour (a) m Thus, its traffic occurrence rate α is determined. m :

[0084] α m =a m s m m∈{residential communities, dormitories, mixed-use commercial and residential buildings} (1)

[0085] The traffic volume of grid i is calculated by summarizing the traffic volume of all POIs (points of origin) within each grid.

[0086]

[0087] Where k is the k-th POI of class m, It is a 0-1 variable. If the k-th m-class POI is in the i-th grid, it is 1, otherwise it is 0.

[0088] Step 2.2: Based on equation (2), calibrate the traffic attraction rate of other POIs and determine the attractiveness Q of the grid. iCompanies, schools, hospitals, and other points of interest (POIs) are considered as travel destinations, and the average area (s) of these destinations is determined according to the Transportation Trip Rate Manual. m Traffic attraction rate per unit area during weekday morning rush hour (b) m Thus, its traffic attraction rate β is determined. m ,

[0089] β m =b m s m m∈{company, school, hospital, etc.} (3)

[0090] The traffic attraction of all travel destinations within each grid is statistically analyzed and summarized, and this is calculated as the attractiveness Q of that grid. i :

[0091]

[0092] Step 2.3: Traffic flow has been generated. The traffic flow decay function with travel distance is:

[0093] D(x ij )=(x ij +x0) -β exp(-x ij / K) (5)

[0094] Where, x ij Let x0 represent the minimum distance, K be the cutoff value, and β be the control parameter. Based on relevant literature and my country's national conditions, x0 = 1.5km, K = 80km, and β = 1.75 are set.

[0095] Step 2.4: Generate the traffic distribution T during the morning rush hour ij The flow rate from grid i to grid j is

[0096]

[0097] Step 2.5: Simulate the daily travel demand chain for individuals based on daily departure time frequency statistics. Assume that each traveler makes exactly two trips a day, meaning the outbound and return journeys during morning and evening rush hours are symmetrical. Estimate the required departure time for each outbound journey based on the daily departure time frequency survey results, and estimate the required departure time for each return journey using a minimum threshold of 4 hours between two trips.

[0098] An individual's daily travel chain, or daily travel plan, includes the departure time, origin, and destination of each travel need. Each person has two travel needs every day: the morning rush hour commute to the office, school, or hospital, etc., and the evening rush hour commute back home, etc.

[0099] The departure time is estimated by probability based on the daily departure frequency survey results (i.e., the higher the departure frequency at a certain time, the higher the probability of departure). In order to ensure that the duration of work and study has a minimum value, if the departure time interval between the outbound and return trips is less than 4 hours according to the probability randomization, it is re-initialized until the requirements are met. 4 hours is set as the minimum threshold.

[0100] The origin and destination points of the morning rush hour commute are: the flow rate T from grid i to grid j based on the traffic distribution calculated in step 2.4. ij Assume that these flows originate from the centroid of grid i and terminate at the centroid of grid j.

[0101] The starting and ending points of the evening rush hour return journey are symmetrical to those of the morning rush hour outward journey. Step 3: Based on the multi-agent transportation simulation platform (MATSim, Multi-Agent Transport Simulation), a multi-modal transportation network structure model and daily travel chain are integrated to construct a multi-modal transportation network functional model, and the model is calibrated based on real data.

[0102] Step 3.1: Obtain the city's average annual GDP, population, and average annual working hours per person from the statistical yearbook, and calculate the average unit time value per person in the city as the hourly utility parameter of the activity.

[0103] Step 3.2: Based on fuel consumption per kilometer and fuel price, determine the travel cost per unit mile of the car; based on the average parking time and parking fee standard for a single parking session, refine the parking cost of the car; and initialize the cost parameters per unit travel time.

[0104] Step 3.3: Determine the fare cost of public transportation based on the single-trip fares of ground buses and rail buses, and initialize the single-transfer cost and unit travel time cost parameters of public transportation.

[0105] Step 3.4: Input the multi-modal transportation network structure data and daily travel chain data into the multi-agent transportation simulation platform (MATSim). Initialize the activity utility parameters, car travel cost parameters, and public transportation travel cost parameters in the configuration file. Through continuous convergence and iteration of the multi-agent transportation simulation platform, an initial transportation network functional model is obtained. Specifically, based on the co-evolutionary algorithm, each individual in the multi-agent transportation simulation platform can simultaneously optimize its own travel chain, including adjusting each individual's departure time, mode of transportation, and travel route. This optimization process aims to improve the daily travel utility of each individual, including travel utility (usually negative; the longer the travel time, the lower the daily travel utility) and activity utility (usually positive; the longer the activity time, such as work or study, the higher the daily travel utility). This optimization process continues until all individuals in the system can no longer benefit from adjusting their own travel plans, i.e., a user equilibrium state is reached.

[0106] Step 3.5: Based on the average travel time, public transport modal share, and average travel distance output by the initial traffic network functional model, adjust the fixed costs of travel utility for cars and public transport until the error is adjusted to a reasonable range. At this point, the total travel volume output by the model is T0, and the average travel time is t0.

[0107] A multimodal transportation network functional model is a simulation of a multimodal transportation system, including but not limited to information such as operating speed and traffic flow on each road segment at each time point; the departure time of each individual's trip, the road segments they pass through, and each mode of transportation;

[0108] The multimodal transportation network functional model is primarily constructed using open-source simulation software (MATSim, a multi-agent transportation simulation platform) that integrates traffic demand (referred to as the daily travel chain in this method) and traffic supply (referred to as the multimodal transportation network structure model in this method). To ensure that this functional model aligns with real-world transportation systems, we select indicators such as average travel time, public transport modal share, and average travel distance. By adjusting the activity utility parameters, private car travel cost parameters, and public transport travel cost parameters, we ensure that the results of these indicators in the functional model are consistent with the actual results of statistical surveys. Traditionally, these parameters are calibrated through random trials, while this method proposes a simple and feasible parameter calibration method (from steps 3.1-3.5). The "adjustment of the fixed costs of private car and public transport travel utility" mentioned in step 3.5 is also part of the model calibration process.

[0109] Multimodal transportation network functional models emphasize both multimodal transport (involving cars, public transport, and rail transit) and network functionality (compared to multimodal transportation network structural models, this functional model involves modeling the dynamic mechanisms within the network, i.e., modeling individual travel movements). Multi-agent traffic simulation models, in terms of modeling methodology, utilize multi-agent modeling methods, considering each individual's travel plan as a single agent continuously seeking their optimal travel plan (travel chain). However, a multi-agent traffic simulation platform (MATSim) is not equivalent to a multi-agent traffic simulation model; the former is a necessary condition for the latter.

[0110] Step 4: Based on the spatial damage characteristics of floods, construct a structural failure model for a multimodal transportation network.

[0111] Step 4.1: Based on the spatial distribution characteristics of flood events, identify and remove directly failed nodes and road segments in the multimodal transportation network structure. Inundation in the road system is concentrated on road segments, while inundation in the rail system is concentrated on stations. Therefore, the road network uses a dual method, treating edges (road segments) as points and points (intersections) as edges, calculating the inundation depth of each node (midpoint of a road segment). The rail network, however, only needs to calculate the inundation depth of the initial network nodes. Inundation thresholds are set for road and rail network nodes. Nodes with flood depths exceeding these thresholds are considered directly failed nodes, and their edges are deleted. The failure criteria for road network nodes are as follows:

[0112]

[0113] The failure criteria for nodes in the track network are as follows:

[0114]

[0115] Among them, h 0 m Let h represent the step height of the subway station, μ represent the submerged water depth of the node, μ represent the submerged threshold of the node, and I is a 0-1 variable, where I = 0 indicates that the node is ineffective and I = 1 indicates that the node is effective.

[0116] Step 4.2: Based on network connectivity, identify and eliminate indirect failures in the multimodal transportation network structure. For the road network, identify the connected subgraph (C) and the maximally connected subgraph (G), and treat nodes in the non-maximally connected subgraph as indirect failure nodes, deleting the indirect failure nodes and their edges, such as... Figure 1 As shown in (A); for a public transport network, the failure of any bus stop will cause the entire bus route to fail. Therefore, bus stops that are not flooded but have no operational bus routes are considered indirectly failed. Delete the indirectly failed nodes and their edges, as shown in (A). Figure 1As shown in (B).

[0117] Step 5: Based on the structural failure model of the multimodal transportation network, construct a functional failure model of the multimodal transportation network, and evaluate the functional resilience of the urban multimodal transportation system under flood disasters from the aspects of accessibility demand and travel delay.

[0118] The structural failure model results in the undamaged network after flooding. These undamaged networks serve as transportation carriers. Travel demands not directly flooded or indirectly interrupted will readjust their travel plans on these undamaged networks (this modeling is called the functional failure model) until the daily travel utility of each individual is maximized. At this point, the state of the system is the state of the transportation system after flooding. Note: The utility parameters of this functional failure model (including activity utility parameters, car travel cost parameters, and public transport travel cost parameters) are basically the same as those of the (normal) functional model; the only difference is that the former represents the result after flooding, while the latter represents the result under normal conditions.

[0119] Step 5.1: Based on the spatial distribution characteristics of flood disasters, identify travel demand directly inundated and indirectly disrupted by floods, such as... Figure 1 (C) and 1(D). Based on the spatial distribution pattern of the flood, the inundation depth of the grid centroid is calculated. Grids with a depth exceeding the inundation threshold are identified as directly inundated by the flood, and all trips originating or ending at that grid are considered directly inundated travel demands. In addition, some trips, although not inundated, cannot be completed because their origins or destinations are surrounded by floodwaters, becoming "islands." These trips are considered indirectly interrupted travel demands. Using the grid centroid as nodes, and connecting any two adjacent grid centroids using the Moore's neighbor rule as edges, a centroid network is constructed. The nodes and edges corresponding to the centroids of directly ineffective grids are deleted. The maximum connected subgraph of the remaining network is calculated. Grid centroids corresponding to nodes not on the maximum connected subgraph are considered "islands" surrounded by floodwaters, and their resulting or attracted travel demands are deleted.

[0120] Step 5.2: Integrate the multimodal transportation network structure model after removing structural failures and the daily travel chain after removing flood-induced demand disruptions to construct a functional failure model of the multimodal transportation network. Input the remaining multimodal transportation network structure and remaining travel demands into the multi-agent transportation simulation platform (MATSim). Based on the co-evolutionary algorithm, each remaining individual aims to maximize its own daily travel utility by continuously adjusting its travel plan on the remaining multimodal transportation network, including departure time, mode of transportation, and travel route, until each individual can no longer benefit from adjusting its travel plan. At this point, the model converges, i.e., the user equilibrium state is reached. Set the simulation time to 0-36 hours.

[0121] Step 5.3: Calculate average travel time and available demand. Due to network disruptions, road congestion, etc., some demand cannot be fulfilled within the stipulated 36 hours; these demands are considered unmet demands. When an individual's daily travel utility converges, the model outputs travel volume T1 and average travel time t1. The available demand ratio is:

[0122]

[0123] The relative travel delay is:

[0124]

[0125] Passable demand ratio F trip The value is between 0 and 1. The closer the value is to 1, the smaller the impact of the flood on the carrying capacity of the city's multimodal transportation system and the stronger its functional resilience; conversely, the smaller the value is, the worse the system is. The relative travel delay τ reflects the change in travel time for the remaining passable demand. A negative relative delay indicates that the remaining demand is more unimpeded on the remaining road network after the disaster; conversely, a positive relative delay indicates that the remaining demand is more congested on the remaining road network.

[0126] Example 1

[0127] The following evaluation focuses on the multimodal transportation network of Nanjing City, located along the lower reaches of the Yangtze River and frequently experiencing floods. This case study verifies the practicality of the proposed method for assessing the functional resilience of urban multimodal transportation networks under flood conditions.

[0128] (1) Modeling results of multimodal transportation network structure in Nanjing

[0129] Nanjing's road system comprises 473 expressways, 1177 rapid transit roads, 2755 arterial roads, 4647 secondary arterial roads, 6886 branch roads, and 1736 ramps (connecting roads of different grades). The road network includes 6865 nodes and 17590 connections. Nanjing's surface public transport system includes 790 routes. Nanjing's rail transit system includes 20 bus routes, with a rail transit network comprising 159 nodes and 328 connections. Figure 3 This is a schematic diagram of the modeling results of a multimodal transportation network structure in Nanjing City provided by an embodiment of the present invention. A: Car (road) network; B: Ground public transport network; C: Rail network; D: Multimodal transportation network, wherein the ground public transport network is embedded in the road network.

[0130] (2) Daily travel chain simulation results in Nanjing

[0131] A total of 182,824 POIs were obtained from the Gaode Map Open Platform, including 4,789 residential communities, 887 dormitories, and 157 mixed-use commercial and residential buildings. The hourly traffic occurrence rates for these three types during the morning peak were 541, 614, and 171 people, respectively. The departure time distribution was obtained from the Nanjing Transportation Development Annual Report, and the departure time for each commute was initialized to obtain the daily travel chain of Nanjing residents.

[0132] (3) Construction and calibration results of the multimodal transportation network functional model in Nanjing

[0133] The road network and rail network of Nanjing were input into the network module of MATSim, the timetables of ground public transport and rail public transport were input into the timetable module of MATSim, and the daily travel chain was input into the demand module of MATSim. The travel utility of each mode was calibrated based on the real results and iterated 50 times. The simulation results were calculated using the average travel time, the proportion of commuting within 45 minutes, the proportion of commuting within 60 minutes (extreme), and the public transport modal share as indicators. The relative errors were all within ±5%.

[0134] Table 1 Calibration Results of Multimodal Transportation Network Functional Model in Nanjing

[0135]

[0136]

[0137] The number of travel requests was 1,847,400, the average travel time was 36.16 minutes, and the standard deviation of travel time was 22.31.

[0138] (4) Results of Structural Failure Model Construction in Nanjing

[0139] The flooding threshold for road network sections is set at 3m, and the flooding threshold for rail transit stations is set at 3.45m (including step height). Using a map of Nanjing's once-in-a-century flood event obtained from the Joint Research Centre Data Catalogue (with a depth of 0.45m), the direct failure rate of Nanjing's road network segments was calculated to be 22.8%, and the indirect failure rate was 20.0%, totaling 42.8%. For the rail network segments, the direct failure rate was 24.4%, and the indirect failure rate was 9.2%, totaling 33.6%. The loss on Nanjing's surface public transport lines was 53.8%, and the loss at stations was 56.3%; the loss on rail public transport lines was 70%, and the loss at stations was 87.2%.

[0140] (5) Assessment results of the functional resilience of Nanjing's multimodal transportation system under flood conditions

[0141] Based on the flood distribution map, demands directly inundated and indirectly disrupted in Nanjing were eliminated. The resulting multi-modal network structure, ground public transport and rail transit timetables, and demand were input into MATSim. Simulation was performed until individual daily travel utility converged. Unreachable demands were then eliminated, yielding a total of 1,036,606 passenger trips with passable demand after flood damage. The average travel time was 92.59 minutes, with a standard deviation of 92.59. The passable demand ratio was 43.89%, the absolute travel delay was 56.37 minutes, and the relative travel delay was 1.56.

[0142] In summary, the method of this invention can assess the functional resilience of urban multimodal transportation systems under flood disasters from two aspects: accessibility demand and travel delays.

[0143] This invention can be applied to assess the functional resilience of urban integrated transportation systems under flood disasters, estimating the economic losses caused by floods to the transportation and socio-economic systems. Based on this method, the actual disaster prevention and mitigation effects of different pre-disaster engineering and management strategies can be measured, including but not limited to roadbed reinforcement and upgrading, subway waterproofing and drainage, vehicle route guidance under disaster conditions, and adjustments to public transportation operation plans. Furthermore, this invention can be used to test the resilience impact of newly constructed roads, bus routes, and rail lines on urban integrated transportation systems, and can also be used to guide and optimize the layout and design of newly constructed roads and bus routes.

[0144] Furthermore, the method of this invention can also be applied to the assessment and optimization of post-disaster integrated transportation infrastructure repair strategies. Specifically, it addresses the priority of restoring road networks, surface public transport, or rail transit networks under conditions of limited cost and resources, and prioritizes the restoration of specific nodes and road sections. Moreover, the data required by this method comes entirely from open-source data or platforms, making it applicable to the resilience assessment of integrated transportation systems in any city under flood disaster conditions. In summary, the method of this invention can be applied to assessing the actual disaster prevention and mitigation effectiveness of pre-disaster engineering and management strategies, evaluating post-disaster infrastructure repair strategies, and implementing emergency management measures, thereby enhancing urban resilience and achieving sustainable development.

[0145] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of one embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing the present invention.

[0146] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0147] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0148] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for assessing the functional resilience of urban multimodal transportation networks under flood disasters, characterized in that, include: Acquire road network data, surface public transport network data, and rail public transport network data of the city under study, and construct a multimodal transportation network structure model; By collecting Points of Interest (POI) data, the traffic generation and attraction of the city are estimated, and the gravity model is improved based on the traffic decay function to simulate the daily travel chain of individuals. By integrating the multimodal transportation network structure model and the individual daily travel chain, a multimodal transportation network functional model of the city is constructed, and the multimodal transportation network functional model is calibrated based on real data. Based on the spatial damage characteristics of floods, a structural failure model for a multimodal transportation network is constructed. Based on the structural failure model of the multimodal transportation network, a failure model of the multimodal transportation network functional model is constructed, and the functional resilience of the multimodal transportation network functional model under flood disaster is evaluated from the aspects of accessibility demand and travel delay. The process of collecting POI data to estimate the city's traffic generation and attraction, improving the gravity model based on the traffic decay function, and simulating individual daily travel chains includes: The study area was divided into grids with a spatial resolution of 1km × 1km. It was assumed that all origin and destination points were located at the centroid of the grid. POI data for the study area was crawled from an open map platform. Residential communities, dormitories, and mixed-use buildings were identified as origin points, and the average area of ​​each type of origin point was determined. Traffic occurrence rate per unit area during weekday morning rush hour Determine the traffic occurrence rate of the three types of travel origins. : (1) The system collects and summarizes the traffic volume of all POIs (Points of Interest) within each grid. i Traffic volume The calculation formula is: (2) in, For the first indivual m POI-like For variables of 0-1, if the first... indivual m Class POI in the first i A value of 1 is assigned to a cell within a given grid, and 0 is assigned to a cell outside the grid. Other Points of Interest (POIs), including companies, schools, and hospitals, are considered as travel destinations. The average area of ​​these other POIs as travel destinations is determined. Traffic attraction rate per unit area during weekday morning rush hour The traffic attraction rate of the other POIs is obtained. ; (3) The traffic attraction of all travel destinations within each grid is statistically analyzed and summarized, and this is calculated as the attractiveness of that grid. : (4) The function for the decrease in traffic flow with travel distance is set as follows: (5) in, Indicates the distance traveled. Indicates the minimum distance. It is the cutoff value. These are control parameters, settings. =1.5 km =80 km and =1.75; From the grid To grid Traffic for: (6) Based on the daily departure time frequency statistics, the daily travel demand chain is simulated. It is assumed that travelers have only two trips a day, that is, the commuting trips during the morning and evening peak hours are symmetrical. The departure time required for each commuting trip is estimated from the daily departure time frequency survey results. The minimum threshold of 4 hours between the two trips is used to estimate the departure time required for each commuting trip. The process of integrating the multimodal transportation network structure model and the individual daily travel chain to construct the city's multimodal transportation network functional model, and calibrating the multimodal transportation network functional model based on real data, includes: Multimodal traffic network structure data and individual daily travel chain data are input into the MATSim multi-agent traffic simulation platform model, and activity utility parameters, car travel cost parameters and public transportation travel cost parameters are initialized in the configuration file respectively. The city’s average annual GDP, population and average annual working hours per person are obtained from statistical yearbooks. The average unit time value per person in the city is calculated as the hourly utility parameter of the activity. The travel cost per unit mile of a car is determined based on fuel consumption per kilometer and fuel price. The parking cost of a car is improved based on the average parking time and parking fee standard, and the cost parameters per unit travel time are initialized. The fare cost of public transportation is determined based on the single-trip fare of ground public transport and rail public transport, and the single-transfer cost and unit travel time cost parameters of public transportation are initialized. Multimodal traffic network structure data and daily travel chain data are input into the multi-agent traffic simulation platform MATSim. Activity utility parameters, car travel cost parameters, and public transportation travel cost parameters are initialized in the configuration file. Based on the co-evolutionary algorithm, each individual in the multi-agent traffic simulation platform simultaneously optimizes its own travel chain, including adjusting the departure time, mode of transportation, and travel route of each individual, with the goal of improving the daily travel utility of each individual, until all individuals in the system can no longer benefit from adjusting their own travel plans. Through the continuous convergence and iteration of the multi-agent traffic simulation platform, a user equilibrium state is reached, and the initial traffic network functional model is obtained. Based on the average travel time, public transport modal share, and average travel distance output from each simulation, adjust the fixed costs of travel utility for both cars and public transport until the error is adjusted to a reasonable range. At this point, the total travel volume output by the model is: T 0 The average travel time is .

2. The method according to claim 1, characterized in that, The process of acquiring road network data, surface public transport network data, and rail public transport network data of the study city, and constructing a multimodal transportation network structure model, includes: The road network dataset of the research city was obtained from the developed street map. This road network dataset includes the geospatial location, topological relationship and road segment attributes of the road network nodes. The road network dataset was cropped using the administrative division data of the research city. Intersections were abstracted as nodes and road segments were abstracted as edges. Improve road segment attribute information, which includes free-flow speed, number of lanes and traffic capacity; set ground public transport network data, including ground public transport station information, route information and departure timetable information; and map ground public transport stations and routes to the road network. The data for setting up the rail transit network includes rail transit station information, route information, and departure timetable information. The station information and route information of the rail transit are obtained, and the rail transit stations are abstracted as nodes, and the topological relationship between two adjacent rail stations of a bus route is abstracted as edges.

3. The method according to claim 2, characterized in that, The mapping of ground public transport stops and routes to the road network includes: Based on an open map platform, obtain the latitude and longitude information of ground public transport stations and the sequence information of public transport stations that the routes pass through in sequence; Based on the latitude and longitude of the station, the nearest neighbor matching method is used to match each bus stop to the nearest road segment; For any two adjacent bus stops in all bus routes, use Dijkstra's algorithm to find the shortest path sequence between the road segments where these two adjacent stops are located, and use this shortest path sequence as the road segments that the bus route passes through in sequence.

4. The method according to claim 1, characterized in that, The aforementioned structural failure model for a multimodal transportation network, based on the spatial damage characteristics of floods, includes: Based on the spatial distribution characteristics of flood events, directly failed nodes and road segments in the multimodal transportation network structure are identified and eliminated. For the road network, a dual method is used, treating edge segments as points and intersections as edges. The inundation depth of midpoint nodes in each road segment is calculated. For the rail network, the inundation depth of initial network nodes is calculated. Inundation thresholds for flood-affected nodes in both the road and rail networks are set. Nodes with flood depths exceeding these thresholds are considered directly failed nodes, and directly failed nodes and their edges are deleted. The node failure criteria for the road network are as follows: (7) The failure criteria for nodes in the track network are as follows: (8) in, h 0 m Indicates the height of the subway station steps. h Indicates the submerged water depth of the node. Indicates the node flooding threshold. It is a 0-1 variable. =0 indicates that the node is invalid. =1 indicates that the node is valid; Based on network connectivity, identify and eliminate indirect failures in the multimodal transportation network structure. Identify the connected subgraphs and the maximum connected subgraphs of the road network. Treat nodes in non-maximum connected subgraphs as indirect failure nodes and delete the nodes and edges of the nodes. Treat bus stops of bus routes that cannot operate as indirect failures and delete the nodes and edges of the nodes.

5. The method according to claim 4, characterized in that, The functional failure model of the multimodal transportation network functional model, constructed based on the structural failure model of the multimodal transportation network, evaluates the functional resilience of the multimodal transportation network functional model under flood disasters from two aspects: accessibility demand and travel delay. This includes: Based on the spatial distribution characteristics of flood disasters, travel demand that is directly inundated and indirectly interrupted by floods is identified. Based on the spatial distribution pattern of floods, the inundation depth of the grid centroid is calculated. Grids with a water depth exceeding the inundation threshold are identified as directly inundated by floods. All travel from grids that serve as the starting or ending point is considered as directly inundated travel demand. Travel from grids whose starting or ending points are surrounded by floods is considered as indirectly interrupted travel demand. Using the centroids of the grid as nodes, and connecting any two adjacent centroids of the grid as edges according to the Moore neighbor rule, construct a centroid network. Delete the nodes and edges corresponding to the centroids of directly invalid grids. Calculate the maximum connected subgraph of the remaining network. The centroids of the grids that are not on the maximum connected subgraph are regarded as "islands" surrounded by floods, and the travel demand they generate or attract is deleted. By integrating the multimodal transportation network structure model after removing structural failures and the daily travel chain after removing flood-induced demand, the remaining multimodal transportation network structure and remaining travel demand are input into the multi-agent transportation simulation platform MATSim. Based on the co-evolutionary algorithm, each remaining individual aims to maximize its own daily travel utility and continuously adjusts its travel plan on the remaining multimodal transportation network, including departure time, mode of transportation, and travel route, until each individual can no longer benefit from adjusting its travel plan. At this point, the model iteratively converges, reaching a user equilibrium state, and the functional failure model of the multimodal transportation network is obtained. Using MATSim to calculate average travel time and available demand, the model outputs the travel volume when an individual's daily travel utility converges. T 1 The average travel time is Passable demand is higher for: (9) The relative travel delay is: (10) Passable demand ratio Between 0 and 1, The closer the value is to 1, the smaller the impact of flooding on the city's multimodal transportation system, and the stronger its functional resilience; conversely, The closer the value is to 0, the greater the impact of the flood on the city's multimodal transportation system, and the worse its functional resilience; the greater the relative travel delay. Changes in travel time, reflecting remaining passable demand, indicate that negative relative delays mean that remaining demand is more readily available on the remaining road network after a disaster; conversely, positive delays mean that remaining demand is more congested on the remaining road network.

Images