A method for perceiving urban traffic status based on residents' travel structure

Abstract

The present invention discloses a method for perceiving urban traffic status based on the travel structure of residents. First, based on the data of boarding and alighting points and remote sensing images in taxi trajectories, machine learning is used to extract the attributes of social functional areas. The attributes of social functional areas of boarding and alighting points in taxi trajectories are calibrated to establish a travel OD matrix for residents. The resident flows in the OD matrix are classified, and the proportion of each type of OD flow in the total travel volume over a period of time is counted, and finally the travel structure of residents in different periods is established. By analyzing the changes in the state of the travel structure of residents during normal social periods and epidemic outbreaks, the threshold of the travel structure of residents is extracted, and the urban traffic status is judged from the perspective of resident travel, providing a quantitative reference for urban managers to formulate targeted travel measures for residents. At the same time, the travel structure in the present invention can be established through other social perception data with resident travel location information, such as mobile phone signaling, bus IC card and other data, and has strong universality.

Description

Technical Field

[0001] The present invention belongs to the technical field of traffic planning, and in particular relates to a method for perceiving urban traffic status based on resident travel structure. Background Art

[0002] Therefore, this patent explores the extreme cases of residents' travel structure, and based on this, extends the residents' travel structure to the era of normalized epidemic situation, so that decision makers can perceive the urban traffic status through the real-time changes of residents' travel structure. Summary of the invention

[0003] Purpose of the invention: In view of the above problems, the purpose of the present invention is to provide a method for perceiving urban traffic status based on residents' travel structure, so as to provide accurate suggestions for decision makers to timely regulate residents' travel in the era of normalized epidemic situation.

[0004] Technical solution: To solve the above technical problems, the technical solution adopted by the present invention is:

[0005] A method for sensing urban traffic status based on residents' travel structure, comprising the following steps:

[0006] S1. First, the taxi trajectory is rectified as a whole, and empty files and abnormal points are deleted. The pick-up and drop-off points are extracted based on the original taxi trajectory data after rectification and cleaning;

[0007] S2. Fusion of trajectory data and remote sensing image data to extract social functional area attributes;

[0008] S3. Establish residents’ travel OD matrix according to the attributes of social functional areas;

[0009] S4. Divide the resident flow types in the OD matrix and extract the resident travel structure;

[0010] S5. Conduct quantitative evaluation of urban traffic status.

[0011] Furthermore, the original taxi trajectory data xls table in step S1 includes license plate number, call sign, time, longitude and latitude, speed, direction and passenger status fields.

[0012] Furthermore, the step S2 is specifically as follows:

[0013] S21. Establish a taxi trajectory sub-model, extract the time frequency series of the taxi trajectory pick-up and drop-off points in each block, and use K-Means++ and KNN algorithms to classify the urban social function attributes of the blocks.

[0014] S22. Build an image sub-model and use MLC-ResNets and YOLO v3 to recognize street images;

[0015] S23. Through the integrated model Γ(STET i ,T i ,I i ), to identify the functional zoning of each block; if STET is greater than or equal to the threshold ∈, the trajectory submodel Ψ(T i ); otherwise, use the image submodel Φ(I i ), the integrated model formula is as follows:

[0016]

[0017] Where n is the number of blocks, T i represents the trajectory data in block i, I i represents the satellite image of block i, is an indicator function, whose value is 1 when the condition is true; otherwise, it is 0, ∈ represents the decision threshold between using the trajectory sub-model or the image sub-model;

[0018] S24. Identify the results based on the model.

[0019] Furthermore, the step S3 is specifically as follows:

[0020] S31. The pick-up point in the taxi trajectory is marked as point O in the OD matrix, and the drop-off point is marked as point D;

[0021] S32. In ArcGIS software, spatial connection is used to determine the social function area attributes of the OD points in each taxi trajectory;

[0022] S33. Label the social functional areas respectively, count the total amount of OD resident flows between different social functional areas, and then establish the OD matrix of residents’ travel in different periods.

[0023] Furthermore, the step S4 is specifically as follows:

[0024] S41. Divide the OD flow according to the social function zone attributes of point O and point D;

[0025] S42. The resident travel structure is defined as the proportion of each type of resident travel volume in the total resident travel volume in a certain period of time;

[0026] S43. Count the number of different types of OD flows in a certain period of time, and calculate the ratio of each type of OD flow to the total OD flow in a certain period, that is, extract the residents' travel structure based on the OD matrix.

[0027] Furthermore, the step S5 is specifically as follows:

[0028] S51. Extract the travel structure of residents in normal social periods and epidemic outbreak periods respectively;

[0029] S52. The travel structures of residents in normal social periods and epidemic outbreak periods are used as the upper and lower limits of the healthy and unhealthy states of the city, respectively. When the travel structure of residents extracted based on real-time updated data is close to the lower limit or the proportion of a certain type of travel in the travel structure of residents shows abnormal fluctuations, targeted measures for a certain type of resident travel can be provided to decision makers from a quantitative perspective.

[0030] Beneficial effects: A method for perceiving urban traffic status based on the travel structure of residents is invented. First, the original taxi trajectory data is cleaned and rectified to extract the boarding and alighting points, and machine learning is used to identify the attributes of social functional areas based on the trajectory sub-model and the remote sensing image sub-model, thereby ensuring the accuracy of the OD matrix and the travel structure of residents in this article to the greatest extent. In addition, the travel structure of residents in the present invention can also be established by inputting other social perception data with resident travel location information, such as mobile phone signaling, public transportation IC card and other social perception data. The travel structure of residents obtained by the present invention through the fusion of multi-source data can help decision makers quantitatively perceive the urban traffic status from the perspective of resident travel, and formulate targeted measures for decision makers for a certain type of resident travel, which is conducive to the sustainable development of the city. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A flow chart of a method for sensing urban traffic status based on residents' travel structure according to the present invention;

[0032] Figure 2 A schematic diagram of passenger pick-up and drop-off point extraction according to the present invention;

[0033] Figure 3 The social function area attribute recognition result diagram of the present invention;

[0034] Figure 4 This is a diagram showing the division results of various types of OD flows according to the present invention. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the present invention will be described clearly and completely below so that those skilled in the art can better understand the advantages and features of the present invention, thereby making a clearer definition of the protection scope of the present invention. The embodiments described in the present invention are only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present invention.

[0036] like Figure 1 As shown, a city state perception method based on resident travel structure of the present invention comprises the following steps:

[0037] S1. Extract the pick-up and drop-off points based on the original taxi trajectory data after correction and cleaning:

[0038] S11. Get the taxi trajectory data of a certain city, where the original taxi trajectory data xls table includes license plate number, call sign, time, longitude and latitude, speed, direction and passenger status fields. Use the pandas library and numpy library in Python to first split the longitude and latitude fields into longitude and latitude fields, and create a new "state" field. Mark empty vehicles as 0 and heavy vehicles as 1. Next, delete empty files and longitude and latitude values ​​with abnormal longitude and latitude values. Because in actual situations, taxi drivers will end the fare before passengers get off the car, resulting in a large difference between the drop-off point recorded in the trajectory point and the actual situation. Therefore, when the vehicle trajectory point changes from a heavy vehicle to an empty vehicle and the distance is less than 50m, the point is marked as the drop-off point. The distance calculation formula is as follows:

[0039] dis=R·arccos[cosβ1cosβ2cos(α1-α2)+sinβ1sinβ2]

[0040] Among them, β1 and β2 are the latitude angles of the loaded and empty points, α1 and α2 are the longitude angles of the loaded and empty points, and R is the radius of the earth.

[0041] S12. After using the Python batch file to make a judgment according to the above formula, a new "delta" field is created, and the boarding point is marked as -1 and the drop-off point is marked as 1. Finally, all xls files containing the original trajectory of the taxi are processed into a csv file containing multiple records.

[0042] S13. Load the csv file in ArcGIS software and extract the pick-up and drop-off points in the taxi trajectory based on the “delta” field.

[0043] S2. Fusion of trajectory data and remote sensing image data to extract social function area attributes:

[0044] S21. First, obtain Landsat8 remote sensing image data, which contains 3 RGB bands and a resolution of 0.5m. Use taxi trajectory data and remote sensing image data to build and optimize two sub-models. Through the integrated model Γ(STET i ,T i ,I i ), as shown below, to identify the social function attributes of each block. If STET is greater than or equal to the threshold ∈, the trajectory submodel Ψ(T i ); otherwise, use the image submodel Φ(I i ).

[0045]

[0046] Where n is the number of blocks, T i represents the trajectory data in block i, I i represents the satellite image of block i, is an indicator function whose value is 1 when the condition is true and 0 otherwise. ∈ represents the decision threshold between using the trajectory sub-model or the image sub-model.

[0047] S22. Based on the model recognition results, this patent selects six types of social functional areas with better recognition effects and that can represent production and life, namely schools, residential areas, transportation centers, hospitals, scenic spots, and shopping centers.

[0048] S3. According to the attributes of social functional areas, establish the resident travel OD matrix:

[0049] S31. Based on the identification results of social functional area attributes, shopping centers, transportation centers, scenic spots, residential areas, schools and hospitals are marked with 1 to 6 respectively to facilitate subsequent work.

[0050] S32. In ArcGIS software, based on the social function zoning vector layer, the spatial connection function is used to determine the social function zone attributes of the starting and ending points of each taxi trajectory, that is, to calibrate the social function zone attributes of point O and point D.

[0051] S33. By counting the total amount of various OD flows in a certain period of time, the OD matrix of residents' travel in the corresponding time period is established.

[0052] S4. Divide the resident flow types in the OD matrix and extract the resident travel structure:

[0053] S41. According to the attributes of the social functional areas of the starting and ending points of residents' travel, the travel purposes of residents in the OD matrix are determined, which are commuting, shopping and entertainment, returning home for official business, seeking medical treatment, tourism and leisure, and random flexible travel.

[0054] S42. The resident flows in the OD matrix are further divided into commuting trips and elastic trips, among which elastic trips can be divided into ordinary elastic trips and random elastic trips. Commuting trips are caused by commuting to get off work and school, and such trips have significant temporal and spatial regularity. Elastic trips are generally caused by shopping and entertainment activities, and are uncertain in time and space. Ordinary elastic trips have certain regularity in space, but are random in time. Random elastic trips have no regularity in time and space.

[0055] S43. Divide the resident flows in the OD matrix into commuting travel, shopping and entertainment in ordinary flexible travel, returning home for official business, seeking medical treatment, tourism and leisure, and random flexible travel.

[0056] S44. Define the resident travel structure as the proportion of resident travel flows with a certain purpose in a certain time period to the total resident travel volume. Finally, calculate the proportion of each type of OD flow in a certain period of time to the total resident travel volume in that period, that is, establish the resident travel structure of the corresponding time period.

[0057] S5. Quantitative evaluation of urban traffic status:

[0058] S51. First, based on the development of the epidemic, it is divided into the normal social period before the epidemic, the epidemic outbreak period, the initial stage of the epidemic, etc.

[0059] S52. Based on the taxi trajectory data and remote sensing image data of this time period, the travel structure of residents during the normal social period and the epidemic outbreak period is extracted respectively, and used as the upper and lower limits of the residents' travel structure threshold.

[0060] S53. If the travel structure of residents extracted in real time based on data is close to the travel structure during the outbreak period, it proves that the city is in a sub-healthy state from the perspective of residents' travel. Decision-makers need to formulate targeted travel measures for residents based on the proportion of various types of resident flows in the travel structure at this time.

Claims

1. A method for sensing urban traffic status based on residents' travel structure, characterized in that: The following steps are involved: S1. First, the taxi trajectory is rectified as a whole, and empty files and abnormal points are deleted. The pick-up and drop-off points are extracted based on the original taxi trajectory data after rectification and cleaning; S2. Fusion of trajectory data and remote sensing image data to extract social functional area attributes; S3. Establish residents’ travel OD matrix according to the attributes of social functional areas; S4. Divide the resident flow types in the OD matrix and extract the resident travel structure; S5. Quantitatively evaluate the urban traffic status; The step S5 is specifically as follows: S51. Extract the travel structure of residents in normal social periods and epidemic outbreak periods respectively; S52. The travel structures of residents in normal social periods and during epidemic outbreaks are used as the upper and lower limits for judging the healthy and unhealthy urban traffic conditions, respectively. When the travel structure of residents extracted based on real-time updated data is close to the lower limit, or when the proportion of a certain type of travel in the travel structure of residents shows abnormal fluctuations, targeted measures for a certain type of resident travel can be provided to decision makers from a quantitative perspective.

2. A method for sensing urban status based on residents' travel structure according to claim 1, characterized in that: The original taxi trajectory data xls table in step S1 includes license plate, call sign, time, longitude and latitude, speed, direction and passenger status fields.

3. A method for sensing urban status based on residents' travel structure according to claim 1, characterized in that: The step S2 is specifically as follows: S21. Establish a taxi trajectory sub-model, extract the time frequency series of the taxi trajectory pick-up and drop-off points in each block, and use K-Means++ and KNN algorithms to classify the urban social function attributes of the blocks; S22. Build an image sub-model and use MLC-ResNets and YOLO v3 to recognize street images; S23. Through the integrated model Γ(STET i ,T i ,I i ), to identify the functional zoning of each block; if STET is greater than or equal to the threshold ∈, the trajectory submodel Ψ(T i ); otherwise, use the image submodel Φ(I i ), the integrated model formula is as follows: Where n is the number of blocks, T i represents the trajectory data in block i, I i represents the satellite image of block i, is an indicator function, whose value is 1 when the condition is true; otherwise, it is 0. ∈ indicates the decision threshold between using the trajectory sub-model or the image sub-model; S24. Identify the results based on the model.

4. A method for sensing urban status based on residents' travel structure according to claim 1, characterized in that: The step S3 is specifically as follows: S31. The pick-up point in the taxi trajectory is marked as point O in the OD matrix, and the drop-off point is marked as point D; S32. In ArcGIS software, spatial connection is used to determine the social function area attributes of the OD points in each taxi trajectory; S33. Label the social functional areas respectively, count the total amount of OD resident flows between different social functional areas, and then establish the OD matrix of residents’ travel in different periods.

5. A method for sensing city status based on residents' travel structure according to claim 1, characterized in that: The corresponding step S4 is specifically as follows: S41. Divide the OD flow according to the social function zone attributes of point O and point D; S42. The resident travel structure is defined as the proportion of each type of resident travel volume in the total resident travel volume in a certain period of time; S43. Count the number of different types of OD flows in a certain period of time, and calculate the ratio of each type of OD flow to the total OD flow in a certain period, that is, extract the residents' travel structure based on the OD matrix.

Images