CN107798869B

CN107798869B - Bus passenger flow acquisition and analysis method based on station WiFi and vehicle-mounted GPS

Info

Publication number: CN107798869B
Application number: CN201710997548.1A
Authority: CN
Inventors: 王礼睿; 李志斌; 于维杰; 杨昊明; 王诗菡; 张应恒; 陆钥
Original assignee: Southeast University
Current assignee: Suzhou Hongtu Smart City Technology Co.,Ltd.
Priority date: 2017-10-24
Filing date: 2017-10-24
Publication date: 2020-09-11
Anticipated expiration: 2037-10-24
Also published as: CN107798869A

Abstract

The invention provides a bus passenger flow acquisition and analysis method based on a station WiFi and a vehicle-mounted GPS, wherein MAC address information of a mobile terminal held by a user in a detection range near each station of a bus line is acquired through a station Wi-Fi detection device, and a real-time position of a bus is acquired through vehicle-mounted GPS positioning; cleaning interference data according to the coincidence relation between the time of getting on or off the bus and the time of getting on or off the bus, respectively obtaining the user information of the passengers getting on or off the bus at each station, and calculating and summarizing to obtain the passenger flow of each station; according to the Wi-Fi detection device based on the bus stop fixed position, the time-space double filtering combining stop Wi-Fi detection and vehicle-mounted GPS positioning is adopted, peripheral sudden interference which cannot be processed by the vehicle-mounted Wi-Fi detection of the bus is effectively eliminated, and errors are greatly reduced. As the coverage of Wi-Fi hotspots becomes wider, the accuracy of the system will also become higher.

Description

Bus passenger flow acquisition and analysis method based on station WiFi and vehicle-mounted GPS

The technical field is as follows:

the invention provides an intelligent traffic management method, and particularly relates to a bus passenger flow acquisition and analysis method based on site WiFi and vehicle-mounted GPS, which is used for screening data by using R language and ArcGIS and visually presenting by using ArcGIS projection.

Background art:

at the present stage, along with the continuous highlighting of urban traffic contradictions, an intelligent public traffic system is developed, the improvement of the operation efficiency of the public traffic system becomes an important means for relieving urban traffic jam, and the research and analysis on public traffic passenger flow also gradually becomes a research hotspot in the current traffic field.

The existing bus passenger flow traveling estimation method comprises an IC card detection method, Wi-Fi identification and other methods. The bus IC card has simple and mature technology, large information collecting quantity and application in many cities, and the quantity of bus passengers can be obtained by utilizing the proportion of the bus passengers using the IC cards, so that the traffic flow characteristic investigation by adopting an IC card method is more common. However, since a large number of buses only need to get on and punch cards, the getting-off information of passengers cannot be accurately obtained. From the precision of the investigation, the IC card investigation is obviously not completely satisfactory. For the existing Wi-Fi identification method, in the Chinese patent application with the publication number of CN103700174A, namely 'a method for collecting and analyzing bus passenger flow data based on Wi-Fi identity identification', the address of a Wi-Fi device held by a passenger is read through a bus AP hotspot, and then bus passenger flow information is obtained through statistics. The above method does not provide a reasonable data screening method to exclude the influence of the Wi-Fi of surrounding vehicles and people on the measured data. Meanwhile, the method does not utilize GPS data to determine the accurate position, and only the appearance and disappearance of Wi-Fi signals in the hotspot network range are used as the basis for passengers to get on or off the vehicle, so that the accuracy of the final result cannot be ensured. In the chinese patent application publication No. CN106297288A, "a method for collecting and analyzing passenger flow data of a bus", stop information is acquired by using a Wi-Fi hotspot and a GPS signal on the bus, interference data is removed by using time and distance as standards, and passenger flow is obtained by statistics. However, the method uses the Wi-Fi carried by the bus to detect, so that the distance from the standard range is continuously moved, a large amount of interference data existing in the moving process of the bus cannot be cleaned, and meanwhile, the error is greatly increased by only using the Wi-Fi carried by the bus as the only judgment standard for getting on and off the bus.

The invention content is as follows:

the technical problem is as follows: in order to overcome the defects of the prior art, the invention aims to provide a passenger flow acquisition and analysis method which can be used for matching MAC address data detected by WiFi and Bluetooth detection equipment of a station with bus real-time GPS data, screening impurity data and effectively extracting passenger information.

The technical scheme is as follows: a bus passenger flow collection and analysis method based on station WiFi and vehicle-mounted GPS comprises the following steps:

(1) data acquisition: recording MAC address information, access time and exit time of a wireless network access point of a user mobile terminal in a scanning range through Wi-Fi equipment arranged on a bus stop; recording longitude and latitude and time information of the bus every second by using a GPS recorder;

(2) extraction and conversion: extracting the data information obtained in the step (1), and converting the time information to obtain a timestamp format;

(3) and (3) screening for duration: screening the data in the step (2) by using R software, wherein the specific operations are as follows:

(31) based on the MAC address data detected in the step (1), acquiring the MAC address information of the mobile equipment acquired by all bus stops in one-time traffic investigation (namely, one-way bus shift) after statistics by software R, and taking the access time of the MAC address as timei and the exit time as timef;

(32) obtaining the duration time of each MAC address data by using the MAC address data access time timei and the exit time timef obtained in the step (31)

(33) Determining a reasonable bus running time threshold according to field investigation, and determining the duration obtained in the step (32)

If the MAC address exceeds the threshold range, the duration of the MAC address is considered not to belong to the reasonable time for passengers to wait, the MAC address is considered as interference data, and the interference data is removed.

(4) Bus stop time determination: the method comprises the following steps of selecting a proper range around a used station by utilizing ArcGIS to set a station buffer area, and obtaining the arrival time and the departure time of a bus at each station along the line by combining the real-time bus position obtained by bus GPS positioning, wherein the method specifically comprises the following steps:

(41) projecting the real-time position acquired by the vehicle-mounted GPS obtained in the step (1) and the position of each bus station to the same coordinate system by utilizing ArcGIS;

(42) according to the maximum detection radius r of Wi-Fi detection equipment arranged at a bus stop_mAs radius threshold r for dividing the station buffer area;

(43) establishing an ArcGIS site buffer area by taking the site as the center of a circle and taking the threshold r as the radius on the basis of the step (42); screening the Time corresponding to the first GPS data point appearing in the buffer sideline or the buffer effective detection range along the bus driving direction as the arrival Time Time _1, and the Time corresponding to the last GPS data point appearing in the buffer sideline or the buffer effective detection range as the arrival Time Time _ 2;

(5) screening passengers getting on the train: combining the access time and the exit time of each MAC address data obtained in the step (3) with the arrival time and the exit time of the bus obtained in the step (4), and screening out MAC address data of which the access time of the MAC address data is before the arrival time of the bus (excluding the interference of passengers existing on the bus entering the bus) and the exit time of the MAC address data is close to the exit time of the bus (excluding the interference of other passengers waiting for the bus at the same station) as the signal of the passengers getting on the bus:

(6) screening of the passengers getting off: combining each MAC address data obtained in the step (3) with the bus arrival and departure time obtained in the step (4), screening out MAC address data which has MAC address signal access time close to the bus arrival time and is not detected at the next station, and regarding the MAC address data as a signal of a passenger getting off;

(7) and (3) OD generation: through the data processing operation, the data of each station are integrated, the station position where each passenger gets on or off the bus is obtained according to the uniqueness of the MAC address of each passenger, and on the basis, a passenger travel OD table is established by using R software;

(8) and (4) correcting the result: comparing the bus passenger flow OD table obtained in the step (7) with the passenger flow obtained by actual statistical investigation, and calculating a passenger flow error; and (4) analyzing the distribution of the passenger flow errors and carrying out hypothesis test, and correcting the number of passengers getting on or off the train at each station obtained in the step (5) and the step (6) by using a proper method.

The step (1) is based on Wi-Fi + Bluetooth detection equipment, the equipment adopts an active scanning mode, a scanner can detect wireless network access points of iPhone and Android equipment, and then MAC addresses and time data of smart phone users are obtained. The equipment performs active scanning every 1s, the scanning radius can reach 30 meters, and mobile terminals using Wi-Fi within the scanning radius range are detected and the MAC addresses of the mobile terminals are recorded.

The step (33) of obtaining the bus based on actual investigation and statisticsThe shortest time for stopping at the bus stop is determined, and the lower threshold t of the duration time of the MAC address detected by the Wi-Fi of the stop is determined according to the shortest time_dDetermining the upper limit t of the duration threshold value according to the maximum value of the arrival time interval of the buses of adjacent shifts_u. Duration less than t_dOr exceeds t_uI.e., considered not to be a passenger, left.

The screening process of the passengers getting on the bus in the step (5) is based on the behavior characteristics of the passengers getting on the bus: passengers getting on the bus arrive at the station before the bus arrives at the station, and get off the station by taking the bus;

the screening process of the get-off passengers in the step (6) is based on the behavior characteristics of the get-off passengers: a passenger getting off enters the station by taking the bus, leaves the station after getting off and can not arrive at the next station of the bus line synchronously with the bus;

in the step (7), the MAC address information only appearing at one station is removed, and the MAC address information does not accord with the behavior characteristics of the passenger, so that the MAC address information does not belong to the passenger data.

In the step (8), the actual recorded data refers to the actual number of people getting on or off the train, and is recorded in the field in the data acquisition process. The method comprises the following steps of determining the Wi-Fi or Bluetooth opening ratio K of a passenger in a questionnaire mode, and finding out that: the value of K remains relatively constant over a certain area.

In the step (8), a data processing error is calculated according to a comparison result between the actual data processing and the data processing. And (4) the influence of the data threshold is comprehensively considered by combining the opening ratio K, and the MAC address duration threshold adopted by the passenger at the bus stop is properly adjusted, so that the data processing result is matched with the actual data as far as possible.

Advantageous effects

The invention collects data by dividing the range of the bus station with fixed coordinates, thereby greatly eliminating the interference to the result caused by the wireless equipment moving along with the bus station in the bus running process; the impurity data are screened out by combining a space-time double-layer filtering mode, the obtained result is more practical, the O-D estimation precision of the bus passenger flow is improved, the bus system dispatching is carried out for traffic control departments, and reliable data support is provided for improving the operation efficiency and the service level of the bus system.

Drawings

FIG. 1 is a schematic illustration of the internal operation and interrelationship of the various systems of the present invention;

FIG. 2 is a flow chart of a passenger flow data collection and analysis process of the present invention;

FIG. 3 is a diagram of a method for determining ArcGIS buffer threshold according to the present invention;

FIG. 4 is a diagram illustrating a MAC address data scan time list according to the present invention;

FIG. 5 is a schematic diagram of the inbound and outbound timing determination screening of the present invention

Detailed Description

The invention will be further described with reference to an embodiment shown in the drawings

The invention provides a passenger flow OD (getting-on station and getting-off station) data acquisition and analysis method based on station Wi-Fi hotspot detection and bus GPS positioning. The data acquisition method is based on Wi-Fi + Bluetooth detection equipment fixedly arranged on a station, can detect the connection and disconnection of wireless equipment, and further obtains the MAC address of the passenger mobile terminal and corresponding connection and disconnection time. The bus real-time position and arrival time information recorded by a GPS recorder are combined, the MAC address is matched by taking the time meeting the bus arrival and departure time as a standard, the O-D acquisition and analysis of bus line passenger flow are realized, and FIG. 1 is a schematic diagram of the internal operation and the mutual relation of each system of the invention.

Firstly, when a user who has wireless equipment and opens a corresponding function of accessing an external wireless signal appears in a Wi-Fi + Bluetooth detection coverage range of a bus stop, a detector can record MAC address information, access time and exit time of a wireless network access point of a user mobile terminal in a scanning range; the method comprises the steps of obtaining the appearance time, disappearance time and duration of each MAC address signal of all bus stops along the bus by utilizing the conversion of a time format and a timestamp format and the processing of R, screening the obtained MAC signals and corresponding information thereof according to a time threshold divided by actual investigation, and reserving MAC address information with the duration meeting the requirements as a primary data processing result; when the screened MAC address time information is consistent with the time positioned by the bus GPS, judging whether the MAC address time information is used as a passenger getting off or a passenger getting on according to the appearance or disappearance of the current station; and obtaining the O-D distribution of each passenger in the traffic investigation according to the uniqueness of the MAC address of the wireless device carried by the passenger, obtaining the number of passengers getting on or off the bus at all stations in the traffic investigation by taking the station as a unit so as to obtain O-D information, comparing with the actual situation and modifying related parameters, and realizing the O-D statistics of the bus passenger flow at one time.

In this example, the O-D acquisition and analysis method (fig. 2 is a flow chart thereof) using the station Wi-Fi detection and the bus GPS data fused with time as a standard mainly includes the following six stages:

a primary collection stage: recording MAC address information, access time and exit time of a wireless network access point of a user mobile terminal in a scanning range through Wi-Fi equipment arranged on a bus stop (a diagram of a MAC address data scanning time list is shown in figure 4), and summarizing the obtained data in a data table named BTtime; and recording the longitude and latitude and time information of the bus once every second by using a GPS recorder, and storing the information in a GPStime table for later use.

Duration screening stage: preprocessing BTtime and GPStime by a conversion formula of a timestamp to convert the BTtime and the GPStime into a timestamp format, counting mobile equipment MAC address information acquired by each bus stop in one-time traffic survey (namely, one-way shift of a bus) by using R software as a unit to obtain access time timei and exit time timef of the MAC address, and thus obtaining duration time delta s which is timef-timei; and deleting the MAC address with the duration not within the range of the threshold preliminarily determined according to the prior investigation (namely, the MAC address does not meet the time that the passenger should meet the condition of waiting for the bus to get on at the bus stop) in the statistics of the fixed stop, and considering the MAC address as the interference data.

A station stopping time determining stage: ArcGIS is utilized to select proper ranges around all stations along the bus to set station buffer areas, and simultaneously, the data table GPStime of bus GPS positioning is combined to project all longitude and latitude of GPS data to the same mapAnd obtaining all recording points of the bus on the whole bus route. The maximum scanning length r of the Wi-Fi equipment is taken as the center of a circle_mFor the radius, a site buffer is established. Screening the Time corresponding to the first GPS data point appearing in the buffer sideline or the buffer effective detection range along the bus driving direction as the arrival Time Time _1, and the Time corresponding to the last GPS data point appearing in the buffer sideline or the buffer effective detection range as the arrival Time Time _ 2; fig. 3 is a diagram of a method for determining an ArcGIS buffer threshold according to the present invention, and fig. 5 is a diagram illustrating a method for determining and screening inbound and outbound timings according to the present invention.

And (3) passenger getting-on and getting-off screening stage: combining each MAC address data access and exit Time timeei and timeef with the bus arrival and exit Time Time _1 and Time _2 obtained in the step (4), and screening out MAC address data of which the MAC address data access Time is before the bus arrival Time (excluding the interference of passengers existing on the bus that arrives at the bus) and the MAC address data exit Time is close to the bus exit Time (excluding the interference of passengers on the same bus) as a passenger signal on the bus; and similarly, MAC address data which is similar to the time of accessing the bus at the time of accessing the MAC address signal and is not detected at the next station is screened out and is regarded as the signal of the passenger getting off the bus.

And (3) OD generation stage: the data of each station are integrated through the data processing operation, the station position where each passenger gets on or off the bus is obtained according to the uniqueness of the MAC address of each passenger, the information of the number of the stations where each passenger gets on or off the bus is obtained by utilizing R software on the basis, the information is respectively represented by Location _1 and Location _2, the repeatability inspection is carried out, and the passenger data of the stations where the MAC addresses appearing and disappearing in the Location _1 and the Location _2 are the same are deleted; and after the repeatability inspection, respectively screening the station Location _1 and the station Location _2 of each passenger to obtain the theoretical number of getting-on and getting-off passengers at each station of the bus line in the traffic survey, and generating the passenger flow O-D of the bus line.

And a data correction stage: comparing the obtained bus passenger flow OD table with the passenger flow obtained by actual statistical investigation, and calculating a passenger flow error; analyzing and hypothesis testing the distribution of passenger flow errors, correcting the obtained number of passengers getting on or off the bus at each station by using a proper method, and comprehensively considering the influence of a data threshold value based on the proportion of the number of the passengers actually using Wi-Fi in the bus, which is obtained by field questionnaire statistics, to the total number of the passengers, so as to dynamically adjust the threshold value and the duration threshold value of the buffer area and reduce the errors as much as possible;

in the preliminary acquisition stage, the specific method for acquiring the bus GPS data is as follows:

the GPS recorder records the data of the bus every 1 second, including the current date and time (accurate to the second), the longitude and latitude of the bus, and the like. The GPS recorder can export the data in a conventional file format (such as CSV format) after the acquisition is finished, and the data format is shown in Table 1

TABLE 1

Wherein, the current time, sat-lon and sat-lat are the key data needed by the invention. sat-lon and sat-lat are longitude and latitude of the bus under a WGS84 coordinate system, and recorded-time is real-time Beijing time recorded by a GPS recorder, and the real-time Beijing time needs to be converted into a time stamp format when data is processed.

Separately representing different MAC addresses detected by Wi-Fi as IDs by using labels, integrating the IDs with the original MAC addresses, and using an R statistical integration output data table as the premise of the subsequent processing steps, as shown in Table 2

TABLE 2

	MAC	BTtime	ID
					1	00:01:7A:6C:83:82	1.48E+09	1180
2	00:01:7A:89:A3:E2	1.48E+09	1169
				3	00:03:0F:39:C2:00	1.48E+09	332
4	00:03:0F:39:C2:00	1.48E+09	332
				5	00:03:0F:42:A5:20	1.48E+09	301

MAC address data corresponding to repeated IDs are removed from R software, the MAC addresses are ensured to be uniquely corresponding to the IDs, and a table 3 is obtained

TABLE 3

The R software is utilized to count each bus stop along the way as an independent statistical unit to obtain the bus stopThe appearing time timei and the disappearing time timef of each MAC address at the station are subtracted to obtain the duration

Determining the duration

Whether within a set time threshold. If not, the user is determined not to be the passenger of the bus, and the data is directly deleted to obtain a table 4

TABLE 4

And matching the MAC address with the GPS data according to the principle that the time when the passenger gets on or off the bus is matched with the time when the bus is out of the station and enters the station. By using the existing appearing Time timei and disappearing Time timef of all MAC addresses of each station and the obtained Time _1 and Time _2, when the timei is less than the Time1 and the timef is approximately equal to the Time2, the corresponding MAC address is considered as the data of the boarding passenger; and a wireless signal in which timef ≈ Time2 and the occurrence of the MAC address is not detected at the next station is processed as the alighting passenger at this station.

And projecting the GPS information (Lat and Lon) of the bus in the whole running process and the position information of the station to the same coordinate system by utilizing ArcGIS, and further screening data according to the relative position of the MAC address and the station. According to the maximum detection radius r of Wi-Fi detection equipment arranged at a bus stop_mThe method comprises the steps of determining a threshold value to establish a buffer area as a radius threshold value r for dividing a bus station buffer area, screening out a GPS signal point which is within the range of the bus station buffer area and is closest to the boundary of the buffer area, outputting corresponding position and Time information of the GPS signal point to be respectively used as an arrival Time Time _1 and an arrival Time Time _2, and independently selecting and counting each station.

And performing repeated inspection on the position information of the appearance and disappearance of the signals near the station obtained in the step, eliminating the MAC address data appearing and disappearing at the same station, and obtaining the getting-on station and the getting-off station of the user by using the station label information Location _1 and Location _2 corresponding to each signal after the repeated inspection to obtain the OD distribution to generate the bus line passenger flow O-D.

Claims

1. A bus passenger flow collection and analysis method based on station WiFi and vehicle-mounted GPS is characterized in that: the method comprises the following steps:

(1) primary collection: recording MAC address information, access time and exit time of a wireless network access point of a user mobile terminal in a scanning range through Wi-Fi equipment arranged on a bus stop; recording real-time position information and time information of the bus by using a vehicle-mounted GPS recorder;

(2) extraction and conversion: extracting the information obtained in the step (1), and converting the time information to obtain a timestamp format;

(3) and (3) screening for duration: counting Wi-Fi detection data of each station, and calculating MAC signal duration according to MAC information access and exit moments detected by Wi-Fi; setting a reasonable duration threshold value, and rejecting MAC information of which the duration is not in the threshold value range;

and (3) running related codes by using an R language, representing the access time of each MAC address as timei and the exit time as timef after conversion in the step (2) on the basis of collecting wireless network detection information of all stations, and calculating to obtain the duration

According to the on-site traffic investigation and a large amount of actual statistics, a reasonable bus passenger waiting time threshold value [ t ] is determined_d,t_u]When the duration is

If the threshold value range is exceeded, the waiting time is not considered to be reasonable waiting time, and the MAC address isThe address belongs to an invalid address and is removed as interference data; determining the lower limit t of the duration threshold value according to the shortest stopping time of the bus at the bus stop obtained by the passenger flow survey of the actual bus stop and the statistics of the actual stopping time of the bus at the stop_dDetermining the upper limit t of the time duration threshold value of the residence at the bus stop according to the maximum value of the arrival time interval of adjacent shifts of the whole bus line_u(ii) a The duration is determined according to the shortest stopping time of the bus at the station in actual investigation, and the lowest requirement is 7 s;

(4) bus stop time determination: according to the maximum detection radius of the used Wi-Fi equipment, a circular station buffer area is set by taking a station as a circle center, and the time of arrival and departure and the time of stop of a bus are determined according to the real-time GPS positioning of the bus and the relative position relation of the station buffer area;

(5) screening passengers getting on the train: combining the access time and the exit time of each MAC signal obtained in the step (3) with the bus arrival and exit time obtained in the step (4), screening out MAC address data of which the access time of each MAC signal is before the bus arrival time and the exit time of each MAC signal is close to the bus exit time as a boarding passenger signal, distinguishing different passengers according to the MAC addresses, counting, and determining the departure time of the boarding passenger at the bus stop according to the exit time in the step (4);

(6) screening of the passengers getting off: combining each MAC address data obtained in the step (3) with the bus arrival and departure time obtained in the step (4), screening out MAC address data which has signal access time close to the bus arrival time and is not detected at the next station, regarding the MAC address data as a signal of a passenger getting off the bus, distinguishing different passengers according to MAC, counting, and determining the arrival time of the passenger getting off the bus at the station according to the arrival time in the step (4);

(7) and (3) OD generation: integrating the information of passengers getting on and off the bus at all stations obtained in the step (5) and the step (6), matching by using unique corresponding MAC addresses to obtain the stations of getting on and off the bus of each MAC signal, further deducing the O-D distribution condition of the passenger flow of the bus passengers in one-way shift, distinguishing different passengers by using the MAC addresses as the basis for matching, summarizing the passenger information of the same MAC address, screening out the MAC addresses detected at only one station, and counting to obtain an O-D table consisting of all effective MAC addresses;

(8) and (3) data correction: comparing the bus passenger flow OD table obtained in the step (7) with the passenger flow obtained by actual investigation, and calculating a passenger flow error; and (3) analyzing the distribution of passenger flow errors and carrying out hypothesis testing, correcting the getting-on and getting-off passenger screening method in the step (5) and the step (6) by using a fitting optimization method, and dynamically adjusting the MAC duration time threshold value by comprehensively considering the influence of the data threshold value based on the proportion of the number of the bus passengers actually using Wi-Fi or Bluetooth to the total number obtained by field questionnaire statistics.

2. The bus passenger flow collection and analysis method of claim 1, wherein: the method comprises the following steps that (1) based on Wi-Fi detection equipment taking RFID and Bluetooth technologies as cores, the MAC address of a user using the Wi-Fi equipment is obtained through a transmission path of wireless signal radio frequency while a station transmits an effective wireless network signal, and the access time and the exit time are obtained; the vehicle-mounted GPS equipment acquires GPS data of the bus in the running process in real time, and the acquisition frequency of the GPS equipment is 1 s/time.

3. The bus passenger flow collection and analysis method of claim 1, wherein: the step (2) facilitates subsequent data processing by converting the time format into a standard Unix time stamp format; the extracted information includes latitude and longitude, time data and MAC address.

4. The bus passenger flow collection and analysis method of claim 1, wherein: the step (4) comprises the following steps:

(41) projecting the position of the selected bus stop on a map through longitude and latitude coordinates;

(42) projecting and displaying all the GPS data of the bus route running process obtained in the step (1) on the same interface in the step (41);

(43) using ArcGIS in the map information in the step (42) and taking the bus stop as a circle center and the maximum Wi-Fi detection length as a radius, and selecting a reasonable circular area around the bus stop as a bus stop buffer area; the time corresponding to the first GPS data point appearing in the buffer sideline or the buffer effective detection range along the bus running direction is the arrival time, and the time corresponding to the last GPS data point appearing in the buffer sideline or the buffer effective detection range is the departure time.