WO2015167169A1 - Method for estimating passenger moving path in subway network - Google Patents

Method for estimating passenger moving path in subway network Download PDF

Info

Publication number
WO2015167169A1
WO2015167169A1 PCT/KR2015/004069 KR2015004069W WO2015167169A1 WO 2015167169 A1 WO2015167169 A1 WO 2015167169A1 KR 2015004069 W KR2015004069 W KR 2015004069W WO 2015167169 A1 WO2015167169 A1 WO 2015167169A1
Authority
WO
WIPO (PCT)
Prior art keywords
passenger
station
passengers
vehicle
group
Prior art date
Application number
PCT/KR2015/004069
Other languages
French (fr)
Korean (ko)
Inventor
홍성필
김경민
고석준
민윤홍
Original Assignee
서울대학교 산학협력단
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 서울대학교 산학협력단 filed Critical 서울대학교 산학협력단
Publication of WO2015167169A1 publication Critical patent/WO2015167169A1/en

Links

Images

Classifications

    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16ZINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
    • G16Z99/00Subject matter not provided for in other main groups of this subclass

Definitions

  • the present invention generally relates to techniques for estimating passenger travel routes within a train network. More specifically, the present invention relates to a technique for estimating a travel route from a passenger's boarding station to an unloading station from smart traffic card tag information for payment of a passenger.
  • the passenger tags the transportation card to the terminal when passing through the gate at the boarding station and when passing through the gate at the getting off station. Therefore, in the case of the train network, there is a problem in that it is not possible to know which vehicle the passenger boarded, which station it was transferred to, and the like.
  • this logit model has a problem that the number of samples used for model development is very small, less than 1%. Seoul's train network, for example, has about 400 train stations, about 4,500 vehicles a day, and more than 5 million passengers per day (i.e.
  • the accuracy of the route estimation of the logit model is reported to be 70-80%, and the allocation of charges by the route estimation with this accuracy can cause disputes among train operators.
  • an object of the present invention is to provide a method for estimating a passenger travel path including a vehicle on which the passenger as well as the train route traveled by the passenger from the passenger's traffic card tag information.
  • the present invention extracts some passengers whose travel path is simple and reliable, and uses the extracted passenger information as a criterion for estimating more complicated movement paths of other passengers, thereby making the most of the passengers traveling paths. Its purpose is to provide a method for accurately estimating.
  • the above objects are provided by a method for estimating a passenger route in a train network provided according to the present invention.
  • a method for estimating a passenger's movement route in a train network which is a method of estimating a passenger's movement route in a train network in which a plurality of vehicles operate along a plurality of routes, Generating a reference group by estimating a train in which each passenger boards from the passenger data including; Generating at least one possible physical route made by the train line information between each passenger's boarding and unloading station from each passenger's getting on and off information; Generating at least one logical path for each of the possible physical paths, based on the reference group, by which a vehicle running on each train line is connected; And checking the consistency of the logical path based on the reference group to determine a passenger's movement path.
  • the generating of the reference group may include: generating a reference riding group by estimating passengers boarding a specific first vehicle at a specific boarding station; Estimating passengers who get off from a specific second vehicle at a specific getting off station and generating a reference dropping group; And generating a reference transfer group by estimating passengers transferred from the specific third vehicle to the specific fourth vehicle at the specific transfer station.
  • the step of generating the reference drop-off group may include: extracting, from the passenger data, a passenger having one possible physical path and no transfer; Selecting, for the extracted passengers, passengers including only one of the same vehicle by comparing vehicles departing after the gate entry time of each boarding station with vehicles arriving before the gate entry time of each boarding station; And classifying passengers as having boarded the vehicle including only one, and classifying passengers having a specific discharging station among the selected passengers as passengers getting off at the specific discharging station from the estimated vehicle And generating as a reference drop group, respectively.
  • the step of creating a reference ride group includes: extracting passengers having one possible physical path and no transfer from the passenger data; Selecting, for the extracted passengers, passengers including only one of the same vehicle by comparing vehicles departing after the gate entry time of each boarding station with vehicles arriving before the gate entry time of each boarding station; And classify passengers who have boarded the vehicle including only one, and classify passengers having a specific boarding station among the selected passengers as passengers who boarded at the specific boarding station to the vehicle estimated to be boarded. And generating each as a reference ride group.
  • the step of creating a reference transit group may include: extracting, from the passenger data, a passenger having one possible physical path and having one transit station; For each of the extracted passengers, a corresponding reference riding group having a reference entry time interval including a gate entry time of each boarding station is extracted from the reference riding groups, and the boarding station is assigned to the boarding vehicle of the corresponding reference riding group. Estimating that the passenger rides in the vehicle; For each of the extracted passengers, a corresponding reference disembarkation group having a reference entry time period including a gate entry time of each discharging station is extracted from the reference discharging groups, and the disembarkation station is displayed on the getting-off vehicle of the corresponding reference discharging group.
  • Estimating to be off For each of the extracted passengers, checking whether the estimated riding vehicle arrived at the transit station and whether the estimated getting off vehicle departed from the transit station; And when it is determined that transfer is possible in the checking of the transfer, as a reference transfer group transferred from the estimated riding vehicle to the estimated getting off vehicle at the transfer station.
  • the step of generating the logical path may include: determining only one vehicle to get off at the discharging station for each passenger.
  • the generating of the logical path may include: determining one or two vehicles boarding at the boarding station for each passenger.
  • the generating of the logical path may include: determining whether one or two vehicles are transferred at the transfer station when there is a transfer in the physical path.
  • a computer readable storage medium comprising computer executable instructions coded to perform a particular task when executed by a processor of a computing device, wherein the computer executable instructions are executed when executed by the processor.
  • a method of estimating a travel route of a passenger in a train network in which a plurality of vehicles are operated is performed, and the method comprises: estimating a train on which each passenger boards from passenger data including getting on and off information of a plurality of passengers.
  • Creating a reference group Generating at least one possible physical route made by the train line information between each passenger's boarding and unloading station from each passenger's getting on and off information; Generating at least one logical path for each of the possible physical paths, based on the reference group, by which a vehicle running on each train line is connected; And checking the consistency of the logical paths based on the reference group to determine a travel path of a passenger.
  • the passenger data includes getting on and off information of passengers accumulated for a certain period of time in the train network
  • the getting on and off information includes at least a boarding station on which the passenger boards, a gate entry time of the boarding station, and the passengers get off It may include the getting off station, and the gate advance time of the getting off station.
  • the passenger movement path in the train network can be accurately estimated based on the traffic card tag information for the passenger's transportation fee payment. Accordingly, since a separate sampling operation is unnecessary in addition to the traffic card tag information, the cost and time required for estimating the passenger travel path may be reduced. In addition, passenger travel routes can be estimated with greater than about 95% accuracy, thereby eliminating disputes related to fare settlement between train operators.
  • a passenger movement path including a vehicle on which the passenger as well as a train route traveled by the passenger can be estimated from the passenger's transportation card tag information. That is, for each passenger, a movement route including a vehicle boarding at the boarding station, a transit station, and a vehicle getting off at the getting off station may be estimated. Therefore, it is possible to know how many people have boarded each vehicle and to know who has moved at each transit station, thereby providing useful data for traffic usage analysis.
  • the present invention extracts some passengers whose passengers have simple and reliable movement paths, and uses the extracted passengers' information as a criterion for estimating more complicated movement paths of other passengers. Accordingly, it is possible to estimate the movement route based on the information of the actual passengers.
  • FIGS. 1 and 2 are schematic diagrams illustrating portions of a train network to which the present invention may be applied.
  • FIG. 3 is a flow chart showing the steps of a method for estimating a passenger travel path in a train network according to an embodiment of the present invention.
  • FIG. 4 is a flowchart illustrating a process of creating a reference ride group in a method for estimating a passenger movement path of a train network according to an embodiment of the present invention.
  • FIG. 5 is a flowchart illustrating a process of creating a reference dropoff group in a method for estimating a passenger movement path of a train network according to an embodiment of the present invention.
  • FIG. 6 is a flowchart illustrating a process of creating a reference transit group in a method for estimating a passenger's movement path of a train network according to an embodiment of the present invention.
  • FIG. 7 is a schematic diagram illustrating a process of generating a possible logical path in a movement route including one transfer in a method for estimating a passenger movement route of a train network according to an embodiment of the present invention.
  • FIG. 8 is a schematic diagram for explaining a process of generating a possible logical path in a movement route including two transfers in a method for estimating a passenger movement route of a train network according to an embodiment of the present invention
  • FIG. 9 is a view illustrating a process of determining a travel route for two passengers each having two possible physical routes including one transfer in the method for estimating a passenger travel route of a train network according to an embodiment of the present invention.
  • FIG. 1 a schematic diagram illustrating a portion of a train network to which the present invention may be applied is shown.
  • the train network shown is a very brief overview of some of the train networks south of the Han River in Seoul.
  • most passengers pay for transportation using smart cards. That is, the passenger tags the traffic card to the terminal for payment provided at the gate while passing through the gate of the boarding station, and similarly tags the traffic card while passing through the gate of the boarding station, thereby depending on the distance between the boarding station and the getting off station. You can pay the fee.
  • the getting on and off gate tag information of the transportation card for the fare payment may provide information of getting on and off the passenger related to the fare payment.
  • the getting on and off information may include a boarding station, a boarding station gate passing time, a getting off station, and a getting off station gate passing time.
  • the time between getting on and getting off can be regarded as the time spent by a passenger for movement within the train network, also referred to as "inter-gate time" below.
  • inter-gate time typically the estimation of the passenger travel path of the train network may be made using the inter-gate time.
  • the inter-gate time is different from the time the passenger actually boarded the train. This is because the inter-gate time further includes a travel time from the boarding gate to the boarding platform, a vehicle waiting time at the boarding platform, a travel time from the boarding platform to the boarding gate, and the like, in addition to the vehicle boarding time.
  • the transit route includes a transit station
  • the inter-gate time should further include the transit time from one platform to the other platform for transit and the vehicle waiting time on the platform after the transit.
  • the inter-gate time may further include time for passengers to use amenities (eg, toilets, convenience stores, shopping malls, restaurants, etc.) located inside the gate.
  • amenities eg, toilets, convenience stores, shopping malls, restaurants, etc.
  • the passenger's route can be estimated only by inter-gate time. For example, if the station and the station are located on one line, there is only one transfer station between the station and the station, and there is a significant difference in the number of stations or the number of stations between the possible physical routes of the passenger. The travel path of the passenger can be estimated relatively easily.
  • FIG. 2 a schematic diagram illustrating a portion of a train network to which the present invention may be applied is shown.
  • the passenger travel path estimation is described in terms of the route
  • FIG. 2 the passenger travel path estimation is described in terms of the vehicle.
  • the illustrated example illustrates a travel path of a passenger that can know the boarding station (BS) M2, the boarding gate entry time Bt, the boarding station (AS) M6, and the boarding gate entrance time At. It can be estimated that this passenger can travel via route R1 transferred at M5 and route R2 transferred at M4 based on the inter-gate time.
  • the former is referred to as the 'physical path' and the latter is referred to as the 'logical path' in order to distinguish the movement path including the vehicle and the movement path estimated by the route only.
  • the physical path means a movement path including elements such as a boarding station, a getting off station, a line, and a transfer station.
  • Logical paths refer to a travel path that further includes, as elements, the particular vehicle on which the passenger boards.
  • the method according to the invention is a method of estimating a passenger movement route of a train network, and in particular, makes it possible to estimate a logical route specifying a passenger vehicle.
  • the method uses the passenger traffic card tag information, the driving log information of the train vehicle, and the topology of the train network (i.e., route map) as the given information.
  • Train network topology or train topology is line 1, line 2,.... It includes a pattern of lines, and defines the train network environment including the train station, the train station, and the train station.
  • the passenger's route of travel is therefore defined as a line in which the metro stations are connected to each other, as illustrated in FIGS. 1 and 2.
  • the driving log information of the train vehicle is data having time information when the specific vehicle arrives and / or departs at the specific subway station.
  • the passenger traffic card tag information is also referred to as passenger's getting on and off information in this document, and includes four pieces of information of a boarding station (BS), a gate entry (Bt), a drop off station (AS), and a gate entry (At). Can be.
  • the boarding station BS and gate entry Bt inform the position and time of the starting point of the passenger travel path.
  • the stop station (AS) and gate entry (At) inform the position and time of the arrival point in the passenger travel path.
  • the term 'quadruple' is often used below.
  • the method provided according to the present invention utilizes, in addition to quadrople information and vehicle driving log information, several presuppositions based on experience or observed facts.
  • the inventors observed from the passengers' exit gate entrance data that the departure points have a constant repetitive pattern.
  • the gate entry pattern is a section in which passengers advance for a predetermined time after detecting the fastest gate entry after a specific vehicle arrives at the platform, and then enter the gate for a predetermined time interval before the next vehicle arrives at the platform. This interrupted section is repeated in the following manner.
  • This entry pattern is consistent with the general premise that most passengers who get off a particular vehicle will have the purpose of entering the gate as quickly as possible.
  • the inventors have found that knowing the gate entry time At of a particular passenger, it is possible to estimate relatively accurately which of those vehicles the passenger has arrived at the platform. With this discovery, a group of passengers who boarded a particular vehicle at various stations and then got off at a particular station could be extracted. It is one of the most important assumptions in the method of the present invention that it is possible to accurately estimate the vehicle a passenger got off based on the passenger's gate entry time.
  • the method of the present invention considering the fact that almost half of the passengers are relatively simple and simple, proposes a technique for estimating the travel paths of passengers whose travel paths are uncertain, using passengers having a certain travel path. do.
  • a reference group is a set of passengers that provides criteria for estimating the travel paths for passengers with complex travel paths.
  • a reference group includes a group of people who ride a specific vehicle at a specific station (reference getting off group), a group of people who ride a specific vehicle at a specific station (reference riding group), and a group of people who ride a specific vehicle and transfer to another specific vehicle at a specific station. (Reference transit group) may be included.
  • a possible vehicle may be extracted into a reference group (AG) by extracting passengers who have a specific vehicle and get off at a specific station.
  • AG reference group
  • BG reference boarding group
  • the reference riding group and the reference getting off group each include two parameters, namely, the vehicle X which got on / off and the boarding / unloading station N.
  • the drop-off gate exit times At of the passengers in the reference drop-off group may correspond to the reference departure time intervals At1 to At2 having an arbitrary length.
  • This reference entry time period may be defined based on data of passengers in the reference dropoff group.
  • the reference departure time period may be defined as the earliest time value to the slowest time value among gate entry times of the passengers in the reference descent group.
  • a passenger's drop-off gate exit time (at) is located between the reference departure time intervals At1-At2 (At1 ⁇ at ⁇ At2), this passenger can be assumed to be one of the passengers in the reference dropoff group, It can be assumed that the drop-off group got off the vehicle that got off.
  • the boarding station gate entry time Bt of the passengers in the reference riding group may correspond to the reference entry time interval Bt1 to Bt2 having any length.
  • This reference entry time interval may be defined based on data of passengers in the reference riding group.
  • the reference entry time period may be defined as the earliest time value to the slowest time value among gate entry times of the passengers in the reference riding group.
  • a passenger's boarding station gate entry time (bt) is located between the reference entry time intervals Bt1 to Bt2 (Bt1 ⁇ bt ⁇ Bt2), this passenger may be assumed to be one of the passengers in the reference riding group, It can be assumed that the riding group has boarded the vehicle.
  • a simple transfer route may be expressed as including a boarding station (O)-a transfer station (A)-a drop off station (D).
  • a particular passenger at the boarding station O can board any one of two consecutive vehicles X1 and X2 arriving for the first time based on the gate entry time, and the drop-off station D In FIG. 3, it may be estimated that a passenger gets off from one vehicle Y based on a gate advance time.
  • a transfer group that boards X1 at O station gets off at Y station, gets on Y vehicle, and gets off at Y station at D station
  • the transit groups X2, Y and A can be distinguished from the X2, Y, A) and O stations, from the A station to the Y vehicle, and from the D station.
  • it may be checked whether the X1, X2, Y vehicle is physically connectable according to information such as the time when the vehicle arrives at the A station. If the arrival time of vehicle X1 at station A is later than the departure time of vehicle Y, the passenger will not be able to physically transfer from vehicle X1 to vehicle Y.
  • the transit groups combine information of passengers of a specific vehicle X1 departing from a plurality of O stations and a Y vehicle arriving at a plurality of D stations, from the X1 vehicle to the Y vehicle at a specific transit station A.
  • Transit passengers can be bundled and created as a reference transfer group (X1, Y, A).
  • the reference transfer group includes three parameters: the boarded vehicle X, the boarded vehicle Y, and the transfer station A.
  • the logical route can be estimated in a similar manner. For example, two ride vehicles may be estimated at a boarding station, one or two transfer cars may be estimated at a first transfer station, and one vehicle may be estimated at a second transfer station.
  • one getting off vehicle Y can be estimated according to the getting off station and the gate entry time At for one physical route, and the starting station and the gate entry time According to Bt, two riding vehicles X1 and X2 can be estimated. Accordingly, at least two different logical paths may be created for one physical path.
  • one possible transit station is included in the possible physical path, for each of the possible logical paths, it may be checked whether a reference transit group exists. If there is a logical path in which the reference transit group exists, this logical path may be determined as the passenger's travel path. If no reference transit group exists in any logical path, the passenger's route is not determined.
  • the riding vehicle X of the reference transfer group may be determined based on the reference riding group.
  • the getting off vehicle Y of the reference transfer group may be determined based on the reference getting off group. And when the arrival time of the boarding vehicle X which arrived at the transit station T was earlier than the departure time of the disembarkation vehicle Y which departed from the transit station T, the transfer from the boarding vehicle X to the getting-off vehicle Y is carried out. It can be determined that this is possible.
  • FIG. 3 there is shown a flowchart showing the steps of a method for estimating a passenger's movement route in a train network according to an embodiment of the present invention.
  • the method according to the invention is a method 300 for estimating a passenger's movement path of a train network, and in particular, makes it possible to estimate a logical path specifying a passenger's vehicle.
  • the method 300 first generates reference groups 301, generates possible physical paths 303, then generates possible logical paths 305, and finally determines a passenger's movement path 307. It can proceed to.
  • the step of creating reference groups 301 is a step of creating a reference group by specifying trains that get on or off for passengers whose movement paths are simple and reliable. More specific embodiments of this step 301 are shown in FIGS. 4, 5, 6.
  • Passenger data includes boarding and getting information generated by tagging when the traffic card of passengers using the train network passes through the gates of the boarding station and the unloading station.
  • a boarding station BS For each extracted passenger, a boarding station BS, a boarding station gate entry time Bt, a boarding station AS, and a boarding station gate entry time At are given.
  • trains departed after Bt at the boarding station may be bundled to generate a departure train group P (403, 503).
  • the departure train group P can be said to be a set of trains with the possibility of passengers to board, but only considering that passengers can board the train after entering the gate.
  • arrival train groups Q may be generated by tying trains arriving before At at the drop off station (405 and 505).
  • Arrival train group (Q) can be said to be a set of trains that passengers may have boarded, but only consider that passengers can get off the train before entering the gate.
  • the passenger can only board one of the trains included in the departure train group P and the arrival train group Q. In other words, if there is only one train involved in the intersection of P and Q, the passenger can safely assume that he was on this train. Therefore, among the trains included in P and Q generated for each passenger, passengers having one identical train X may be selected (407 and 507). Passengers thus selected can be assumed to have boarded this particular train (X).
  • the reference boarding group BG (X, O) may be generated (409), and similarly, only passengers having information getting off at a specific station (D) may be bundled and generated as the reference boarding group AG (X, D) (509). .
  • a process 600 of creating a reference transit group is shown.
  • passengers via the specific transit station T are extracted (601).
  • the riding vehicle X of which a particular passenger has boarded at the boarding station may be determined based on the reference riding group found for this passenger (603).
  • the disembarkation vehicle Y that the passenger has disembarked at the disembarkation station may be determined 605 based on the reference disembarkation group found for this passenger.
  • whether or not the passenger can transit can be determined 607.
  • the passenger gets off the vehicle Y from the boarding vehicle X. It may be determined that transfer to is possible.
  • passengers transitable from a specific riding vehicle X to a specific getting off vehicle Y are bundled and generated as a reference transfer group TG (X, Y, T) (609).
  • a possible physical path may be generated from the vehicle topology, that is, the train network topology, based on the passenger's getting on and off information (303).
  • Each physical route includes a boarding station, a transit station (if present), and a disembarkation station.
  • the physical path may be selected when there are no transfers, one transfer, two transfers, etc. among the possible routes by the inter-gate time based on the passenger's getting on and off information. Other possible cases are exceptional and are outside the scope of the present invention.
  • possible physical paths for each passenger may be generated in one or two or more numbers.
  • the physical path generation process 303 will be described in more detail as follows.
  • each physical path will have a finite number of stations, since each physical path is in the form of a line connected to each other.
  • PR1 contains 10 stations
  • PR2 contains 12 stations
  • PR3 contains 40 stations, and so on. etc.
  • PR1 and PR2 have two differences
  • PR3 has 30 and 28 differences for PR1 and PR2, respectively. Therefore, PR3 has more than 8 tolerances than PR1, which is the smallest inverse, so it can be excluded from the possible physical paths.
  • one transit station may be regarded as having the same value as a plurality of general stations (for example, five general stations).
  • a possible physical route for passengers who boarded at Sadang Station (transfer stations of M3, Line 2 and 4 in Fig. 1) and got off at the interchange (transfer stations of M4, Line 2 and 3 in Fig. 1) was created.
  • a physical path may be generated by separating a path having a transfer number of 0, a path having a transfer number of 1, and a path having a transfer number of 2 times.
  • a route having two transfers may combine two transfer stations between Line 4 and Line 3, so that a plurality of routes PR4, PR5, PR6, and PR7 may be generated.
  • PR4 may be the second transfer route from Line 4 to Line 7 and then transfer again from Line 6 to Line 3. Since there are four general stations and two transfer stations, it has a total of 14 stations.
  • PR5 via Line 9, PR6 via Line 6, and PR7 via Line 1 can have 16, 26, and 35 reciprocals, respectively.
  • the logical path is a path including a passenger vehicle on which the passenger boards and a disembarked vehicle on which the passenger gets off.
  • the riding vehicle may be determined based on the reference riding group created according to the process described with reference to FIG.
  • the riding vehicle can be estimated to be about two vehicles.
  • the getting off vehicle may be determined based on the reference getting off group created according to the process described with reference to FIG.
  • the getting off vehicle can be assumed to be a single vehicle in accordance with the present invention.
  • the gate entry / exit time of the passenger can be checked to see if there is only one reasonably consistent logical path. If there is one coherent logical path, this logical path may be determined as the passenger's travel path (307). If there is more than one coherent logical path, if the coherent logical paths are all the same physical path, then the physical path can be determined.
  • FIGS. 7 and 8 are schematic diagrams for explaining a process of generating a possible logical path in a movement route including a first and second transfer in a method for estimating a passenger movement route of a train network according to an embodiment of the present invention.
  • a process of generating a possible logic route in a movement route including one transfer is illustrated.
  • a passenger q having a boarding station O, a transit station A, and a discharging station D as a physical path
  • AG (Y, D) may be extracted from the reference discharging group corresponding to the departure time of the passenger.
  • This reference disengagement group AG (Y, D) is a set of passengers who get off at station D from vehicle Y.
  • the entry time of the passenger q at the boarding station O corresponds to the two boarding reference groups BG (X1, O) and BG (X2, O).
  • two logical paths may be made, which are obtained either by the X1 vehicle or the X2 vehicle, then transferred from the A station to the Y vehicle and then disembarked from the D station.
  • the possibility of transfer can be determined after the presence of TG (X1, Y, A) and TG (X2, Y, A) is checked. have.
  • a process of generating a possible logic route in a movement route including two transfers is shown.
  • a passenger q having a boarding station (O), a first transfer station (A), a second transfer station (B), and an unloading station (D) as a physical path
  • first the exit time of the stop station D (Exit Time) Based on at D)
  • AG Z, D
  • Passenger q can safely assume that he got off of vehicle Z of AG (Z, D).
  • the entry time of the passenger q at the boarding station O corresponds to the two boarding reference groups BG (X1, O) and BG (X2, O). Furthermore, among the trains (Y1, Y2, ...) departing after the time when the riding vehicles X1, X2 arrived at station A, the train arriving at station B before the time when the disembarkation vehicle Z at the station D departed from station B It can be seen from the train operation information that they are Y1 and Y2.
  • the train network as shown in FIG. 2 includes two persons having two possible physical paths including one transfer.
  • a schematic diagram is shown for explaining the process of determining the movement route for the passengers p and q, respectively.
  • the entry time and the exit time are expressed as numbers counting a specific time of day from 0 am to seconds.
  • the passengers p and q have the same boarding station (M2 in FIG. 2) and the getting off station (M6 in FIG. 2), and the inter-gate time is also similar, so that it is difficult to determine the physical path only with the inter-gate time. That is, the passengers p and q may both have a path R1 translating in Jamsil station and a path R2 translating in the interband based on the inter-gate time. Because paths R1 and R2 have similar times, the inter-gate time alone is difficult to distinguish.
  • This possible physical path is shown up and down on the left in FIG. 8. Above is a route (R2) starting at Sinrim Station (M2) on Line 2 and transferring to Line 3 at Miao-Bang (M4), and arriving at Garak Market Station (M6), and from Jamlim Station (M2) on Line 2 below.
  • the route R1 arriving at the Garak market station M6 after the transfer to the line 8 in M5) is shown.
  • the reference drop-off group is found for R2 via the alternating band M4. That is, the reference departure time interval including the gate departure time 29,813 of the passenger p [29,524; 29,874] and find the reference drop-off group AG (Y1, Garak Market Station).
  • the first number of reference entry time intervals may be the earliest time value among gate entry times of the passengers in the reference dropoff group.
  • the second number of reference entry time intervals may be the slowest time value among gate entry times of the passengers in the reference drop-off group. It is determined that the riding vehicle Y1 of the found reference getting off group is the getting off vehicle of the passenger p.
  • the route R1 may be determined to be a route not used by the passenger p.
  • the first number of reference entry time periods may be the earliest time value among gate entry times of the passengers of the reference riding group.
  • the second number of reference entry time intervals may be the slowest time value among gate entry times of the passengers of the reference riding group. It may be determined that the riding vehicle X1 of the found corresponding reference riding group is the riding vehicle of the passenger p.
  • route R2 may be determined to be a route that passenger q did not use.
  • a reference entry time period 27,231 comprising a gate entry time 27,295 of q for passenger q; 27,411] and find the reference riding group BG (X2, Sillim Station). And the riding vehicle X2 of passenger q can be determined by the found reference riding group.
  • the path R1 may be determined to be the movement route used by the passenger q.
  • the method according to the present invention it is possible to accurately estimate the vehicle traveling path including the vehicle information. Accordingly, under the integrated fare collection system, it can be used in the field of passenger travel path estimation for distributing fare among different operators on the train network. In addition, according to the present invention, since the vehicle used by each passenger can be estimated, it provides an advantage that it is possible to provide data data analyzing the passenger use pattern.
  • the method may be implemented by a computing device that may include a processor, a memory, and the like.
  • a memory is a computer readable storage medium that stores computer executable software, applications, program modules, routines, instructions, and / or data that are coded to be executed by a processor to perform a particular task.
  • the processor may read and execute from the memory, computer executable software, applications, program modules, routines, instructions, and / or data, which, when executed by the processor, are configured to perform the methods of the present invention described above.
  • the above-described methods of the present invention may be implemented by being coded as computer executable instructions and executed by a processor of a computing device.
  • Computer-executable instructions may include software, applications, modules, procedures, plug-ins, programs, instructions, and / or data structures, and the like.
  • Computing devices may include various devices, such as wearable computing devices, hand-held computing devices, smartphones, tablets, laptops, desktops, personal computers, servers, and the like.
  • the computing device may be a stand-alone type device.
  • the computing device may include a number of computing devices that cooperate with each other via a communication network.

Landscapes

  • Train Traffic Observation, Control, And Security (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A method for estimating a moving path of a passenger in a subway network is disclosed. According to an example, a method for estimating a moving path of a passenger in a subway network in which a plurality of subway vehicles are running along a plurality of subway lines, comprises the steps of: generating a reference group by estimating a train which each of the passengers boards, from passenger data including boarding and deboarding information of a plurality of passengers; generating, from the boarding and deboarding information of each passenger, at least one possible physical path configured by the subway line information between the departure station and the arrival station of each passenger; generating, for each of the possible physical paths, at least one logical path formed by connecting subway vehicles running on the individual subway line; and determining the moving path of a passenger by checking the consistency of the logical path.

Description

전철 네트워크의 승객 이동경로 추정 방법Estimation of Passenger Routes in Train Networks
본 발명은 일반적으로 전철 네트워크 내의 승객 이동경로를 추정하는 기술에 관련된다. 보다 상세하게는, 승객의 요금 결제를 위한 스마트 교통 카드 태그 정보로부터 승객의 승차역으로부터 하차역까지의 이동경로를 추정하는 기술에 관련된다.The present invention generally relates to techniques for estimating passenger travel routes within a train network. More specifically, the present invention relates to a technique for estimating a travel route from a passenger's boarding station to an unloading station from smart traffic card tag information for payment of a passenger.
근래 서울 및 대도시에서는 버스나 전철 등의 대중교통의 요금 결제를 위해 스마트 카드 형태의 교통 카드가 널리 사용되고 있다. 스마트 교통 카드의 태그 정보로부터 승객의 이동경로를 파악하는 일은 서로 다른 교통기관들 사이의 요금정산, 대중교통 이용 분석 등을 위해 중요하다.In recent years, smart cards have been widely used in Seoul and large cities for payment of public transport such as buses and trains. Knowing the passenger's route from the tag information of the smart transit card is important for fare settlement and public transportation usage analysis among different transit agencies.
일반적으로 승객은 대중교통 중 버스를 이용할 때 버스 승차시와 버스 하차시에 교통 카드를 요금 단말기에 태그 한다. 이에 따라, 교통 카드의 태그 정보에는 승객이 이용한 각각의 버스에 대해서 승차시간 및 하차시간이 기록될 수 있다. 하지만, 전철을 이용할 때 승객은 승차역에서 게이트를 통과할 때와 하차역에서 게이트를 통과할 때 교통 카드를 요금 단말기에 태그 한다. 그러므로, 전철 네트워크의 경우 승객이 어느 차량에 탑승하였는지, 어느 역에서 환승하였는지 등에 대해서 알 수 없다는 문제가 있다.In general, passengers tag their farecards at the terminal when they board the bus and when they get off the bus. Accordingly, in the tag information of the traffic card, the boarding time and the getting off time may be recorded for each bus used by the passenger. However, when using the train, the passenger tags the transportation card to the terminal when passing through the gate at the boarding station and when passing through the gate at the getting off station. Therefore, in the case of the train network, there is a problem in that it is not possible to know which vehicle the passenger boarded, which station it was transferred to, and the like.
이러한 문제 때문에, 버스 및 전철을 이용한 승객의 요금을 각 운영주체들 사이에 요금을 분배하는 경우, 먼저 버스 운영주체들과 전철 운영주체들 사이에 요금을 분배한 후, 이후 전철 운영주체들 사이에서 별도로 요금을 분배하고 있다. 예를 들어, 국내 특허 등록번호 10-0777217 (2007년 11월 12일 등록)의 등록공보에는 교통 카드의 승차 태그 정보로부터 환승여부를 판단하고, 환승일 때 소정 비율로 요금을 배분하는 기술이 개시된다. 그러나, 이 기술은 전철과 전철 사이의 환승에는 적용되지 않는다는 한계가 있었다. 그러므로, 승객이 전철 네트워크에서 어떤 이동경로를 움직였는지를 결정함으로써 전철 운영주체들 사이에 요금정산이 합리적으로 이루어지도록 해야 한다는 요구가 있었다.Because of this problem, in the case of distributing the fare of passengers using buses and trains between the operators, first distribute the fare between the bus operators and train operators, and then between the train operators We are distributing fees separately. For example, the registration publication of Korean Patent Registration No. 10-0777217 (registered on November 12, 2007) discloses a technology for determining whether to transfer from a ride tag information of a traffic card and allocating a fare at a predetermined rate when transferring. do. However, there was a limitation that this technique does not apply to the transfer between trains. Therefore, there was a demand for reasonable fare settlement among train operators by deciding which route the passengers traveled on the train network.
이 요구에 부응하기 위해 종래에 다항 로짓 모형(Multinomial Logit)을 사용하여 승객의 이동경로를 추정하는 기법이 사용되고 있다. 다항 로짓 모형은, 전철 노선들의 연결 형태인 물리적인 경로를 추정하기 위한 것으로서, 차내 이동시간(IVT), 환승 횟수(NT), 환승 시간(TT) 등의 파라미터를 이용하여 C=f(IVT, NT, TT)와 같이 구성하며, 통상 샘플을 기초로 최우추정법(MLE, Maximum Likelihood estimation)을 사용한다. 그러나 이러한 로짓 모형은 모형 개발을 위해 사용되는 샘플의 수가 1% 이하로 매우 적다는 문제가 있다. 예컨대, 서울의 전철 네트워크는 약 400개의 전철역을 가지며, 하루에 약 4,500개의 차량이 운행되고, 하루의 승객 이용건(즉, 승차역부터 하차역까지의 이동)은 약 5백만건이 넘는다. 그러므로 충분한 수의 샘플을 얻기 위해서는 너무나 과다한 비용과 시간이 필요하다는 문제가 있다. 더 나아가 이러한 로짓 모형의 경로 추정의 정확도는 70~80% 정도로 보고되고 있는데, 이런 정도의 정확도를 가진 경로 추정에 의한 요금 배분은 전철 운영주체들 사이에 분쟁을 발생시킬 수 있다.In order to meet this demand, a technique of estimating a passenger's route using a multinomial logit has been conventionally used. The polynomial logit model is for estimating the physical route, which is the connection form of the train lines, and uses C = f (IVT,) using parameters such as the in-vehicle travel time (IVT), the number of transfers (NT), and the transfer time (TT). NT, TT), and usually uses a maximum likelihood estimation (MLE) based on a sample. However, this logit model has a problem that the number of samples used for model development is very small, less than 1%. Seoul's train network, for example, has about 400 train stations, about 4,500 vehicles a day, and more than 5 million passengers per day (i.e. travel from the station to the station). Therefore, there is a problem that it takes too much cost and time to obtain a sufficient number of samples. Furthermore, the accuracy of the route estimation of the logit model is reported to be 70-80%, and the allocation of charges by the route estimation with this accuracy can cause disputes among train operators.
본 발명은 전철 네트워크의 승객 이동경로를 정확히 추정하는 방법을 제공하는 것을 목적으로 한다.It is an object of the present invention to provide a method for accurately estimating a passenger travel path of a train network.
또한 본 발명은 승객의 교통 카드 태그 정보로부터 승객이 이동한 전철 노선뿐만 아니라 승객이 탑승한 차량을 포함하는 승객 이동경로를 추정하는 방법을 제공하는 것을 목적으로 한다.In addition, an object of the present invention is to provide a method for estimating a passenger travel path including a vehicle on which the passenger as well as the train route traveled by the passenger from the passenger's traffic card tag information.
더 나아가 본 발명은 승객들 중 이동경로가 단순하고 확실한 일부 승객들을 추출하고, 추출된 승객들의 정보를 그 외 다른 승객들의 더 복잡한 이동경로를 추정하는 기준으로 사용함으로써, 이용 승객들의 대부분의 이동경로를 정확히 추정하는 방법을 제공할 것을 그 목적으로 한다.Furthermore, the present invention extracts some passengers whose travel path is simple and reliable, and uses the extracted passenger information as a criterion for estimating more complicated movement paths of other passengers, thereby making the most of the passengers traveling paths. Its purpose is to provide a method for accurately estimating.
상기 목적들은 본 발명에 따라 제공되는 전철 네트워크 내 승객 이동경로 추정 방법에 의하여 제공된다.The above objects are provided by a method for estimating a passenger route in a train network provided according to the present invention.
본 발명의 일 양상에 따라 제공되는 전철 네트워크 내 승객 이동경로 추정 방법은, 복수의 노선을 따라 복수의 차량이 운행되는 전철 네트워크에서 승객의 이동경로를 추정하는 방법으로서, 복수의 승객의 승하차정보를 포함하는 승객 데이터로부터 각 승객이 탑승한 열차를 추정하여 참조그룹을 생성하는 단계; 각 승객의 승하차정보로부터 각 승객의 승차역 및 하차역 사이에서 전철 노선 정보에 의해 이루어지는 적어도 하나의 가능 물리경로를 생성하는 단계; 상기 가능 물리경로 각각에 대하여, 상기 참조그룹에 기초하여, 각 전철 노선 상에서 운행하는 차량이 연결되어 이루어지는 적어도 하나의 논리경로를 생성하는 단계; 및 상기 논리경로의 일관성을 상기 참조그룹에 기초하여 체크하여 승객의 이동경로를 결정하는 단계를 포함할 수 있다.According to an aspect of the present invention, there is provided a method for estimating a passenger's movement route in a train network, which is a method of estimating a passenger's movement route in a train network in which a plurality of vehicles operate along a plurality of routes, Generating a reference group by estimating a train in which each passenger boards from the passenger data including; Generating at least one possible physical route made by the train line information between each passenger's boarding and unloading station from each passenger's getting on and off information; Generating at least one logical path for each of the possible physical paths, based on the reference group, by which a vehicle running on each train line is connected; And checking the consistency of the logical path based on the reference group to determine a passenger's movement path.
일 실시예에 있어서, 상기 참조그룹을 생성하는 단계는: 특정 승차역에서 특정한 제1차량에 승차한 승객들을 추정하여 참조 승차그룹을 생성하는 단계; 특정 하차역에서 특정한 제2차량으로부터 하차한 승객들을 추정하여 참조 하차그룹을 생성하는 단계; 특정 환승역에서 특정한 제3차량으로부터 특정한 제4차량으로 환승한 승객들을 추정하여 참조 환승그룹을 생성하는 단계를 포함할 수 있다.In an embodiment, the generating of the reference group may include: generating a reference riding group by estimating passengers boarding a specific first vehicle at a specific boarding station; Estimating passengers who get off from a specific second vehicle at a specific getting off station and generating a reference dropping group; And generating a reference transfer group by estimating passengers transferred from the specific third vehicle to the specific fourth vehicle at the specific transfer station.
다른 실시예에 있어서, 상기 참조 하차그룹을 생성하는 단계는: 상기 승객 데이터로부터, 가능 물리경로가 1개이며 환승이 없는 승객을 추출하는 단계; 추출된 승객에 대해, 각 승차역의 게이트 진입시간 이후에 출발한 차량들과 각 하차역의 게이트 진출시간 이전에 도착한 차량들을 비교하여, 동일한 차량을 오직 1개만 포함하는 승객을 선별하는 단계; 및 선별된 승객들이 상기 오직 1개만 포함하는 차량에 탑승한 것으로 추정하고, 상기 선별된 승객들 중에서 특정 하차역을 가지는 승객들을 이 탑승한 것으로 추정된 차량으로부터 상기 특정 하차역에서 하차한 승객들로서 분류하여 각각 참조 하차그룹으로서 생성하는 단계를 포함할 수 있다.In another embodiment, the step of generating the reference drop-off group may include: extracting, from the passenger data, a passenger having one possible physical path and no transfer; Selecting, for the extracted passengers, passengers including only one of the same vehicle by comparing vehicles departing after the gate entry time of each boarding station with vehicles arriving before the gate entry time of each boarding station; And classifying passengers as having boarded the vehicle including only one, and classifying passengers having a specific discharging station among the selected passengers as passengers getting off at the specific discharging station from the estimated vehicle And generating as a reference drop group, respectively.
또 다른 실시예에 있어서, 상기 참조 승차그룹을 생성하는 단계는: 상기 승객 데이터로부터, 가능 물리경로가 1개이며 환승이 없는 승객을 추출하는 단계; 추출된 승객에 대해, 각 승차역의 게이트 진입시간 이후에 출발한 차량들과 각 하차역의 게이트 진출시간 이전에 도착한 차량들을 비교하여, 동일한 차량을 오직 1개만 포함하는 승객을 선별하는 단계; 및 선별된 승객들이 상기 오직 1개만 포함하는 차량에 탑승한 것으로 추정하고, 상기 선별된 승객들 중에서 특정 승차역을 가지는 승객들을 이 탑승한 것으로 추정된 차량에 상기 특정 승차역에서 승차한 승객들로서 분류하여 각각 참조 승차그룹으로서 생성하는 단계를 포함할 수 있다.In another embodiment, the step of creating a reference ride group includes: extracting passengers having one possible physical path and no transfer from the passenger data; Selecting, for the extracted passengers, passengers including only one of the same vehicle by comparing vehicles departing after the gate entry time of each boarding station with vehicles arriving before the gate entry time of each boarding station; And classify passengers who have boarded the vehicle including only one, and classify passengers having a specific boarding station among the selected passengers as passengers who boarded at the specific boarding station to the vehicle estimated to be boarded. And generating each as a reference ride group.
또 다른 실시예에 있어서, 상기 참조 환승그룹을 생성하는 단계는: 상기 승객 데이터로부터, 가능 물리경로가 1개이며 1개의 환승역을 가지는 승객을 추출하는 단계; 추출된 승객 각각에 대해, 상기 참조 승차그룹들 중에서, 각 승차역의 게이트 진입시간을 포함하는 참조 진입시간구간을 가지는 대응 참조 승차그룹을 추출하고, 이 대응 참조 승차그룹의 승차 차량에 상기 승차역에서 승차한 것으로 추정하는 단계; 추출된 승객 각각에 대해, 상기 참조 하차그룹들 중에서, 각 하차역의 게이트 진출시간을 포함하는 참조 진출시간구간을 가지는 대응 참조 하차그룹을 추출하고, 이 대응 참조 하차그룹의 하차 차량에 상기 하차역에서 하차한 것으로 추정하는 단계; 추출된 승객 각각에 대해, 상기 추정된 승차 차량이 상기 환승역에 도착한 시간 및 상기 추정된 하차 차량이 상기 환승역에서 출발한 시간에 기초하여 환승여부를 체크하는 단계; 상기 환승여부를 체크하는 단계에서 환승 가능하다고 판정된 경우를, 상기 추정된 승차 차량으로부터 상기 환승역에서 상기 추정된 하차 차량으로 환승한 참조 환승그룹으로서 생성하는 단계를 포함할 수 있다.In another embodiment, the step of creating a reference transit group may include: extracting, from the passenger data, a passenger having one possible physical path and having one transit station; For each of the extracted passengers, a corresponding reference riding group having a reference entry time interval including a gate entry time of each boarding station is extracted from the reference riding groups, and the boarding station is assigned to the boarding vehicle of the corresponding reference riding group. Estimating that the passenger rides in the vehicle; For each of the extracted passengers, a corresponding reference disembarkation group having a reference entry time period including a gate entry time of each discharging station is extracted from the reference discharging groups, and the disembarkation station is displayed on the getting-off vehicle of the corresponding reference discharging group. Estimating to be off; For each of the extracted passengers, checking whether the estimated riding vehicle arrived at the transit station and whether the estimated getting off vehicle departed from the transit station; And when it is determined that transfer is possible in the checking of the transfer, as a reference transfer group transferred from the estimated riding vehicle to the estimated getting off vehicle at the transfer station.
또 다른 실시예에 있어서, 상기 논리경로를 생성하는 단계는: 각 승객에 대해 하차역에서 하차한 차량을 오직 하나로 결정할 수 있다.In yet another embodiment, the step of generating the logical path may include: determining only one vehicle to get off at the discharging station for each passenger.
또 다른 실시예에 있어서, 상기 논리경로를 생성하는 단계는: 각 승객에 대해 승차역에서 승차한 차량을 1개 또는 2개로 결정할 수 있다.In another embodiment, the generating of the logical path may include: determining one or two vehicles boarding at the boarding station for each passenger.
또 다른 실시예에 있어서, 상기 논리경로를 생성하는 단계는: 물리적 경로에 환승이 있는 경우 환승역에서 환승한 차량을 1개 또는 2개로 결정할 수 있다.In another embodiment, the generating of the logical path may include: determining whether one or two vehicles are transferred at the transfer station when there is a transfer in the physical path.
다른 양상에 따라, 컴퓨팅 장치의 프로세서에 의해 실행되면 특정 태스크를 수행하도록 코딩된 컴퓨터 실행가능 인스트럭션을 포함하는 컴퓨터 판독가능 스토리지 매체로서, 상기 컴퓨터 실행가능 인스트럭션은 상기 프로세서에 의해 실행될 때, 복수의 노선을 따라 복수의 차량이 운행되는 전철 네트워크에서 승객의 이동경로를 추정하는 방법을 수행하도록 구성되고, 상기 방법은: 복수의 승객의 승하차정보를 포함하는 승객 데이터로부터 각 승객이 탑승한 열차를 추정하여 참조그룹을 생성하는 단계; 각 승객의 승하차정보로부터 각 승객의 승차역 및 하차역 사이에서 전철 노선 정보에 의해 이루어지는 적어도 하나의 가능 물리경로를 생성하는 단계; 상기 가능 물리경로 각각에 대하여, 상기 참조그룹에 기초하여, 각 전철 노선 상에서 운행하는 차량이 연결되어 이루어지는 적어도 하나의 논리경로를 생성하는 단계; 및 상기 논리경로의 일관성을 상기 참조그룹에 기초하여 체크하여 승객의 이동경로를 결정하는 단계를 포함하는, 컴퓨터 판독가능 스토리지 매체가 제공될 수 있다.According to another aspect, a computer readable storage medium comprising computer executable instructions coded to perform a particular task when executed by a processor of a computing device, wherein the computer executable instructions are executed when executed by the processor. A method of estimating a travel route of a passenger in a train network in which a plurality of vehicles are operated is performed, and the method comprises: estimating a train on which each passenger boards from passenger data including getting on and off information of a plurality of passengers. Creating a reference group; Generating at least one possible physical route made by the train line information between each passenger's boarding and unloading station from each passenger's getting on and off information; Generating at least one logical path for each of the possible physical paths, based on the reference group, by which a vehicle running on each train line is connected; And checking the consistency of the logical paths based on the reference group to determine a travel path of a passenger.
여기서, 상기 승객 데이터는 상기 전철 네트워크에서 일정 시간 기간 동안 축적된 승객의 승하차정보를 포함하며, 상기 승하차정보는 적어도 상기 승객이 승차한 승차역, 상기 승차역의 게이트 진입시간, 상기 승객이 하차한 하차역, 및 상기 하차역의 게이트 진출시간을 포함할 수 있다.Here, the passenger data includes getting on and off information of passengers accumulated for a certain period of time in the train network, and the getting on and off information includes at least a boarding station on which the passenger boards, a gate entry time of the boarding station, and the passengers get off It may include the getting off station, and the gate advance time of the getting off station.
상술한 바와 같은 본 발명에 따르면, 승객의 교통 요금 지불을 위한 교통 카드 태그 정보에 기초하여 전철 네트워크 내에서의 승객 이동경로가 정확히 추정될 수 있다. 이에 따라, 교통 카드 태그 정보 이외에 별도의 샘플링 작업이 불필요하므로 승객 이동경로 추정에 필요한 비용과 시간이 절감될 수 있다. 또한 약 95% 이상의 정확도를 가지고 승객 이동경로가 추정될 수 있기 때문에, 전철 운영주체들 사이의 요금 정산과 관련된 분쟁이 해소될 수 있다.According to the present invention as described above, the passenger movement path in the train network can be accurately estimated based on the traffic card tag information for the passenger's transportation fee payment. Accordingly, since a separate sampling operation is unnecessary in addition to the traffic card tag information, the cost and time required for estimating the passenger travel path may be reduced. In addition, passenger travel routes can be estimated with greater than about 95% accuracy, thereby eliminating disputes related to fare settlement between train operators.
또한 본 발명에 따르면, 승객의 교통 카드 태그 정보로부터 승객이 이동한 전철 노선뿐만 아니라 승객이 탑승한 차량을 포함하는 승객 이동경로가 추정될 수 있다. 즉, 각각의 승객에 대해 승차역에서 탑승한 차량, 환승한 역, 하차역에서 하차한 차량을 포함하는 이동경로가 추정될 수 있다. 그러므로, 각각의 차량에 얼마나 많은 인원이 탑승하였는지를 알 수 있고, 각 환승역에서 이동한 인원을 알 수 있어, 교통 이용 분석을 위해 유용한 데이터를 제공할 수 있다.In addition, according to the present invention, a passenger movement path including a vehicle on which the passenger as well as a train route traveled by the passenger can be estimated from the passenger's transportation card tag information. That is, for each passenger, a movement route including a vehicle boarding at the boarding station, a transit station, and a vehicle getting off at the getting off station may be estimated. Therefore, it is possible to know how many people have boarded each vehicle and to know who has moved at each transit station, thereby providing useful data for traffic usage analysis.
더 나아가 본 발명은 승객들 중 이동경로가 단순하고 확실한 일부 승객들을 추출하고, 추출된 승객들의 정보를 그 외 다른 승객들의 더 복잡한 이동경로를 추정하는 기준으로 사용한다. 이에 따라, 실제 승객들의 정보에 기초하여 이동경로를 추정할 수 있게 한다.Furthermore, the present invention extracts some passengers whose passengers have simple and reliable movement paths, and uses the extracted passengers' information as a criterion for estimating more complicated movement paths of other passengers. Accordingly, it is possible to estimate the movement route based on the information of the actual passengers.
도 1 및 도 2는 본 발명이 적용될 수 있는 전철 네트워크의 일부를 예시하는 개략도. 1 and 2 are schematic diagrams illustrating portions of a train network to which the present invention may be applied.
도 3은 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법의 단계들을 보여주는 흐름도.3 is a flow chart showing the steps of a method for estimating a passenger travel path in a train network according to an embodiment of the present invention.
도 4는 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 참조 승차그룹을 생성하는 프로세스를 보여주는 흐름도.4 is a flowchart illustrating a process of creating a reference ride group in a method for estimating a passenger movement path of a train network according to an embodiment of the present invention.
도 5는 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 참조 하차그룹을 생성하는 프로세스를 보여주는 흐름도. FIG. 5 is a flowchart illustrating a process of creating a reference dropoff group in a method for estimating a passenger movement path of a train network according to an embodiment of the present invention. FIG.
도 6은 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 참조 환승그룹을 생성하는 프로세스를 보여주는 흐름도.6 is a flowchart illustrating a process of creating a reference transit group in a method for estimating a passenger's movement path of a train network according to an embodiment of the present invention.
도 7은 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 1번의 환승을 포함하는 이동경로에서 가능 논리경로를 생성하는 과정을 설명하기 위한 개략도. 7 is a schematic diagram illustrating a process of generating a possible logical path in a movement route including one transfer in a method for estimating a passenger movement route of a train network according to an embodiment of the present invention.
도 8은 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 2번의 환승을 포함하는 이동경로에서 가능 논리경로를 생성하는 과정을 설명하기 위한 개략도.8 is a schematic diagram for explaining a process of generating a possible logical path in a movement route including two transfers in a method for estimating a passenger movement route of a train network according to an embodiment of the present invention;
도 9는 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 1번의 환승을 포함하는 2개의 가능 물리경로를 가지는 2명의 승객에 대해 각각 이동경로를 결정하는 과정을 설명하기 위한 개략도.FIG. 9 is a view illustrating a process of determining a travel route for two passengers each having two possible physical routes including one transfer in the method for estimating a passenger travel route of a train network according to an embodiment of the present invention. schematic.
아래에서 정보 검색 시스템 및 방법이 도면들을 참조하여 예시로서 기술된다.The information retrieval system and method is described below by way of example with reference to the drawings.
도 1을 참조하면, 본 발명이 적용될 수 있는 전철 네트워크의 일부를 예시하는 개략도가 도시된다. 도시된 전철 네트워크는 서울의 한강 이남의 전철 네트워크 중 일부를 매우 간단히 개략적으로 보여준다. 근래에 대부분의 전철 승객들은 스마트 카드 형태의 교통 카드를 이용하여 교통 요금을 결제하고 있다. 즉, 승객은 승차역의 게이트를 통과하면서 게이트에 구비되어 있는 요금 결제용 단말기에 교통 카드를 태그하며, 하차역의 게이트를 통과하면서 마찬가지로 교통 카드를 태그 함으로써, 승차역과 하차역 사이의 거리에 따른 요금을 결제할 수 있다.Referring to FIG. 1, a schematic diagram illustrating a portion of a train network to which the present invention may be applied is shown. The train network shown is a very brief overview of some of the train networks south of the Han River in Seoul. In recent years, most passengers pay for transportation using smart cards. That is, the passenger tags the traffic card to the terminal for payment provided at the gate while passing through the gate of the boarding station, and similarly tags the traffic card while passing through the gate of the boarding station, thereby depending on the distance between the boarding station and the getting off station. You can pay the fee.
요금 결제를 위한 교통 카드의 승하차 게이트 태그 정보는 요금 결제와 관련된 승객의 승하차 정보를 제공할 수 있다. 승하차 정보에는 승차역, 승차역 게이트 통과 시각, 하차역, 및 하차역 게이트 통과 시각 등을 포함할 수 있다. 승차시와 하차시 사이의 시간은 승객이 전철 네트워크 내에서 이동을 위해 소비한 시간으로서 간주될 수 있으며, 이하 "인터-게이트 시간"이라고도 지칭한다. 통상적으로 전철 네트워크의 승객 이동 경로의 추정은 인터-게이트 시간을 이용하여 이루어질 수 있다.The getting on and off gate tag information of the transportation card for the fare payment may provide information of getting on and off the passenger related to the fare payment. The getting on and off information may include a boarding station, a boarding station gate passing time, a getting off station, and a getting off station gate passing time. The time between getting on and getting off can be regarded as the time spent by a passenger for movement within the train network, also referred to as "inter-gate time" below. Typically the estimation of the passenger travel path of the train network may be made using the inter-gate time.
일반적으로 인터-게이트 시간은 승객이 실제로 전철 차량에 탑승한 시간과는 다르다. 인터-게이트 시간은, 차량 탑승 시간 이외에, 승차 게이트로부터 승차 플랫폼까지 이동 시간, 승차 플랫폼에서의 차량 대기 시간, 하차 플랫폼에서 하차 게이트까지의 이동 시간 등이 더 포함되기 때문이다. 그 이외에도, 이동경로에 환승역이 포함되는 경우, 인터-게이트 시간에는 환승을 위해 한 플랫폼에서 다른 한 플랫폼까지 이동 시간 및 이동한 후 플랫폼에서의 차량 대기 시간이 더 포함되어야 한다. 또한 선택적으로, 인터-게이트 시간에는 승객이 게이트 내부에 위치하는 편의시설(예컨대, 화장실, 편의점, 쇼핑몰, 식당 등)을 이용하는 시간이 더 포함될 경우도 있다.In general, the inter-gate time is different from the time the passenger actually boarded the train. This is because the inter-gate time further includes a travel time from the boarding gate to the boarding platform, a vehicle waiting time at the boarding platform, a travel time from the boarding platform to the boarding gate, and the like, in addition to the vehicle boarding time. In addition, if the transit route includes a transit station, the inter-gate time should further include the transit time from one platform to the other platform for transit and the vehicle waiting time on the platform after the transit. Optionally, the inter-gate time may further include time for passengers to use amenities (eg, toilets, convenience stores, shopping malls, restaurants, etc.) located inside the gate.
상대적으로 간단한 경로의 경우 승객의 이동 경로는 인터-게이트 시간만으로 추정할 수 있다. 예를 들어, 승차역과 하차역이 하나의 노선 안에 위치하는 경우, 승차역과 하차역 사이에 하나의 환승역만이 존재하는 경우, 승객의 가능한 물리 경로들 사이에 역의 수나 환승역 수가 현저히 차이가 있는 경우 등에는 비교적 쉽게 승객의 이동 경로가 추정될 수 있다.In the case of a relatively simple route, the passenger's route can be estimated only by inter-gate time. For example, if the station and the station are located on one line, there is only one transfer station between the station and the station, and there is a significant difference in the number of stations or the number of stations between the possible physical routes of the passenger. The travel path of the passenger can be estimated relatively easily.
그런데, 예컨대 도 1에서 2호선의 신림역(M2)에서 승차한 승객이 1호선과 4호선이 교차하는 서울역(M7)에서 하차한 경우, 또는 8호선과 3호선이 교차하는 가락시장역(M6)에서 하차한 경우와 같이, 인터-게이트 시간만으로는 경로 추정이 어려운 경우가 많이 존재한다. 신림역(M2)에서 서울역(M7)까지 이동한 승객은 신도림역(M1)에서 1호선으로 환승한 후 서울역(M7)으로 가는 경로 R3와 사당역(M3)에서 4호선으로 환승한 후 서울역(M7)으로 가는 경로 R4를 이용할 수 있다. 경로 R3 및 R4 둘 모두, 거의 비슷한 인터-게이트 시간을 가질 수 있기 때문에, 이 승객이 어느 경로로 이동하였는지를 인터-게이트 시간만으로 결정하기 어렵다. 마찬가지로, 신림역(M2)에서 가락시장(M6)으로 이동한 승객의 경우에도, 잠실역(M5)에서 8호선으로 환승한 후 가락시장역(M6)으로 가는 경로 R1과 교대역(M4)에서 3호선으로 환승한 후 가락시장역(M6)으로 가는 경로 R2를 이용할 수 있다. 경로 R1 및 R2 둘 모두, 거의 비슷한 인터-게이트 시간을 가질 수 있기 때문에, 이 승객이 어느 경로로 이동하였는지를 인터-게이트 시간만으로 결정하기 어렵다.However, for example, when the passenger who boarded at Sillim Station (M2) of Line 2 gets off at Seoul Station (M7) where Line 1 and Line 4 cross, or at Garak Market Station (M6) where Line 8 and Line 3 intersect. As in the case of getting off, there are many cases where it is difficult to estimate the path using only the inter-gate time. Passengers who moved from Sinlim Station (M2) to Seoul Station (M7) transfer to Line 1 at Sindorim Station (M1), transfer to Route 4 from Seoul Station (M7) and then to Line 4 at Sadang Station (M3) and then to Seoul Station (M7). You can use the route R4. Since both paths R3 and R4 can have nearly similar inter-gate times, it is difficult to determine only by which inter-gate time this passenger traveled. Similarly, in the case of passengers moving from Sinrim Station (M2) to Garak Market (M6), transfer to Line 8 at Jamsil Station (M5) and then to Route 3 from Garak Market Station (M6) to Line 3 at Migyo (M4). You can then take R2 to Garak Market Station (M6). Since both paths R1 and R2 can have nearly similar inter-gate times, it is difficult to determine only by which inter-gate time this passenger traveled.
이와 같이, 승하차정보로부터 알려지는 승차역과 하차역 사이에서 인터-게이트 시간에 기초하여 판단할 때, 서로 구별하기 어려운 가능한 경로들이 복수로 존재할 수 있다. 이런 가능한 복수의 경로들을 가지는 경우의 비율은 전철 네트워크가 복잡할수록 더 증가할 것이 예상된다. 실제로 서울과 같은 대도시의 전철 네트워크에서, 전체 승객들 중 절반 정도의 승객들이, 구별이 어려운 복수 경로들을 가질 수 있는 승차역과 하차역 사이에서 이동하고 있음이 알려져 있다.As such, when judging based on the inter-gate time between the boarding station and the getting off station known from the getting on and off information, there may be a plurality of possible paths that are difficult to distinguish from each other. The proportion of cases with multiple such possible routes is expected to increase as the train network becomes more complex. Indeed, in metro networks like Seoul, it is known that about half of all passengers are traveling between boarding stations and getting off stations, which may have multiple routes that are difficult to distinguish.
도 2를 참조하면, 본 발명이 적용될 수 있는 전철 네트워크의 일부를 예시하는 개략도가 도시된다. 도 1에서는 승객 이동 경로 추정이 노선의 측면에서 기술되었으나, 도 2에서는 승객 이동 경로 추정을 차량의 측면에서 기술한다. 도시된 예는 승하차 정보로부터, 승차역(BS) M2, 승차 게이트 진입시각 Bt, 하차역(AS) M6, 하차 게이트 진출시각 At를 알 수 있는 승객의 이동 경로를 예시한다. 이 승객은 인터-게이트 시간에 기초하여 M5에서 환승한 경로 R1과 M4에서 환승한 경로 R2를 통해 이동할 수 있음이 추정될 수 있다.2, a schematic diagram illustrating a portion of a train network to which the present invention may be applied is shown. In FIG. 1, the passenger travel path estimation is described in terms of the route, while in FIG. 2, the passenger travel path estimation is described in terms of the vehicle. The illustrated example illustrates a travel path of a passenger that can know the boarding station (BS) M2, the boarding gate entry time Bt, the boarding station (AS) M6, and the boarding gate entrance time At. It can be estimated that this passenger can travel via route R1 transferred at M5 and route R2 transferred at M4 based on the inter-gate time.
더 구체적으로 보면, 경로 R1으로 이동한 경우, 승객은 M2에서 2호선 상에서 운행되는 차량 X1, X2, … 중 어느 하나의 차량에 승차한 후 M5에서 하차하고, 그 다음에 M5에서 8호선 상에서 운행되는 차량 Z1, Z2, Z3, … 중 어느 하나의 차량에 승차한 후 M6에서 하차할 수 있다. 이에 비하여, 경로 R2로 이동한 경우, 승객은 M2에서 2호선 상에서 운행되는 차량 X1, X2, … 중 어느 하나의 차량에 승차한 후 M4에서 하차하고, M4에서 3호선 상에서 운행되는 차량 Y1, Y2, … 중 어느 하나의 차량에 승차한 후 M6에서 하차할 수 있다. 이렇게 승객이 탑승한 특정 차량을 요소로서 더 포함하는 이동 경로는 인터-게이트 시간에 의해 추정하기 곤란하며 다른 수단이 필요하다. 예를 들어, 게이트-플랫폼, 플랫폼-플랫폼 이동 시간 추정값 및 차량의 운행 예정시간표 또는 운행 로그 등과 같이 특정 차량이 특정 역의 플랫폼에 도착하고 출발한 시각에 관한 정보가 필요할 수 있다.More specifically, when traveling on route R1, the passengers travel on lines 2 to M1, X2,... After getting on any one of the vehicles, get off at M5, and then the vehicles Z1, Z2, Z3,... You can get on any of the vehicles and get off at M6. In contrast, when traveling on route R2, the passengers are on vehicles X1, X2,... After getting on any one of the vehicles, get off at M4, and run on line 3 at M4 Y1, Y2,. You can get on any of the vehicles and get off at M6. This route of travel further comprising as a component a particular vehicle on which the passenger has boarded is difficult to estimate by inter-gate time and requires other means. For example, information about a time when a specific vehicle arrives and departs on a platform of a specific station may be needed, such as a gate-platform, a platform-platform movement time estimate, and a scheduled timetable or a driving log of the vehicle.
본 문서에서, 노선 만으로 추정되는 이동 경로와 차량을 포함하는 이동 경로를 구별하기 위하여, 전자를 '물리경로'라고 지칭하고 후자를 '논리경로'라고 지칭한다. 다시 말해서, 물리경로는 승차역, 하차역, 노선, 환승역 등의 요소들을 포함하는 이동 경로를 의미한다. 반면에 논리경로는 승객이 탑승한 특정 차량을 요소로서 더 포함하는 이동 경로를 의미한다. 이러한 명칭들은 단지 경로들의 성격에 따라 구별함으로써 설명이 용이해지도록 하기 위한 것이며, 어떠한 면으로도 본 발명을 제한하려는 의도는 없다.In this document, the former is referred to as the 'physical path' and the latter is referred to as the 'logical path' in order to distinguish the movement path including the vehicle and the movement path estimated by the route only. In other words, the physical path means a movement path including elements such as a boarding station, a getting off station, a line, and a transfer station. Logical paths, on the other hand, refer to a travel path that further includes, as elements, the particular vehicle on which the passenger boards. These names are provided for ease of explanation only by distinguishing according to the nature of the paths, and are not intended to limit the present invention in any aspect.
본 발명에 따른 방법은, 전철 네트워크의 승객 이동경로를 추정하는 방법으로서, 특히 승객의 탑승 차량을 특정하는 논리경로를 추정할 수 있게 한다.The method according to the invention is a method of estimating a passenger movement route of a train network, and in particular, makes it possible to estimate a logical route specifying a passenger vehicle.
이를 위하여, 본 발명의 방법은, 승객 교통카드 태그 정보, 전철 차량의 운행 로그 정보, 및 전철 네트워크의 토폴로지(즉, 노선도)를 주어진 정보로서 이용한다. 전철 네트워크 토폴로지 또는 전철 토폴로지는 1호선, 2호선, … 등의 노선들의 패턴을 포함하며, 전철의 승차역 및 하차역과 환승역 등을 포함하는 전철 네트워크 환경을 정의한다. 그러므로 승객의 이동 경로는, 도 1 및 도 2에서 예시된 바와 같이, 전철역들이 서로 연결된 선으로서 정의된다. 전철 차량의 운행 로그 정보는, 특정 전철역에 특정 차량이 도착 및/또는 출발한 시간 정보를 가지는 데이터이다.To this end, the method uses the passenger traffic card tag information, the driving log information of the train vehicle, and the topology of the train network (i.e., route map) as the given information. Train network topology or train topology is line 1, line 2,…. It includes a pattern of lines, and defines the train network environment including the train station, the train station, and the train station. The passenger's route of travel is therefore defined as a line in which the metro stations are connected to each other, as illustrated in FIGS. 1 and 2. The driving log information of the train vehicle is data having time information when the specific vehicle arrives and / or departs at the specific subway station.
한편, 승객 교통카드 태그 정보는 본 문서에서 승객의 승하차정보라고도 지칭되며, 승차역(BS), 게이트 진입시(Bt), 하차역(AS), 게이트 진출시(At)의 4개의 정보를 포함할 수 있다. 승차역(BS)과 게이트 진입시(Bt)는 승객 이동 경로 중 출발점의 위치 및 시간을 알려준다. 하차역(AS) 및 게이트 진출시(At)는 승객 이동 경로 중 도착점의 위치 및 시간을 알려준다. 이 4개의 정보를 집합적으로 지칭하기 위해 이하에서 종종 '쿼드로플(quadruple)'이라는 용어를 사용한다. Meanwhile, the passenger traffic card tag information is also referred to as passenger's getting on and off information in this document, and includes four pieces of information of a boarding station (BS), a gate entry (Bt), a drop off station (AS), and a gate entry (At). Can be. The boarding station BS and gate entry Bt inform the position and time of the starting point of the passenger travel path. The stop station (AS) and gate entry (At) inform the position and time of the arrival point in the passenger travel path. In order to collectively refer to these four pieces of information, the term 'quadruple' is often used below.
본 발명에 따라 제공되는 방법은, 쿼드로플 정보와 차량 운행 로그 정보 이외에, 경험이나 관찰된 사실에 근거하는 몇 가지 전제들을 이용한다.The method provided according to the present invention utilizes, in addition to quadrople information and vehicle driving log information, several presuppositions based on experience or observed facts.
한편, 본 발명자들은 승객들의 하차역 게이트 진출 데이터로부터, 하차역 게이트를 진출 시점들이 일정한 반복적인 패턴을 가진다는 것을 관찰하였다. 즉, 게이트 진출 패턴은, 특정 차량이 플랫폼에 도착한 이후에 가장 빠른 게이트 진출을 검출된 다음 일정 시간 동안 연속하여 승객들이 진출하는 구간과, 그 다음 차량이 플랫폼에 도착하기 전까지 일정한 시간간격 동안 게이트 진출이 중단된 구간이 이어지는 방식으로 반복된다. 이러한 진출 패턴은, 특정 차량에서 하차한 승객은 대부분 게이트를 가능한 한 빠르게 진출하려는 목적을 가질 것이라는 일반적인 전제와 일치한다.On the other hand, the inventors observed from the passengers' exit gate entrance data that the departure points have a constant repetitive pattern. In other words, the gate entry pattern is a section in which passengers advance for a predetermined time after detecting the fastest gate entry after a specific vehicle arrives at the platform, and then enter the gate for a predetermined time interval before the next vehicle arrives at the platform. This interrupted section is repeated in the following manner. This entry pattern is consistent with the general premise that most passengers who get off a particular vehicle will have the purpose of entering the gate as quickly as possible.
이 관찰 사실에 기초하여, 본 발명자들은, 특정 승객의 게이트 진출 시간(At)을 알면, 이 승객이 플랫폼에 도착한 차량들 중 어느 차량에서 하차한 것인지를 비교적 정확하게 추정할 수 있다는 점을 발견하였다. 이 발견에 따라, 다양한 역들에서 특정 차량에 승차한 후 특정 역에서 하차한 승객들의 집단이 추출될 수 있었다. 승객의 게이트 진출 시간에 기초하여 승객이 하차한 차량을 정확히 추정할 수 있다는 것은 본 발명의 방법에 있어서 가장 중요한 전제들 중 하나이다.Based on this observation, the inventors have found that knowing the gate entry time At of a particular passenger, it is possible to estimate relatively accurately which of those vehicles the passenger has arrived at the platform. With this discovery, a group of passengers who boarded a particular vehicle at various stations and then got off at a particular station could be extracted. It is one of the most important assumptions in the method of the present invention that it is possible to accurately estimate the vehicle a passenger got off based on the passenger's gate entry time.
이렇게 추출된 특정 역에서 하차한 승객들은 다시 특정 승차역을 동일하게 가지는 승객들로 더 분리할 수 있다. 이 서브 그룹들을 검토한 결과, 동일한 승차역을 가지는 승객들의 게이트 진입 시간들은, 동일 하차역의 게이트 진출 시간들의 분포에 비하여, 훨씬 비정형적이고 넓게 분산되어 있다는 것이 발견되었다. 다시 말해서, 특정 승객의 게이트 진입 시간(Bt)를 알고 있다고 하더라도, 이 승객이 특정 시점에 승차역 플랫폼에 도착한 특정 차량에 승차하였는지를 알 수 없었다. 그러나 가장 빠르게 이동하고자 원할 것이라는 전제에 기초하면 승객은 플랫폼에서 대기하는 시간을 단축하고자 할 것이다. 그러므로, 본 발명자들은 승차역의 게이트를 진입한 승객은 이후 도착한 복수의 차량들에 대해 균등한 비율로 승차한다기 보다는 먼저 도착한 차량들에 승차할 확률이 더 높다고 추정할 수 있었다.Passengers who get off at this particular station can be further separated into passengers having the same specific station. As a result of reviewing these subgroups, it has been found that the gate entry times of passengers having the same boarding station are much more informal and widely distributed compared to the distribution of gate entry times of the same stop station. In other words, even if the specific passenger's gate entry time (Bt) was known, it was not known whether the passenger had boarded the particular vehicle arriving at the station platform at that particular point in time. But based on the premise that they want to move the fastest, passengers will want to reduce the waiting time on the platform. Therefore, the inventors could assume that the passenger who entered the gate of the boarding station is more likely to board the vehicles that arrived earlier than to ride at an even rate with respect to the plurality of vehicles that arrived later.
이와 같이 승객의 게이트 진입 시간 또는 진출 시간의 패턴과 차량의 플랫폼 도착(또는 출발) 시간을 고려하면, 승객이 실제 승차하거나 하차한 차량을 유일하게 또는 2개와 같은 유한한 수로 추정할 수 있다.In this way, considering the pattern of gate entry or exit time of the passenger and the time of arrival (or departure) of the platform of the vehicle, it is possible to estimate a vehicle in which the passenger actually gets on or gets off in a finite number such as only two or two.
더 나아가 본 발명의 방법은, 승객들 중 거의 절반은 이동 경로가 비교적 간단하고 단순하다는 사실에 착안하여, 이동 경로가 확실한 승객들을 이용하여, 이동 경로가 불확실한 승객들의 이동 경로를 추정하는 기법을 제안한다.Furthermore, the method of the present invention, considering the fact that almost half of the passengers are relatively simple and simple, proposes a technique for estimating the travel paths of passengers whose travel paths are uncertain, using passengers having a certain travel path. do.
본 발명에 따라, 이동 경로가 확실한 승객들은 참조그룹으로서 생성된다. 참조그룹은 이동경로가 복잡한 승객들의 이동 경로를 추정하기 위해 기준을 제공하는 승객들의 집합이다. 참조그룹에는 특정 차량을 타고 특정 역에서 하차한 그룹(참조 하차그룹), 특정 차량을 특정 역에서 승차한 그룹(참조 승차그룹), 및 특정 차량을 타고 가다가 특정 역에서 다른 특정 차량으로 환승한 그룹(참조 환승그룹)이 포함될 수 있다.According to the present invention, passengers having a certain route of travel are created as reference groups. A reference group is a set of passengers that provides criteria for estimating the travel paths for passengers with complex travel paths. A reference group includes a group of people who ride a specific vehicle at a specific station (reference getting off group), a group of people who ride a specific vehicle at a specific station (reference riding group), and a group of people who ride a specific vehicle and transfer to another specific vehicle at a specific station. (Reference transit group) may be included.
참조그룹의 생성을 위해, 물리적으로 가능한 경로가 1개이면서 환승역이 없는 경우 즉 단순 선형경로를 가진 승객들을 추출할 수 있다. 단순 선형경로를 가진 승객들 중에서, 가능한 탑승 차량이 특정 차량 1개이면서 특정 역에서 하차한 승객들을 추출하여 참조 하차그룹(AG, alighting group)으로 생성할 수 있다. 유사하게, 가능한 탑승 차량이 특정 차량 1개이면서 특정 역에서 승차한 승객들을 추출하여 참조 승차그룹(BG, boarding group)으로 생성할 수 있다. 이와 같이 참조 승차그룹 및 참조 하차그룹은 각각 2개의 파라미터 즉 승차/하차한 차량(X) 및 승차/하차 역(N)을 포함한다.For the creation of a reference group, it is possible to extract passengers with one physically possible route and no transit station, ie a simple linear route. Among passengers having a simple linear path, a possible vehicle may be extracted into a reference group (AG) by extracting passengers who have a specific vehicle and get off at a specific station. Similarly, it is possible to extract passengers who ride on a specific station while having one possible vehicle, and generate a reference boarding group (BG). In this way, the reference riding group and the reference getting off group each include two parameters, namely, the vehicle X which got on / off and the boarding / unloading station N.
참조 하차그룹의 승객들의 하차역 게이트 진출 시간(At)들은 임의의 길이를 가진 참조 진출 시간구간 At1 ~ At2 에 대응할 수 있다. 이러한 참조 진출 시간구간은 참조 하차그룹의 승객들의 데이터에 기초하여 정의될 수 있다. 예를 들어, 참조 진출 시간구간은 참조 하차그룹의 승객들의 게이트 진출 시간들 중 가장 빠른 시간 값부터 가장 느린 시간 값까지로 정의될 수 있다.The drop-off gate exit times At of the passengers in the reference drop-off group may correspond to the reference departure time intervals At1 to At2 having an arbitrary length. This reference entry time period may be defined based on data of passengers in the reference dropoff group. For example, the reference departure time period may be defined as the earliest time value to the slowest time value among gate entry times of the passengers in the reference descent group.
만약 어떤 승객의 하차역 게이트 진출 시간(at)이 참조 진출 시간구간 At1 ~ At2 사이에 위치된다면(At1 < at < At2), 이 승객은 참조 하차그룹의 승객들 중 하나라고 추정될 수 있고, 참조 하차그룹이 하차한 차량에서 하차하였다고 추정될 수 있다.If a passenger's drop-off gate exit time (at) is located between the reference departure time intervals At1-At2 (At1 <at <At2), this passenger can be assumed to be one of the passengers in the reference dropoff group, It can be assumed that the drop-off group got off the vehicle that got off.
유사하게, 참조 승차그룹의 승객들의 승차역 게이트 진입 시간(Bt)은 임의의 길이를 가진 참조 진입 시간구간 Bt1 ~ Bt2 에 대응할 수 있다. 이러한 참조 진입 시간구간은 참조 승차그룹의 승객들의 데이터에 기초하여 정의될 수 있다. 예를 들어, 참조 진입 시간구간은 참조 승차그룹의 승객들의 게이트 진입 시간들 중 가장 빠른 시간 값부터 가장 느린 시간 값까지로 정의될 수 있다.Similarly, the boarding station gate entry time Bt of the passengers in the reference riding group may correspond to the reference entry time interval Bt1 to Bt2 having any length. This reference entry time interval may be defined based on data of passengers in the reference riding group. For example, the reference entry time period may be defined as the earliest time value to the slowest time value among gate entry times of the passengers in the reference riding group.
만약 어떤 승객의 승차역 게이트 진입 시간(bt)이 참조 진입 시간구간 Bt1 ~ Bt2 사이에 위치된다면(Bt1 < bt < Bt2), 이 승객은 참조 승차그룹의 승객들 중 하나라고 추정될 수 있고, 참조 승차그룹이 승차한 차량에 승차하였다고 추정될 수 있다.If a passenger's boarding station gate entry time (bt) is located between the reference entry time intervals Bt1 to Bt2 (Bt1 <bt <Bt2), this passenger may be assumed to be one of the passengers in the reference riding group, It can be assumed that the riding group has boarded the vehicle.
한편, 참조 환승그룹(TG, transfer group)을 생성하기 위하여, 승객들 중에서 물리적으로 가능한 경로가 1개이면서 환승역이 1개만 있는 경우 즉 단순 환승경로를 가진 승객들을 추출할 수 있다. 단순 환승경로를 승차역(O) - 환승역(A) - 하차역(D)을 포함하는 것으로 표현할 수 있다.Meanwhile, in order to create a reference transfer group (TG), it is possible to extract passengers having a single transit path, that is, when there is only one transit station and one physically possible route among passengers. A simple transfer route may be expressed as including a boarding station (O)-a transfer station (A)-a drop off station (D).
이 경우, 승차역(O)에서 특정 승객은 게이트 진입 시간을 기준으로 처음 도착하는 2개의 연속적인 차량들(X1, X2) 중 어느 하나에 승차할 수 있다고 추정할 수 있고, 하차역(D)에서 승객은 게이트 진출 시간을 기준으로 하나의 차량(Y)에서 하차한다고 추정될 수 있다.In this case, it is possible to assume that a particular passenger at the boarding station O can board any one of two consecutive vehicles X1 and X2 arriving for the first time based on the gate entry time, and the drop-off station D In FIG. 3, it may be estimated that a passenger gets off from one vehicle Y based on a gate advance time.
이러한 추정에 기초하여, 단순 환승경로(O-A-D)를 가진 승객들 중에서, O역에서 X1 차량에 승차하고, A역에서 하차한 후 Y차량에 승차하고, D역에서 Y 차량으로부터 하차한 환승그룹(X1, Y, A)과 O역에서 X2 차량에 승차하고, A역에서 하차한 후 Y차량에 승차하고, D역에서 Y 차량으로부터 하차한 환승그룹(X2, Y, A)을 구분할 수 있다. 이렇게 차량이 특정된 환승 그룹들 중에서, X1, X2, Y 차량이 A역에 도착한 시간 등의 정보에 따라 물리적으로 연결 가능한 것인지 여부가 체크될 수 있다. 만약 A역에서 X1 차량 도착시간이 Y 차량의 출발시간보다 이후라면, 승객은 물리적으로 X1 차량으로부터 Y 차량으로 환승할 수 없을 것이다.Based on these estimates, among the passengers with a simple transit route (OAD), a transfer group that boards X1 at O station, gets off at Y station, gets on Y vehicle, and gets off at Y station at D station ( The transit groups X2, Y and A can be distinguished from the X2, Y, A) and O stations, from the A station to the Y vehicle, and from the D station. Among the transfer groups in which the vehicle is specified, it may be checked whether the X1, X2, Y vehicle is physically connectable according to information such as the time when the vehicle arrives at the A station. If the arrival time of vehicle X1 at station A is later than the departure time of vehicle Y, the passenger will not be able to physically transfer from vehicle X1 to vehicle Y.
이와 같이, 환승 가능 여부가 체크되면, 환승그룹들은 다수의 O 역들에서 출발하는 특정 차량 X1과 다수의 D 역들에 도착한 Y 차량의 승객들 정보를 합침으로써, 특정 환승역 A에서 X1차량으로부터 Y차량으로 환승한 승객들을 묶어 참조 환승 그룹(X1, Y, A)으로서 생성될 수 있다. 이와 같이, 참조 환승 그룹은 3개의 파라미터 즉 승차한 차량(X), 하차한 차량(Y), 및 환승역(A)을 포함한다.As such, if the transit availability is checked, the transit groups combine information of passengers of a specific vehicle X1 departing from a plurality of O stations and a Y vehicle arriving at a plurality of D stations, from the X1 vehicle to the Y vehicle at a specific transit station A. Transit passengers can be bundled and created as a reference transfer group (X1, Y, A). As such, the reference transfer group includes three parameters: the boarded vehicle X, the boarded vehicle Y, and the transfer station A. FIG.
위에서는 승차역과 하차역 사이에 1회의 환승이 존재할 때를 기술하였다. 승차역과 하차역 사이에 2회의 환승이 이루어진다면, 유사한 방식으로 논리적인 경로가 추정될 수 있다. 예컨대, 승차역에서 2개의 승차차량이 추정되고, 1번째 환승역에서 1개 또는 2개의 환승차량이 추정될 수 있고, 2번째 환승역에서 하차 차량은 1개로 추정될 수 있다.Above, we describe when there is a single transfer between the station and the station. If two transfers are made between the station and the station, the logical route can be estimated in a similar manner. For example, two ride vehicles may be estimated at a boarding station, one or two transfer cars may be estimated at a first transfer station, and one vehicle may be estimated at a second transfer station.
본 발명에 따라, 어떤 승객의 승하차정보가 주어지면, 하나의 물리적 경로에 대하여 하차역 및 게이트 진출 시간(At)에 따라 1개의 하차 차량(Y)이 추정될 수 있고, 승차역 및 게이트 진입 시간(Bt)에 따라 2개의 승차 차량(X1, X2)이 추정될 수 있다. 이에 따라, 하나의 물리적 경로에 대하여 적어도 서로 다른 2개의 논리경로가 만들어질 수 있다.According to the present invention, given the getting on and off information of a passenger, one getting off vehicle Y can be estimated according to the getting off station and the gate entry time At for one physical route, and the starting station and the gate entry time According to Bt, two riding vehicles X1 and X2 can be estimated. Accordingly, at least two different logical paths may be created for one physical path.
이후 만약 가능한 물리경로가 1개이고, 환승역이 존재하지 않는 경우라면, 2개의 논리경로 각각에 대해, 하차 차량(Y)이 승차 차량들(X1, X2)과 동일한지가 체크될 수 있다. 동일한 차량이 있는 논리경로가 있으면, 이 논리경로를 승객의 이동 경로로서 결정할 수 있다.Then, if there is one possible physical path and no transfer station exists, for each of the two logical paths, it can be checked whether the getting off vehicle Y is the same as the riding vehicles X1, X2. If there is a logical path with the same vehicle, this logical path can be determined as the passenger's travel path.
한편, 만약 가능한 물리경로에 1개의 환승역이 포함되는 경우, 가능한 논리경로 각각에 대해, 참조 환승그룹이 존재하는지가 체크될 수 있다. 참조 환승그룹이 존재하는 논리경로가 있으면, 이 논리경로를 승객의 이동 경로로서 결정할 수 있다. 만약 어느 논리경로에도 참조 환승그룹이 존재하지 않으면, 승객의 이동 경로는 결정되지 않는다.On the other hand, if one possible transit station is included in the possible physical path, for each of the possible logical paths, it may be checked whether a reference transit group exists. If there is a logical path in which the reference transit group exists, this logical path may be determined as the passenger's travel path. If no reference transit group exists in any logical path, the passenger's route is not determined.
참조 환승그룹의 승차 차량(X)은 참조 승차그룹에 기초하여 결정될 수 있다. 마찬가지로 참조 환승그룹의 하차 차량(Y)은 참조 하차그룹에 기초하여 결정될 수 있다. 그리고 환승역(T)에 도착한 승차 차량(X)의 도착시간이 환승역(T)에서 출발한 하차 차량(Y)의 출발시간보다 앞선 경우에, 승차 차량(X)으로부터 하차 차량(Y)으로의 환승이 가능하다고 결정될 수 있다.The riding vehicle X of the reference transfer group may be determined based on the reference riding group. Similarly, the getting off vehicle Y of the reference transfer group may be determined based on the reference getting off group. And when the arrival time of the boarding vehicle X which arrived at the transit station T was earlier than the departure time of the disembarkation vehicle Y which departed from the transit station T, the transfer from the boarding vehicle X to the getting-off vehicle Y is carried out. It can be determined that this is possible.
이하 도 3 내지 도 9을 참조하여, 본 발명의 방법의 실시예들이 기술된다. 여기서 기술되는 방법들은 단지 예시에 불과하다. 해당 기술분야의 통상의 지식을 가진 자는, 청구항들의 범위 내에서 다양한 조합의 다른 방법들이 가능하다는 것을 쉽게 이해할 것이다. 기술된 방법의 전부 또는 일부는, 컴퓨팅 장치의 프로세서에 의해 수행되면 특정 태스크를 실행할 수 있도록 하는 컴퓨터-실행가능 인스트럭션, 모듈, 소프트웨어, 데이터, 알고리즘, 프로시저 등으로 코딩될 수 있다. 컴퓨터-실행가능 인스트럭션 등은 소프트웨어 개발자에 의해 예를 들어 베이직, 포트란, C, C++ 등과 같은 프로그래밍 언어에 의해 코딩된 후, 기계언어로 컴파일될 수 있다. 3-9, embodiments of the method of the present invention are described. The methods described herein are merely examples. One of ordinary skill in the art will readily appreciate that other methods of various combinations are possible within the scope of the claims. All or part of the described methods can be coded into computer-executable instructions, modules, software, data, algorithms, procedures, and the like, which, when performed by a processor of a computing device, enable a particular task to be executed. Computer-executable instructions and the like can be coded by a software developer, for example, in a programming language such as Basic, Fortran, C, C ++, etc., and then compiled into a machine language.
도 3을 참조하면, 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법의 단계들을 보여주는 흐름도가 도시된다. 본 발명에 따른 방법은, 전철 네트워크의 승객 이동경로를 추정하는 방법(300)으로서, 특히 승객의 탑승 차량을 특정하는 논리경로를 추정할 수 있게 한다.Referring to FIG. 3, there is shown a flowchart showing the steps of a method for estimating a passenger's movement route in a train network according to an embodiment of the present invention. The method according to the invention is a method 300 for estimating a passenger's movement path of a train network, and in particular, makes it possible to estimate a logical path specifying a passenger's vehicle.
이 방법(300)은 먼저 참조그룹들을 생성(301)하고, 가능 물리경로(303)를 생성하며, 이후 가능 논리경로 생성(305)한 다음, 최종적으로 승객의 이동경로를 결정(307)하는 방식으로 진행될 수 있다.The method 300 first generates reference groups 301, generates possible physical paths 303, then generates possible logical paths 305, and finally determines a passenger's movement path 307. It can proceed to.
참조그룹들을 생성하는 단계(301)는 이동경로가 단순하고 확실한 승객들을 대상으로 승차 또는 하차한 열차를 특정함으로써 참조그룹을 생성하는 단계이다. 이 단계(301)의 더 구체적인 실시예들이 도 4, 도5, 도 6에 도시된다.The step of creating reference groups 301 is a step of creating a reference group by specifying trains that get on or off for passengers whose movement paths are simple and reliable. More specific embodiments of this step 301 are shown in FIGS. 4, 5, 6.
먼저 도 4 및 도 5를 참조하면, 참조 승차그룹의 생성 프로세스(400) 및 참조 하차그룹의 생성 프로세스(500)가 도시된다. 먼저 승객 데이터로부터 단순 선형 경로를 가진 승객들을 추출할 수 있다(401, 501). 승객 데이터는 전철 네트워크를 이용한 승객들의 교통카드가 승차역 및 하차역의 게이트를 통과할 때 태그 함으로써 생성된 승하차정보를 포함한다.Referring first to FIGS. 4 and 5, a process of generating a reference ride group 400 and a process of generating a reference ride group 500 is illustrated. First, passengers having a simple linear path may be extracted from the passenger data (401, 501). Passenger data includes boarding and getting information generated by tagging when the traffic card of passengers using the train network passes through the gates of the boarding station and the unloading station.
추출된 각각의 승객에 대해, 승차역(BS), 승차역 게이트 진입시간(Bt), 하차역(AS), 하차역 게이트 진출시간(At)가 주어진다. 이러한 승객 정보와 함께 열차 운행정보를 참조함으로써, 각 승객에 대해, 승차역에서 Bt 이후에 출발한 열차들을 묶어 출발열차그룹(P)을 생성할 수 있다(403, 503). 출발열차그룹(P)은 승객이 탑승할 수 있는 가능성이 있는 열차들의 집합이라고 할 수 있지만, 승객은 게이트 진입 이후에 열차에 승차할 수 있다는 점만을 고려한다. 한편, 각 승객에 대해 하차역에서 At 이전에 도착한 열차들을 묶어 도착열차그룹(Q)를 생성할 수 있다(405, 505). 도착열차그룹(Q)은 승객이 탑승하였던 가능성이 있는 열차들의 집합이라고 할 수 있지만, 승객이 게이트 진출 이전에 열차에서 하차할 수 있다는 점만을 고려한다.For each extracted passenger, a boarding station BS, a boarding station gate entry time Bt, a boarding station AS, and a boarding station gate entry time At are given. By referring to the train operation information together with such passenger information, for each passenger, trains departed after Bt at the boarding station may be bundled to generate a departure train group P (403, 503). The departure train group P can be said to be a set of trains with the possibility of passengers to board, but only considering that passengers can board the train after entering the gate. Meanwhile, for each passenger, arrival train groups Q may be generated by tying trains arriving before At at the drop off station (405 and 505). Arrival train group (Q) can be said to be a set of trains that passengers may have boarded, but only consider that passengers can get off the train before entering the gate.
물리적으로, 승객은 출발열차그룹(P)과 도착열차그룹(Q)에 포함된 열차들 중 오직 하나의 열차에만 탑승할 수 있다. 다시 말해서, P와 Q의 교집합에 포함된 열차가 오직 하나만 존재한다면, 승객은 이 열차에 탑승하였다고 안전하게 추정할 수 있다. 그러므로, 각 승객에 대해 생성된 P와 Q에 포함되어 있는 열차들 중 서로 동일한 열차(X)가 1개인 승객들을 선별할 수 있다(407, 507). 이렇게 선별된 승객은 이 특정 열차(X)에 탑승한 것으로 추정될 수 있다.Physically, the passenger can only board one of the trains included in the departure train group P and the arrival train group Q. In other words, if there is only one train involved in the intersection of P and Q, the passenger can safely assume that he was on this train. Therefore, among the trains included in P and Q generated for each passenger, passengers having one identical train X may be selected (407 and 507). Passengers thus selected can be assumed to have boarded this particular train (X).
단계(407, 507)에서 선별된 승객들이 특정 열차(X)에 탑승한 점이 확실하므로, 이 특정 열차(X)에 탑승한 승객들 중에서 특정 역(O)에서 승차한 정보를 가진 승객들만을 묶어 참조 승차그룹 BG(X,O)으로 생성하고(409), 유사하게 특정 역(D)에서 하차한 정보를 가진 승객들만을 묶어 참조 하차그룹 AG(X, D)으로 생성할 수 있다(509).Since it is certain that the passengers selected in steps 407 and 507 have boarded a specific train X, only the passengers who have information on boarding at a specific station O among the passengers in this specific train X are tied up. The reference boarding group BG (X, O) may be generated (409), and similarly, only passengers having information getting off at a specific station (D) may be bundled and generated as the reference boarding group AG (X, D) (509). .
도 6을 참조하면, 참조 환승그룹의 생성 프로세스(600)가 도시된다. 먼저 특정 환승역(T)을 경유한 승객들을 추출한다(601). 이들 중 특정 승객이 승차역에서 승차한 승차 차량(X)은 이 승객에 대해 발견된 참조 승차그룹에 기초하여 결정될 수 있다(603). 그리고, 승객이 하차역에서 하차한 하차 차량(Y)은 이 승객에 대해 발견된 참조 하차그룹에 기초하여 결정될 수 있다(605). 그런 다음, 승객의 승차 차량(X)과 하차 차량(Y)이 특정 환승역(T)에 도착하고 출발한 시간에 기초하여, 승객의 환승 가능 여부가 결정될 수 있다(607). 예컨대, 환승역(T)에 도착한 승차 차량(X)의 도착시간이 환승역(T)에서 출발한 하차 차량(Y)의 출발시간보다 앞선 경우에, 승객은 승차 차량(X)으로부터 하차 차량(Y)으로의 환승이 가능하다고 결정될 수 있다. 마지막으로, 특정 환승역(T)에서 특정 승차 차량(X)으로부터 특정 하차 차량(Y)으로 환승가능한 승객들을 묶어, 참조 환승그룹 TG(X,Y,T)으로 생성한다(609).Referring to FIG. 6, a process 600 of creating a reference transit group is shown. First, passengers via the specific transit station T are extracted (601). The riding vehicle X of which a particular passenger has boarded at the boarding station may be determined based on the reference riding group found for this passenger (603). Then, the disembarkation vehicle Y that the passenger has disembarked at the disembarkation station may be determined 605 based on the reference disembarkation group found for this passenger. Then, based on the time when the passenger's riding vehicle X and the discharging vehicle Y arrives and departs at the specific transit station T, whether or not the passenger can transit can be determined 607. For example, when the arrival time of the boarding vehicle X arriving at the transit station T is earlier than the departure time of the getting off vehicle Y departing from the transit station T, the passenger gets off the vehicle Y from the boarding vehicle X. It may be determined that transfer to is possible. Finally, in a specific transit station T, passengers transitable from a specific riding vehicle X to a specific getting off vehicle Y are bundled and generated as a reference transfer group TG (X, Y, T) (609).
이러한 참조그룹들을 이용하여 본 발명에 따라 승객의 이동 경로가 정확하게 추정될 수 있으며, 도 3의 단계들을 참조하여 아래에서 설명된다.Using these reference groups the passenger's route of travel can be accurately estimated in accordance with the invention and is described below with reference to the steps of FIG.
다시 도 3을 참조하면, 승객의 승하차 정보에 기초하여 차량 토폴로지 즉 전철 네트워크 토폴로지로부터 가능한 물리경로가 생성될 수 있다(303). 각각의 물리경로는 승차역, 환승역(만약 존재하는 경우), 하차역을 포함한다. 이러한 물리경로는 승객의 승하차정보에 기초한 인터-게이트 시간에 의해 가능한 경로들 중에서 환승이 없는 경우, 환승이 1회인 경우, 환승이 2회인 경우 등이 선택될 수 있다. 그 외의 가능한 경우는 예외적 것으로서 본 발명의 취급 범위 외의 것이다.Referring back to FIG. 3, a possible physical path may be generated from the vehicle topology, that is, the train network topology, based on the passenger's getting on and off information (303). Each physical route includes a boarding station, a transit station (if present), and a disembarkation station. The physical path may be selected when there are no transfers, one transfer, two transfers, etc. among the possible routes by the inter-gate time based on the passenger's getting on and off information. Other possible cases are exceptional and are outside the scope of the present invention.
본 발명에 따라, 각각의 승객에 대해 가능한 물리경로가 1개 또는 2개 이상의 수로 생성될 수 있다. 물리경로 생성 과정(303)을 좀 더 구체적으로 설명하면 다음과 같다.According to the invention, possible physical paths for each passenger may be generated in one or two or more numbers. The physical path generation process 303 will be described in more detail as follows.
먼저 생성된 물리경로들 사이의 가능한 허용오차를 정할 수 있다. 예컨대, 어떤 승객에 대해 복수의 물리경로들 PR1, PR2, PR3, … 가 가능하다고 하면, 각 물리경로는 전철역들이 서로 연결되어 있는 선의 형태를 하므로, 각 물리경로는 유한한 수의 역을 포함할 것이다. 예컨대 PR1은 10개의 역을 포함하고, PR2는 12개의 역을 포함하고, PR3는 40개의 역을 포함하고… 등등. 이때, 승객은 될 수 있는 한 짧은 경로를 선택할 것이라는 전제에 따라, 너무 많은 수의 역을 포함하는 물리경로는 가능한 물리경로가 아닌 것으로 결정될 수 있다. 예컨대, 물리경로의 역 허용오차를 8개 정하면, PR1과 PR2는 2개의 차이를 가지고, PR3는 PR1 및 PR2에 대해 각각 30개 및 28개의 차이를 가진다. 그러므로, PR3는 가장 적은 역수인 PR1보다 허용오차 8개를 넘는 수의 역을 가지고 있으므로 가능한 물리경로에서 제외할 수 있다.First, we can determine the possible tolerances between the generated physical paths. For example, for a passenger a plurality of physical paths PR1, PR2, PR3,... If possible, each physical path will have a finite number of stations, since each physical path is in the form of a line connected to each other. For example, PR1 contains 10 stations, PR2 contains 12 stations, PR3 contains 40 stations, and so on. etc. At this time, depending on the premise that the passenger will choose the shortest possible route, it may be determined that the physical path including too many stations is not a possible physical path. For example, if eight inverse tolerances of the physical path are determined, PR1 and PR2 have two differences, and PR3 has 30 and 28 differences for PR1 and PR2, respectively. Therefore, PR3 has more than 8 tolerances than PR1, which is the smallest inverse, so it can be excluded from the possible physical paths.
또한, 물리경로 상에 포함된 역의 수를 카운트할 때, 환승역의 가치가 정의될 필요가 있다. 환승역을 포함한 물리경로와 환승역을 포함하지 않은 물리경로가 혼합되어 있는 경우, 환승역을 통과하는 것은 상대적으로 시간이 많이 걸리므로, 환승역과 일반역을 1:1의 동일 가치로 평가하는 것은 합리적이지 않다. 그러므로, 예컨대 환승역 1개를 복수의 일반역(예컨대, 5개 일반역)과 동일한 가치를 가지는 것으로 간주할 수 있다.In addition, when counting the number of stations included on the physical path, the value of the transfer station needs to be defined. If there is a mixture of physical paths including transit stations and non-transit stations, passing through a transit station is relatively time consuming, so it is not reasonable to evaluate a transit station and a regular station at the same value of 1: 1. . Thus, for example, one transit station may be regarded as having the same value as a plurality of general stations (for example, five general stations).
이러한 두 가지 전제하에서, 예컨대 사당역(도 1의 M3, 2호선과 4호선의 환승역)에서 승차하여 교대역(도 1의 M4, 2호선과 3호선의 환승역)에서 하차한 승객의 가능한 물리 경로를 생성하면 다음과 같다. 먼저, 환승 횟수가 0번인 경로, 환승 횟수가 1번인 경로, 환승 횟수가 2번인 경로로 분리하여 물리경로를 생성할 수 있다. 환승 횟수가 0번인 경로는 사당역과 교대역 사이의 순환선인 2호선을 따라 이동한 것으로서, 내선순환경로 상의 PR1(역수=4), 외선순환경로 상의 PR2(역수=40)가 생성될 수 있다. 환승 횟수가 1회인 경로는 4호선과 3호선의 환승역인 충무로역(미도시)에서 환승한 경로 PR3(역수 = 일반역 18개 + 환승1회(5개역) = 23)가 생성될 수 있다. 환승 횟수가 2번인 경로는 4호선과 3호선 사이의 환승역들 2개를 조합할 수 있으므로 다수의 경로들 PR4, PR5, PR6, PR7이 생성될 수 있다. PR4는 4호선에서 7호선으로 1차 환승한 후 다시 6호선에서 3호선으로 2차 환승한 경로일 수 있으며, 일반역이 4개이고 환승역이 2개 이므로 총 14개의 역수를 가진다. 유사하게 계산하면 9호선을 경유한 PR5, 6호선을 경유한 PR6, 1호선을 경유한 PR7은 각각 16개, 26개, 35개의 역수를 가질 수 있다.Under these two premises, for example, a possible physical route for passengers who boarded at Sadang Station (transfer stations of M3, Line 2 and 4 in Fig. 1) and got off at the interchange (transfer stations of M4, Line 2 and 3 in Fig. 1) was created. Is as follows. First, a physical path may be generated by separating a path having a transfer number of 0, a path having a transfer number of 1, and a path having a transfer number of 2 times. The route with the number of transfers as 0 moves along Line 2, which is a circulation line between Sadang Station and the intersection, and may generate PR1 (reverse = 4) on the inward route environment path and PR2 (reverse = 40) on the outward route environment path. In the route having one transfer, a route PR3 (transverse = 18 general stations + one transfer (5 stations) = 23) transferred from Chungmuro Station (not shown), which is a transfer station of Line 4 and Line 3, may be generated. A route having two transfers may combine two transfer stations between Line 4 and Line 3, so that a plurality of routes PR4, PR5, PR6, and PR7 may be generated. PR4 may be the second transfer route from Line 4 to Line 7 and then transfer again from Line 6 to Line 3. Since there are four general stations and two transfer stations, it has a total of 14 stations. Similarly, PR5 via Line 9, PR6 via Line 6, and PR7 via Line 1 can have 16, 26, and 35 reciprocals, respectively.
그런 다음, 생성된 물리경로들 중에서 허용오차 8을 이용하여 체크하면, 가장 역수가 작은 경로 PR1(역수=4)에 비하여, 다른 경로들 PR2~PR7은 모두 8개보다 큰 역수를 가지고 있다는 것을 알 수 있다. 그러므로, PR2~PR7는 가능한 물리경로가 아니라고 결정할 수 있다.Then, if you check using tolerance 8 of the generated physical paths, you can see that all other paths PR2 ~ PR7 have more than 8 inverse of the smallest inverse path PR1 (reverse = 4). Can be. Therefore, it can be determined that PR2 to PR7 are not possible physical paths.
결과적으로, 사당역(M3)과 교대역(M4) 사이를 이동한 승객에 대해 생성되는 가능한 물리 경로는 PR1 하나만이 출력될 수 있다. 이에 따라, 이 승객의 이동경로는 환승이 없는 1개의 물리 경로만을 가지는 단순 선형경로라는 것도 결정될 수 있다.As a result, only one PR1 can be output as a possible physical path generated for the passenger who has moved between the Saddang station M3 and the inter-band M4. Accordingly, it can also be determined that this passenger's travel path is a simple linear path having only one physical path without a transfer.
이와 같이 1개 이상의 물리경로가 생성되면, 각각의 물리경로에 대해서, 가능한 논리경로가 생성될 수 있다(305). 논리경로는 승객이 승차한 승차 차량 및 승객이 하차한 하차 차량을 포함하는 경로이다. 승차 차량은 도 4를 참조하여 기술된 프로세스에 따라 생성된 참조 승차그룹에 기초하여 결정될 수 있다. 승차 차량은 2개 정도의 차량으로 추정될 수 있다. 유사하게, 하차 차량은 도 5를 참조하여 기술된 프로세스에 따라 생성된 참조 하차그룹에 기초하여 결정될 수 있다. 하차 차량은 본 발명에 따라 단일한 차량으로 추정될 수 있다. 승차 차량과 하차 차량이 서로 다른 환승의 경우에는, 도 6을 참조하여 기술된 프로세스에 따라 생성된 참조 환승그룹에 기초하여 환승 가능 여부가 체크될 수 있다. 이와 같은 과정에 의해, 각 승객의 물리적인 경로에서 가능한 논리경로가 생성될 수 있다.When one or more physical paths are generated in this way, for each physical path, possible logical paths may be generated (305). The logical path is a path including a passenger vehicle on which the passenger boards and a disembarked vehicle on which the passenger gets off. The riding vehicle may be determined based on the reference riding group created according to the process described with reference to FIG. The riding vehicle can be estimated to be about two vehicles. Similarly, the getting off vehicle may be determined based on the reference getting off group created according to the process described with reference to FIG. The getting off vehicle can be assumed to be a single vehicle in accordance with the present invention. In the case of a transfer where the riding vehicle and the discharging vehicle are different from each other, whether or not the transfer is possible may be checked based on the reference transfer group generated according to the process described with reference to FIG. 6. By this process, possible logical paths can be generated in the physical path of each passenger.
그런 다음, 복수의 가능한 논리경로들 중에서, 특히 승객의 게이트 진입/진출시간이 체크됨으로써 합리적으로 일관성 있는 논리경로가 하나만 존재하는 경우인지가 체크될 수 있다. 만약 일관성 있는 논리경로가 하나라면, 이 논리경로가 승객의 이동경로로서 결정될 수 있다(307). 만약 일관성 있는 논리경로가 두 개 이상일지라도 일관성 있는 논리경로가 모두 같은 물리적 경로이면, 물리적 경로가 결정될 수 있다.Then, among the plurality of possible logic paths, in particular the gate entry / exit time of the passenger can be checked to see if there is only one reasonably consistent logical path. If there is one coherent logical path, this logical path may be determined as the passenger's travel path (307). If there is more than one coherent logical path, if the coherent logical paths are all the same physical path, then the physical path can be determined.
도 7 및 도 8은 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 1번 및 2번의 환승을 포함하는 이동경로에서 가능 논리경로를 생성하는 과정을 설명하기 위한 개략도이다.7 and 8 are schematic diagrams for explaining a process of generating a possible logical path in a movement route including a first and second transfer in a method for estimating a passenger movement route of a train network according to an embodiment of the present invention.
도 7을 참조하면, 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 1번의 환승을 포함하는 이동경로에서 가능 논리경로를 생성하는 과정이 도시된다. 도시된 예에서, 승차역(O), 환승역(A), 하차역(D)를 물리경로로 가지는 승객 q에 있어서, 먼저 하차역 D의 진출 시간(Exit Time at D)에 기초하여, 미리 결정되어 있는 참조 하차그룹들 중에서, 승객의 진출시간에 대응하는 참조 하차그룹을 AG(Y, D)를 추출할 수 있다. 이 참조 하차그룹 AG(Y, D)는 Y 차량으로부터 D 역에서 하차한 승객들의 집합이다. 만약 AG(Y, D)의 승객들의 진출 시간들에 의해 정해지는 참조 진출 시간구간 안에 승객 q의 진출 시간이 위치된다면, AG(Y, D)는 승객 q의 진출 시간에 대응하는 참조 하차그룹이다. 따라서 승객 q는 AG(Y, D)의 하차 차량 Y에서 하차하였다고 안전하게 추정할 수 있다. 한편, 도시된 예에서, 승객 q의 승차역 O에서의 진입시간은 두 개의 승차 참조그룹 BG(X1, O) 및 BG(X2, O)에 대응하고 있다.Referring to FIG. 7, in the method of estimating a passenger movement route of a train network according to an embodiment of the present invention, a process of generating a possible logic route in a movement route including one transfer is illustrated. In the example shown, for a passenger q having a boarding station O, a transit station A, and a discharging station D as a physical path, first, based on the exit time at D, Among the reference discharging groups, AG (Y, D) may be extracted from the reference discharging group corresponding to the departure time of the passenger. This reference disengagement group AG (Y, D) is a set of passengers who get off at station D from vehicle Y. If the departure time of passenger q is located within the reference departure time interval defined by the departure times of the passengers of AG (Y, D), AG (Y, D) is the reference dropout group corresponding to the departure time of passenger q. . Therefore, passenger q can safely be estimated that he got off of vehicle Y of AG (Y, D). On the other hand, in the example shown, the entry time of the passenger q at the boarding station O corresponds to the two boarding reference groups BG (X1, O) and BG (X2, O).
이에 따라, 승객 q에 대해, X1 차량 또는 X2 차량 중 어느 한 차량에 승차한 후 A 역에서 Y 차량으로 환승한 다음에 D 역에서 하차하였다는 2개의 논리경로가 만들어질 수 있다. 환승역 A에서 X1, X2의 도착 시간과 Y 차량의 출발 시간이 정렬됨으로써, 환승 가능 여부는 TG(X1,Y,A)와 TG(X2,Y,A)의 존재여부가 체크된 다음에 결정될 수 있다.Accordingly, for the passenger q, two logical paths may be made, which are obtained either by the X1 vehicle or the X2 vehicle, then transferred from the A station to the Y vehicle and then disembarked from the D station. By arranging the arrival times of X1 and X2 at the transfer station A and the departure time of the vehicle Y, the possibility of transfer can be determined after the presence of TG (X1, Y, A) and TG (X2, Y, A) is checked. have.
도 8을 참조하면, 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 2번의 환승을 포함하는 이동경로에서 가능 논리경로를 생성하는 과정이 도시된다. 도시된 예에서, 승차역(O), 제1환승역(A), 제2환승역(B), 하차역(D)를 물리경로로 가지는 승객 q에 있어서, 먼저 하차역 D의 진출 시간(Exit Time at D)에 기초하여, 미리 결정되어 있는 참조 하차그룹들 중에서, 승객의 진출시간에 대응하는 참조 하차그룹을 AG(Z, D)를 추출할 수 있다. 승객 q는 AG(Z, D)의 하차 차량 Z에서 하차하였다고 안전하게 추정할 수 있다. 한편, 도시된 예에서, 승객 q의 승차역 O에서의 진입시간은 두 개의 승차 참조그룹 BG(X1, O) 및 BG(X2, O)에 대응하고 있다. 더 나아가, 승차 차량 X1, X2가 A역에 도착한 시간 이후에 출발한 열차들(Y1, Y2, …) 중에서, 하차역 D의 하차 차량 Z가 B역에서 출발한 시간 이전에 B역에 도착한 열차들은 Y1, Y2라는 것을 열차 운행 정보로부터 알 수 있다.Referring to FIG. 8, in the method of estimating a passenger movement route of a train network according to an embodiment of the present invention, a process of generating a possible logic route in a movement route including two transfers is shown. In the example shown, for a passenger q having a boarding station (O), a first transfer station (A), a second transfer station (B), and an unloading station (D) as a physical path, first the exit time of the stop station D (Exit Time) Based on at D), AG (Z, D) may be extracted from the reference disengagement groups determined in advance, and the reference disembarkation group corresponding to the departure time of the passenger. Passenger q can safely assume that he got off of vehicle Z of AG (Z, D). On the other hand, in the example shown, the entry time of the passenger q at the boarding station O corresponds to the two boarding reference groups BG (X1, O) and BG (X2, O). Furthermore, among the trains (Y1, Y2, ...) departing after the time when the riding vehicles X1, X2 arrived at station A, the train arriving at station B before the time when the disembarkation vehicle Z at the station D departed from station B It can be seen from the train operation information that they are Y1 and Y2.
이에 따라, 승객 q에 대해, X1 차량 또는 X2 차량 중 어느 한 차량에 승차한 후 A 역에서 Y1 차량 또는 Y2 차량으로 환승한 다음에, B역에서 Z 차량으로 환승하고, D 역에서 하차하였다는 3개의 논리경로(X1-Y1-Z; X1-Y2-Z; X2-Y2-Z)가 만들어질 수 있다. 3개의 논리경로는 환승역 A에서 X1, X2의 도착 시간과 Y1, Y2 차량의 출발 시간이 정렬되고, 환승역 B에서 Y1, Y2 차량의 도착 시간과 Z 차량의 출발시간이 정렬됨으로써, 환승 가능 여부는 TG(X1,Y1,A), TG(X1,Y2,A), TG(X2,Y2,A), TG(Y1,Z,B), TG(Y2,Z,B)의 존재여부가 체크된 다음에 결정된다.Accordingly, for passenger q, after boarding either vehicle X1 or vehicle X2, transfer to Y1 or Y2 at station A, then transfer to vehicle Z at station B, and get off at station D. Three logical paths (X1-Y1-Z; X1-Y2-Z; X2-Y2-Z) can be made. The three logical routes align the arrival times of X1 and X2 at the transit station A with the departure times of the Y1 and Y2 vehicles, and the arrival times of the Y1 and Y2 vehicles at the transit station B and the departure times of the Z vehicles at the transfer station B. TG (X1, Y1, A), TG (X1, Y2, A), TG (X2, Y2, A), TG (Y1, Z, B), TG (Y2, Z, B) It is decided next.
도 9를 참조하면, 도 2에 도시된 바와 같은 전철 네트워크 상에서, 본 발명의 일 실시예에 따른 전철 네트워크의 승객 이동경로 추정 방법에서, 1번의 환승을 포함하는 2개의 가능 물리경로를 가지는 2명의 승객 p, q에 대해 각각 이동경로를 결정하는 과정을 설명하기 위한 개략도가 도시된다.Referring to FIG. 9, in the method for estimating the passenger movement path of a train network according to an embodiment of the present invention, the train network as shown in FIG. 2 includes two persons having two possible physical paths including one transfer. A schematic diagram is shown for explaining the process of determining the movement route for the passengers p and q, respectively.
도시된 예에서, 승객 p에 대해 승차역 = 신림역, 게이트 진입 시간 = 27,227 (초), 하차역 = 가락시장역, 게이트 진출 시간 = 29,813 (초)의 승하차정보가 주어진다. 여기서 진입 시간 및 진출 시간은 하루 중 특정 시점을 새벽 0시부터 초 단위로 카운트한 숫자로 표시되어 있다. 한편 승객 q에 대해서는 승차역과 하차역이 승객 p와 동일하고, 게이트 진입/진출 시간이 유사하게 주어진다. 즉 승객 q에 대해 승차역 = 신림역, 게이트 진입 시간 = 27,295 (초), 하차역 = 가락시장역, 게이트 진출 시간 = 29,991 (초)의 승하차정보가 주어진다.In the example shown, the passengers p are given boarding information of boarding station = Sillim station, gate entry time = 27,227 (seconds), getting off station = Garak market station, gate entrance time = 29,813 (seconds). In this case, the entry time and the exit time are expressed as numbers counting a specific time of day from 0 am to seconds. On the other hand, for the passenger q, the boarding station and the unloading station are the same as the passenger p, and the gate entry / exit time is similarly given. That is, passengers q are given the getting on and off information of boarding station = Sillim station, gate entry time = 27,295 (sec), disembarkation station = Garak market station, and gate advance time = 29,991 (sec).
이와 같이 승객 p, q는 동일한 승차역(도 2의 M2) 및 하차역(도 2의 M6)을 가지며, 인터-게이트 시간도 유사하기 때문에, 물리적인 경로를 인터-게이트 시간만으로 확정하기 어렵다. 즉, 승객 p, q는 둘 모두, 인터-게이트 시간에 기초하면, 잠실역에서 환승하는 경로 R1과 교대역에서 환승하는 경로 R2를 가질 수 있다. 경로 R1과 R2는 소요 시간이 유사하기 때문에 인터-게이트 시간만으로는 구별하기 어렵다.As such, the passengers p and q have the same boarding station (M2 in FIG. 2) and the getting off station (M6 in FIG. 2), and the inter-gate time is also similar, so that it is difficult to determine the physical path only with the inter-gate time. That is, the passengers p and q may both have a path R1 translating in Jamsil station and a path R2 translating in the interband based on the inter-gate time. Because paths R1 and R2 have similar times, the inter-gate time alone is difficult to distinguish.
이에 비하여, 본 발명에 따르면, 승객 p, q에 대하여 먼저, 가능 물리경로가 생성된다. 즉 경로 R1 = M2-M5-M6, 경로 R2 = M2-M4-M6 가 생성될 수 있다. 이 가능 물리경로는 도 8에서 왼쪽에 위 아래로 도시되어 있다. 위에는 2호선의 신림역(M2)에서 출발하여 교대역(M4)에서 3호선으로 환승한 후 가락시장역(M6)에 도착하는 경로(R2)이며, 아래쪽에는 2호선의 신림역(M2)에서 출발하여 잠실역(M5)에서 8호선으로 환승한 후 가락시장역(M6)에서 도착하는 경로(R1)가 도시된다.In contrast, according to the present invention, a possible physical path is first generated for passengers p and q. That is, paths R1 = M2-M5-M6 and paths R2 = M2-M4-M6 may be generated. This possible physical path is shown up and down on the left in FIG. 8. Above is a route (R2) starting at Sinrim Station (M2) on Line 2 and transferring to Line 3 at Miao-Bang (M4), and arriving at Garak Market Station (M6), and from Jamlim Station (M2) on Line 2 below. The route R1 arriving at the Garak market station M6 after the transfer to the line 8 in M5) is shown.
먼저, 승객 p를 보면, 교대역(M4)을 경유하는 R2에 대해서 참조 하차그룹을 찾는다. 즉, 승객 p의 게이트 진출 시간 29,813을 포함하는 참조 진출 시간구간 [29,524; 29,874]을 가지는 참조 하차그룹 AG(Y1, 가락시장역)을 발견한다. 참조 진출 시간구간의 첫 번째 숫자는 참조 하차그룹의 승객들의 게이트 진출 시간들 중 가장 빠른 시간 값일 수 있다. 그리고 참조 진출 시간구간의 두 번째 숫자는 참조 하차그룹의 승객들의 게이트 진출 시간들 중 가장 느린 시간 값일 수 있다. 발견된 참조 하차그룹의 승차 차량 Y1은 승객 p의 하차 차량이라고 결정된다. 반면에, 잠실역을 경유하는 R1에 대해서는 대응 참조 하차그룹이 발견되지 않았으므로, 경로 R1은 승객 p가 이용하지 않은 경로라고 결정될 수 있다.First, looking at the passenger p, the reference drop-off group is found for R2 via the alternating band M4. That is, the reference departure time interval including the gate departure time 29,813 of the passenger p [29,524; 29,874] and find the reference drop-off group AG (Y1, Garak Market Station). The first number of reference entry time intervals may be the earliest time value among gate entry times of the passengers in the reference dropoff group. The second number of reference entry time intervals may be the slowest time value among gate entry times of the passengers in the reference drop-off group. It is determined that the riding vehicle Y1 of the found reference getting off group is the getting off vehicle of the passenger p. On the other hand, for R1 via Jamsil station, since no corresponding reference drop group was found, the route R1 may be determined to be a route not used by the passenger p.
그런 다음, 승객 p에 대해 p의 게이트 진입 시간 27,227을 포함하는 참조 진입 시간구간[27,090; 27,237]을 가지는 참조 승차그룹 BG(X1, 신림역)을 발견한다. 참조 진입 시간구간의 첫 번째 숫자는 참조 승차그룹의 승객들의 게이트 진입 시간들 중 가장 빠른 시간 값일 수 있다. 그리고 참조 진입 시간구간의 두 번째 숫자는 참조 승차그룹의 승객들의 게이트 진입 시간들 중 가장 느린 시간 값일 수 있다. 발견된 대응 참조 승차그룹의 승차 차량 X1은 승객 p의 승차 차량이라고 결정될 수 있다.Then, the reference entry time interval 27,090 including the gate entry time 27,227 of p for the passenger p; 27, 237] and find the reference riding group BG (X1, Sillim Station). The first number of reference entry time periods may be the earliest time value among gate entry times of the passengers of the reference riding group. The second number of reference entry time intervals may be the slowest time value among gate entry times of the passengers of the reference riding group. It may be determined that the riding vehicle X1 of the found corresponding reference riding group is the riding vehicle of the passenger p.
그리고, 논리경로 X1-Y1에 대해 환승역 교대역에 대해 미리 생성되어 있는 참조 환승그룹(X1, Y1, 교대역)이 존재하는 것이 체크됨으로써, 경로 R2는 승객 p에 의해 이용된 이동경로라고 결정될 수 있다.Then, it is checked that the reference transfer groups (X1, Y1, alternating bands) previously generated for the transfer station alternate bands for the logical paths X1-Y1 exist, so that the route R2 can be determined to be the movement route used by the passenger p.
한편, 승객 q를 보면, 잠실역을 경유하는 R1에 대해서 승객 q의 게이트 진출 시간 29,991을 포함하는 참조 진출 시간구간 [29,975; 30,162]을 가지는 참조 하차그룹 AG(Z2, 가락시장역)을 발견한다. 그리고 발견된 참조 하차그룹에 의해 승객 q의 하차 차량 Z2가 결정된다. 반면에, 교대역을 경유하는 R2에 대해서는 대응 참조 하차그룹이 발견되지 않았으므로, 경로 R2는 승객 q가 이용하지 않은 경로라고 결정될 수 있다.On the other hand, when looking at the passenger q, the reference departure time interval including the gate departure time 29,991 of the passenger q for R1 via Jamsil station [29,975; 30, 162] and find the reference drop-off group AG (Z2, Garak Market Station). And the getting off vehicle Z2 of the passenger q is determined by the found reference getting off group. On the other hand, since no corresponding reference drop group was found for R2 via the interband, route R2 may be determined to be a route that passenger q did not use.
그런 다음, 승객 q에 대해 q의 게이트 진입 시간 27,295을 포함하는 참조 진입 시간구간[27,231; 27,411]을 가지는 참조 승차그룹 BG(X2, 신림역)을 발견한다. 그리고 발견된 참조 승차그룹에 의해 승객 q의 승차 차량 X2이 결정될 수 있다.Then, a reference entry time period 27,231 comprising a gate entry time 27,295 of q for passenger q; 27,411] and find the reference riding group BG (X2, Sillim Station). And the riding vehicle X2 of passenger q can be determined by the found reference riding group.
그리고, 논리경로 X2-Z2에 대해 환승역 잠실역에 대해 미리 생성되어 있는 참조 환승그룹(X2, Z2, 잠실역)이 존재하는 것이 체크됨으로써, 경로 R1는 승객 q에 의해 이용된 이동경로라고 결정될 수 있다.Then, by checking that there exists a reference transfer group (X2, Z2, Jamsil station) previously generated for the transfer station Jamsil station for the logical paths X2-Z2, the path R1 may be determined to be the movement route used by the passenger q.
상술한 바와 같이, 본 발명에 따른 방법에 의하면, 승객의 이동 경로를 탑승한 차량 정보를 포함하여 정확하게 추정할 수 있다. 이에 따라 통합 요금 징수 체계하에서, 전철 네트워크 상의 서로 다른 운영주체들 사이에 요금을 배분하기 위한 승객 이동경로 추정 분야에서 이용될 수 있다. 또한 본 발명에 따라 각 승객이 이용하는 차량이 추정될 수 있기 때문에, 승객 이용 형태를 분석하는 자료 데이터를 제공할 수 있다는 장점을 제공한다.As described above, according to the method according to the present invention, it is possible to accurately estimate the vehicle traveling path including the vehicle information. Accordingly, under the integrated fare collection system, it can be used in the field of passenger travel path estimation for distributing fare among different operators on the train network. In addition, according to the present invention, since the vehicle used by each passenger can be estimated, it provides an advantage that it is possible to provide data data analyzing the passenger use pattern.
본 방법은, 프로세서 및 메모리 등을 포함할 수 있는 컴퓨팅 장치에 의해 구현될 수 있다. 메모리는, 프로세서에 의해 실행되면 특정 태스크를 수행할 있도록 코딩되어 있는 컴퓨터 실행가능 소프트웨어, 애플리케이션, 프로그램 모듈, 루틴, 인스트럭션(instructions), 및/또는 데이터 등을 저장하는 컴퓨터 판독가능 스토리지 매체이다. 프로세서는 메모리로부터, 프로세서에 의해 실행되면 상술한 본 발명의 방법을 수행하도록 구성된, 컴퓨터 실행가능 소프트웨어, 애플리케이션, 프로그램 모듈, 루틴, 인스트럭션, 및/또는 데이터 등을 판독하여 실행할 수 있다. 상술한 본 발명의 방법은, 컴퓨터 실행가능 인스트럭션으로서 코딩되어 컴퓨팅 장치의 프로세서에 의해 실행됨으로써 구현될 수 있다. 컴퓨터 실행가능 인스트럭션은 소프트웨어, 애플리케이션, 모듈, 프로시저, 플러그인, 프로그램, 인스트럭션, 및/또는 데이터 구조 등을 포함할 수 있다. 컴퓨팅 장치는 웨어러블 컴퓨팅 장치, 핸드-헬드 컴퓨팅 장치, 스마트폰, 타블렛, 랩탑, 데스크탑, 개인용 컴퓨터, 서버 등의 다양한 장치를 포함할 수 있다. 컴퓨팅 장치는 스탠드-얼론 타입의 장치일 수 있다. 컴퓨팅 장치는 통신망을 통하여 서로 협력하는 다수의 컴퓨팅 장치들을 포함할 수 있다.The method may be implemented by a computing device that may include a processor, a memory, and the like. A memory is a computer readable storage medium that stores computer executable software, applications, program modules, routines, instructions, and / or data that are coded to be executed by a processor to perform a particular task. The processor may read and execute from the memory, computer executable software, applications, program modules, routines, instructions, and / or data, which, when executed by the processor, are configured to perform the methods of the present invention described above. The above-described methods of the present invention may be implemented by being coded as computer executable instructions and executed by a processor of a computing device. Computer-executable instructions may include software, applications, modules, procedures, plug-ins, programs, instructions, and / or data structures, and the like. Computing devices may include various devices, such as wearable computing devices, hand-held computing devices, smartphones, tablets, laptops, desktops, personal computers, servers, and the like. The computing device may be a stand-alone type device. The computing device may include a number of computing devices that cooperate with each other via a communication network.

Claims (9)

  1. 복수의 노선을 따라 복수의 차량이 운행되는 전철 네트워크에서 승객의 이동경로를 추정하는 방법으로서,A method of estimating a travel route of a passenger in a train network in which a plurality of vehicles operate along a plurality of routes,
    복수의 승객의 승하차정보를 포함하는 승객 데이터로부터 각 승객이 탑승한 열차를 추정하여 참조그룹을 생성하는 단계;Generating a reference group by estimating trains on which each passenger boards from passenger data including getting on and off information of the plurality of passengers;
    각 승객의 승하차정보로부터 각 승객의 승차역 및 하차역 사이에서 전철 노선 정보에 의해 이루어지는 적어도 하나의 가능 물리경로를 생성하는 단계;Generating at least one possible physical route made by the train line information between each passenger's boarding and unloading station from each passenger's getting on and off information;
    상기 가능 물리경로 각각에 대하여, 상기 참조그룹에 기초하여, 각 전철 노선 상에서 운행하는 차량이 연결되어 이루어지는 적어도 하나의 논리경로를 생성하는 단계; 및Generating at least one logical path for each of the possible physical paths, based on the reference group, by which a vehicle running on each train line is connected; And
    상기 논리경로의 일관성을 상기 참조그룹에 기초하여 체크하여 승객의 이동경로를 결정하는 단계Checking the consistency of the logical path based on the reference group to determine a passenger's movement path
    를 포함하는, 전철 네트워크 내 승객 이동경로 추정 방법.A method of estimating a passenger's movement route in a train network.
  2. 제 1 항에 있어서,The method of claim 1,
    상기 참조그룹을 생성하는 단계는: The step of creating the reference group is:
    특정 승차역에서 특정한 제1차량에 승차한 승객들을 추정하여 참조 승차그룹을 생성하는 단계;Generating a reference boarding group by estimating passengers boarding a specific first vehicle at a specific boarding station;
    특정 하차역에서 특정한 제2차량으로부터 하차한 승객들을 추정하여 참조 하차그룹을 생성하는 단계;Estimating passengers who get off from a specific second vehicle at a specific getting off station and generating a reference dropping group;
    특정 환승역에서 특정한 제3차량으로부터 특정한 제4차량으로 환승한 승객들을 추정하여 참조 환승그룹을 생성하는 단계를 Generating a reference transfer group by estimating passengers transferred from a specific third vehicle to a specific fourth vehicle at a specific transfer station;
    포함하는, 전철 네트워크 내 승객 이동경로 추정 방법.A method of estimating passenger movement paths in a train network, the method comprising: a.
  3. 제 2 항에 있어서,The method of claim 2,
    상기 참조 하차그룹을 생성하는 단계는:The step of creating the reference drop group is:
    상기 승객 데이터로부터, 가능 물리경로가 1개이며 환승이 없는 승객을 추출하는 단계;Extracting passengers having one possible physical path and no transfer from the passenger data;
    추출된 승객에 대해, 각 승차역의 게이트 진입시간 이후에 출발한 차량들과 각 하차역의 게이트 진출시간 이전에 도착한 차량들을 비교하여, 동일한 차량을 오직 1개만 포함하는 승객을 선별하는 단계; 및Selecting, for the extracted passengers, passengers including only one of the same vehicle by comparing vehicles departing after the gate entry time of each boarding station with vehicles arriving before the gate entry time of each boarding station; And
    선별된 승객들이 상기 오직 1개만 포함하는 차량에 탑승한 것으로 추정하고, 상기 선별된 승객들 중에서 특정 하차역을 가지는 승객들을 이 탑승한 것으로 추정된 차량으로부터 상기 특정 하차역에서 하차한 승객들로서 분류하여 각각 참조 하차그룹으로서 생성하는 단계It is assumed that the selected passengers have boarded the vehicle including only one, and the passengers having a specific stop station among the selected passengers are classified as the passengers who get off at the specific stop station from the vehicle estimated to have boarded the vehicle. Generating as reference reference groups, respectively
    를 포함하는, 전철 네트워크 내 승객 이동경로 추정 방법. A method of estimating a passenger's movement route in a train network.
  4. 제 2 항에 있어서,The method of claim 2,
    상기 참조 승차그룹을 생성하는 단계는:Creating the reference ride group:
    상기 승객 데이터로부터, 가능 물리경로가 1개이며 환승이 없는 승객을 추출하는 단계;Extracting passengers having one possible physical path and no transfer from the passenger data;
    추출된 승객에 대해, 각 승차역의 게이트 진입시간 이후에 출발한 차량들과 각 하차역의 게이트 진출시간 이전에 도착한 차량들을 비교하여, 동일한 차량을 오직 1개만 포함하는 승객을 선별하는 단계; 및Selecting, for the extracted passengers, passengers including only one of the same vehicle by comparing vehicles departing after the gate entry time of each boarding station with vehicles arriving before the gate entry time of each boarding station; And
    선별된 승객들이 상기 오직 1개만 포함하는 차량에 탑승한 것으로 추정하고, 상기 선별된 승객들 중에서 특정 승차역을 가지는 승객들을 이 탑승한 것으로 추정된 차량에 상기 특정 승차역에서 승차한 승객들로서 분류하여 각각 참조 승차그룹으로서 생성하는 단계It is assumed that the selected passengers have boarded the vehicle including only one, and the passengers having a specific boarding station among the selected passengers are classified as passengers who boarded at the specific boarding station to the vehicle estimated to be boarded. Creating each as a reference ride group
    를 포함하는, 전철 네트워크 내 승객 이동경로 추정 방법.A method of estimating a passenger's movement route in a train network.
  5. 제 2 항에 있어서,The method of claim 2,
    상기 참조 환승그룹을 생성하는 단계는:Creating the reference transit group may include:
    상기 승객 데이터로부터, 가능 물리경로가 1개이며 1개의 환승역을 가지는 승객을 추출하는 단계;Extracting, from the passenger data, passengers having one possible physical path and having one transit station;
    추출된 승객 각각에 대해, 상기 참조 승차그룹들 중에서, 각 승차역의 게이트 진입시간을 포함하는 참조 진입시간구간을 가지는 대응 참조 승차그룹을 추출하고, 이 대응 참조 승차그룹의 승차 차량에 상기 승차역에서 승차한 것으로 추정하는 단계; For each of the extracted passengers, a corresponding reference riding group having a reference entry time interval including a gate entry time of each boarding station is extracted from the reference riding groups, and the boarding station is assigned to the boarding vehicle of the corresponding reference riding group. Estimating that the passenger rides in the vehicle;
    추출된 승객 각각에 대해, 상기 참조 하차그룹들 중에서, 각 하차역의 게이트 진출시간을 포함하는 참조 진출시간구간을 가지는 대응 참조 하차그룹을 추출하고, 이 대응 참조 하차그룹의 하차 차량에 상기 하차역에서 하차한 것으로 추정하는 단계;For each of the extracted passengers, a corresponding reference disembarkation group having a reference entry time period including a gate entry time of each discharging station is extracted from the reference discharging groups, and the disembarkation station is displayed on the getting-off vehicle of the corresponding reference discharging group. Estimating to be off;
    추출된 승객 각각에 대해, 상기 추정된 승차 차량이 상기 환승역에 도착한 시간 및 상기 추정된 하차 차량이 상기 환승역에서 출발한 시간에 기초하여 환승여부를 체크하는 단계;For each of the extracted passengers, checking whether the estimated riding vehicle arrived at the transit station and whether the estimated getting off vehicle departed from the transit station;
    상기 환승여부를 체크하는 단계에서 환승 가능하다고 판정된 경우를, 상기 추정된 승차 차량으로부터 상기 환승역에서 상기 추정된 하차 차량으로 환승한 참조 환승그룹으로서 생성하는 단계Generating a case where it is determined that transfer is possible in the checking of the transfer, as a reference transfer group transferred from the estimated riding vehicle to the estimated getting off vehicle at the transfer station.
    를 포함하는 전철 네트워크 내 승객 이동경로 추정 방법.Passenger movement path estimation method in a train network comprising a.
  6. 제 1 항에 있어서,The method of claim 1,
    상기 논리경로를 생성하는 단계는:Generating the logical path includes:
    각 승객에 대해 하차역에서 하차한 차량을 오직 하나로 결정하는, 전철 네트워크 내 승객 이동경로 추정 방법.A method of estimating a passenger's movement route in a train network that determines, for each passenger, only one vehicle to be disembarked at the discharging station.
  7. 제 1 항에 있어서, The method of claim 1,
    상기 논리경로를 생성하는 단계는:Generating the logical path includes:
    각 승객에 대해 승차역에서 승차한 차량을 1개 또는 2개로 결정하는, 전철 네트워크 내 승객 이동경로 추정 방법. A method for estimating passenger travel routes in a train network, for each passenger, determining one or two vehicles boarded at a boarding station.
  8. 제 1 항에 있어서, The method of claim 1,
    상기 논리경로를 생성하는 단계는:Generating the logical path includes:
    각 승객에 대해 환승역에서 환승한 차량을 1개 또는 2개로 결정하는, 전철 네트워크 내 승객 이동경로 추정 방법.A method for estimating passenger travel routes in a train network, which determines, for each passenger, one or two vehicles transferred at a transit station.
  9. 컴퓨팅 장치의 프로세서에 의해 실행되면 특정 태스크를 수행하도록 코딩된 컴퓨터 실행가능 인스트럭션을 포함하는 컴퓨터 판독가능 스토리지 매체로서,A computer readable storage medium comprising computer executable instructions coded to execute a particular task when executed by a processor of a computing device,
    상기 컴퓨터 실행가능 인스트럭션은 상기 프로세서에 의해 실행될 때, 복수의 노선을 따라 복수의 차량이 운행되는 전철 네트워크에서 승객의 이동경로를 추정하는 방법을 수행하도록 구성되고, 상기 방법은:The computer executable instructions, when executed by the processor, are configured to perform a method of estimating a travel route of a passenger in a train network in which a plurality of vehicles travel along a plurality of routes, the method comprising:
    복수의 승객의 승하차정보를 포함하는 승객 데이터로부터 각 승객이 탑승한 열차를 추정하여 참조그룹을 생성하는 단계;Generating a reference group by estimating trains on which each passenger boards from passenger data including getting on and off information of the plurality of passengers;
    각 승객의 승하차정보로부터 각 승객의 승차역 및 하차역 사이에서 전철 노선 정보에 의해 이루어지는 적어도 하나의 가능 물리경로를 생성하는 단계;Generating at least one possible physical route made by the train line information between each passenger's boarding and unloading station from each passenger's getting on and off information;
    상기 가능 물리경로 각각에 대하여, 상기 참조그룹에 기초하여, 각 전철 노선 상에서 운행하는 차량이 연결되어 이루어지는 적어도 하나의 논리경로를 생성하는 단계; 및Generating at least one logical path for each of the possible physical paths, based on the reference group, by which a vehicle running on each train line is connected; And
    상기 논리경로의 일관성을 상기 참조그룹에 기초하여 체크하여 승객의 이동경로를 결정하는 단계Checking the consistency of the logical path based on the reference group to determine a passenger's movement path
    를 포함하는, 컴퓨터 판독가능 스토리지 매체.And a computer readable storage medium.
PCT/KR2015/004069 2014-05-02 2015-04-23 Method for estimating passenger moving path in subway network WO2015167169A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR1020140053590A KR101582283B1 (en) 2014-05-02 2014-05-02 Method for estimating route of passenger in a metro network
KR10-2014-0053590 2014-05-02

Publications (1)

Publication Number Publication Date
WO2015167169A1 true WO2015167169A1 (en) 2015-11-05

Family

ID=54358843

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2015/004069 WO2015167169A1 (en) 2014-05-02 2015-04-23 Method for estimating passenger moving path in subway network

Country Status (2)

Country Link
KR (1) KR101582283B1 (en)
WO (1) WO2015167169A1 (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3690843A4 (en) * 2017-09-29 2020-08-12 Fujitsu Limited Estimation program, estimation device, and estimation method
CN111584091A (en) * 2020-04-29 2020-08-25 北京交通大学 Method and device for identifying cross infection risk of urban rail line level close contact person
CN112101640A (en) * 2020-08-28 2020-12-18 通号城市轨道交通技术有限公司 Passenger path determining method and device and readable storage medium
CN112347596A (en) * 2020-11-05 2021-02-09 浙江非线数联科技有限公司 Urban public transport network optimization method
CN112966218A (en) * 2021-02-26 2021-06-15 佳都新太科技股份有限公司 Real-time calculation method and device for number of passengers in carriage
CN114723163A (en) * 2022-04-26 2022-07-08 深圳市城市交通规划设计研究中心股份有限公司 Time-sharing time-space passenger flow distribution method and system, electronic equipment and storage medium
CN114866969A (en) * 2022-04-12 2022-08-05 江苏欣网视讯软件技术有限公司 Subway clearing method and system based on mobile phone signaling big data
CN116933958A (en) * 2023-07-18 2023-10-24 西南交通大学 Subway transfer station transfer passenger flow estimation method based on card swiping data

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR102074894B1 (en) * 2019-05-08 2020-03-18 서울대학교산학협력단 Estimation Method for Traveled Routes of Transit Users, and Its Program Implementation and Performance System

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005234710A (en) * 2004-02-17 2005-09-02 Sharp Corp Charge adjustment system and charge adjustment device
KR20090128309A (en) * 2008-10-06 2009-12-15 김도한 Subway profit dividing system and method, subway passing control system and method, bus boarding control system and method based on rfid
JP2013241109A (en) * 2012-05-21 2013-12-05 Ntt Docomo Inc Server device, notification method, and program

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005234710A (en) * 2004-02-17 2005-09-02 Sharp Corp Charge adjustment system and charge adjustment device
KR20090128309A (en) * 2008-10-06 2009-12-15 김도한 Subway profit dividing system and method, subway passing control system and method, bus boarding control system and method based on rfid
JP2013241109A (en) * 2012-05-21 2013-12-05 Ntt Docomo Inc Server device, notification method, and program

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
LEE, MI YEONG ET AL.: "A Link-Based Label Correcting Multi-Objective Shortest Paths Algorithm in Multi-Modal Transit Networks", JOURNAL OF KOREAN SOCIETY OF TRANSPORTATION, vol. 28, no. 2, February 2008 (2008-02-01), XP055234173 *

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3690843A4 (en) * 2017-09-29 2020-08-12 Fujitsu Limited Estimation program, estimation device, and estimation method
US11462102B2 (en) 2017-09-29 2022-10-04 Fujitsu Limited Storage medium, estimation device, and estimation method
CN111584091A (en) * 2020-04-29 2020-08-25 北京交通大学 Method and device for identifying cross infection risk of urban rail line level close contact person
CN111584091B (en) * 2020-04-29 2023-10-24 北京交通大学 Cross infection risk identification method and device for urban rail line-level close contact person
CN112101640A (en) * 2020-08-28 2020-12-18 通号城市轨道交通技术有限公司 Passenger path determining method and device and readable storage medium
CN112347596A (en) * 2020-11-05 2021-02-09 浙江非线数联科技有限公司 Urban public transport network optimization method
CN112966218A (en) * 2021-02-26 2021-06-15 佳都新太科技股份有限公司 Real-time calculation method and device for number of passengers in carriage
CN114866969A (en) * 2022-04-12 2022-08-05 江苏欣网视讯软件技术有限公司 Subway clearing method and system based on mobile phone signaling big data
CN114866969B (en) * 2022-04-12 2023-08-01 江苏欣网视讯软件技术有限公司 Metro sorting and calculating method and system based on mobile phone signaling big data
CN114723163A (en) * 2022-04-26 2022-07-08 深圳市城市交通规划设计研究中心股份有限公司 Time-sharing time-space passenger flow distribution method and system, electronic equipment and storage medium
CN116933958A (en) * 2023-07-18 2023-10-24 西南交通大学 Subway transfer station transfer passenger flow estimation method based on card swiping data
CN116933958B (en) * 2023-07-18 2024-03-29 西南交通大学 Subway transfer station transfer passenger flow estimation method based on card swiping data

Also Published As

Publication number Publication date
KR20150126212A (en) 2015-11-11
KR101582283B1 (en) 2016-01-04

Similar Documents

Publication Publication Date Title
WO2015167169A1 (en) Method for estimating passenger moving path in subway network
Zhu et al. A probabilistic passenger-to-train assignment model based on automated data
Zhao et al. Estimation of passenger route choice pattern using smart card data for complex metro systems
CN107886723B (en) Traffic travel survey data processing method
Kusakabe et al. Estimation method for railway passengers’ train choice behavior with smart card transaction data
US20150286936A1 (en) Transportation analysis system
CN201936369U (en) Accurate urban rail transit passenger flow statistics system
JP6875311B2 (en) Transportation congestion prediction system and congestion prediction method
CN107590239B (en) Method for measuring connection radius of public bicycle at subway station based on IC card data
CN105335795B (en) A kind of subway Public Transport Transfer problem automatic diagnosis method based on IC card data
JP6454222B2 (en) Data processing system and data processing method
CN109903555B (en) Bus passenger getting-off data prediction method and system based on big data
CN109872536A (en) A kind of passenger flow guidance system and its control method
CN107578619B (en) Method for measuring public bicycle service range of subway station based on IC card data
CN103985265B (en) Bus priority control method and system for avoiding bus clustering
Zhao et al. Isolating high-priority metro and feeder bus transfers using smart card data
CN109816085A (en) A kind of metro passenger flow information real-time publishing system and method based on outpost display
CN111126689A (en) Subway station passenger flow line management and control method, device, equipment and storage medium
Godachevich et al. Does the measured performance of bus operators depend on the index chosen to assess reliability in contracts? An analysis of bus headway variability
CN108269400A (en) A kind of trunk roads signalized crossing delay evaluation method and device
WO2023027361A1 (en) Method and apparatus for determining movement path of vehicle in consideration of movement flow of passengers
WO2015186875A1 (en) System and method for providing service for public transportation means guidance, device for same, and recording medium having computer program recorded therein
CN110738768A (en) method for identifying ticket of multiple rail transit standards based on human face
JP7425680B2 (en) Navigation device and navigation method
CN113657725A (en) Bus route and scheduling optimization method, system, device and medium

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15785432

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15785432

Country of ref document: EP

Kind code of ref document: A1