A method for dividing traffic blocks based on resident travel characteristics

Abstract

The application provides a traffic cell division method based on resident travel characteristics. Administrative division boundaries and hierarchical road network vector data are acquired and data preprocessing is performed; basic clustering units are divided based on the road network and the administrative division; a clustering index system is established, and resident travel characteristics in the units are acquired through mobile phone signaling in combination with the clustering units; according to the location of the clustering units in the administrative division, a k-means clustering algorithm with spatial characteristics is adopted for preliminary clustering, and clustering results and clustering centers corresponding to each clustering number are obtained; clustering effectiveness indexes of the clustering results of each clustering number are calculated; the clustering effectiveness indexes of each clustering number are compared to determine a proper clustering number; final clustering division is performed through a FCM clustering algorithm with comprehensive spatial characteristics; the clustering results are modified in combination with main hierarchical roads, and traffic cell division is completed. The application improves the robustness of the traffic cell division results and can support the construction of subsequent traffic planning models.

Description

Technical Field

[0001] This invention mainly relates to the field of intelligent transportation technology, specifically to a method for dividing traffic zones based on residents' travel characteristics. Background Technology

[0002] Highway transportation is currently the largest mode of transportation and is most closely linked to people's production and daily life. However, traffic congestion and chaos on highways are manifestations of the mismatch between rapidly growing transportation demand and existing transportation supply. If reasonable planning and adjustments to transportation are not carried out, this problem will become increasingly prominent.

[0003] Traffic zones are the primary task in traffic demand forecasting within transportation planning. A traffic zone is an aggregate analysis unit created in transportation planning based on the similarity of travel origins. This unit serves as the carrier of traffic survey data and the foundation for forecasting in transportation planning, significantly influencing the construction of the final transportation planning model.

[0004] Traditional traffic zone delineation relies heavily on survey data regarding socioeconomic attributes, population, and land use, neglecting to consider the travel characteristics of residents within each traffic zone. With the development of big data and computer technology, improved techniques for traffic zone delineation are increasingly incorporating residents' travel features, such as using public transport card swipe data to obtain travel characteristics and then dividing zones based on travel similarity. However, the clustering units in these techniques are primarily grids, resulting in significant discrepancies between the boundaries of the delineated traffic zones and the actual urban road network and layout, which hinders the construction of subsequent traffic planning models.

[0005] In summary, the main drawbacks of the existing traffic zone division system are:

[0006] (1) The travel characteristics of traffic sources were not considered;

[0007] (2) The boundaries of the traffic zones are significantly different from the actual roads and administrative divisions, which affects the construction of subsequent traffic planning models. Summary of the Invention

[0008] To address the aforementioned issues, this invention provides a method for dividing traffic zones based on residents' travel characteristics. This method divides clustering units using road networks and administrative boundaries, and then divides traffic zones by combining the travel characteristics of residents within each unit. While taking into account the characteristics of the main traffic demand subjects, it ensures that the boundaries of the divided traffic zones match the actual road network and administrative boundaries, thus satisfying the needs of subsequent traffic planning model construction.

[0009] To achieve the above objectives, the technical solution adopted by this invention is a method for dividing traffic zones based on residents' travel characteristics, and the specific steps are as follows:

[0010] Step 1: Obtain administrative division boundaries and hierarchical road network vector data, and perform data preprocessing;

[0011] Step 2: Divide the data into basic clustering units based on the road network and administrative divisions;

[0012] Step 3: Establish a clustering indicator system and, in conjunction with the clustering units from Step 2, obtain the travel characteristics of residents within the units through mobile phone signaling;

[0013] Step 4: Based on the administrative region where the clustering units are located, the k-means clustering algorithm is used for preliminary clustering to obtain the clustering results and cluster centers corresponding to each cluster number;

[0014] Step 5: Calculate the clustering effectiveness index based on the clustering results obtained in Step 4 for each cluster number;

[0015] Step 6: Compare the clustering effectiveness indices of each cluster number to determine the appropriate number of clusters;

[0016] Step 7: Based on the number of clusters determined in Step 6, perform the final clustering using the FCM clustering algorithm;

[0017] Step 8: Based on the main road levels, correct the clustering results obtained in Step 7 to complete the traffic zone division.

[0018] Preferably, in step 1, based on the acquired basic data, basic data processing is completed using ArcGIS professional software, including road network topology processing and intersection simplification.

[0019] Preferably, in step 2, the administrative divisions are segmented using the road network to divide the basic clustering units at the street scale for subsequent steps.

[0020] Preferably, in step 4, before using the k-means algorithm for clustering, it is necessary to determine the range of the number of clusters based on the distribution of residents' travel distances within the district / county.

[0021] Clustering algorithms describe the clustering process using a cost function, which is:

[0022]

[0023] in, For the sample size, The number of clusters is the number of divisions. For the first One data sample, For the first The center of each cluster, Indicates sample Family membership degree For the first The sample to the first The "distance" between the centers of each cluster. Considering cluster units as geospatial objects, this "distance," in addition to basic attribute distance, should also take into account spatial distance, as expressed below.

[0024]

[0025] In the formula, This represents the spatial distance between the sample and the cluster center. If the sample is adjacent to the cluster, its value is 0; otherwise, it represents the spatial distance from the sample centroid to the cluster centroid. This is the spatial effect coefficient; The attribute distance from a sample to the cluster center is expressed as follows:

[0026]

[0027] In the formula, For the sample The Clustering factors, For clusters The Clustering factors, For the first The weights of each clustering factor, For the sample with cluster The attribute distance.

[0028] Preferably, in step 5, an index that comprehensively considers both intra-group similarity and inter-group differences is used to evaluate the clustering results. The expression for the clustering effectiveness index is as follows:

[0029]

[0030] in, Representing clustering units to cluster Distance from the center Indicates the first The number of clustering units, Indicates the first The average distance from the cluster center to the cluster interior. The total number of cluster units. Used to characterize the similarity of clustering units within a cluster. For clusters From cluster center to cluster Distance from the center It is the sum of the pairwise distances between all cluster centers, used to characterize the differences between clusters.

[0031] Preferably, in step 6, the optimal number of clusters is selected by referring to the "elbow method" based on the plot of each cluster number and cluster effectiveness index.

[0032] Preferably, in step 7, the FCM algorithm incorporating fuzzy concepts is used for secondary clustering, where the membership degree... ,and .

[0033] Preferably, in step 8, the correction of the clustering results involves filtering the main roads in the hierarchical road network and performing secondary segmentation on the clustering results in step 7 accordingly to obtain the final traffic zone division results.

[0034] The beneficial effects of this invention are reflected in:

[0035] First, this invention divides traffic zones into cluster units based on the road network, ensuring that the boundaries of the traffic zone division results match the actual road network, which is more conducive to the construction and application of subsequent traffic planning models. Second, by leveraging big data to extract residents' travel characteristics within the cluster units, this invention incorporates residents' travel characteristics into the traffic zone division considerations, satisfying the principle that traffic zones have similar travel characteristics. Finally, by coupling the clustering algorithm, this invention solves the problem of the difficulty in determining the number of clusters in traditional clustering algorithms and improves the robustness of the traffic zone division results. Traffic zones divided based on this method can support the construction of subsequent traffic planning models. Attached Figure Description

[0036] Figure 1 : This is a flowchart of the present invention;

[0037] Figure 2 : This is a line graph illustrating the cluster number versus cluster validity value in an embodiment of the present invention;

[0038] Figure 3 This is a schematic diagram of the clustering results of various districts and counties in Ningbo City in an embodiment of the present invention;

[0039] Figure 4 This is a schematic diagram showing the final traffic zone division results in Ningbo City in an embodiment of the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] The first embodiment of the present invention is as follows:

[0042] S1. Data Acquisition and Preprocessing: This embodiment acquires the administrative boundaries of the 11 districts and counties of Ningbo City, as well as road network data within Ningbo City, including highways, expressways, primary roads, secondary roads, tertiary roads, township roads, and other roads. ArcGIS is used to preprocess the road network, including simplifying intersections, converting two-lane roads to single-lane roads, and reconstructing the topology, ensuring the road network data meets the needs of subsequent analysis.

[0043] S2. Clustering Unit Division: Using ArcGIS, the 11 districts and counties of Ningbo City were divided into administrative divisions based on road networks, resulting in clustering units bounded by road networks or administrative boundaries. In this embodiment, Ningbo City was divided into 9426 clustering units, and the number of clustering units for each district and county is shown in Table 1 below.

[0044] Table 1. Number of Cluster Units in Each District / County

[0045]

[0046] S3. Acquisition of Resident Travel Characteristics: Before acquiring resident travel characteristics, a clustering indicator system needs to be established to purposefully acquire and process relevant data. Each indicator also needs to be assigned a corresponding weight for calculating weighted distances in subsequent clustering. When establishing the clustering indicator system, based on the spatial and travel characteristics of the cluster units, seven indicators were selected: centroid distance, number of trips, travel distance, travel time, and the proportion of trips during morning, evening, and off-peak hours. Spatial characteristic indicators ensure the continuity of traffic zones in the clustering results, while travel characteristics follow the principle of dividing traffic zones based on similar traffic characteristics of traffic sources within the traffic zone. Specifically, the morning peak hour is defined as 6:30-9:30, the off-peak hour as 9:30-17:00, and the evening peak hour as 17:00-20:00. The proportion of trips during each time period is defined as follows.

[0047]

[0048] In the formula, express The proportion of trips during different time periods express Number of travelers during a given time period This represents the total number of travelers throughout the day. Using an expert system, different weights are assigned to the importance of clustering for different indicators; for example, adjacent cluster units are prioritized for clustering, and their spatial characteristics are given the highest weight. Specific clustering indicator types and their weights are shown in Table 2.

[0049] Table 2 Clustering Index System and Weight Table

[0050]

[0051] S4. Preliminary Clustering: Combining the clustering indicators and weights obtained in S3, a k-means clustering algorithm is designed and implemented using Python. This algorithm is used to cluster the units divided in S2 according to their respective districts and counties. Before clustering, it is necessary to determine the range of the number of clusters in each district and county. This invention takes into account the distribution of residents' travel distances in each district and county, and sets the range of the number of clusters in each district and county as follows: In this embodiment, the range of cluster numbers for each district and county in Ningbo City is shown in Table 3 below.

[0052] Table 3. Clustering Upper and Lower Limits for Each District / County

[0053]

[0054] S5. Determining the number of clusters: Based on the clustering results corresponding to each cluster number obtained in S4, calculate the clustering validity value of each clustering result according to the clustering validity index formula.

[0055] S6. Determine the appropriate number of clusters. Based on the cluster validity values ​​corresponding to each cluster number calculated in S5, plot a line graph of cluster number versus cluster validity value. In this embodiment, Jiangdong District is used as an example; see the appendix for the line graph. Figure 2 The optimal number of clusters was determined by referring to the "elbow method". The optimal number of clusters for each district and county was obtained by calculating the value, as shown in Table 4 below.

[0056] Table 4 Optimal Cluster Numbers for Each District / County

[0057]

[0058] S7. Based on the cluster count of each district / county determined in S6, the FCM clustering algorithm is implemented using Python. Similarly, using the clustering index system and weights obtained in S3, secondary clustering is performed on each district / county to obtain the final traffic zone clustering results. For the clustering results of each district / county in Ningbo City in this embodiment, please refer to the appendix. Figure 3 .

[0059] S8. Clustering Result Correction: To further optimize the clustering results, the boundaries of traffic zones are defined by the road classification (highways, expressways, Class I roads, Class II roads, and Class III roads) and administrative district boundaries. In this embodiment, the clustering results obtained in S7 are re-segmented and corrected using road classification, ultimately completing the traffic zone division. The number of traffic zones in each district and county of Ningbo City is shown in Table 5 below. See the attached map for the final traffic zone division results of Ningbo City. Figure 4 .

[0060] Table 5. Number of Traffic Communities in Each District / County

[0061]

[0062] The second embodiment of the present invention is as follows:

[0063] Step 1: Obtain administrative division boundaries and hierarchical road network vector data, and perform data preprocessing;

[0064] In step 1, based on the acquired basic data, basic data processing is completed using ArcGIS professional software, including road network topology processing and intersection simplification.

[0065] Step 2: Divide the data into basic clustering units based on the road network and administrative divisions;

[0066] In step 2, the road network is used to segment the administrative divisions, and basic clustering units at the street scale are delineated for subsequent steps.

[0067] Step 3: Establish a clustering indicator system and, in conjunction with the clustering units from Step 2, obtain the travel characteristics of residents within the units through mobile phone signaling;

[0068] Step 4: Based on the administrative region where the clustering units are located, the k-means clustering algorithm is used for preliminary clustering to obtain the clustering results and cluster centers corresponding to each cluster number;

[0069] In step 4, before using the k-means algorithm for clustering, it is necessary to determine the range of the number of clusters based on the distribution of residents' travel distances within the district / county.

[0070] Clustering algorithms describe the clustering process using a cost function, which is:

[0071]

[0072] in, For the sample size, The number of clusters, the sample size and the number of clusters for each district / county are shown in Table 6. For the first One data sample, For the first The center of each cluster, Indicates sample Family membership degree For the first The sample to the first The "distance" between the centers of each cluster.

[0073] Table 6 Sample size of each district / county ( ) and cluster number range ( )

[0074]

[0075] Considering cluster units as geospatial objects, this "distance" should take into account not only the basic attribute distance but also spatial distance, as expressed below.

[0076]

[0077] In the formula, This represents the spatial distance between the sample and the cluster center. If the sample is adjacent to the cluster, its value is 0; otherwise, it represents the spatial distance from the sample centroid to the cluster centroid. This is the spatial effect coefficient; The attribute distance from a sample to the cluster center is expressed as follows:

[0078]

[0079] In the formula, For the sample The Clustering factors, For clusters The Clustering factors, For the first The weights of each clustering factor, For the sample with cluster The attribute distance.

[0080] Step 5: Calculate the clustering effectiveness index based on the clustering results obtained in Step 4 for each cluster number;

[0081] In step 5, an index that comprehensively considers both intra-group similarity and inter-group differences is used to evaluate the clustering results. The expression for the clustering effectiveness index is as follows:

[0082]

[0083] in, Representing clustering units to cluster Distance from the center Indicates the first The number of clustering units, Indicates the first The average distance from the cluster center to the cluster interior. The total number of cluster units. Used to characterize the similarity of clustering units within a cluster. For clusters From cluster center to cluster Distance from the center It is the sum of the pairwise distances between all cluster centers, used to characterize the differences between clusters.

[0084] Step 6: Compare the clustering effectiveness indices of each cluster number to determine the appropriate number of clusters;

[0085] In step 6, the optimal number of clusters is selected based on the plot of the number of clusters and the clustering effectiveness index, with reference to the "elbow method".

[0086] Step 7: Based on the number of clusters determined in Step 6, perform the final clustering using the FCM clustering algorithm;

[0087] In step 7, the FCM algorithm, which introduces fuzzy concepts, is used for secondary clustering, where the membership degree... ,and .

[0088] Step 8: Based on the main road levels, correct the clustering results obtained in Step 7 to complete the traffic zone division.

[0089] In step 8, the clustering results are corrected by filtering the main roads in the hierarchical road network, and the clustering results in step 7 are then divided a second time to obtain the final traffic zone division results.

[0090] The above embodiments are only used to illustrate the design concept and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, all equivalent changes or modifications made based on the principles and design ideas disclosed in the present invention are within the protection scope of the present invention.

[0091] It should be understood that the above description of the preferred embodiments is quite detailed, but it should not be considered as a limitation on the scope of protection of this invention. Those skilled in the art, under the guidance of this invention, can make substitutions or modifications without departing from the scope of protection of the claims of this invention, and all such substitutions or modifications fall within the scope of protection of this invention. The scope of protection of this invention should be determined by the appended claims.

Claims

1. A method for dividing traffic zones based on residents' travel characteristics, characterized in that: Step 1: Obtain administrative division boundaries and hierarchical road network vector data, and perform data preprocessing; Step 2: Divide the data into basic clustering units based on the road network and administrative divisions; Step 3: Establish a clustering indicator system and, in conjunction with the clustering units from Step 2, obtain the travel characteristics of residents within the units through mobile phone signaling; Step 4: Based on the administrative region where the clustering units are located, the k-means clustering algorithm is used for preliminary clustering to obtain the clustering results and cluster centers corresponding to each cluster number; Step 5: Calculate the clustering effectiveness index based on the clustering results obtained in Step 4 for each cluster number; Step 6: Compare the clustering effectiveness indices of each cluster number to determine the appropriate number of clusters; Step 7: Based on the number of clusters determined in Step 6, perform the final clustering using the FCM clustering algorithm; Step 8: Based on the main road levels, correct the clustering results obtained in Step 7 to complete the traffic zone division; In step 1, based on the acquired basic data, basic data processing is completed using ArcGIS professional software, including road network topology processing and intersection simplification. In step 2, the road network is used to segment the administrative divisions and divide them into basic clustering units at the street scale for subsequent steps. In step 4, before using the k-means algorithm for clustering, it is necessary to determine the range of the number of clusters based on the distribution of residents' travel distances within the district / county. Clustering algorithms describe the clustering process using a cost function, which is: in, For the sample size, The number of clusters is the number of divisions. For the first One data sample, For the first The center of each cluster, Indicates sample Family membership degree For the first The sample to the first The "distance" between the centers of each cluster; considering cluster units as geospatial objects, this "distance" should take into account not only the basic attribute distance but also spatial distance, as expressed below. In the formula, This represents the spatial distance between the sample and the cluster center. If the sample is adjacent to the cluster, its value is 0; otherwise, it represents the spatial distance from the sample centroid to the cluster centroid. This is the spatial effect coefficient; The attribute distance from a sample to the cluster center is expressed as follows: In the formula, For the sample The Clustering factors, For clusters The Clustering factors, For the first The weights of each clustering factor, For the sample with cluster The attribute distance; In step 5, an index that comprehensively considers both intra-group similarity and inter-group differences is used to evaluate the clustering results. The expression for the clustering effectiveness index is as follows: in, Representing clustering units to cluster Distance from the center Indicates the first The number of clustering units, Indicates the first The average distance from the cluster center to the cluster interior. The total number of cluster units. Used to characterize the similarity of clustering units within a cluster; For clusters From cluster center to cluster Distance from the center It is the sum of pairwise distances between all cluster centers, used to characterize the differences between clusters; In step 6, the optimal number of clusters is selected based on the plot of the number of clusters and the clustering effectiveness index, with reference to the "elbow method". In step 7, the FCM algorithm, which introduces fuzzy concepts, is used for secondary clustering, where the membership degree... ,and ; In step 8, the clustering results are corrected by filtering the main roads in the hierarchical road network, and the clustering results in step 7 are then divided a second time to obtain the final traffic zone division results.

Images