CN116504050A - A Special Event City Partitioning Method Based on Multimodal Public Transit Trip Data - Google Patents

Abstract

The invention discloses a special event city partitioning method based on multi-mode public transport travel data, which divides traffic areas according to urban road data, and then inputs processed urban multi-mode public transport travel data, divides multi-mode public transport according to the division of traffic areas The travel data is aggregated into the traffic area to obtain a joint data set. Then construct a spatial interaction network based on the joint dataset to reflect the spatio-temporal characteristics of urban public transport travel. Then, the Leiden community detection algorithm is used to divide the spatial interaction network, and a preliminary urban partition scheme is obtained. Finally, according to the spatial constraints, the spatially disconnected areas in the preliminary zoning scheme are corrected to obtain the final urban zoning scheme. The invention can effectively reduce the impact of urban zoning for special events on travel of residents, and provide refined and sustainable management reference for cities to deal with special events.

Description

A Special Event City Partitioning Method Based on Multimodal Public Transit Trip Data

technical field

The invention relates to the technical field of urban epidemic prevention and control, in particular to a special event city partitioning method based on multi-mode public transport travel data

Background technique

Existing urban zoning methods are mainly based on administrative divisions or geographical intervals of cities, and the boundaries defined by these two belong to historical evolution. The urban zoning obtained by the above method does not take into account the travel characteristics and rules of urban residents, and at the same time brings a lot of inconvenience to the operation and management of public transportation after special events. The urban public transportation under the urban zoning has been affected to varying degrees. Some cities have completely suspended public transport services for administrative reasons. In some areas, public transportation is only allowed in times of great demand or the frequency of departures has been reduced and some stations have been closed accordingly. This has led to the continuous sluggish passenger flow of public transport in various countries after the special event. Even after the public transport system resumes normal operation, due to changes in people's travel habits, the passenger flow of public transport is still difficult to return to the previous level. severe challenge.

Therefore, in order to improve the method of urban zoning, ensure the basic travel needs of urban residents and basic public transport services when emergency events occur, and improve the resilience of public transport systems, it is necessary to sum up experience in contemporary society and explore urban zoning under special events from the perspective of transportation. traffic adaptability, and research more scientific and reasonable urban zoning strategies.

Contents of the invention

The technical problem to be solved by the present invention is: aiming at the deficiencies of the prior art, a special event city partitioning method based on multi-mode public transport travel data is proposed.

In order to solve the above technical problems, the present invention provides the following technical solutions: a special event city partitioning method based on multi-mode public transport travel data, comprising the following steps:

S1. Based on the urban traffic network, divide the urban traffic network into several traffic districts;

S2. Acquiring urban multi-mode user travel data, and preprocessing the user multi-mode travel data;

S3. Based on the traffic districts obtained in step S1 and the preprocessed user multi-mode travel data, match the travel starting point and travel end point in the preprocessed user multi-mode travel data with their corresponding traffic districts respectively, construct Urban multi-modal travel space interactive network, used to statistically generate joint data sets about travel flow among various traffic areas;

S4. According to the joint data set about the travel flow between each traffic area, use the Leiden algorithm to divide the spatial interaction network, and obtain the community division plan with the highest modulus value and its corresponding preliminary divided urban community;

S5. Using the modulus value to measure the preliminarily divided urban communities obtained in step S4, and adding constraints to correct the preliminary divided urban communities according to the measurement results, to obtain the final urban divisions.

Further, the aforementioned step S2 includes the following sub-steps:

S2.1. Obtaining urban multi-mode user travel data includes: separately obtaining subway data, bus data, shared bicycle data, taxi data, user number in online car-hailing data, order date, order start time and the location corresponding to the start time , order completion time and the location corresponding to the completion time, a total of six fields of data;

S2.2. According to the order start time and the location corresponding to the start time, the order completion time and the location corresponding to the completion time, perform order latitude and longitude matching conversion to obtain user travel data: user travel starting point latitude and longitude coordinates, user travel end point latitude and longitude coordinates, user travel Starting time, end time of user travel;

S2.3. Deduplicate the user travel data, delete null data, and delete abnormal data.

Further, the joint data set about the traffic flow between each traffic zone generated in the aforementioned step S3 includes the travel flow between each traffic zone, the shortest network distance between the centroids of each traffic zone, and the population in each zone.

Further, the aforementioned step S4 includes the following sub-steps:

S4.1. Number the centroids of each traffic area, corresponding to each node;

S4.2. Taking each node as the initial community, move the initial community to another community according to the attributes of the joint data set about traffic flow between each traffic area, and obtain a class of community in this algorithm; and calculate according to the modulus value calculation formula The modulus value of the current division result;

S4.3. Refine the interior of the obtained type of community, obtain the corresponding sub-community, and increase the modulus value of S4.2;

S4.4. Coagulate the nodes in each sub-community to obtain new independent nodes, each independent node represents a type of community in step S4.3;

S4.5. Based on the new independent node obtained in step S4.4, return to step S4.1 to step S4.3, and calculate the modulus value in the iterative process, and keep the classification result of modulus value promotion until the modulus The degree value cannot be further improved, and the community division obtained corresponding to the modulus value corresponding to the last iteration is the final community division result.

Further, the aforementioned step S5 specifically includes;

Calculate the modulus value according to the following formula:

Among them, A _ij represents the weight of the edge between nodes i and j; the parameter K _i is the sum of the weights of the edges connected to node i; K _j is the sum of the weights of the edges connected to node j; the parameter m represents the weight of the edge Quantity; parameter C _i is the community to which node i is assigned, if C _i =C _j , modularity δ(C _i , C _j ) is equal to 1, otherwise it is equal to 0;

The constraints are: for a node with one adjacent node, merge it directly with the adjacent node; for a surrounding node with multiple adjacent nodes with connecting edges, combine it with the node that can achieve the maximum modulus value gain Node merging; for an isolated node that does not have any edges with adjacent nodes, group it with the node with the smallest area.

Further, in the aforementioned step S2.2, the minimum entropy rate travel chain method is used to perform order latitude and longitude matching conversion to obtain the user's travel starting point and user's travel destination.

Further, in the aforementioned step S2.3, deleting the abnormal data includes deleting the longitude and latitude drift data, the abnormal travel distance data, and the abnormal travel time data.

Compared with the prior art, the present invention adopts the above technical scheme and has the following technical effects:

1. In response to special urban events, the traffic-adaptive urban zoning management model is becoming more and more important. Therefore, it is urgent for us to explore scientific and reasonable urban zoning methods. This technology can divide more traffic-adaptable urban subregions on the basis of effectively identifying urban residents' public transportation trips, reduce the travel connections of urban residents cut off due to partition management through community discovery algorithms, and cover most of the trips within urban subdivisions . Thereby reducing the impact of the urban zoning management mode on the travel of urban residents.

2. This urban zoning method redesignates the urban emergency management unit, which has a certain guiding significance for the urban scientific and sustainable response to special events. At the same time, because the divided urban zoning is more traffic-adaptable, it is also suitable for urban special events. The strategic adjustment of public transport organization provides a certain reference value.

Description of drawings

Fig. 1 is a flow chart of a special event city partitioning method based on multi-mode public transport travel data according to the present invention.

Fig. 2 is a step diagram of the Leiden community detection algorithm used in the present invention.

Fig. 3 is a schematic diagram of the preliminary division of cities in an embodiment of the present invention.

Fig. 4 is a schematic diagram of the final division of the city in the embodiment of the present invention.

Fig. 5 is a schematic diagram of comparison of partition results in the implementation cases of the present invention.

Detailed ways

In order to better understand the technical content of the present invention, specific embodiments are given together with the attached drawings for description as follows.

Aspects of the invention are described herein with reference to the accompanying drawings, in which a number of illustrative embodiments are shown. Embodiments of the present invention are not limited to those described in the drawings. It should be understood that the present invention can be realized by any one of the various concepts and embodiments described above, as well as the concepts and embodiments described in detail below, because the disclosed concepts and embodiments of the present invention are not limited to any implementation Way. In addition, some aspects of the present disclosure may be used alone or in any suitable combination with other aspects of the present disclosure.

As shown in Figure 1, a flow chart of a special event city partition method based on multi-mode public transport travel data, including the following steps:

S1. Based on the urban transportation network, the urban transportation network is divided into several traffic areas; the traffic area refers to the combination of nodes or connections with a certain degree of traffic correlation and traffic similarity, and the coverage area of a city's traffic network is divided into several It is easy to understand the traffic travel rules from the mesoscopic level, which is helpful to formulate reasonable traffic management measures.

S2. Acquire urban multi-mode user travel data, and preprocess the user multi-mode travel data; user travel data includes subways, buses, shared bicycles, taxis, online car-hailing, etc.

S3. Based on the various traffic districts obtained in step S1 and the preprocessed user multimodal travel data, field data such as the longitude and latitude coordinates of the travel start point, the longitude and latitude coordinates of the travel end point, and the travel start and end point time are extracted, and the preprocessed user multimode The travel start point and travel end point in the travel data are matched with their corresponding traffic districts, and an urban multi-mode travel space interactive network is constructed to statistically generate a joint data set of travel flow among various traffic districts;

S4. According to the joint data set about the travel flow between each traffic area, use the Leiden algorithm to divide the spatial interaction network, and obtain the community division plan with the highest modulus value and its corresponding preliminary divided urban community;

S5. Using the modulus value to measure the preliminarily divided urban communities obtained in step S4, and adding constraints to correct the preliminary divided urban communities according to the measurement results, to obtain the final urban divisions.

In this embodiment, a certain city A is selected as a representative comprehensive transportation hub. By the end of a certain year, there are 11 districts in city A, with a total area of nearly 80,000 square kilometers and a permanent population of nearly 10 million. City A has a complete multimodal transport system, with 11 lines and 191 stations, with a total length of 427.1.10 kilometers, forming a subway network covering the whole city.

Input the data of urban roads (arterial roads, secondary roads, and branch roads) within the research scope of city A, and divide traffic zones (TAZ) based on the nature of land use, land use conditions and other factors, and at the same time, based on the road network topology and traffic flow characteristics . Through division, 3028 traffic districts are obtained, as shown in Figure 3.

Input urban multi-modal travel data: including subway, bus, shared bicycles, taxis, online car-hailing, etc., and preprocess the urban multi-modal travel data, extracting the longitude and latitude coordinates of the starting point of travel, the longitude and latitude coordinates of the end point of travel, and the start and end of travel Field data such as time.

There are mainly two types of datasets used in this case study. The first type of data is the multi-modal travel data of city A residents, and the other type is the traffic area (TAZ) information. The first type of data includes four types of travel data: Free Floating Bike Share (FFBS) travel data, city A public bicycle data, city A subway travel data, city A smart card data from a bus. These four types of data include the following fields: ID, date, origin location, destination location, origin time, and destination time, as shown in Table 1, the multi-modal public transport travel data set table. The second type of data set is the traffic zone (TAZ) data of city A, including ID, population, area, and employment, as shown in Table 2 Traffic Zone (TAZ) Information Table.

Among them, the bicycle sharing supervision platform of city A provided data on 3,647,984 shared bicycle trips in a week. The data of 459,609 public bicycle trips in city A in the same week are provided by the public bicycle company of city A. The subway company of city A provided data on 18,614,830 subway trips for a period of one week. The bus company of city A provided the IC card data of 46,514,543 bus trips in a week. The multi-modal public transport trip data set of city A contains a total of 16,701,097 trips.

Table 1

Table 2

The processing of the data includes matching and converting the latitude and longitude of the origin and destination of the trip. The bus trip data does not include alighting locations, so it is necessary to use the minimum entropy rate modified trip chain method for origin-destination estimation. Data cleaning mainly includes deduplication, deletion of null data, and deletion of abnormal data. Anomaly data mainly include latitude and longitude drift, travel distance anomalies and travel time anomalies.

3. Build an urban multi-modal travel space interactive network, and summarize the travel of users into a travel flow at the traffic zone (TAZ) level. Through the spatial connection function of ArcGIS software, the user travel data is aggregated into OD traffic flow according to the traffic zone (TAZ) data set. A joint data set, as shown in Table 3 Joint Data Set Statistics Table. The data set of this table records travel-related information from one traffic zone (TAZ) to another traffic zone (TAZ). In addition to travel flows, the dataset also records population, employment, and the distance between any pair of traffic zones (TAZs). Wherein, the distance used refers to the shortest network distance from the centroid of one traffic zone (TAZ) to the centroid of another traffic zone (TAZ).

table 3

As shown in Figure 2, the Leiden algorithm is used to divide the spatial interaction network, and the community division scheme with the highest modulus value is obtained. The specific steps include the following steps S4.1 to S4.5. The value range of the modulus value is (0,1).

S4.1. Number the centroids of each traffic area, corresponding to each node;

S4.2. Taking each node as the initial community, move the initial community to another community according to the attributes of the joint data set about traffic flow between each traffic area, and obtain a class of community in this algorithm; and calculate according to the modulus value calculation formula The modulus value of the current division result;

S4.3. Subdivide the inside of the obtained type of community, obtain the corresponding sub-community, and increase the modulus value of the algorithm result in the previous step.

S4.4. Coagulate the nodes in each sub-community to obtain new independent nodes, and each independent node represents a type of community in the last round of algorithm;

S4.5. Based on the new independent node obtained in step S4.4, return to step S4.1 to step S4.3, and calculate the modulus value in each round of algorithm iteration, and keep the classification result of modulus value promotion , until the modulus value cannot continue to increase, and the community division obtained corresponding to the modulus value corresponding to the last iteration is the final community division result.

Among them, the modulus value is calculated according to the following formula:

Among them, A _ij represents the weight of the edge between nodes i and j; the parameter K _i is the sum of the weights of the edges connected to node i; K _j is the sum of the weights of the edges connected to node j; the parameter m represents the weight of the edge Quantity; parameter C _i is the community to which node i is assigned, if C _i =C _j , modularity δ(C _i , C _j ) is equal to 1, otherwise it is equal to 0;

In general, higher modularity values will correspond to more stable community structures. A good community detection result is expressed in the form of higher similarity of nodes within the community and lower similarity of nodes outside the community.

Measure the preliminarily divided urban divisions through constraints, and modify the initially divided urban communities by adding constraints based on the measurement results, and obtain the final urban divisions to correct communities with poor spatial connections. The final urban division plan is shown in Figure 4. . where the constraints are: for a node with one adjacent node, merge it directly with the adjacent node; for a surrounding node with multiple adjacent nodes with connecting edges, merge it with the node that can achieve the maximum modulus value gain Merge; for an isolated node that does not have any edges with neighboring nodes, group it with the node with the smallest area.

Comparative analysis of the invention effect: the travel network of city A is divided based on the multi-mode public transportation travel data of city A through the Leiden community detection algorithm, and the community classification scheme with the smallest modularity is obtained. This method minimizes the connection between communities and maximizes the The connection within the community, so the delineation of urban divisions by this method greatly reduces the traffic connection between urban divisions, making the division of urban divisions as small as possible to cut off the traffic and travel connections, thereby reducing the prevention and control after special events occur Cost, providing refined and sustainable prevention and control reference for cities to deal with special events. Carry out the program of city division with routine with the administrative district as the boundary and compare the result as shown in table 4 city division cutting off travel connection statistical table, table 5 administrative district division plan cutting off travel connection statistical table, it can be seen that this invention has reduced and cut off 8.2% of the traffic Contact for travel. The left side of Fig. 5 is the district of city A divided by the present invention, and the right side is the division map of the administrative district of city A. By comparison, the present invention provides a refined and sustainable disposal reference for dealing with special events.

Table 4

ID intra-zone travel total trips by division cut ratio ID intra-zone travel total trips by division cut ratio 1 814001 1445322 43.7% 2 793516 1565796 49.3% 3 203003 323750 37.3% 4 1216952 2765344 56.0% 5 626018 1409669 55.6% 6 660339 1221855 46.0% 7 181320 364057 50.2% 8 786739 1453103 45.9% 9 712767 1119151 36.3% 10 965260 1561400 38.2% 11 907676 1674548 45.8% 12 830472 1111477 25.3% 13 36331 63556 42.8% 14 339629 510470 33.5% 15 65767 111599 41.1% Sum 9139790 16701097 45.3%

table 5

Administrative area number District travel statistics All travel statistics in the district Cut off travel ratio ① 1027089 1965139 47.7% ② 1136900 2742648 58.5% ③ 1268809 2418767 47.5% ④ 990459 2701024 63.3% ⑤ 497849 1387883 64.1% ⑥ 917562 1787173 48.7% ⑦ 752001 1521008 50.6% ⑧ 262370 690744 40.0% ⑨ 746350 1486711 49.8% total 7765393 16701097 53.5%

Although the present invention has been described above with preferred embodiments, it is not intended to limit the present invention. Those skilled in the art of the present invention may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be defined by the claims.

Claims (7)

1. a special event city partition method based on multi-mode public transport trip data, is characterized in that, comprises the steps: S1. Based on the urban traffic network, divide the urban traffic network into several traffic districts; S2. Acquiring urban multi-mode user travel data, and preprocessing the user multi-mode travel data; S3. Based on the traffic districts obtained in step S1 and the preprocessed user multi-mode travel data, match the travel starting point and travel end point in the preprocessed user multi-mode travel data with their corresponding traffic districts respectively, construct Urban multi-modal travel space interactive network, used to statistically generate joint data sets about travel flow among various traffic areas; S4. According to the joint data set about the travel flow between each traffic area, use the Leiden algorithm to divide the spatial interaction network, and obtain the community division plan with the highest modulus value and its corresponding preliminary divided urban community; S5. Use the modulus value to measure the preliminary urban division obtained in step S4, and add constraints to correct the preliminary divided urban community according to the measurement result to obtain the final urban division. 2. a kind of special event urban partition method based on multi-mode public transportation travel data according to claim 1, is characterized in that, step S2 comprises the following sub-steps: S2.1. Obtaining urban multi-mode user travel data includes: separately obtaining subway data, bus data, shared bicycle data, taxi data, user number in online car-hailing data, order date, order start time and the location corresponding to the start time , order completion time and the location corresponding to the completion time, a total of six fields of data; S2.2. According to the order start time and the location corresponding to the start time, the order completion time and the location corresponding to the completion time, perform order latitude and longitude matching conversion to obtain user travel data: user travel starting point latitude and longitude coordinates, user travel end point latitude and longitude coordinates, user travel Starting time, end time of user travel; S2.3. Deduplicate the user travel data, delete null data, and delete abnormal data. 3. a kind of special event urban partition method based on multi-mode public transport travel data according to claim 2, it is characterized in that, in step S3, generate the joint data set about traffic flow between each traffic sub-district including between each traffic sub-district travel flow, the shortest network distance between the centroids of each traffic area, and the population in each area. 4. a kind of special event urban partition method based on multi-mode public transport travel data according to claim 3, is characterized in that, step S4 comprises the following sub-steps: S4.1. Number the centroids of each traffic area, corresponding to each node; S4.2. Taking each node as the initial community, move the initial community to another community according to the attributes of the joint data set about traffic flow between each traffic area, and obtain a class of community in this algorithm; and calculate according to the modulus value calculation formula The modulus value of the current division result; S4.3. Refine the interior of the obtained type of community, obtain the corresponding sub-community, and increase the modulus value of S4.2; S4.4. Coagulate the nodes in each sub-community to obtain new independent nodes, each independent node represents a type of community in step S4.3; S4.5. Based on the new independent node obtained in step S4.4, return to step S4.1 to step S4.3, and calculate the modulus value in the iterative process, and keep the classification result of modulus value promotion until the modulus The degree value cannot be further improved, and the community division obtained corresponding to the modulus value corresponding to the last iteration is the final community division result. 5. A kind of special event city partition method based on multi-mode public transport trip data according to claim 4, it is characterized in that, step S5 specifically comprises; Calculate the modulus value according to the following formula: Among them, A _ij represents the weight of the edge between nodes i and j; the parameter K _i is the sum of the weights of the edges connected to node i; K _j is the sum of the weights of the edges connected to node j; the parameter m represents the weight of the edge Quantity; parameter C _i is the community to which node i is assigned, if C _i =C _j , modularity δ(C _i , C _j ) is equal to 1, otherwise it is equal to 0; The constraints are: for a node with one adjacent node, merge it directly with the adjacent node; for a surrounding node with multiple adjacent nodes with connecting edges, combine it with the node that can achieve the maximum modulus value gain Node merging; for an isolated node that does not have any edges with adjacent nodes, group it with the node with the smallest area. 6. A special event city partitioning method based on multi-mode public transportation travel data according to claim 5, characterized in that, in step S2.2, use the minimum entropy rate travel chain method to perform order latitude and longitude matching conversion to obtain user Travel origin, user travel destination. 7. A kind of special event city partition method based on multi-mode public transportation travel data according to claim 6, characterized in that, in step S2.3, deleting abnormal data includes deleting longitude and latitude drift data, abnormal travel distance data, abnormal Travel time data.