A passenger flow OD algorithm based on multiple matching of public transportation passengers' trips

Abstract

The present invention discloses a passenger flow OD algorithm based on multiple matching of public transportation passengers, which is characterized by combining the initial bus card swiping data with the historical card swiping data, considering factors such as time, distance and travel rules on the basis of routes and platforms, and matching the boarding station and the alighting station for the initial bus card swiping data. Based on these data that complete the matching of the boarding station and the alighting station, the passenger flow OD data of different times, different routes and different stations can be obtained. The present invention can obtain the passenger flow OD data of different time periods, different routes and different platforms. The analysis method based on various factors can also realize the rapid analysis of the data, improve the utilization rate of the bus card swiping data, and facilitate the planning and adjustment of the actual bus routes, platforms, stations and other factors in combination with the operation information.

Description

Technical Field

[0001] The invention relates to the technical field of public transport passenger flow analysis, in particular to a passenger flow OD algorithm based on multiple matching of public transport passengers. Background Art

[0002] Ground public transportation is an important part of the urban transportation system. Accurate and quantitative evaluation of the operating status of ground public transportation is an urgent and practical need for urban transportation planning, organization and management, and is also the basis for providing public transportation information services to the public. The acquisition of passenger flow OD data can greatly reduce the difficulty of overall analysis of passenger flow data and improve the accuracy of overall evaluation of passenger flow data, providing a data basis for the analysis and improvement of ground public transportation operations. The traditional passenger flow OD data calculation method has the problem of low accuracy. Summary of the invention

[0003] The purpose of the present invention is to provide a passenger flow OD algorithm based on multiple matching of bus passenger travel, combining the initial bus card swiping data with the historical card swiping data, considering factors such as time, distance and travel rules on the basis of routes and platforms, and matching the boarding station and the alighting station for the initial bus card swiping data. Based on these data that complete the matching of the boarding station and the alighting station, the passenger flow OD data of different times, different routes and different stations can be obtained to solve the problems raised in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A passenger flow OD algorithm based on multiple matching of bus passenger trips, including bus card swiping data matching boarding stations and bus card swiping data matching alighting stations.

[0006] 1. Bus card swipe data matches the boarding station

[0007] The vehicle GPS data is combined with the actual platform data. When the difference between the vehicle GPS data and the actual platform data is within a certain range, the vehicle is considered to have stopped at the platform. The vehicle's entry and exit times can then be obtained. These data are used as comparison data to match the boarding point.

[0008] Initial card swiping data: The IC card ID number, swiping time, line identifier and vehicle identifier can be obtained from the original IC card swiping data. These data are used as basic data to participate in the matching of boarding points.

[0009] Platform GPS data: According to the actual operation of the bus, the GPS data of different line tracks and different platforms can be obtained. These data are used as comparison data to participate in the matching of boarding points.

[0010] Vehicle GPS data: Based on the actual operation of the bus, the GPS data of different vehicles on different routes can be obtained. These data are used as comparison data to match the boarding point.

[0011] The process of matching bus card swiping data with boarding stations is as follows:

[0012] (3) Data preprocessing

[0013] d. Data extraction: extract the existing data and put them into designated containers. For example, the initial card swipe data should be extracted from the original IC card swipe data, and the station GPS data should be extracted from the bus operation data. The extracted data is placed in the designated container and awaits subsequent use.

[0014] e. Data standardization: Standardize the data extracted and placed in the specified container, such as unifying the form and meaning of the line identifiers in the initial card swiping data, platform GPS data, and vehicle entry and exit time data. For the convenience of display, you can also create a table to specify the correspondence between the line identifier and the line name.

[0015] f. Data cleaning: Due to the influence of bus hardware and the actual operating environment, there may be erroneous data and redundant data in the initial IC card swiping data. These data have a great impact on the accuracy of passenger flow OD data acquisition and analysis. Therefore, the data needs to be cleaned before use to exclude erroneous data and redundant data from the actual analysis data; these erroneous data or redundant data have many types and need to be analyzed according to the actual operation of the bus. For example, due to the influence of the card swiping environment or card swiping equipment, some card swiping data keywords are missing. These data often cannot be matched to the boarding and alighting stations through comparison, so OD matching is often not performed, but because these data really exist, they are included in the passenger flow statistics.

[0016] (4) Boarding station matching

[0017] When matching the boarding point, it is necessary to use the initial card swiping data, platform GPS data and vehicle entry and exit time data. Generally, the initial card swiping data and vehicle entry and exit time data are sorted according to the card swiping time and vehicle entry time respectively. When matching, classification and comparison are performed based on the line identifier, platform identifier, vehicle identifier and other information.

[0018] 2. Bus card swipe data matches the alighting station

[0019] Based on the actual situation of passengers taking the bus, the passengers' travel is divided into four scenarios. The first one is that the passenger gets on the bus at the second to last stop of his / her line; the second one is that the card swiping data is the passenger's last trip of the day, and the passenger's number of trips on that day is not unique; the third one is that the card swiping data is not the passenger's last trip of the day; the fourth one is the card swiping data of the passenger's only trip of the day.

[0020] New integrated data can be obtained by using the boarding and alighting stations matched based on the initial card swiping data, platform GPS data, and vehicle entry and exit time data. Using this integrated data, data at different times, different routes, and different stations can be analyzed in combination with specific needs.

[0021] Compared with the prior art, the present invention can obtain passenger flow OD data of different time periods, different routes and different platforms. The analysis method based on various factors can also realize rapid analysis of the data, improve the utilization rate of bus card swiping data, and facilitate planning and adjustment of actual bus routes, platforms, stations and other factors in combination with operational information. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a flow chart of boarding station matching in a passenger flow OD algorithm based on multiple matching of bus passengers' travel.

[0023] Figure 2 This is a flow chart of getting-off station matching in a passenger flow OD algorithm based on multiple matching of bus passengers' travel.

[0024] Figure 3 This is a schematic diagram of the application of passenger flow OD data.

[0025] Figure 4 It is a principle block diagram of the passenger flow OD algorithm with multiple matching of the present invention.

[0026] Figure 5 This is a schematic diagram of an embodiment of the present invention in which a passenger gets on a bus at the second to last stop of the line on which the passenger is located.

[0027] Figure 6 It is a schematic diagram of a travel route that is opposite or a circular route in an embodiment of the present invention.

[0028] Figure 7 This is a schematic diagram of a travel route in which the distance between subsequent platforms meets the requirements in an embodiment of the present invention.

[0029] Figure 8 This is a schematic diagram of comparing the last trip of the day with the first trip of the next day in an embodiment of the present invention.

[0030] Fig. 9 This is a schematic diagram of two consecutive trips with the same route and direction in an embodiment of the present invention.

[0031] Fig.10 This is a schematic diagram of two consecutive trips with the same routes and opposite directions in an embodiment of the present invention.

[0032] Fig.11 This is a schematic diagram of two consecutive trips with inconsistent routes according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The technical solution of this patent is further described in detail below in conjunction with specific implementation methods.

[0034] See also Figure 4 , a passenger flow OD algorithm based on multiple matching of bus passenger travel, based on the existing bus card swiping data, combines the current card swiping data with the historical card swiping data, and matches the corresponding boarding station and alighting station for the current card swiping data. The boarding station matching of the current card swiping data utilizes the GPS, line and other information in the IC card swiping data, and compares and analyzes with the actual platform information to complete the matching; the alighting station matching of the current card swiping data fully considers the passenger alighting situation and travel rules under different circumstances, and compares and analyzes the current card swiping data with the historical data to complete the matching. Based on these card swiping data matched to the boarding station and the alighting station, the passenger flow OD data can be obtained. Analyzing the passenger flow OD data according to different factors can realize the rapid application of the passenger flow OD data.

[0035] A passenger flow OD algorithm based on multiple matching of bus passenger trips, including bus card swiping data matching boarding stations and bus card swiping data matching alighting stations.

[0036] 1. Bus card swipe data matches the boarding station

[0037] Initial card swiping data: The IC card ID number, swiping time, line identifier and vehicle identifier can be obtained from the original data of IC card swiping. These data are used as basic data to participate in the matching of boarding points. The data format is shown in Table 1.

[0038] Table 1 Initial card swiping data

[0039]

[0040] Platform GPS data: According to the actual operation of the bus, the GPS data of different line tracks and different platforms can be obtained. These data are used as comparison data to participate in the matching of boarding points. The data format is shown in Table 2.

[0041] Table 2 Station GPS data

[0042]

[0043] Vehicle GPS data: According to the actual operation of the bus, the GPS data of different vehicles on different routes can be obtained, and these data are used as comparison data to participate in the matching of boarding points. The data format is shown in Table 3.

[0044] Table 3 Vehicle GPS data

[0045]

[0046] The vehicle GPS data is combined with the actual platform data. When the difference between the vehicle GPS data and the actual platform data is within a certain range, it is considered that the vehicle is parked at the platform. Then the vehicle's entry and exit time can be obtained. These data are used as comparison data to participate in the matching of boarding points. The vehicle entry and exit time data are shown in Table 4 below:

[0047] Table 4 Vehicle entry and exit time data

[0048]

[0049] like Figure 1 As shown in the figure, the process of matching bus card swiping data with boarding stations is as follows:

[0050] (1) Data preprocessing

[0051] 1) Data extraction: extract the existing data and put them into designated containers. For example, the initial card swipe data should be extracted from the original IC card swipe data, and the station GPS data should be extracted from the bus operation data. The extracted data is placed in the designated container and awaits subsequent use.

[0052] 2) Data standardization: Standardize the data extracted and placed in the specified container, such as unifying the form and meaning of the line identifiers in the initial card swiping data, platform GPS data, and vehicle entry and exit time data. For the convenience of display, you can also create a table to specify the correspondence between line identifiers and line names. In the initial card swiping data, there are many types of bus IC cards, such as ordinary cards, student cards, senior citizen cards, love cards, etc., and the ID number of the IC card is also divided into the ID number inside the card and the ID number on the card surface. Therefore, it is necessary to standardize the data before using it, determine the logic for obtaining the IC card ID number, and unify the type of IC card ID number.

[0053] 3) Data cleaning: Due to the influence of bus hardware and the actual operating environment, there may be erroneous data and redundant data in the initial IC card swiping data. These data have a great impact on the accuracy of passenger flow OD data acquisition and analysis. Therefore, the data needs to be cleaned before use to exclude erroneous data and redundant data from the actual analysis data. There are many types of these erroneous data or redundant data, which need to be analyzed according to the actual operation of the bus. For example, due to the influence of the card swiping environment or card swiping equipment, some card swiping data keywords are missing. These data often cannot be matched to the boarding and alighting stations through comparison, so OD matching is often not performed, but because these data really exist, they are included in the passenger flow statistics.

[0054] (2) Boarding station matching

[0055] When matching the boarding station, it is necessary to use the initial card swiping data, platform GPS data and vehicle entry and exit time data. Generally, the initial card swiping data and vehicle entry and exit time data are sorted according to the card swiping time and vehicle entry time respectively. When matching, they are classified and compared according to the line identifier, platform identifier, vehicle identifier and other information. The specific matching is divided into the following situations:

[0056] 1) The vehicle is dispatched for the first time on the day

[0057] For the first departure of the vehicle on the day, if the card swiping time T of some data in the initial card swiping data is earlier than the time t that the vehicle entered the departure station s ,Right now

[0058] T<t s ,

[0059] The boarding station of these initial card swiping data is the starting station of the line.

[0060] It should be noted that a time threshold t needs to be set during data preprocessing. y , if the time t at which the vehicle enters the departure station s The time difference with the card swiping time T is greater than t y ,Right now

[0061] t s -T>t y ,

[0062] Then this data needs to be excluded from the actual analysis data during data cleaning.

[0063] 2) Card swiping between stations

[0064] In the initial card swiping data, for the card swiping time T of a known vehicle identifier, there is an entry time interval (t s1 ,t s2 ), where t s1 With t s2 is the arrival time of two adjacent platforms.

[0065] t s1 ≤T<t s2 ,

[0066] Then t s1 The corresponding stations are the boarding stations of these card swiping data.

[0067] 3) Swipe your card across multiple lines

[0068] After the vehicle arrives at the last stop, all passengers will get off according to the travel rules. At this time, there should be no boarding data, that is, the boarding data at the last stop should be 0. If in the initial card swiping data, for the card swiping time T of a known vehicle identifier, there is an entry time interval (t s1 ,t s2 ), where t s1 With t s2 are the time when the vehicle arrives at the last stop and the time when the vehicle starts again and enters the first stop.

[0069] t s1 ≤T<t s2 ,

[0070] The boarding stations of these initial card swiping data are the starting stations of the route where the vehicle will start again, that is, t s2 The corresponding site.

[0071] It should be noted that during data preprocessing, it is necessary to determine whether the vehicle has the next trip after arriving at the terminal. If the vehicle does not travel on the same day after arriving at the terminal, then for the card swiping time T, if there is

[0072] T>t s1 ,

[0073] Then this data needs to be excluded from the actual analysis data during data cleaning.

[0074] 2. Bus card swipe data matches the alighting station

[0075] like Figure 2As shown in the figure, based on the actual situation of passengers taking the bus, the passengers' travel is divided into four scenarios. The first scenario is that the passenger's boarding station is the second to last station of the line; the second scenario is that the passenger's card swiping data is the last trip of the day, and the passenger's number of trips on the day is not unique; the third scenario is that the passenger's card swiping data is not the last trip of the day; the fourth scenario is the passenger's only trip data on the day. The following analyzes the four travel scenarios of passengers respectively:

[0076] (1) The passenger boarding station is the second to last station on the route

[0077] like Figure 5 As shown, when the passenger gets on the bus at the second to last stop of the line, the alighting stop is directly matched as the last stop of the line for the card swiping data.

[0078] (2) The passenger’s last trip of the day, and the passenger’s number of trips on the day is not unique

[0079] 1) If Figure 6 As shown in the figure, it is known that the route of the vehicle taken by the passenger for the last trip of the day belongs to route A. If the passenger also has route A on other trips on the same day, and the other trips are in the opposite direction of the last trip route or route A is a circular route, then the passenger's last trip card swiping data of the day The stop corresponding to the trip boarding stop that meets the conditions and is closest to the last trip time.

[0080] It should be noted that since the departure and destination directions of the current data and the comparison data are opposite, when the last travel route and other travel routes that meet the conditions are both circular routes, the station corresponding to the boarding station is the boarding station itself; when the last travel route and other travel routes that meet the conditions are in opposite directions, the station corresponding to the boarding station is the opposite station of the boarding station, and this is the case below.

[0081] 2) If Figure 7 As shown in the figure, it is known that the route of the vehicle that the passenger took on the last trip of the day is assumed to be route B. If the passenger's other travel routes on the same day do not have route B, but the calculation starts from the passenger's boarding station, a distance threshold d is set according to the travel pattern. y , among the subsequent stations of line B, there are stations that are at a distance D from the other boarding stations of passengers.

[0082] D<d y ,

[0083] The passenger's last trip card swiping data alighting station on that day is the boarding station of the trip card swiping data closest to the last trip time that meets the conditions.

[0084] It should be noted that the distance threshold d herey It needs to be set according to the general travel rules of passengers. y The setting of has a great relationship with the accuracy of the matching of the alighting station. y =800m.

[0085] 3) If both conditions 1) and 2) are not met, the passenger's historical card swipe data is considered. If the card swipe data of the previous day or the previous few days contains the same route, the same direction of the card swipe data matched to the get-off station, and the travel time period is similar, then the passenger's last trip swipe data of the day The get-off station is the closest to the trip time that meets the conditions.

[0086] 4) If Figure 8 As shown, if conditions 1), 2), and 3) are not met, then the travel data of the passenger on the next day is considered. If the passenger's first trip route on the next day is the same as the last trip route on the same day, the direction is opposite, or it is the same circular route, then the passenger's last trip card swiping data of the day The alighting station is the corresponding station of the first trip boarding station on the next day; if the consideration starts from the time when the passenger swipes the card to board the bus, if there is a station in the subsequent stations of the last trip route on the same day and the distance D between the passenger's first trip boarding station on the next day meets

[0087] D<d y ,

[0088] The passenger's last trip card swiping data of the day's alighting station will be the corresponding station for the first trip on the next day.

[0089] (3) The passenger is not travelling for the last time on the day

[0090] Based on the initial card swiping data, the boarding station of each passenger's trip can be determined. Therefore, we can consider whether the passenger's two consecutive trips are on the same route. If they are on the same route, we can analyze the passenger's alighting station whether the two consecutive trips are in the same direction. The specific process is as follows:

[0091] 1) If the passenger takes the same route for two consecutive trips, there are two cases to discuss:

[0092] a. Fig. 9 As shown, if a passenger travels in the same direction twice in a row, let a1 be the boarding stop for the first trip and a2 be the boarding stop for the second trip, and if a2 is after a1 in that direction of the route, then a2 is the alighting stop for the first trip.

[0093] b. Fig.10As shown, if a passenger travels in opposite directions twice in a row, let the first boarding station be b1 and the second boarding station be b2, then the alighting station for the first trip is the station opposite to b2.

[0094] 2) If Fig.11 As shown in the figure, if the passenger's two consecutive trips are not the same, assuming that the passenger's first trip route is A, the card boarding station is a, and the second trip route is B, the card boarding station is b. If route A starts from station a, there is a subsequent station where the distance D between platform a1 and station b is less than the set distance threshold d y , then a1 is the passenger's first get-off stop.

[0095] (4) The passenger's only trip on that day

[0096] 1) Consider the passenger's historical card swipe data. If the card swipe data from the previous day or a few days contains the same route and the same direction of the card swipe data matched to the get-off station, and the travel time period is similar, then the get-off station of the passenger's last trip swipe data on that day is the get-off station closest to the travel time that meets the conditions.

[0097] 2) Consider the passenger's travel data for the next day. If the passenger's first trip route the next day is the same as the last trip route of the day, or the direction is opposite or the same circular route, then the passenger's last trip card swiping data of the day's alighting station is the corresponding station for the passenger's first trip boarding station the next day; if the passenger swipes the card to board the bus from the beginning, if there is a station in the subsequent stations of the last trip route of the day and the passenger's first trip boarding station the next day that satisfies the distance D

[0098] D<d y ,

[0099] The passenger's last trip card swiping data of the day's alighting station will be the corresponding station for the first trip on the next day.

[0100] See the matching process diagram for details. Figure 2 .

[0101] 3. Some uses of passenger flow OD data

[0102] New integrated data can be obtained by using the boarding and alighting stations matched based on the initial card swiping data, platform GPS data, and vehicle entry and exit time data. Using these integrated data, data at different times, different routes, and different stations can be analyzed in combination with specific needs. The integrated data is shown in Table 5 below:

[0103] Table 5 Integrated data based on initial data and matching of upper and lower sites

[0104]

[0105] (1) Route passenger flow analysis

[0106] To analyze the passenger flow of a route, first determine the time period for analysis. According to the travel characteristics, most of the time for going to work and school is concentrated in the morning and evening. The passenger flow during these times accounts for a large proportion of the whole day, which is often referred to as the morning peak and evening peak. By dividing the daily time into morning peak, evening peak and all day, the travel patterns of different time periods can be analyzed more clearly. Here, according to the travel pattern, the morning peak is set to 07:00-09:00 and the evening peak is set to 17:00-19:00.

[0107] On the basis of dividing different time periods, data can be aggregated according to routes to obtain complete card swiping data and passenger flow OD data of different routes, and analyze the passenger flow of each route; based on the passenger flow OD data, the boarding and alighting data of each station can be obtained, and whether the setting of the route stations is reasonable can be determined according to the passenger flow of getting on and off the bus; on this basis, the alighting passenger flow of each platform at other platforms can be analyzed. For two platforms with larger passenger flow, consider whether to add routes between the platforms; for platforms with smaller passenger flow, consider whether to remove duplicate routes between the two platforms.

[0108] (2) Cross-sectional passenger flow analysis

[0109] For the cross-sectional passenger flow analysis, first determine the time unit for cross-sectional division. The time unit should not be too large or too small. If the time unit is too large, the analysis of different time periods will not be clear enough and cannot be accurate to the required time period; if the time unit is too small, the passenger flow within the time unit will be small, making it difficult to perform data analysis. For the analysis of bus passenger flow, it is better to divide the time unit into half an hour or one hour.

[0110] On the basis of the time unit division, the OD passenger flow in different routes and different time periods can be considered. According to the card swiping time and alighting time in Table 5, the passenger flow on and off the bus in different routes and different time periods can be obtained. By comparing with the passenger flow data in different time periods of the route, the departure frequency of the route in a certain time period can be analyzed to consider whether it is necessary to increase or reduce the vehicle frequency in a specific time period.

[0111] (3) Station passenger flow analysis

[0112] For the passenger flow analysis of the station, we first consider dividing the analysis time period into morning peak, evening peak and all day. By aggregating the passenger flow OD data of different time periods, we can obtain the passenger flow data of boarding and alighting at a certain platform in different time periods. Combined with the line, we can analyze the rationality of the platform setting; further, different platforms can be divided into different areas according to geographical characteristics, and the passenger flow OD data between different areas can be analyzed. Studying the passenger flow trend in different time periods in different areas can plan and adjust the line platforms in different areas.

[0113] It should be noted that this is only part of the application of passenger flow OD data. Other application methods can be selected according to actual conditions.

[0114] The above describes in detail the preferred implementation of this patent, but this patent is not limited to the above implementation. Various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of this patent.

Claims

1. A passenger flow OD algorithm based on multiple matching of public transportation passengers, characterized in that: Combine the initial bus card swipe data with the historical card swipe data, consider time, distance and travel rules based on the route and platform, and match the boarding and alighting stations for the initial bus card swipe data; Based on these data, the matching of boarding stations and alighting stations can be completed, that is, the passenger flow OD data of different times, different routes and different stations can be obtained; including bus card swiping data matching boarding stations and bus card swiping data matching alighting stations; (1) Bus card swipe data matches the boarding station Combine the vehicle GPS data with the actual platform data. When the difference between the vehicle GPS data and the actual platform data is within a certain range, the vehicle is considered to have stopped at the platform. Then the vehicle's entry and exit times can be obtained. These data are used as comparison data to match the boarding point. (2) Bus card swipe data matches the alighting station Based on the actual situation of passengers taking public transportation, the passengers' trips are divided into four scenarios. The first scenario is that the passenger's boarding station is the second to last station of the line he / she is on; the second scenario is that the card swiping data is the passenger's last trip of the day, and the passenger's number of trips on the day is not unique; the third scenario is that the passenger's card swiping data is not the last trip of the day; the fourth scenario is the passenger's only trip data on the day; The process of matching bus card swiping data with boarding stations is as follows: 1) Data preprocessing a. Data extraction: extract existing data and put them into designated containers for subsequent use; b. Data standardization: standardize the data extracted and placed in the specified container; c. Data cleaning: Due to the influence of bus hardware and the actual operating environment, there may be erroneous data and redundant data in the initial IC card swiping data. These data have a great impact on the accuracy of passenger flow OD data acquisition and analysis. Therefore, the data needs to be cleaned before use to exclude erroneous data and redundant data from the actual analysis data; There are many types of these erroneous or redundant data, which need to be analyzed based on the actual operation of public transportation; 2) Boarding station matching When matching the boarding point, it is necessary to use the initial card swiping data, platform GPS data and vehicle entry and exit time data. Generally, the initial card swiping data and vehicle entry and exit time data are sorted according to the card swiping time and vehicle entry time respectively. When matching, they are classified and compared according to the line identifier, platform identifier, and vehicle identifier information.

2. According to claim 1, a passenger flow OD algorithm based on multiple matching of public transportation passengers is characterized in that: The boarding station matching is divided into the following situations (1) The vehicle departs for the first time on the day For the first departure of the vehicle on the day, if the card swiping time T of some data in the initial card swiping data is earlier than the time t when the vehicle enters the departure station s , that is, T<t s , then the boarding station of these initial card swiping data is the starting station of the line; It should be noted that a time threshold t needs to be set during data preprocessing. y , if the time t at which the vehicle enters the departure station s The time difference with the card swiping time T is greater than t y , that is, t s -T>t y , then the data needs to be excluded from the actual analysis data during data cleaning; (2) Card swiping between stations In the initial card swiping data, for the card swiping time T of a known vehicle identifier, there is an entry time interval (t s1 ,t s2 ), where t s1 With t s2 is the arrival time of two adjacent platforms. If t s1 ≤T<t s2 , then t s1 The corresponding stations are the boarding stations of these card swiping data; (3) Swipe your card across multiple lines After the vehicle arrives at the last stop, all passengers will get off according to the travel rules. At this time, there should be no boarding data, that is, the boarding data at the last stop should be 0; if in the initial card swiping data, for the card swiping time T of a known vehicle identifier, there is an entry time interval (t s1 ,t s2 ), where t s1 With t s2 are the time when the vehicle arrives at the last stop and the time when the vehicle starts again and enters the first stop, respectively. If there is t s1 ≤T<t s2 , then the boarding stations of these initial card swiping data are the starting stations of the route where the vehicle re-departs, that is, t s2 The corresponding site; It should be noted that during data preprocessing, it is necessary to determine whether the vehicle has the next trip after arriving at the terminal. If the vehicle does not travel on the same day after arriving at the terminal, then for the card swiping time T, if T>t s1 , then the data needs to be excluded from the actual analysis data during data cleaning.

3. The passenger flow OD algorithm based on multiple matching of public transportation passengers according to claim 1 is characterized in that: When the passenger gets on the bus at the second-to-last stop on his / her line: For this card swiping data, the alighting station is directly matched as the last stop of the line.

4. The passenger flow OD algorithm based on multiple matching of public transportation passengers according to claim 1 is characterized in that: When the passenger's last trip of the day, and the number of trips the passenger has made on that day is not unique: (1) If the route of the vehicle taken by the passenger for the last trip of the day is known, and if the passenger has taken the same route on other trips of the day, and the other trips are in the opposite direction of the last trip or are circular routes, then the alighting station of the passenger’s last trip card swiping data on the day shall be the station corresponding to the boarding station of the trip that meets the conditions and is closest to the last trip time; Since the departure and destination of the current data and the comparison data are in opposite directions, when the last trip route and other trip routes that meet the conditions are both circular routes, the station corresponding to the boarding station is the boarding station itself; when the last trip route and other trip routes that meet the conditions are in opposite directions, the station corresponding to the boarding station is the station opposite to the boarding station; (2) If the route of the vehicle that the passenger took on the last trip of the day is known, and if the passenger does not have this route on other trips that day, but starts from the passenger's boarding station, a distance threshold d is set according to the travel pattern. y , among the subsequent stations of this line, there are stations with a distance D from the other boarding stations of passengers. If D<d y , then the passenger's last trip card swipe data alighting station on that day is the closest to the last trip card swipe data boarding station that meets the conditions; the distance threshold d here y It needs to be set according to the general travel rules of passengers. y The setting of is closely related to the accuracy of the matching of the alighting station; (3) If conditions (1) and (2) are not met, the passenger's historical card swipe data is considered; if the card swipe data from the previous day or a few days contains card swipe data for the same route and direction that matches the alighting station, and the travel time period is similar, then the alighting station of the passenger's last trip swipe data on that day shall be the alighting station of the trip closest to the travel time that meets the conditions. (4) If conditions (1), (2), and (3) are not met, then the passenger's travel data for the next day is considered; if the passenger's first trip route on the next day is the same as the last trip route on the same day, or is in the opposite direction or is the same circular route, then the passenger's last trip card swiping data alighting station on the same day is the corresponding station for the passenger's first trip boarding station on the next day; if the passenger starts to consider the station from the time when he swiped his card to board the bus, if there is a station among the subsequent stations of the last trip route on the same day and the passenger's first trip boarding station on the next day, the distance D satisfies D < d y , then the passenger's last trip card swiping data alighting station of the day is the corresponding station of the first trip boarding station the next day.

5. The passenger flow OD algorithm based on multiple matching of public transportation passengers according to claim 1 is characterized in that: When the passenger is not travelling for the last time that day: Based on the initial card swiping data, the passenger's boarding station for each trip can be determined, so we can consider whether the passenger's two consecutive trips are on the same route. If they are the same route, whether the two consecutive trips are in the same direction can be analyzed to determine the passenger's alighting station.

6. The passenger flow OD algorithm based on multiple matching of public transportation passengers according to claim 1 is characterized in that: When the passenger is travelling for the only time that day: (1) Consider the passenger's historical card swiping data; if the card swiping data from the previous day or a few days contains the same route and the same direction of the card swiping data that matches the get-off station, and the travel time period is similar, then the get-off station of the passenger's last trip swiping data on that day is the get-off station of the trip closest to the travel time that meets the conditions; (2) Consider the passenger's travel data for the next day; if the passenger's first trip route on the next day is the same as the last trip route on the day, or the direction is opposite to the last trip route on the day, or the route is the same loop route, then the passenger's last trip card swiping data alighting station on the day is the corresponding station for the passenger's first trip boarding station on the next day; if the passenger swipes the card to board the bus from the time the passenger boards the bus, if there is a station in the subsequent stations of the last trip route on the day whose distance D from the passenger's first trip boarding station on the next day satisfies D < d y , then the passenger's last trip card swiping data alighting station of the day is the corresponding station of the first trip boarding station the next day.

7. A passenger flow OD algorithm based on multiple matching of public transportation passengers according to any one of claims 1 to 6, characterized in that: New integrated data can be obtained by using the boarding and alighting stations matched based on the initial card swiping data, platform GPS data, and vehicle entry and exit time data. By utilizing these integrated data and combining them with specific needs, we can analyze data at different times, different routes, and different stations; the analysis includes but is not limited to route passenger flow analysis, section passenger flow analysis, and station passenger flow analysis.

8. The passenger flow OD algorithm based on multiple matching of public transportation passengers according to claim 7 is characterized in that: The specific processes of route passenger flow analysis, section passenger flow analysis, and station passenger flow analysis are as follows: (1) Route passenger flow analysis To analyze the passenger flow of a route, first determine the time period for analysis; on the basis of dividing different time periods, aggregate the data according to the route, obtain the complete card swiping data and passenger flow OD data of different routes, and analyze the passenger flow of each route; based on the passenger flow OD data, obtain the boarding and alighting data of each station, and determine whether the setting of the route station is reasonable according to the passenger flow of getting on and off the bus; on this basis, analyze the passenger flow of getting off each platform at other platforms, and consider whether to add a line between two platforms with large passenger flow, and consider whether to remove the duplicate lines between two platforms for platforms with small passenger flow; (2) Cross-sectional passenger flow analysis For the passenger flow analysis of the section, first determine the time unit for section division; on the basis of the completion of the time unit division, the OD passenger flow in different routes and different time periods can be considered; according to the card swiping time and the alighting time, the passenger flow in different routes and different time periods can be obtained, and by comparing with the passenger flow data in different time periods of the route, the departure frequency of the route in a certain time period can be analyzed to consider whether it is necessary to increase or reduce the vehicle frequency in a specific time period; (3) Station passenger flow analysis For the passenger flow analysis of the station, we first consider dividing the analysis time period into morning peak, evening peak and whole day; by aggregating the passenger flow OD data of different time periods, we can obtain the passenger flow data of getting on and off a platform in different time periods, and analyze the rationality of the platform setting in combination with the line; further, different platforms can be divided into different areas according to geographical characteristics, and the passenger flow OD data between different areas can be analyzed. By studying the passenger flow flow trends in different time periods in different areas, we can plan and adjust the line platforms in different areas.

Images