CN109035770A - The real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment - Google Patents
The real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment Download PDFInfo
- Publication number
- CN109035770A CN109035770A CN201810860244.5A CN201810860244A CN109035770A CN 109035770 A CN109035770 A CN 109035770A CN 201810860244 A CN201810860244 A CN 201810860244A CN 109035770 A CN109035770 A CN 109035770A
- Authority
- CN
- China
- Prior art keywords
- bus
- card
- public
- icid
- data
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
Landscapes
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
- Devices For Checking Fares Or Tickets At Control Points (AREA)
Abstract
Public bus network time sharing segment is obtained the road section where public transit vehicle different moments using the space-time trajectory that public transport vehicle-mounted GPS data extracts bus by the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment provided by the invention;The time data for obtaining user and taking bus are recorded using bus card-reading, obtain station data after matching with the GPS data of bus, and the travel behaviour feature that individual takes bus is extracted with this;Space clustering is carried out to the website of public bus network, it would be possible to transfer point merge, prevent the difference due to site name and transfer information can not be extracted;According to the card-swiping mode of different public bus networks, statistics calculates individual and takes the probability of a certain public bus network in a certain period, and the probability distribution that each website is got off along the line after getting on the bus;When statistical in each public bus network of pavement branch sections the carrying situation of bus historical data, and it is predicted.
Description
Technical field
The present invention relates to one kind based on magnanimity encryption public transport barcode scanning or to swipe the card (hereinafter, IC card is swiped the card, intelligent terminal sweeps two
Dimension code, NFC such as swipe the card at all kinds of means of payment of taking pubic transport, and are referred to as swiping the card) at times-section bus carryings of record data
Prediction technique is measured, the real-time spatial position of bus is obtained according to the GPS data of bus, is extracted according to bus card-reading record a
Body takes bus the basic act feature of trip, and therefrom building individual takes the moment of probability distribution of certain public bus network at times
Battle array, the get-off stop for taking certain public bus network to individual are predicted, are taken bus possible O-D points to obtain individual
Cloth, thus obtains at times-section bus passenger capacity and its overload situations, provides number for the scheduling and optimization of public bus network
According to support.
Background technique
With the quickening of urbanization process, public transport is heavy between node as maintaining in city hinge-network structure
Traffic delivery means are wanted, are that one of most important selection of city dweller's daily trip and city maintain normal daily operation
Important leverage.As the continuous expansion of area is built up in city, public transit system becomes to become increasingly complex, the passenger flow pressure of carrying
Also increasing, and this pressure changes over time obviously.Among one day, peak time morning and evening is public transit system passenger flow pressure
Peak, and remaining period passenger flow pressure is relatively small;On long terms, with the continuous expansion in city, script passenger compared with
Few public bus network also can with a large amount of moving into for population pressure multiplier.Therefore, the short-term and long-term change of bus passenger flow demand
Change and shift to public bus network and at times is required to have reasonable configuration and optimization again, this just needs to have the demand of bus passenger flow
One monitoring in real time and accurate prediction on this basis.
In recent years, as explosive growth is presented in the development of information technology, data information amount, data source is more and more,
Data volume is also more and more huger.Wherein, by the data of the information sensors such as mobile phone, WIFI, Internet of Things, GPS, IC card record
As data source most important in big data analysis, more complete individual trip is recorded as big data, especially traffic
Big data, analysis provide good data and support.By taking public transport barcode scanning or IC card as an example, the record of passenger getting on/off is formed
User uses the volume of data collection of bus trip, generates for bus passenger flow and its point sequentially the extraction of changing features mentions
Important data source is supplied.
Summary of the invention
The purpose of the present invention is: public transport barcode scanning is encrypted by magnanimity or brushing card data obtains at times-section bus
Passenger capacity and its overload situations
In order to achieve the above object, overall technological scheme of the invention is: extracting bus using public transport vehicle-mounted GPS data
Space-time trajectory public bus network time sharing segment is obtained into the road section where public transit vehicle different moments;Utilize bus card-reading
Record obtains the time data that user takes bus, and obtains station data after matching with the GPS data of bus, is extracted with this
The travel behaviour feature that individual takes bus;Space clustering is carried out to the website of public bus network, it would be possible to transfer point merge,
It prevents the difference due to site name and transfer information can not be extracted;According to the card-swiping mode of different public bus networks, statistics meter
It calculates individual and takes the probability of a certain public bus network in a certain period, and the probability distribution that each website is got off along the line after getting on the bus;
When statistical in each public bus network of pavement branch sections the carrying situation of bus historical data, and it is predicted.
Specifically, the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment provided by the invention, special
Sign is, comprising the following steps:
Step 1 obtains all continuous bus GPS data over time and space, and bus GPS data includes at least public affairs
Longitude station Long and public affairs where handing over circuit number LID, bus number BID, communication operation that moment TIME1, bus occurs
Position Latitude Lat where handing over vehicle, different buses correspond to different public bus network number LID and bus number BID, extract
The at the appointed time GPS data in section of bus corresponding to each bus number BID, constitutes the bus of each bus
Trip Time-space serial;
Step 2, the bus card-reading for obtaining anonymous encryption record data, in time sequencing, often swipe the card and once just generate one
Bus card-reading records data, and every bus card-reading record data include at least IC number ICID, public bus network number LID, public affairs
Hand over vehicle number BID, communication operation occur moment TIME2, type of getting on or off the bus TYPE and by bus cost COST, different barcode scanning terminals or
IC card corresponds to different IC number ICID, different IC number ICID and corresponds to Different Individual, extracts each IC number ICID specified
Bus in period is swiped the card record, and brushing card data collection individual corresponding to different IC number ICID is constituted;
Step 3, the bus station for obtaining all public bus networks within the scope of designated space, using Spatial Clustering to all
Bus station is clustered, and the bus station that merging spatial position is closed on is as the transfer node being likely to occur, to all public transport
Bus station in route is adjusted, the bus station before being clustered with the bus station replacement after cluster, so as to subsequent
The transfer behavior of body extracts;
Step 4 arranges the brushing card data collection for all individuals that step 2 obtains, obtain all individuals it is all it is upper,
Get-off stop information forms complete passenger's bus card-reading time series data of all individuals, for swiping the card corresponding to current individual
For current bus card-reading record data in data set, the space that moment TIME2 bus occurs for present communications movement is obtained
Position to obtain the site information of getting on the bus of current bus card-reading record data, and is added to current bus card-reading record data
In;
Bus card-reading mode is divided into three kinds by step 5: unit price after segmentation of getting on or off the bus after swiping the card, getting on the bus swipes the card and gets on the bus
It swipes the card, for different card-swiping modes, the when and where of getting on or off the bus of individual is analyzed and determined respectively, including following step
It is rapid:
The corresponding complete passenger's bus card-reading time series data of all IC number ICID of step 5.1, order traversal reads its public affairs
Friendship, which is swiped the card, records data, according to public bus network number LID swiped the card timing split;
Step 5.2, for every public bus network in each IC number ICID, its form of swiping the card is judged, according to different
Whether card-swiping mode uses get on or off the bus information of the different method statistic individuals on current public bus network, infer it to get off a little and may be used
To calculate, comprising the following steps:
Step 5.2.1, it is directed to current IC number ICID, reads site information of getting on the bus, the public transport that it takes bus every time
Moment TIME2 occurs for circuit number LID and pick-up time, i.e. communication operation;
Step 5.2.2, card-swiping modes different for different public bus networks estimates that current IC is numbered using different methods
ICID is in the information of getting off to take bus every time;
If the card-swiping mode of current public bus network is to get on or off the bus to be intended to swipe the card, involved in public transport barcode scanning terminal or IC card when
The information of getting off for the bus for taking current public bus network every time has been contained in the record of swiping the card of preceding public bus network, therefore,
For current IC number ICID obtain its it is all known to get off record, comprising the following steps: if it is in current public bus network
On swipe the card and record data R1, and corresponding bus station S1 is judged according to the type TYPE that gets on or off the bus recorded in data R1 that swipes the card
For website of getting on the bus, then judgement swipe the card record next of data R1 swipe the card record in data R2 get on or off the bus type TYPE whether be
It gets off, records the corresponding bus station S2 of data R2 if so, swiping the card as get-off stop, will swipe the card and record data R1 and note of swiping the card
It records data R2 to merge, is designated as a known record of getting off, if the type TYPE that gets on or off the bus recorded in data R2 that swipes the card is to get on the bus, table
Bright current IC number ICID does not get off in a upper stroke and swipes the card, then discarded swipe the card records data R1, reacquires next
Item, which is swiped the card, to be recorded data and is judged;
If the card-swiping mode of current public bus network is swiped the card to get on the bus, and block meter rate, then all IC number ICID institute is obtained
There is the record that can be calculated get off a little and can not calculate the record got off a little, comprising the following steps:
If current IC number ICID swipes the card on current public bus network during period T1 records data R3, corresponding
Bus station is S3, then:
Step 5.2.2.1, the possible get-off stop area of passenger is calculated according to the COST of cost by bus for recording data R3 that swipes the card
Between;
Step 5.2.2.2, it reads to swipe the card to record next of data R3 and swipe the card and records data R4, record data R4 if swiping the card
Corresponding starting bus station S4 is in the get-off stop section that step 5.2.2.1 is calculated, then it is assumed that current IC number
The get-off stop that ICID takes the bus of current public bus network is S4, while will swipe the card and record data R3 and be recorded as to calculate
The record got off a little;The corresponding starting bus station S4 of data R4 is recorded not in the case where step 5.2.2.1 is calculated if swiping the card
Within the point section of station, then will swipe the card records data R3 and is recorded as not calculating the record got off a little;
If the card-swiping mode of current public bus network is swiped the card to get on the bus, and unified charging, then all IC number ICID institute is obtained
There is the record that can be calculated get off a little and can not calculate the record got off a little, comprising the following steps:
If current IC number ICID swipes the card on current public bus network during period T1 records data R5, note of swiping the card
The corresponding bus stop of getting on the bus record data R5 is S5, reads to swipe the card to record next of data R5 and swipe the card and records data R6:
If the starting bus station S6 recorded in data R6 that swipes the card records data R5 from swiping the card in current public bus network
Along afterwards in station, and swipes the card and record the communication operation generation moment TIME2 and current public bus network to starting public affairs of data R6
Hand over the difference of website S6 time in threshold range T_Thrh, then it is assumed that current IC number ICID is in the case where originating bus station S6
Vehicle, while will swipe the card and recording data R5 and be recorded as to calculate the record got off a little;
It does not record data R5 from swiping the card in current public bus network if swiping the card and recording starting bus station S6 in data R6 and goes out
Along after hair in station, then will swipe the card records data R6 and is recorded as not calculating the record got off a little;
Current IC number ICID all frequencys occurred of getting on the bus in current public bus network are counted, as can not calculate
A foundation for a statistics of getting off;
Step 6, for individual corresponding to each IC number ICID, count individual one week and daily take every at times
The probability of public bus network, and its probability that each website is got off on the way after getting on the bus obtain the probability point in its section O-D of going on a journey
Cloth;
Step 7, the GPS data that each car in real-time bus card-reading data and every public bus network is obtained from data source,
The situation by bus of the potential passenger of public bus network is excavated, the probability distribution in the trip section O-D according to step 6 acquisition calculates every public affairs
The possible real-time passenger capacity of each car and the degree of crowding in intersection road, to the carrying demand of bus in the following designated time period and
The degree of crowding is predicted;
Preferably, in the step 1, the GPS data of a bus is a space-time trajectory record, according to bus
Public bus network number LID and bus number BID, inquire each bus at the appointed time all space-time trajectories notes in section
Record, by space-time trajectory record in warp, latitude be converted into geographical coordinate, thus construct bus trip Time-space serial.
Preferably, in the step 2, it further includes spending COST by bus that bus card-reading, which records data,.
Preferably, the step 3 includes:
Step 3.1, the position letter for obtaining the bus station and each bus station of all public bus networks within the scope of designated space
Breath, converts XY coordinate for location information, and XY coordinate is mapped in the geographical space with traffic route;
Step 3.2, using spatial clustering method, using the traffic distance between bus station as standard, to bus station carry out
Cluster merges very close bus station on space length, comprising the following steps:
Step 3.2.1, it sets cluster standard and is less than d meters as the distance between two bus stations;
Step 3.2.2, using each bus station as a cluster core, space clustering: the sky currently to cluster core is obtained
Meta position is set to the bus station that its periphery is searched in the center of circle, is less than d meters of bus station with its traffic distance if it exists, then by the public affairs
Website is handed over to be put into the Cluster space of current cluster core;
Step 3.2.3, the space clustering obtained by step 3.2.2 is merged, is constituted spatially relatively independent
Biggish bus station space clustering, merging condition is: if there are identical bus stations for any two space clustering, will work as
Preceding two spaces Cluster merging;
Step 3.2.4, the space center for extracting each bus station space clustering, maps that on map, acquisition is worked as
The spatial position at front space center and geographic name, after the bus station in each bus station space clustering is merged, with every
Bus station after the geographic name name merging of the space center of a bus station space clustering, the XY at current spatial center are sat
It is designated as the mean value of all bus station XY coordinates in corresponding bus station space clustering;
Step 3.3 rearranges every public bus network, the bus station before being clustered with the bus station replacement after cluster,
So that the transfer behavior to subsequent individual extracts.
Preferably, in the step 4, the current bus card-reading note that brushing card data corresponding to current individual is concentrated is obtained
The website information of getting on the bus for recording data includes following steps:
Step 4.1, to the bus stop for having already passed through space clustering within the scope of designated space, it is raw according to road traffic net
At Thiessen polygon, the spatial dimension of each bus stop is divided;
According to the communication operation in bus card-reading record data moment TIME2, public bus network number LID occur for step 4.2
With bus number BID, read from the GPS data of bus corresponding to public bus network number LID and bus number BID
The location information X-IC and Y-IC of the position of moment TIME2 bus occur for communication operation;
Location information X-IC and Y-IC that step 4.2 obtains are mapped to the bus stop that step 4.1 generates by step 4.3
Spatial dimension in, obtain communication operation occur moment TIME2 bus where bus stop, to obtain current public transport brush
The site information of getting on the bus of card record data.
Preferably, the step 6 includes:
Step 6.1 is directed to each IC number ICID, counts it at the appointed time section is interior and take public transport line in specific time period T
The number of road L, if the IC number ICID public transport barcode scanning terminal or IC card for IC1, in bus stop, S multiplies in period T
The probability of route L of taking transit bus be P_U (T, S, L, IC1) is N_U (T, S, L, IC1)/N_Day, in formula: N_U (T, S, L, IC1)
It gets on the bus the number of route L of taking pubic transport in daily period T in bus stop S for IC1 in designated time period, N_Day is
Number of days N_Day in period T;
Step 6.2, step 5 obtain to every time by bus record get off estimation on the basis of, for each public bus network
Get off probability of the different card-swiping modes using different method statistic individuals in each bus station;
If the card-swiping mode of public bus network L is to get on or off the bus to be intended to swipe the card, counting IC1, at the appointed time section T is interior in T1
The number that each station is got off on the way under conditions of bus stop S1 gets on the bus in section, then IC1 is in period T1 in bus
The S1 that stands gets on the bus take pubic transport route L after, S2 gets off in bus station probability be N_D (S1, S2, L, IC1)/N_U (T1, S1, L,
IC1), in formula, N_U (T1, S1, L, IC1) gets on the bus for IC1 S1 in bus stop in period T1 in designated time period and takes public affairs
The number of intersection road L;
If the card-swiping mode of public bus network L is swiped the card to get on the bus, and block meter rate, then to the frequency got off a little depending on its note of swiping the card
Record gets off a little whether can calculate separated statistics: for that can not calculate the record got off a little, it is assumed that an IC number ICID takes
Bus generally has a continuity, i.e., its get on the bus be a little its it is another wade out it is capable get off a little, therefore count current IC number ICID
All frequencys occurred of getting on the bus in public bus network L, as can not calculate a foundation for statistics of getting off:
For that can calculate the record got off a little, statistics IC1 S3 from bus stop within the T1 period gets on the bus line of taking pubic transport
The frequency N_D (S3, S4, D, L, IC1) that bus stop S4 gets off along it after the L of road, then IC1 is in period T1 from bus
The S3 that stands take pubic transport route L to the probability that bus stop S4 gets off be N_D (S3, S4, D, L, IC1)/N_U (T1, S3, L,
IC1), in formula: N_U (T1, S3, L, IC1) gets on the bus for IC1 S3 in bus stop in period T1 in designated time period and takes public affairs
The number of intersection road L;
For that can not calculate the record got off a little, statistics IC1 take pubic transport route L historical record in bus stop S4
Get on the bus frequency N_U (S4, L, IC1), then IC1 is after in bus stop, S3 takes public bus network L in period T1, in residue
The probability that bus stop S4 gets off on route is N_U (S4, L, IC1)/sum (N_U (SN, L, IC1)), and in formula, SN is IC1 in public affairs
Website Hosting after handing over station S3 to get on the bus in public bus network L residual paths, sum (N_U (SN, L, IC1)) are IC1 in bus stop
The frequency that website in residual paths is got on the bus after S3 gets on the bus, for substituting the frequency got off;
If the card-swiping mode of public bus network L is swiped the card to get on the bus, and unified charging, then to the frequency got off a little depending on its note of swiping the card
Whether the getting off of record, which can a little calculate, also needs separated statistics:
For that can calculate the record got off a little, statistics IC1 S5 from bus stop in period T1 gets on the bus line of taking pubic transport
The frequency N_D (S5, S6, L, IC1) that bus stop S6 after the L of road along it gets off, then IC1 is in period T1 from bus
The S5 that stands take pubic transport route L to the probability that bus stop S6 gets off be N_D (S5, S6, L, IC1)/N_U (T1, S5, L, IC1),
In formula: N_U (T1, S5, L, IC1) is that IC1 S5 in bus stop in period T1 gets on the bus line of taking pubic transport in designated time period
The number of road L;
For that can not calculate the record got off a little, statistics IC1 take pubic transport route L historical record in bus stop S6
The frequency N_S6_U_IC1 that gets on the bus, then IC1 is after in bus stop, S5 takes public bus network L in period T1, on remaining road
The probability that bus stop S6 on line gets off is N_U (S6, L, IC1)/sum (N_U (SN, L, IC1)), and in formula, SN is IC1 in public affairs
Website Hosting after handing over station S5 to get on the bus in public bus network L residual paths, sum (N_U (SN, L3, IC1)) are IC1 in bus
The frequency that website in residual paths is got on the bus after the S5 that stands gets on the bus, for substituting the frequency got off;
It gets off under step 6.3, the three kinds of card-swiping modes obtained according to step 6.2 for that can calculate to get off a little and can not calculate
It gets off the statistics of probability in the record of swiping the card of point for each website, obtains IC1 in period T1 by the way of record number weighting
The probability that SS gets off in bus stop again after inherent bus stop S gets on the bus:
For the public bus network for being intended to swipe the card of getting on or off the bus, all effective records have information of completely getting on or off the bus, therefore
What step 6.2.1 was obtained be exactly final IC1 got off after in bus stop, S gets on the bus in period T1 in bus stop SS it is general
Rate P_D (T1, S, SS, L, IC1);
For the public bus network swiped the card of only getting on the bus, record judgement of swiping the card be divided into get off a deducibility why not deducibility two
Kind, statistics can calculate the record got off a little and the quantity N_C (T1, S, L, IC1) and N_ that can not calculate the record got off a little respectively
NC (T1, S, L, IC1) and total number of records N (T1, S, L, IC1)=N_C (T1, S, L, IC1)+N_NC (T1, S, L, IC1), is adopted
With record number weighting pattern, IC1 takes public bus network L after bus stop S gets on the bus again in bus stop SS in period T1
Probability P _ the D (T1, S, SS, L, IC1) to get off=N_D (T1, S, SS, L, IC1)/N_U (T1, S, L, IC1) * N_C (T1, S, L,
IC1)/N (T1, S, L, IC1)+N_U (SS, L, IC1)/sum (N_U (SN, L, IC1)) * N_NC (T1, S, L, IC1)/N (T1, S,
L, IC1);
Step 6.4, all IC number ICID of traversal obtain corresponding each individual in period T1 in bus stop S
Website takes the probability of public bus network L, and the probability that each bus stop is got off on the way thereafter;And it counts based on this
Calculating each individual, in bus stop, S takes the probability of all public bus networks in period T1, and thereafter in each public affairs on the way
The probability for handing over station to get off;And calculate based on this each individual take in period T1 in all bus stops it is all
The probability of public bus network, and the probability that each bus stop is got off on the way thereafter;Each individual is finally calculated one
The probability of all public bus networks is taken in all bus stops in it all periods, and thereafter in each bus stop on the way
The probability got off.
Preferably, the step 7 includes:
Step 7.1 obtains real-time bus card-reading data and bus GPS data that time interval is TM from data source, right
It swipes the card and records data and arranged, carry out classification and ordination according to public bus network number LID and IC number ICID;
Step 7.2, the last item for reading each IC number ICID, which are swiped the card, records data, obtains current IC number ICID and multiplies
Public bus network number LID, bus number BID, type of the getting on or off the bus TYPE of seat pass through bus according to the method for step 4
GPS data obtains the site information of getting on the bus of current IC number ICID, and then judges it according to different public bus network card-swiping modes
In the situation by bus of current time node:
If current public bus network is to get on or off the bus to swipe the card, and last swiping the card records under data are in current IC number ICID
Vehicle, then it is assumed that current IC number ICID is currently without taking bus, then it takes the bus in current public bus network in real time
Be desired for 0;
If current public bus network is to get on or off the bus to swipe the card, and it is upper that last swiping the card, which records data, in current IC number ICID
Vehicle, then it is assumed that current IC number ICID currently takes the bus in current public bus network, then it takes current public transport line in real time
Bus in road is desired for 1;
If current public bus network is to get on the bus to swipe the card, the vehicle time communicates each ICID obtained according to step 6 on it
Movement occurs in the period of moment TIME2 after the bus that bus stop SL takes current public bus network, at each station on the way
The probability got off obtains the current bus stop position of current bus according to GPS data, calculates current IC number ICID and working as
The probability that preceding bus stop SN does not get off yet, probability value subtract current IC number ICID in bus stop SL and bus stop for 1
The sum of probability that bus stop SM between SN gets off PT (TR, SN, L, ICID)=1-sum (P_D (TU, SL, SM, L,
ICID)), TR is indicated in real time, then current IC number ICID take pubic transport in real time the bus in route be desired for E (TR, L,
ICID)=PT (TR, SN, L, ICID);
Step 7.3 is directed to each bus, counts all last times and swipes the card behavior generation in the current of current bus
The expectation of the real-time seating current bus of IC number ICID, it is expected that the sum of be that current bus carries phase of number in real time
It hopes;
Step 7.4, each IC number ICID probability that branch website takes bus at times obtained according to step 6, from
Predict the carrying demand expectation of every time bus routes between sites in designated time period TPJ, including following step backward in real time
It is rapid:
Step 7.4.1, according to the GPS data of bus and its arrangement of dispatching a car, predict every time public bus network in TPJ first
Between dispatch a car in section the time interval of situation and each bus stop of each Public Transit Bus Stopping;
Step 7.4.2, it is directed to each bus, traverses all IC number ICID, is searched at the appointed time section TPJ at it
There is the IC number ICID for record of getting on the bus in subsequent website each on the way;
Step 7.4.3, each IC number ICID obtained for step 7.4.2, according to its in period T in bus
Probability P _ U (T, S, L, ICID) that the S that stands gets on the bus, calculating each IC number ICID, in bus stop, S takes pubic transport in period T
The expectation of vehicle is calculated according to the getting off probability P _ D_prj (TU, SL, S, L, ICID) in bus stop S of the passenger in vehicle
The expectation that each passenger that may have taken current bus gets off in period T in bus stop S;Due in passenger capacity
It is a probability distribution for whether the IC number ICID of prediction gets on the bus in prediction, and carries statistics difference in real time, therefore getting on the bus
Behavior is that the passenger of prediction in the probability of getting off of bus stop S is passenger in the bus stop SL probability got on the bus and in bus stop
S gets off the product of probability: P_D_prj (TU, SL, S, L, ICID)=P_U (TU, SL, L, ICID) * P_D (TU, SL, S, L,
ICID), current IC number ICID bus stop S get off be desired for E_D (T, S, L, ICID)=P_D_prj (TU, SL, S,
L, ICID);
Step 7.4.4, each IC number ICID is traversed, is summed in period T in the expectation that bus stop S gets on the bus to it,
As current bus volume of passenger traffic in the expectation of bus stop S seeks each IC number ICID in the expectation that bus stop S gets off
With as current bus volume of passenger traffic under the expectation of bus stop S, the two subtract each other working as in the period T predicted
The expectation passenger capacity of preceding bus;
Step 7.5, the design load number data for obtaining each bus, by the expectation of itself and each bus at times
Passenger capacity compares, and obtains the real-time expectation degree of crowding statistics and prediction result of each bus at times.
The present invention handles the GPS data of bus, the space-time trajectory of the daily traveling of bus is obtained, to public transport
IC, which swipes the card, to be recorded big data and is handled and screened, by swiping the card between the held public transport barcode scanning terminal of individual or IC card and bus
Record constructs the brushing card data collection that individual takes bus, and handles bus station spatial information, using space clustering
The method bus station of closing on spatial position merge, determine that the space of each bus station takes by dividing Thiessen polygon
Business range;By the GPS data of bus and bus card-reading behavior synchronization, obtains public transport barcode scanning terminal or IC card is swiped the card behavior
When geospatial location, excavate swipe the card behavior occur when bus station information;Successively traverse each public transport barcode scanning terminal or
The history of IC card is swiped the card record, and public bus network is divided into according to different card-swiping modes to get on or off the bus and swiped the card, got on the bus and swipe the card point
Section charging and three kinds of modes of unified charging of swiping the card of getting on the bus calculate passenger using different methods for different types of public bus network
O-D point of getting on or off the bus on every time public bus network, the holder for calculating each public transport barcode scanning terminal or IC card based on this exist
The probability got on the bus in each period at each station, and the probability that each station is got off on the way after getting on the bus;It is real-time obtaining
On the basis of bus card-reading record, using the holder of each public transport barcode scanning terminal or IC card at times on every public bus network
Probability distribution of getting on or off the bus, calculate the real-time seating capacity of each bus;Use holding for each public transport barcode scanning terminal or IC card
The probability that someone gets off after each website takes the probability of every public bus network and gets on the bus at each station at times predicts each public affairs
Hand over seating capacity and the degree of crowding of the vehicle in the following designated time period.
The invention has the advantages that leverage fully on bus card-reading record and bus GPS data, can low cost, automation,
The time and location information that every record of swiping the card occurs easily is obtained, complete individual is constructed and takes public bus network time series data
Collection, handles it, obtains after individual branch website at times takes the probability distribution of every time public bus network and gets on the bus on the way
The probability distribution that each website is got off, thus it is convenient, efficiently to real-time and following bus seating capacity and the degree of crowding into
Row calculates and prediction.
Detailed description of the invention
Figure 1A and Figure 1B is flow chart of the invention.
Specific embodiment
In order to make the present invention more obvious and understandable, hereby with preferred embodiment, and attached drawing is cooperated to be described in detail below.
Step 1, system read the bus GPS data obtained from transit operator, and bus GPS data theoretically exists
Time and be all spatially it is continuous, different buses corresponds to different public bus network number LID and bus number BID, mentions
GPS data of each BID at the appointed time in section is taken, bus is constituted and goes on a journey Time-space serial.
The GPS data of bus is the vehicle GPS of each bus obtained in real time from bus operation quotient to bus
The record data of spatial position.
Step 1.1, system read the bus GPS data obtained from transit operator, theoretically bus GPS data
All should be continuous in the time and space, comprising: unique public bus network number LID, unique bus number BID,
Position Latitude Lat where longitude station Long, bus where moment TIME1, bus occur for communication operation, wherein public transport
Circuit number LID, bus number BID constitute bus number;
Step 1.2, a bus GPS data are a space-time trajectory record;
Step 1.3, public bus network number LID and bus number BID according to bus, inquire its at the appointed time section
Interior all GPS records, convert geographical coordinate for its longitude and latitude, construct the spatiotemporal motion track of bus.
In this example, the part GPS data of the bus SB of public bus network SL is shown in Table 1:
LID | BID | TIME | Long | Lat | X | Y |
...... | ...... | ...... | ...... | ...... | ...... | |
SL | SB | 2017-06-20 08:05:57 | 121.6981 | 31.1289 | 22706.9313 | -11185.0718 |
SL | SB | 2017-06-20 08:08:23 | 121.7038 | 31.1227 | 23253.0972 | -11869.8001 |
SL | SB | 2017-06-20 08:11:43 | 121.6995 | 31.1151 | 22846.4016 | -12714.8964 |
SL | SB | 2017-06-20 08:14:25 | 121.6933 | 31.1099 | 22257.4919 | -13296.5478 |
SL | SB | 2017-06-20 08:18:19 | 121.6862 | 31.1042 | 21577.8252 | -13931.1146 |
...... | ...... | ...... | ...... | ...... | ...... |
Step 2, system read from the bus card-reading that transit operator obtains anonymous encryption and record data, in time sequencing
On, often swiping the card once generate one and swipe the card records data, and different barcode scanning terminals or IC card correspond to different IC number ICID,
And bus number BID is had recorded, it extracts bus of each IC number ICID at the appointed time in section and swipes the card record, constituting should
The brushing card data collection of IC number ICID;
Step 2.1, system read the bus card-reading record data of the anonymous encryption obtained from transit operator, are recorded
Data include: that the moment occurs for barcode scanning terminal or bus card unique number ICID, bus unique number BID, communication operation
TIME2, type of getting on or off the bus TYPE, COST is spent by bus;
Anonymity encryption bus card-reading data are the encrypted anonymous public transport barcode scannings that obtains and desensitize in real time from bus operation quotient
The encryption card using information of terminal and IC card user time sequence, content include: ICID, LID, BID, TIME2, TYPE, COST.Tool
Body is described below:
ICID is to carry out unidirectional irreversible encryption to each public transport barcode scanning terminal or IC card user, so that unique identification is every
A public transport barcode scanning terminal or IC card user, it is desirable that the ICID after each public transport barcode scanning terminal or IC card user encryption keeps unique
Property.
LID is the number of public bus network.
BID is the number of each specific bus.
TIME2 is that the moment occurs for the behavior of swiping the card of current record, and unit is millisecond.
TYPE, is public transport barcode scanning terminal or IC card is swiped the card type, is divided to get on the bus to swipe the card and get off and swipe the card two kinds.
COST is the expense by bus of bus, if public transport barcode scanning terminal or IC card are to get on or off the bus to swipe the card, in record of getting off
In withhold, swipe the card if getting on the bus, get on the bus record in withhold.
Step 2.2, a brushing card data are a signaling record, and every signaling record is decrypted;
Step 2.3, according to IC number ICID, inquire its it is at the appointed time all in section swipe the card, constitute individual take it is public
Hand over the brushing card data collection of vehicle;
In this example, the part of public transport barcode scanning terminal or IC card that number is IC1 record of swiping the card is shown in Table 2:
Swiping the card for 2 IC1 of table records data
RID | ICID | LID | BID | TIME | TYPE | COST |
...... | ...... | ...... | ...... | ...... | ...... | ...... |
R1 | IC1 | RL1 | B11 | 2017-06-20 08:30:31 | 1 | -1 |
R2 | IC1 | RL1 | B11 | 2017-06-20 09:04:65 | 2 | 3 |
R3 | IC1 | RL2 | B13 | 2017-06-20 09:17:22 | 1 | 2 |
R4 | IC1 | RL3 | B42 | 2017-06-20 14:15:56 | 1 | 2 |
R5 | IC1 | RL4 | B5 | 2017-06-20 17:31:43 | 1 | 7 |
R6 | IC1 | RL1 | B21 | 2017-06-21 08:28:12 | 1 | -1 |
...... | ...... | ..... | ..... | ...... | ...... | ...... |
Step 3 obtains public bus network website all within the scope of designated space, is gathered using Spatial Clustering to it
Class, the website that merging spatial position is closed on adjust the website in all public bus networks as the transfer node being likely to occur
It is whole, so that the transfer behavior to individual is extracted and is analyzed;
Step 3.1 obtains all public bus network website and its location information, converts XY coordinate for its longitude and latitude, and
XY coordinate is mapped in the geographical space with traffic route;
In this example, the spatial positional information of part bus station is shown in Table 3:
The spatial information of 3 bus station of table
Step 3.2 clusters bus station using the traffic distance between website as standard using spatial clustering method,
Merge very close website on space length:
Step 3.2.1, it sets cluster standard and is less than d meters as the distance between two websites;
Step 3.2.2, regard each website as a cluster core, be circle with its spatial position by taking a of bus stop as an example
The heart searches for the website on its periphery, is less than d meters of bus stop b with its traffic distance if it exists, then using bus stop b as public affairs
In the cluster for handing over station a;
Step 3.2.3, the space clustering obtained by step 3.2.2 is merged, if merging condition is cluster x and cluster
Y is then merged there are identical website, constitutes spatially relatively independent biggish bus station space clustering;
Step 3.2.4, the space center for extracting each cluster, maps that on map, obtains the space of the central point
Position and geographic name, after the bus station in each cluster is merged, after being merged with the geographic name name of the central point
Bus station, the XY coordinate of cluster centre are the mean value of all bus station XY coordinates in cluster;
In this example, the part bus station after space clustering is shown in Table 4:
Step 3.3 rearranges every public bus network, before being clustered with the bus station replacement after cluster in public bus network
Website, so that the transfer information to subsequent individual extracts;
Step 4 arranges the record of swiping the card of public transport barcode scanning terminal or IC card, according to charge time and the time public transport
The spatial position of vehicle obtains the site information that individual is got on or off the bus;
Step 4.1, the record of swiping the card for extracting all bus passengers in designated time period, according to public transport barcode scanning terminal or IC card
The record of swiping the card of each ICID is sorted according to charge time, constitutes swiping the card for each bus passenger by unique identification number ICID
Time of the act sequence;
According to the communication operation in record of swiping the card moment TIME2, public bus network number LID and bus occur for step 4.2
Number BID reads the position X-IC and Y-IC of bus when behavior of swiping the card occurs from the GPS data of the bus;
In this example, the public transport barcode scanning terminal or IC card that number is IC1 swipe the card behavior occur when spatial position be shown in Table 5:
5 IC1 of table swipe the card behavior occur when spatial position
RID | ICID | LID | BID | TIME | TYPE | X | Y | COST |
...... | ...... | ...... | ...... | ...... | ...... | ...... | ||
R1 | IC1 | RL1 | B11 | 2017-06-20 08:30:31 | 1 | -3706.4625 | -18.1212 | -1 |
R2 | IC1 | RL1 | B11 | 2017-06-20 09:04:65 | 2 | -2618.1421 | -733.5432 | 3 |
R3 | IC1 | RL2 | B13 | 2017-06-20 09:17:22 | 1 | -2614.1322 | -706.8132 | 2 |
R4 | IC1 | RL3 | B42 | 2017-06-20 14:15:56 | 1 | -2883.4211 | -1063.6231 | 2 |
R5 | IC1 | RL4 | B5 | 2017-06-20 17:31:43 | 1 | -2313.4212 | -1350.4321 | 7 |
R6 | IC1 | RL1 | B21 | 2017-06-21 08:28:12 | 1 | -3707.4324 | -19.4321 | -1 |
...... | ...... | ...... | ...... | ...... | ...... | ...... |
Step 4.3, due to swipe the card behavior occur when bus may be in initial state, it is therefore desirable to be every
A bus station divides spatial dimension, for the bus stop for having already passed through space clustering within the scope of designated space, according to road
The network of communication lines generates Thiessen polygon, divides the spatial dimension of each bus stop;
The spatial position X-IC and Y-IC of bus are mapped to step when step 4.4, the individual for obtaining step 4.2 are swiped the card
In 4.3 Thiessen polygons generated, website when behavior of swiping the card occurs where bus is obtained;
In step 4.5, the site information that will acquire are added to the public transport barcode scanning terminal of passenger or IC card is swiped the card record, constitute
Complete passenger's bus card-reading time series data;
In this example, the public transport barcode scanning terminal or IC card that number is IC1 swipe the card behavior occur when where bus station see
Table 6:
6 IC1 of table swipes the card behavior locating bus station when occurring
RID | ICID | LID | BID | TIME | TYPE | X | Y | SID | COST |
...... | ...... | ...... | ...... | ...... | ...... | ...... | ...... | ...... | ...... |
R1 | IC1 | RL1 | B11 | 2017-06-20 08:30:31 | 1 | -3706.4625 | -18.1212 | The market L | -1 |
R2 | IC1 | RL1 | B11 | 2017-06-20 09:04:65 | 2 | -2618.1421 | -733.5432 | The crossing B | 3 |
R3 | IC1 | RL2 | B13 | 2017-06-20 09:17:22 | 1 | -2614.1322 | -706.8132 | The crossing B | 2 |
R4 | IC1 | RL3 | B42 | 2017-06-20 14:15:56 | 1 | -2883.4211 | -1063.6231 | The road D | 2 |
R5 | IC1 | RL4 | B5 | 2017-06-20 17:31:43 | 1 | -2313.4212 | -1350.4321 | The road G | 7 |
R6 | IC1 | RL1 | B21 | 2017-06-21 08:28:12 | 1 | -3707.4324 | -19.4321 | The market L | -1 |
...... | ...... | ...... | ...... | ...... | ...... | ...... | ...... | ...... | ...... |
The card-swiping mode of bus is divided into three kinds by step 5: segmentation after swiping the card, getting on the bus of getting on or off the bus, which is swiped the card, (only brushes one
It is secondary, but different according to purpose ground price) and unit price is swiped the card (uniform price) after getting on the bus, it is right respectively for different card-swiping modes
The when and where of getting on or off the bus of individual is analyzed and determined;
The each ICID of step 5.1, order traversal, reading it includes charge time, public bus network of swiping the card, bus of swiping the card volume
Record of swiping the card including number is numbered timing of being swiped the card according to public bus network and is split;
Step 5.2, for every public bus network in each ICID, its form of swiping the card is judged, according to the different sides of swiping the card
Formula uses get on or off the bus information of the different method statistic individuals on this route, and whether it is inferred to get off a little can calculate;
Step 5.2.1, be directed to each ICID, read website of getting on the bus, public bus network number that it takes bus every time and
Pick-up time;
Step 5.2.2, card-swiping mode different for different public bus networks, using different method estimation ICID each
The information of getting off to take bus;
If the card-swiping mode of public bus network L1 is to get on or off the bus to be intended to swipe the card, LID1 involved in public transport barcode scanning terminal or IC card
Swipe the card record in contained the information of getting off to take bus every time, therefore, for IC1, if it swipes the card on L1
R1, bus station S1 are recorded, then swipes the card to record and has recorded get off place S2 and time in R2 for its next, by R1 and R2 conjunction
And it is designated as a known record of getting off;If R2 is also record of getting on the bus, shows that IC1 does not get off in a upper stroke and swipe the card, then
Discarded R1;
If the card-swiping mode of public bus network L2 is swiped the card to get on the bus, and block meter rate, if then IC1 is during period T1 in L2
On swipe the card and record R3, bus station S3 calculates that passenger is possible and gets off then first according to the expense COST of the record of swiping the card
Website section swipes the card for next after then reading IC1 swipes the card on L2 and records R4;If the start site S4 in R3 is being calculated
To get-off stop section in, then it is assumed that the get-off stop that IC1 takes L2 is S4, while R3 being recorded as to calculate and is got off a little
Record;If R4's gets on the bus a little not within the possible get-off stop section R3, R3 is recorded as not calculating and is got off a little
Record;
If the card-swiping mode of public bus network L3 is swiped the card to get on the bus, and unified charging, if then IC1 is during period T1 in L3
On swipe the card and record R5, station of getting on the bus is S5, reads next after ICI swipes the card on L3 and swipes the card and record R6: if rising in R6
Initial station point S6 is in L3 along after R5 in station, and the difference of the charge time of R6 and public bus network L3 to S6 time exist
In threshold range T_Thrh, then it is assumed that IC1 gets off in S6, while R3 being recorded as to calculate the record got off a little;If R6's is upper
R6 in the station along the line of R5, then is not recorded as not calculating the record got off a little by vehicle point;
IC1 all frequencys occurred of getting on the bus in L3 are counted, as can not calculate a foundation for statistics of getting off.
In the present example it is assumed that time threshold T_Thrh be 1 hour, IC1 on June 20th, 2017 taken RL1, RL2 and
RL3 tri- plows bus, and wherein RL1 is to get on or off the bus to swipe the card, therefore get off a little it can be extrapolated that it is got off a little just in the crossing B, RL2 and RL3
For unified valuation of getting on the bus, RL4 is pricing for segment of getting on the bus, and the deduction got off a little, which needs to rely on, changes to information, wherein R3's is upper
Vehicle charge time is at 9 points in the morning 17 minutes, and the pick-up time of swiping the card of R4 is to divide 2 pm 15, and time interval is greater than T_Thrh, therefore
L2, which can not be calculated, to get off a little, and the pick-up time of swiping the card of R5 is 31 minutes at 5 points in afternoons, takes 46 on theoretical time of the L3 from the road D to the road G
Minute, therefore getting off for RL3 can not a little calculate, RL4 is pricing for segment, and swiping the card for R6 is got on the bus website as the market L, gone out from the road G
Cost in the range on the way of the RL4 of hair, and from the crossing G to the market L is consistent with the COST of R5, therefore R5 can be calculated and be got off a little,
It is got off a little as the market L.
Step 6 is directed to each ICID, counts the probability that individual takes every public bus network for one week at times daily, and
Its probability that each website is got off on the way after getting on the bus obtains the probability distribution in its section O-D of going on a journey, comprising the following steps:
Step 6.1 is directed to each ICID, counts its at the appointed time interior time that public bus network L is taken in specific time period T of section
Number, if No. ICID public transport barcode scanning terminal or IC card for IC1, in bus stop, S takes pubic transport route L's in period T
Probability P _ U (T, S, L, IC1) is N_U(T, S, L, IC1)/N_Day, i.e., IC1 is in daily T time section in designated time period
AT STATION S get on the bus take pubic transport route L times N _ U (T, S, L, IC1) divided by designated time period number of days N_Day (distinguish
Working day and nonworkdays);
In this example, if time interval is 1 hour, data (30 days) is recorded according to long-term the swiping the card of IC1, obtain its work
Day every morning, 8-9 point in the number that RL1 is taken at the market L station was 22 times, and during which working day amounts to 22 days, then IC1 is on weekdays
Taking the probability of RL1 at the market L station in 8-9 point is 1, and nonworkdays 8-9 every morning point takes the number of RL1 at the market L station
It is 1 time, during which nonworkdays number of days is 8 days, then IC1 is in the probability that RL1 is taken at the market L station in 8-9 point in nonworkdays
0.125;
IC1 on weekdays 8-9 in morning point the crossing B station take RL2 number be 2 times, then it exists in 8-9 point on weekdays
The probability that RL2 is taken at the crossing B station is 0.091;Its on weekdays 9-10 in morning point the crossing B station take RL2 number be 20 times,
Then it is on weekdays 0.909 in the probability that RL2 is taken at the crossing B station in 9-10 point;It is in nonworkdays 8-9 in morning point on the road B
The number that RL2 is taken at mouth station is 0 time, then it is 0 in the probability that RL2 is taken at the crossing B station in 8-9 point in nonworkdays;It is non-
Morning on working day 9-10 point the crossing B station take RL2 number be 2 times, then it takes in 8-9 point at the crossing B station in nonworkdays
The probability for multiplying RL2 is 0.125;
IC1 on weekdays 14-15 point D way station take RL3 number be 3 times, then it is on weekdays in 14-15 point in D
The probability that RL3 takes in way station is 0.136;Its nonworkdays 14-15 point D way station take RL3 number be 0 time, then its
Working day is 0 in the probability that RL3 takes in D way station in 14-15 point;17-18 point takes the number of RL4 in G way station to IC1 on weekdays
It is 22 times, then it is on weekdays 1 in the probability that RL4 takes in G way station in 14-15 point;It is in nonworkdays 14-15 point on the road G
The number that RL4 is taken at station is 0 time, then it is on weekdays 0 in the probability that RL4 takes in G way station in 14-15 point.
Step 6.2, step 5 obtain to every time by bus record get off estimation on the basis of, for each public bus network
Different card-swiping modes is got off probability using different method statistic individuals in each website, comprising the following steps:
If the card-swiping mode of public bus network L1 is to get on or off the bus to be intended to swipe the card, counting IC1, at the appointed time section is interior in T1
The number that each station (such as S2) is got off on the way under conditions of S1 gets on the bus AT STATION in section, then IC1 is in period T1 in public transport
Station S1 gets on the bus take pubic transport route L1 after, the probability that website S2 gets off be N_D (S1, S2, L1, IC1)/N_U (T1, S1,
L1, IC1), wherein N_U (T1, S1, L1, IC1) is that S1 gets on the bus and takes pubic transport IC1 AT STATION within the T1 period in designated time period
The number of route L1;
In this example, IC1 is after 8-9 point takes RL1 at the market L station on weekdays, each station is got off on the way number and under
Vehicle probability is shown in Table 7:
7 IC1 of table on weekdays get off number and probability by 8-9 point each station on the way after RL1 is taken at the market L station
SID | Number | Probability |
The road M | 0 | 0 |
The crossing N | 0 | 0 |
O organ | 0 | 0 |
...... | ...... | ...... |
The crossing B | 22 | 1 |
The building P | 0 | 0 |
The road Q | 0 | 0 |
The main road S | 0 | 0 |
...... | ...... | ...... |
If step 6.2.2, the card-swiping mode of public bus network L2 is swiped the card to get on the bus, and block meter rate, then to the frequency got off a little
It is secondary to get off a little whether calculate separated statistics depending on its record of swiping the card, for that can not calculate the record got off a little, it will be assumed that one
A ICID, which takes bus, generally has continuity, i.e., its get on the bus a little while also have it is very big may be its it is another wade out it is capable under
Che Dian, therefore count ICID all frequencys occurred of getting on the bus in L, as can not calculate a foundation for statistics of getting off:
For that can calculate the record got off a little, statistics IC1 gets on the bus in period T1 from S3 take L2 after it is each along it
Stand the frequency N_D (S3, S4, D, L2, IC1) that (such as S4) get off, then IC1 in period T1 from S3 take L2 to S4 get off it is general
Rate is N_D (S3, S4, D, L2, IC1)/N_U (T1, S3, L2, IC1), and wherein N_U (T1, S3, L2, IC1) is designated time period
S3's interior IC1 gets on the bus the number of route L2 of taking pubic transport AT STATION within the T1 period;
For that can not calculate that the record got off a little, statistics IC1 are taken in the historical record of L2 in the upper of each website (such as S4)
Vehicle frequency N_U (S4, L2, IC1), then in IC1 period T1 after S3 takes public bus network L2, each website on remaining route
The probability that (such as S4) gets off is N_U (S4, L2, IC1)/sum (N_U (SN, L2, IC1)), and wherein SN is IC1 L2 after S3 gets on the bus
Website Hosting in residual paths, sum (N_U (SN, L2, IC1)) are that website of the IC1 after S3 gets on the bus in residual paths is got on the bus
The frequency, for substituting the frequency got off;
In this example, IC1 on weekdays 17-18 point after RL4 takes in G way station, can calculate get off a little be recorded as 19,
It can not calculate that the record got off a little is 3, according to that can calculate the record got off a little, number and get off general that on the way, each station is got off
Rate is shown in Table 8:
Number and probability get off (under can calculating in table 8 IC1 working day 17-18 point each station on the way after RL4 takes in G way station
The record of vehicle point)
17-18 point is after RL4 takes to IC1 in G way station on weekdays, according to that can not calculate in the record got off a little, on the way
The number that each station is got off and probability of getting off are shown in Table 9:
Table 9 IC1 working day 17-18 point each station on the way after RL4 takes in G way station number of getting on the bus (can not be calculated with probability
The record got off a little)
SID | Number | Probability |
The road R | 0 | 0 |
The mansion S | 5 | 0.161 |
...... | ...... | ...... |
The market T | 4 | 0.129 |
The road U | 22 | 0.710 |
The market L | 0 | 0 |
V bridge | 0 | 0 |
...... | ...... | ...... |
If step 6.2.3, the card-swiping mode of public bus network L3 is swiped the card to get on the bus, and unified charging, then to the frequency got off a little
It is secondary depending on its record of swiping the card get off whether can a little calculate also need separate statistics;
Step 6.2.3.1, for that can calculate the record got off a little, statistics IC1, which gets on the bus in period T1 from S5, takes L3
Each station (such as S6) frequency N_D (S5, S6, L3, IC1) that gets off along it afterwards, then IC1 in period T1 from S5 take L3 to
The probability that S6 gets off is N_D (S5, S6, L3, IC1)/N_U (T1, S5, L3, IC1), and wherein N_U (T1, S5, L3, IC1) is to refer to
S5's IC1 gets on the bus the number of route L3 of taking pubic transport AT STATION within the T1 period in section of fixing time;
Step 6.2.3.2, for that can not calculate that the record got off a little, statistics IC1 are taken in the historical record of L3 at each station
The frequency N_S6_U_IC1 that gets on the bus of point (such as S6), then in IC1 period T1 after S5 takes public bus network L3, in remaining route
The probability that upper each website (such as S6) is got off is N_U (S6, L3, IC1)/sum (N_U (SN, L3, IC1)), and wherein SN is IC1 in S5
Website Hosting after getting on the bus in L3 residual paths, sum (N_U (SN, L3, IC1)) be IC1 after S5 gets on the bus in residual paths
The frequency that website is got on the bus, for substituting the frequency got off;
In this example, IC1 is after 9-10 point takes RL2 at the crossing B station on weekdays, can calculate get off a little be recorded as 15
Item can not calculate that the record got off a little is 5, and according to that can calculate the record got off a little, the number that each station is got off on the way is under
Vehicle probability is shown in Table 10:
Table 10 IC1 working day 9-10 point each station on the way after RL2 is taken at the crossing B station number of getting off (can be calculated with probability
The record got off a little)
SID | Number | Probability |
The road W | 0 | 0 |
...... | ...... | ...... |
The square X | 15 | 1 |
Y bridge | 0 | 0 |
...... | ...... | ...... |
IC1 is after 9-10 point takes RL2 at the crossing B station on weekdays, according to that can not calculate in the record got off a little, on the way
The number that each station is got off and probability of getting off are shown in Table 11:
Table 11 IC1 working day 9-10 point each station on the way after RL2 is taken at the crossing B station number of getting on the bus (can not be pushed away with probability
Calculate the record got off a little)
SID | Number | Probability |
The road W | 0 | 0 |
...... | ...... | ...... |
The square X | 17 | 1 |
Y bridge | 0 | 0 |
...... | ...... | ...... |
It gets off under step 6.3, the three kinds of card-swiping modes obtained according to step 6.2 for that can calculate to get off a little and can not calculate
It gets off the statistics of probability in the record of swiping the card of point for each website, obtains IC1 within the T1 period by the way of record number weighting
The probability that SS website is got off again after S website is got on the bus;
For the public bus network (such as L1) for being intended to swipe the card of getting on or off the bus, all effective records have letter of completely getting on or off the bus
Breath, therefore the probability that the exactly final IC1 of step 6.2.1 acquisition gets off after S website is got on the bus in SS website within the T1 period
P_D (T1, S, SS, L1, IC1);
For the public bus network (such as L2 and L3) swiped the card of only getting on the bus, record judgement of swiping the card be divided into get off a deducibility what
Two kinds of not deducibility, respectively statistics can calculate get off a little record with can not calculate the record got off a little quantity N_C (T1, S,
L2, IC1) and N_NC (T1, S, L2, IC1) and total number of records N (T1, S, L2, IC1)=N_C (T1, S, L2, IC1)+N_NC
(T1, S, L2, IC1), using record number weighting pattern, IC1 takes L2 within the T1 period, and SS website is got off again after S website is got on the bus
Probability P _ D (T1, S, SS, L2, IC1)=N_D (T1, S, SS, L2, IC1)/N_U (T1, S, L2, IC1) * N_C (T1, S, L2,
IC1)/N (T1, S, L2, IC1)+N_U (SS, L2, IC1)/sum (N_U (SN, L2, IC1)) * N_NC (T1, S, L2, IC1)/N
(T1, S, L2, IC1);
In this example, IC1 on weekdays be shown in by the 9-10 point probability of recombination that each station is got off on the way after RL2 is taken at the crossing B station
Table 12:
Table 12 IC1 working day 9-10 point is got off the probability of recombination after RL2 is taken at the crossing B station at each station
SID | Probability |
The road W | 0 |
...... | ...... |
The square X | 1 |
Y bridge | 0 |
...... | ...... |
The 17-18 point probability of recombination that each station is got off on the way after RL4 takes in G way station is shown in Table 13 to IC1 on weekdays:
Table 13 IC1 working day 17-18 point each station on the way after RL4 takes in G way station is got off the probability of recombination
SID | Probability |
The road R | 0 |
The mansion S | 0.158 |
...... | ...... |
The market T | 0.108 |
The road U | 0.733 |
The market L | 0 |
V bridge | 0 |
...... | ...... |
Step 6.4, all ICID of traversal, obtain each user and take the general of public bus network L in S website within the T1 period
Rate, and the probability that each website is got off on the way thereafter;Each user is calculated based on and to take within the T1 period in S website
Multiply the probability of all public bus networks, and the probability that each website is got off on the way thereafter;Each user is calculated based on and
The probability of all public bus networks is taken in all bus stations within the T1 period, and thereafter on the way each website get off it is general
Rate;Each user is finally calculated and took the general of all public bus networks in all bus stations within one day all period
Rate, and the probability that each website is got off on the way thereafter;
Step 7, the GPS data that each car in real-time bus card-reading data and every public bus network is obtained from data source,
The situation by bus for excavating the potential passenger of public bus network, calculates the possible real-time passenger capacity of each car and crowded in every public bus network
Degree predicts the carrying demand and the degree of crowding of bus in the following designated time period;
Step 7.1, time interval is obtained from data source is that TM (is recorded, TM needs are greater than by bus to construct complete individual
Equal to 3 hours) real-time bus card-reading data and bus GPS data, brushing card data is arranged, according to public bus network and
ICID number carries out classification and ordination;
Step 7.2 reads each ICID the last item and swipes the card record, obtains public bus network number LID, the public affairs of ICID seating
Vehicle number BID, type of getting on or off the bus TYPE are handed over, according to the method for step 4, the bus loading zone of ICID is obtained by bus GPS data
Then point information judges it in the situation by bus of current time node according to different public bus network card-swiping modes:
Step 7.2.1, it swipes the card if getting on or off the bus, and last swiping the card is recorded as getting off in ICID, then it is assumed that the ICID
Currently without taking bus, then its bus BID to be taken pubic transport in route LID in real time is desired for 0;
Step 7.2.2, it swipes the card if getting on or off the bus, and last swiping the card is recorded as getting on the bus in ICID, then it is assumed that the ICID
The bus BID currently to take pubic transport in route LID, then its (TR) takes pubic transport phase of the bus BID in route LID in real time
Hope that E (TR, L, ICID) is 1;
Step 7.2.3, it swipes the card if getting on the bus, then vehicle time TIME is (i.e. on it by each ICID obtained according to step 6
TIME2 in period) website SL take pubic transport route LID bus BID after, the probability that each station is got off on the way, according to
The current site location of BID is obtained according to GPS data, calculates the probability that ICID does not get off yet in current site (SN), probability value 1
Subtract the sum of the probability that website (such as SM) of the ICID between SL and SN is got off PT (TR, SN, L, ICID)=1-sum (P_D (TU,
SL, SM, L, ICID)), then the ICID take pubic transport in real time the bus BID in route LID be desired for E (TR, L, ICID)=
PT (TR, SN, L, ICID);
Step 7.3, be directed to each bus BID, count all last times swipe the card behavior occur BID ICID reality
When take bus the expectation of BID, it is expected that the sum of sum (E (TR, L, ICID)) be that bus BID carries number in real time
Expectation;
In this example, if TM is 3 hours, starting point is 7 points of the morning of certain working day, and the public transport of public bus network SL1 is calculated
Vehicle RB1 on the day of 10 points pass through the market L attachment section when, it is expected that carrying number be 32 people, according to the Ministry of Construction " city printed and distributed in 2001
City's construction system index explanation ", the calculation formula of rated passenger capacity are as follows: the fixed passenger seat digit+compartment in rated passenger capacity=compartment has
Effect standing area (square metre) × every square metre allows station number, and every square metre allows 8 people of standing number, then the maximum lotus of RB1
Manned number is 80 people or so, and RB1 is 0.4 in 10 points of the degree of crowding of this day morning, not crowded;
Step 7.4, according to step 6 obtain each ICID probability that branch website takes bus at times, from real time to
The carrying demand expectation of every time bus routes between sites in designated time period TPJ is predicted afterwards;
Step 7.4.1, according to the GPS data of bus and its arrangement of dispatching a car, predict every time public bus network in TPJ first
Between dispatch a car in section the time interval of situation and each platform of each Public Transit Bus Stopping;
Step 7.4.2, it is directed to each bus (such as B1), traverses all ICID, is searched at the appointed time section TPJ at it
There is the ICID for record of getting on the bus in subsequent website each on the way;
Step 7.4.3, it is directed to each ICID, probability P _ U for being got on the bus in period T in website S according to it (T, S, L,
ICID), calculate each ICID to take bus the expectation of B1 at the T moment in website S, according to the passenger in vehicle in website S
Get off probability P _ D_prj (TU, SL, S, L, ICID), the passenger for calculating each B1 that may take bus exists at the T moment
The expectation that website S gets off;It is a probability distribution due to whether being got on the bus in the prediction of passenger capacity for the ICID of prediction, and
It carries that statistics is different in real time, therefore is that passenger gets on the bus at the station SL in the probability of getting off of website S in the passenger that the behavior of getting on the bus is prediction
It probability and gets off the product of probability at the station S: P_D_prj (TU, SL, S, L, ICID)=P_U (TU, SL, L, ICID) * P_D (TU,
SL, S, L, ICID), which is desired for E_D (T, S, L, ICID)=P_D_prj (TU, SL, S, L, ICID) what the station S was got off;
Step 7.4.4, each ICID is traversed, is summed at the T moment in the expectation that website S gets on the bus to it, as bus B1
The volume of passenger traffic in the expectation of website S sums to each ICID in the expectation that website S gets off, as expectation of the bus B1 in website S
Lower volume of passenger traffic, the two subtract each other the expectation passenger capacity as the T moment bus predicted;
Step 7.5, the design load number data for obtaining each bus, by the passenger of itself and each bus at times
It is expected that several comparisons, obtain the real-time expectation degree of crowding statistics and prediction result of each bus at times.
In this example, according to taking L1 at the market L station and its other subsequent websites at 10 points in the morning certain and later on working day
Probability and the probability got off at the market L station and its other subsequent websites of the passenger that has taken, calculate that the expectation of bus RB1 carries
Guest's number and the degree of crowding are shown in Table 14:
14 RB1 of table gets on or off the bus number and crowding after 10 points of the morning of certain working day by the expectation of each website
The purpose of the present invention is utilize swipe the card record and the bus between public transport barcode scanning terminal or IC card and bus
GPS data obtains the space operation data set that individual takes public bus network between sites, excavates dividing at times for a large amount of individuals
Place public bus network takes behavior;By carrying out space clustering to bus station, the neighbouring bus station in geographical location is merged,
Original bus station is replaced in record of swiping the card, provides geographical basis for the transfer behavior during identification bus trip;It is right
Bus station after merging divides Thiessen polygon, determines the spatial dimension of each bus station;Remembered using a large amount of bus card-readings
Record obtains each individual and working day and nonworkdays is divided to take every public bus network in each bus station within each period
Probability, the card-swiping mode different according to public bus network on this basis, be classified as getting on or off the bus swipe the card, get on the bus block meter rate and
It gets on the bus unified charging Three models, calculates separately acquisition individual after different time sections difference website takes different public bus networks
The probability got off in its on the way remaining each website;According to this probability, on the basis for obtaining bus card-reading record in real time
On, the passengers quantity on each current bus is estimated, and to each bus in a period of time later at each station
Point get on or off the bus number and seating capacity and crowding is predicted.The present invention is anonymous public using having magnanimity in public transit system
Swipe the card record and lasting bus GPS location information are handed over, it can low cost, automation, easily acquisition specified time range
The public bus network of interior a large amount of individuals takes behavior, to individual after different periods and place take different bus and get on the bus
The probability that other stations are got off on the way is inferred;To realize quickly and efficiently to real-time and following bus seating capacity
It is calculated and is predicted with the degree of crowding.
Claims (7)
1. the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment, which comprises the following steps:
Step 1 obtains all continuous bus GPS data over time and space, and bus GPS data includes at least public transport line
Longitude station Long and bus where moment TIME1, bus occur for road number LID, bus number BID, communication operation
Place Position Latitude Lat, different buses correspond to different public bus network number LID and bus number BID, extract each
The at the appointed time GPS data in section of bus corresponding to bus number BID, constitutes the bus trip of each bus
Time-space serial;
Step 2, the bus card-reading for obtaining anonymous encryption record data, in time sequencing, often swipe the card and once just generate a public affairs
Friendship, which is swiped the card, records data, and every bus card-reading record data include at least IC number ICID, public bus network number LID, bus
Number BID, communication operation occur moment TIME2, type of getting on or off the bus TYPE and spend COST, different barcode scanning terminals or IC card by bus
Corresponding different IC number ICID, different IC number ICID correspond to Different Individual, extract each IC number ICID at the appointed time
Bus in section is swiped the card record, and brushing card data collection individual corresponding to different IC number ICID is constituted;
Step 3, the bus station for obtaining all public bus networks within the scope of designated space, using Spatial Clustering to all public transport
Website is clustered, and the bus station that merging spatial position is closed on is as the transfer node being likely to occur, to all public bus networks
In bus station be adjusted, with after cluster bus station replacement cluster before bus station, so as to it is subsequent individual
Transfer behavior extracts;
Step 4 arranges the brushing card data collection for all individuals that step 2 obtains, and obtains all all of individual and gets on and off
Site information forms complete passenger's bus card-reading time series data of all individuals, for brushing card data corresponding to current individual
For the current bus card-reading record data concentrated, the spatial position that moment TIME2 bus occurs for present communications movement is obtained,
To obtain the site information of getting on the bus of current bus card-reading record data, and it is added in current bus card-reading record data;
Bus card-reading mode is divided into three kinds by step 5: unit price is swiped the card after segmentation of getting on or off the bus after swiping the card, getting on the bus swipes the card and gets on the bus,
For different card-swiping modes, the when and where of getting on or off the bus of individual is analyzed and determined respectively, comprising the following steps:
The corresponding complete passenger's bus card-reading time series data of all IC number ICID of step 5.1, order traversal, reads its public transport brush
Card record data, according to public bus network number LID swiped the card timing fractionation;
Step 5.2, for every public bus network in each IC number ICID, judge its form of swiping the card, swiped the card according to different
Mode uses get on or off the bus information of the different method statistic individuals on current public bus network, and whether it is inferred to get off a little can push away
It calculates, comprising the following steps:
Step 5.2.1, it is directed to current IC number ICID, reads site information of getting on the bus, the public bus network that it takes bus every time
Moment TIME2 occurs for number LID and pick-up time, i.e. communication operation;
Step 5.2.2, card-swiping modes different for different public bus networks estimates current IC number ICID using different methods
In the information of getting off to take bus every time;
If the card-swiping mode of current public bus network is to get on or off the bus to be intended to swipe the card, current public involved in public transport barcode scanning terminal or IC card
The information of getting off for the bus for taking current public bus network every time has been contained in the record of swiping the card on intersection road, therefore, for
Current IC number ICID obtains its all known record of getting off, comprising the following steps: if it has on current public bus network
It swipes the card and records data R1, and judge corresponding bus station S1 to be upper according to the type TYPE that gets on or off the bus recorded in data R1 that swipes the card
Station point, then judgement swipe the card record next of data R1 swipe the card record in data R2 get on or off the bus type TYPE whether be under
Vehicle records the corresponding bus station S2 of data R2 if so, swiping the card as get-off stop, will swipe the card and record data R1 and record of swiping the card
Data R2 merges, and is designated as a known record of getting off, if the type TYPE that gets on or off the bus recorded in data R2 that swipes the card is to get on the bus, shows
Current IC number ICID does not get off in a upper stroke to swipe the card, then discarded swipe the card records data R1, reacquires next
It swipes the card and records data and judged;
If the card-swiping mode of current public bus network is to get on the bus to swipe the card, and block meter rate, then obtaining that all IC number ICID are all can
Reckoning gets off and record a little and can not calculate the record got off a little, comprising the following steps:
If current IC number ICID swipes the card on current public bus network during period T1 records data R3, corresponding public transport
Website is S3, then:
Step 5.2.2.1, the possible get-off stop section of passenger is calculated according to the COST of cost by bus for recording data R3 that swipes the card;
Step 5.2.2.2, it reads to swipe the card to record next of data R3 and swipe the card and records data R4, to record data R4 corresponding if swiping the card
Starting bus station S4 in the get-off stop section that step 5.2.2.1 is calculated, then it is assumed that current IC number ICID multiplies
The get-off stop for sitting the bus of current public bus network is S4, while will swipe the card to record data R3 and be recorded as to calculate and get off a little
Record;The get-off stop that the corresponding starting bus station S4 of data R4 is not calculated in step 5.2.2.1 is recorded if swiping the card
Within section, then will swipe the card records data R3 and is recorded as not calculating the record got off a little;
If the card-swiping mode of current public bus network is to get on the bus to swipe the card, and unified charging, then obtaining that all IC number ICID are all can
Reckoning gets off and record a little and can not calculate the record got off a little, comprising the following steps:
If current IC number ICID swipes the card on current public bus network during period T1 records data R5, swipes the card and record number
It is S5 according to the corresponding bus stop of getting on the bus R5, reads to swipe the card to record next of data R5 and swipe the card and record data R6:
Starting bus station S6 in data R6 is recorded in current public bus network after swiping the card and recording data R5 if swiping the card
Along the line in station, and swipes the card and record the communication operation generation moment TIME2 and current public bus network to starting bus station of data R6
The difference of point S6 time is in threshold range T_Thrh, then it is assumed that current IC number ICID gets off in starting bus station S6, together
When will swipe the card and record data R5 and be recorded as to calculate the record got off a little;
Starting bus station S6 in data R6 is recorded not in current public bus network after swiping the card and recording data R5 if swiping the card
Along in station, then will swipe the card records data R6 and is recorded as not calculating the record got off a little;
The current IC number ICID frequency occurred of getting on the bus all in current public bus network is counted, is got off as that can not calculate
The foundation of point statistics;
Step 6, for individual corresponding to each IC number ICID, count individual one week and daily take every public transport at times
The probability of route, and its probability that each website is got off on the way after getting on the bus obtain the probability distribution in its section O-D of going on a journey;
Step 7, the GPS data that each car in real-time bus card-reading data and every public bus network is obtained from data source are excavated
The situation by bus of the potential passenger of public bus network, the probability distribution in the trip section O-D according to step 6 acquisition calculate every public transport line
The possible real-time passenger capacity of each car and the degree of crowding in road, carrying demand to bus in the following designated time period and crowded
Degree is predicted;
2. the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment as described in claim 1, which is characterized in that
In the step 1, the GPS data of a bus is a space-time trajectory record, is numbered according to the public bus network of bus
LID and bus number BID inquires each bus at the appointed time all space-time trajectories records in section, space-time trajectory is remembered
Warp, latitude in record are converted into geographical coordinate, to construct bus trip Time-space serial.
3. the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment as described in claim 1, which is characterized in that
In the step 2, it further includes spending COST by bus that bus card-reading, which records data,.
4. the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment as described in claim 1, which is characterized in that
The step 3 includes:
Step 3.1 obtains the bus station of all public bus networks and the location information of each bus station within the scope of designated space, will
Location information is converted into XY coordinate, and XY coordinate is mapped in the geographical space with traffic route;
Step 3.2 clusters bus station using the traffic distance between bus station as standard using spatial clustering method,
Merge very close bus station on space length, comprising the following steps:
Step 3.2.1, it sets cluster standard and is less than d meters as the distance between two bus stations;
Step 3.2.2, using each bus station as a cluster core, space clustering: the space bit currently to cluster core is obtained
It is set to the bus station that its periphery is searched in the center of circle, is less than d meters of bus station with its traffic distance if it exists, then by the bus station
Point is put into the Cluster space of current cluster core;
Step 3.2.3, the space clustering obtained by step 3.2.2 is merged, is constituted spatially relatively independent larger
Bus station space clustering, merging condition is: if any two space clustering there are identical bus station, will be current
Two spaces Cluster merging;
Step 3.2.4, the space center for extracting each bus station space clustering, maps that on map, obtains current empty
Between center spatial position and geographic name, by each bus station space clustering bus station merge after, with each public affairs
Bus station after handing over the geographic name name of the space center of website space clustering to merge, the XY coordinate at current spatial center are
The mean value of all bus station XY coordinates in corresponding bus station space clustering;
Step 3.3 rearranges every public bus network, the bus station before being clustered with the bus station replacement after cluster, so as to
The transfer behavior of subsequent individual is extracted.
5. the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment as described in claim 1, which is characterized in that
In the step 4, the bus loading zone for the current bus card-reading record data that brushing card data corresponding to current individual is concentrated is obtained
Point information the following steps are included:
Step 4.1, to the bus stop for having already passed through space clustering within the scope of designated space, generated according to road traffic net safe
Gloomy polygon divides the spatial dimension of each bus stop;
According to the communication operation in bus card-reading record data moment TIME2, public bus network number LID and public affairs occur for step 4.2
Vehicle number BID is handed over, communication is read from the GPS data of bus corresponding to public bus network number LID and bus number BID
The location information X-IC and Y-IC of the position of moment TIME2 bus occur for movement;
Step 4.3, the sky that location information X-IC and Y-IC that step 4.2 obtains are mapped to the bus stop that step 4.1 generates
Between in range, obtain communication operation and bus stop where moment TIME2 bus occur, to obtain current bus card-reading note
Record the site information of getting on the bus of data.
6. the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment as described in claim 1, which is characterized in that
The step 6 includes:
Step 6.1 is directed to each IC number ICID, counts it at the appointed time section is interior and take public bus network L's in specific time period T
Number, if the IC number ICID public transport barcode scanning terminal or IC card for IC1, in bus stop, S takes pubic transport in period T
The probability of route L is that P_U (T, S, L, IC1) is N_U (T, S, L, IC1)/N_Day, and in formula: N_U (T, S, L, IC1) is specified
IC1 gets on the bus the number of route L of taking pubic transport in daily period T in bus stop S in period, and N_Day is period T
Interior number of days N_Day;
Step 6.2, step 5 obtain to every time by bus record get off estimation on the basis of, it is different for each public bus network
Card-swiping mode using different method statistic individuals each bus station probability of getting off;
If the card-swiping mode of public bus network L is to get on or off the bus to be intended to swipe the card, counting IC1, at the appointed time section T is interior within the T1 period
The number that each station is got off on the way under conditions of S1 gets on the bus in bus stop, then IC1 is in period T1 in bus stop S1
Get on the bus after taking pubic transport route L, the probability that S2 gets off in bus station be N_D (S1, S2, L, IC1)/N_U (T1, S1, L,
IC1), in formula, N_U (T1, S1, L, IC1) gets on the bus for IC1 S1 in bus stop in period T1 in designated time period and takes public affairs
The number of intersection road L;
If the card-swiping mode of public bus network L is swiped the card to get on the bus, and block meter rate, then to the frequency got off a little depending on its record of swiping the card
Whether a little can calculate separated statistics: for that can not calculate the record got off a little if getting off, it is assumed that an IC number ICID takes pubic transport
Vehicle generally has a continuity, i.e., its get on the bus be a little its it is another wade out it is capable get off a little, therefore count current IC number ICID in public affairs
All frequencys occurred of getting on the bus in the L of intersection road, as can not calculate a foundation for statistics of getting off:
For that can calculate the record got off a little, statistics IC1 S3 from bus stop within the T1 period gets on the bus take pubic transport route L after
The frequency N_D (S3, S4, D, L, IC1) that bus stop S4 gets off along it, then IC1 is in period T1 from bus stop S3
Take pubic transport route L to the probability that bus stop S4 gets off be N_D (S3, S4, D, L, IC1)/N_U (T1, S3, L, IC1), formula
In: N_U (T1, S3, L, IC1) is that IC1 S3 in bus stop in period T1 gets on the bus the route L that takes pubic transport in designated time period
Number;
For that can not calculate the record got off a little, statistics IC1 take pubic transport route L historical record in the upper of bus stop S4
Vehicle frequency N_U (S4, L, IC1), then IC1 is after in bus stop, S3 takes public bus network L in period T1, in remaining route
The probability that upper bus stop S4 gets off is N_U (S4, L, IC1)/sum (N_U (SN, L, IC1)), and in formula, SN is IC1 in bus
Website Hosting after the S3 that stands gets on the bus in public bus network L residual paths, sum (N_U (SN, L, IC1)) are IC1 on the S3 of bus stop
The frequency that website after vehicle in residual paths is got on the bus, for substituting the frequency got off;
If the card-swiping mode of public bus network L is swiped the card to get on the bus, and unified charging, then to the frequency got off a little depending on its record of swiping the card
Get off whether can a little calculate also need separate statistics:
For that can calculate the record got off a little, statistics IC1 S5 from bus stop in period T1 gets on the bus take pubic transport route L after
The frequency N_D (S5, S6, L, IC1) that bus stop S6 along it gets off, then IC1 is in period T1 from bus stop S5
Take pubic transport route L to the probability that bus stop S6 gets off be N_D (S5, S6, L, IC1)/N_U (T1, S5, L, IC1), formula
In: N_U (T1, S5, L, IC1) is that IC1 S5 in bus stop in period T1 gets on the bus the route L that takes pubic transport in designated time period
Number;
For that can not calculate the record got off a little, statistics IC1 take pubic transport route L historical record in the upper of bus stop S6
Vehicle frequency N_S6_U_IC1, then IC1 is after in bus stop, S5 takes public bus network L in period T1, on remaining route
The probability got off of bus stop S6 be N_U (S6, L, IC1)/sum (N_U (SN, L, IC1)), in formula, SN is IC1 in bus
Website Hosting after the S5 that stands gets on the bus in public bus network L residual paths, sum (N_U (SN, L3, IC1)) are IC1 in bus stop S5
The frequency that website after getting on the bus in residual paths is got on the bus, for substituting the frequency got off;
It gets off a little under step 6.3, the three kinds of card-swiping modes obtained according to step 6.2 for that can calculate to get off a little and can not calculate
Swipe the card and got off the statistics of probability in record for each website, obtained by the way of record number weighting IC1 in period T1
The probability that SS gets off in bus stop again after bus stop S gets on the bus:
For the public bus network for being intended to swipe the card of getting on or off the bus, all effective records have completely get on or off the bus information, therefore step
6.2.1 obtain be exactly final IC1 get off in bus stop SS after in bus stop, S gets on the bus in period T1 probability P _
D (T1, S, SS, L, IC1);
For the public bus network swiped the card of only getting on the bus, record judgement of swiping the card be divided into get off a deducibility why not two kinds of deducibility,
Statistics can calculate the record got off a little and the quantity N_C (T1, S, L, IC1) and N_NC that can not calculate the record got off a little respectively
(T1, S, L, IC1) and total number of records N (T1, S, L, IC1)=N_C (T1, S, L, IC1)+N_NC (T1, S, L, IC1) are used
Record number weighting pattern, IC1 takes public bus network L after bus stop S gets on the bus again at the SS of bus stop in period T1
Probability P _ D (T1, S, SS, L, IC1) of vehicle=N_D (T1, S, SS, L, IC1)/N_U (T1, S, L, IC1) * N_C (T1, S, L,
IC1)/N (T1, S, L, IC1)+N_U (SS, L, IC1)/sum (N_U (SN, L, IC1)) * N_NC (T1, S, L, IC1)/N (T1, S,
L, IC1);
Step 6.4, all IC number ICID of traversal obtain corresponding each individual in period T1 in bus stop S website
The probability of public bus network L is taken, and the probability that each bus stop is got off on the way thereafter;And it calculates based on this every
In bus stop, S takes the probability of all public bus networks to individual in period T1, and thereafter in each bus on the way
The probability that station is got off;And each individual is calculated based on this and takes all public transport in all bus stops in period T1
The probability of route, and the probability that each bus stop is got off on the way thereafter;Each individual is finally calculated at one day
The probability of all public bus networks is taken in all bus stops in all periods, and each bus stop is got off on the way thereafter
Probability.
7. the real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment as described in claim 1, which is characterized in that
The step 7 includes:
Step 7.1 obtains real-time bus card-reading data and bus GPS data that time interval is TM from data source, to swiping the card
Record data are arranged, and carry out classification and ordination according to public bus network number LID and IC number ICID;
Step 7.2, the last item for reading each IC number ICID, which are swiped the card, records data, obtains what current IC number ICID took
Public bus network number LID, bus number BID, type of getting on or off the bus TYPE pass through bus GPS number according to the method for step 4
According to the site information of getting on the bus for obtaining current IC number ICID, then judge it current according to different public bus network card-swiping modes
The situation by bus of timing node:
If current public bus network is to get on or off the bus to swipe the card, and it is to get off that last swiping the card, which records data, in current IC number ICID,
Then think current IC number ICID currently without taking bus, then its phase for taking the bus in current public bus network in real time
Hope to be 0;
If current public bus network is to get on or off the bus to swipe the card, and it is to get on the bus that last swiping the card, which records data, in current IC number ICID,
Then think that current IC number ICID currently takes the bus in current public bus network, then it is taken in real time in current public bus network
Bus be desired for 1;
If current public bus network is to get on the bus to swipe the card, each ICID obtained according to step 6 vehicle time, that is, communication operation on it
Occur in the period of moment TIME2 after the bus that bus stop SL takes current public bus network, on the way, each station is got off
Probability, obtain the current bus stop position of current bus according to GPS data, calculate current IC number ICID in current public affairs
Hand over the probability do not got off yet of station SN, probability value be 1 subtract current IC number ICID bus stop SL and bus stop SN it
Between the sum of the probability got off of bus stop SM PT (TR, SN, L, ICID)=1-sum (P_D (TU, SL, SM, L, ICID)), TR
It indicates in real time, then what current IC number ICID took pubic transport the bus in route in real time is desired for E (TR, L, ICID)=PT
(TR, SN, L, ICID);
Step 7.3, be directed to each bus, count all last times swipe the card behavior occur current bus current IC compile
The expectation of the real-time seating current bus of number ICID, it is expected that the sum of be expectation that current bus carries number in real time;
Step 7.4, each IC number ICID probability that branch website takes bus at times obtained according to step 6, from real-time
The carrying demand expectation of every time bus routes between sites in designated time period TPJ is predicted backward, comprising the following steps:
Step 7.4.1, according to the GPS data of bus and its arrangement of dispatching a car, predict every time public bus network in the TPJ period first
The time interval of situation of inside dispatching a car and each bus stop of each Public Transit Bus Stopping;
Step 7.4.2, it is directed to each bus, traverses all IC number ICID, searches and continues behind at the appointed time section TPJ
There is the IC number ICID for record of getting on the bus in each website on the way;
Step 7.4.3, each IC number ICID obtained for step 7.4.2, according to its in period T in bus stop S
Probability P _ the U (T, S, L, ICID) to get on the bus, calculating each IC number ICID, in bus stop, S takes bus in period T
Expectation calculated every according to the getting off probability P _ D_prj (TU, SL, S, L, ICID) in bus stop S of the passenger in vehicle
The expectation that a passenger that may have taken current bus gets off in period T in bus stop S;Due in the pre- of passenger capacity
It is a probability distribution for whether the IC number ICID of prediction gets on the bus in survey, and carries statistics difference in real time, therefore in upper garage
It in the probability of getting off of bus stop S is passenger in the bus stop SL probability got on the bus and in bus stop S for the passenger for prediction
It gets off the product of probability: P_D_prj (TU, SL, S, L, ICID)=P_U (TU, SL, L, ICID) * P_D (TU, SL, S, L,
ICID), current IC number ICID bus stop S get off be desired for E_D (T, S, L, ICID)=P_D_prj (TU, SL, S,
L, ICID);
Step 7.4.4, each IC number ICID is traversed, is summed in period T in the expectation that bus stop S gets on the bus to it, as
Current bus volume of passenger traffic in the expectation of bus stop S sums to each IC number ICID in the expectation that bus stop S gets off,
As current bus volume of passenger traffic under the expectation of bus stop S, the two subtract each other the current public affairs in the period T predicted
Hand over the expectation passenger capacity of vehicle;
Step 7.5, the design load number data for obtaining each bus, by the expectation carrying of itself and each bus at times
Amount compares, and obtains the real-time expectation degree of crowding statistics and prediction result of each bus at times.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810860244.5A CN109035770B (en) | 2018-07-31 | 2018-07-31 | Real-time analysis and prediction method for bus passenger capacity in big data environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810860244.5A CN109035770B (en) | 2018-07-31 | 2018-07-31 | Real-time analysis and prediction method for bus passenger capacity in big data environment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109035770A true CN109035770A (en) | 2018-12-18 |
CN109035770B CN109035770B (en) | 2022-01-04 |
Family
ID=64648252
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810860244.5A Active CN109035770B (en) | 2018-07-31 | 2018-07-31 | Real-time analysis and prediction method for bus passenger capacity in big data environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109035770B (en) |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110009124A (en) * | 2019-01-03 | 2019-07-12 | 阿里巴巴集团控股有限公司 | A kind of data processing method, server and the system of determining bus station |
CN110211379A (en) * | 2019-05-27 | 2019-09-06 | 南京航空航天大学 | A kind of public transport method for optimizing scheduling based on machine learning |
CN110223116A (en) * | 2019-06-06 | 2019-09-10 | 武汉元光科技有限公司 | Public transport network information questionnaire method and device |
CN110322694A (en) * | 2019-07-16 | 2019-10-11 | 青岛海信网络科技股份有限公司 | A kind of method and device of urban traffic control piece Division |
CN110647929A (en) * | 2019-09-19 | 2020-01-03 | 京东城市(北京)数字科技有限公司 | Method for predicting travel destination and method for training classifier |
CN111339159A (en) * | 2020-02-24 | 2020-06-26 | 交通运输部科学研究院 | Analysis and mining method for one-ticket public transportation data |
CN111400419A (en) * | 2019-01-03 | 2020-07-10 | 腾讯科技(深圳)有限公司 | Method and equipment for fusing homonymous bus stations in electronic map |
CN111402618A (en) * | 2020-03-27 | 2020-07-10 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining boarding station, storage medium and electronic equipment |
CN111476494A (en) * | 2020-04-11 | 2020-07-31 | 重庆交通开投科技发展有限公司 | Method for accurately analyzing geographic distribution of public transport population based on multi-source data |
CN111491261A (en) * | 2020-04-24 | 2020-08-04 | 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) | Individual movement track extraction method based on intelligent card swiping data |
CN111741051A (en) * | 2020-04-14 | 2020-10-02 | 腾讯科技(深圳)有限公司 | Method and device for determining full load rate of vehicle, storage medium and electronic device |
CN112288131A (en) * | 2020-09-24 | 2021-01-29 | 和智信(山东)大数据科技有限公司 | Bus stop optimization method, electronic device and computer-readable storage medium |
CN112949939A (en) * | 2021-03-30 | 2021-06-11 | 福州市电子信息集团有限公司 | Taxi passenger carrying hotspot prediction method based on random forest model |
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 |
CN112991722A (en) * | 2021-02-03 | 2021-06-18 | 浙江浙大中控信息技术有限公司 | Method and system for predicting real-time intersection of bus at high-frequency gps point |
CN113658433A (en) * | 2021-08-18 | 2021-11-16 | 苏州工业园区测绘地理信息有限公司 | Method for extracting passenger flow characteristics based on bus card swiping and code scanning data |
CN113780851A (en) * | 2021-09-16 | 2021-12-10 | 上海世脉信息科技有限公司 | Expected driving cost-based bus parking lot site selection method |
CN113781822A (en) * | 2021-09-24 | 2021-12-10 | 湖北惠诚共创科技有限公司 | Bus dispatching system based on big data |
CN114241770A (en) * | 2021-12-21 | 2022-03-25 | 杭州图软科技有限公司 | Bus scheduling method and system based on accurate real-time information |
CN115186049A (en) * | 2022-09-06 | 2022-10-14 | 深圳市城市交通规划设计研究中心股份有限公司 | Intelligent bus alternative station site selection method, electronic equipment and storage medium |
CN117542181A (en) * | 2024-01-10 | 2024-02-09 | 四川三思德科技有限公司 | Real-time abnormality early warning method and system for multi-mode deep learning |
Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101615340A (en) * | 2009-07-24 | 2009-12-30 | 北京工业大学 | Real-time information processing method in the bus dynamic dispatching |
JP2010061321A (en) * | 2008-09-03 | 2010-03-18 | Railway Technical Res Inst | Passenger flow prediction system |
US20100299177A1 (en) * | 2009-05-22 | 2010-11-25 | Disney Enterprises, Inc. | Dynamic bus dispatching and labor assignment system |
CN103198565A (en) * | 2013-04-12 | 2013-07-10 | 王铎源 | Charge and passenger flow information acquisition method for bus IC (integrated circuit) cards |
CN103730008A (en) * | 2014-01-15 | 2014-04-16 | 汪涛 | Bus congestion degree analysis method based on real-time data of bus GPS (Global Position System) and IC (Integrated Circuit) cards |
CN103778778A (en) * | 2013-12-19 | 2014-05-07 | 银江股份有限公司 | Rapid bus station service information system and rapid bus arrival information measuring and calculating method |
US20140289003A1 (en) * | 2013-03-25 | 2014-09-25 | Amadeus S.A.S. | Methods and systems for detecting anomaly in passenger flow |
CN104463364A (en) * | 2014-12-04 | 2015-03-25 | 中国科学院深圳先进技术研究院 | Subway passenger real-time distribution and subway real-time density prediction method and system |
CN105427594A (en) * | 2015-11-23 | 2016-03-23 | 青岛海信网络科技股份有限公司 | Bidirectional-getting-on-passenger-flow-based obtaining method and system for bus section passenger flow volume |
CN105528457A (en) * | 2015-12-28 | 2016-04-27 | 重庆云途交通科技有限公司 | Traffic information extraction and inquiry method based on big data technology |
CN105550789A (en) * | 2016-02-19 | 2016-05-04 | 上海果路交通科技有限公司 | Method for predicting bus taking passenger flow |
CN105868861A (en) * | 2016-04-08 | 2016-08-17 | 青岛海信网络科技股份有限公司 | Bus passenger flow evolution analysis method based on time-space data fusion |
CN106297288A (en) * | 2016-08-23 | 2017-01-04 | 同济大学 | A kind of bus passenger passenger flow data gathers and the method for analysis |
CN106448155A (en) * | 2016-08-29 | 2017-02-22 | 北京工业大学 | System for real-time transmitting crowded level inside bus based on public transport IC card |
CN107563540A (en) * | 2017-07-25 | 2018-01-09 | 中南大学 | A kind of public transport in short-term based on random forest is got on the bus the Forecasting Methodology of the volume of the flow of passengers |
CN108242149A (en) * | 2018-03-16 | 2018-07-03 | 成都智达万应科技有限公司 | A kind of big data analysis method based on traffic data |
-
2018
- 2018-07-31 CN CN201810860244.5A patent/CN109035770B/en active Active
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010061321A (en) * | 2008-09-03 | 2010-03-18 | Railway Technical Res Inst | Passenger flow prediction system |
US20100299177A1 (en) * | 2009-05-22 | 2010-11-25 | Disney Enterprises, Inc. | Dynamic bus dispatching and labor assignment system |
CN101615340A (en) * | 2009-07-24 | 2009-12-30 | 北京工业大学 | Real-time information processing method in the bus dynamic dispatching |
US20140289003A1 (en) * | 2013-03-25 | 2014-09-25 | Amadeus S.A.S. | Methods and systems for detecting anomaly in passenger flow |
CN103198565A (en) * | 2013-04-12 | 2013-07-10 | 王铎源 | Charge and passenger flow information acquisition method for bus IC (integrated circuit) cards |
CN103778778A (en) * | 2013-12-19 | 2014-05-07 | 银江股份有限公司 | Rapid bus station service information system and rapid bus arrival information measuring and calculating method |
CN103730008A (en) * | 2014-01-15 | 2014-04-16 | 汪涛 | Bus congestion degree analysis method based on real-time data of bus GPS (Global Position System) and IC (Integrated Circuit) cards |
CN104463364A (en) * | 2014-12-04 | 2015-03-25 | 中国科学院深圳先进技术研究院 | Subway passenger real-time distribution and subway real-time density prediction method and system |
CN105427594A (en) * | 2015-11-23 | 2016-03-23 | 青岛海信网络科技股份有限公司 | Bidirectional-getting-on-passenger-flow-based obtaining method and system for bus section passenger flow volume |
CN105528457A (en) * | 2015-12-28 | 2016-04-27 | 重庆云途交通科技有限公司 | Traffic information extraction and inquiry method based on big data technology |
CN105550789A (en) * | 2016-02-19 | 2016-05-04 | 上海果路交通科技有限公司 | Method for predicting bus taking passenger flow |
CN105868861A (en) * | 2016-04-08 | 2016-08-17 | 青岛海信网络科技股份有限公司 | Bus passenger flow evolution analysis method based on time-space data fusion |
CN106297288A (en) * | 2016-08-23 | 2017-01-04 | 同济大学 | A kind of bus passenger passenger flow data gathers and the method for analysis |
CN106448155A (en) * | 2016-08-29 | 2017-02-22 | 北京工业大学 | System for real-time transmitting crowded level inside bus based on public transport IC card |
CN107563540A (en) * | 2017-07-25 | 2018-01-09 | 中南大学 | A kind of public transport in short-term based on random forest is got on the bus the Forecasting Methodology of the volume of the flow of passengers |
CN108242149A (en) * | 2018-03-16 | 2018-07-03 | 成都智达万应科技有限公司 | A kind of big data analysis method based on traffic data |
Non-Patent Citations (4)
Title |
---|
暨育雄,妥思佳,米苏拉尼莱比,麦寇德马克: "《基于上下车人数的公交客流起迄点估计方法》", 《同济大学学报(自然科学版)》 * |
朱翔希: "《基于IC卡数据的短时公交客流预测》", 《中国优秀硕士学位论文全文数据库 (工程科技Ⅱ辑)》 * |
杨思源: "《成都市快速公交乘客刷卡数据研究》", 《中国优秀硕士学位论文全文数据库 经济与管理科学辑》 * |
葛柯娜,马涛,王晶,陈聪,孙艺萍: "《基于Android平台的车载GPS公交车次与载客量查询系统的设计与实现》", 《科技资讯》 * |
Cited By (32)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111400419A (en) * | 2019-01-03 | 2020-07-10 | 腾讯科技(深圳)有限公司 | Method and equipment for fusing homonymous bus stations in electronic map |
CN110009124A (en) * | 2019-01-03 | 2019-07-12 | 阿里巴巴集团控股有限公司 | A kind of data processing method, server and the system of determining bus station |
CN111400419B (en) * | 2019-01-03 | 2023-10-27 | 腾讯科技(深圳)有限公司 | Method and equipment for fusing same-name bus stops in electronic map |
CN110211379A (en) * | 2019-05-27 | 2019-09-06 | 南京航空航天大学 | A kind of public transport method for optimizing scheduling based on machine learning |
CN110223116B (en) * | 2019-06-06 | 2021-07-06 | 武汉元光科技有限公司 | Public transport network information questionnaire investigation method and device |
CN110223116A (en) * | 2019-06-06 | 2019-09-10 | 武汉元光科技有限公司 | Public transport network information questionnaire method and device |
CN110322694A (en) * | 2019-07-16 | 2019-10-11 | 青岛海信网络科技股份有限公司 | A kind of method and device of urban traffic control piece Division |
CN110647929A (en) * | 2019-09-19 | 2020-01-03 | 京东城市(北京)数字科技有限公司 | Method for predicting travel destination and method for training classifier |
CN111339159A (en) * | 2020-02-24 | 2020-06-26 | 交通运输部科学研究院 | Analysis and mining method for one-ticket public transportation data |
CN111339159B (en) * | 2020-02-24 | 2023-08-18 | 交通运输部科学研究院 | Analysis mining method for one-ticket public transport data |
CN111402618A (en) * | 2020-03-27 | 2020-07-10 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining boarding station, storage medium and electronic equipment |
CN111476494A (en) * | 2020-04-11 | 2020-07-31 | 重庆交通开投科技发展有限公司 | Method for accurately analyzing geographic distribution of public transport population based on multi-source data |
CN111741051A (en) * | 2020-04-14 | 2020-10-02 | 腾讯科技(深圳)有限公司 | Method and device for determining full load rate of vehicle, storage medium and electronic device |
CN111741051B (en) * | 2020-04-14 | 2021-08-27 | 腾讯科技(深圳)有限公司 | Method and device for determining full load rate of vehicle, storage medium and electronic device |
CN111491261B (en) * | 2020-04-24 | 2022-03-01 | 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) | Individual movement track extraction method based on intelligent card swiping data |
CN111491261A (en) * | 2020-04-24 | 2020-08-04 | 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) | Individual movement track extraction method based on intelligent card swiping data |
CN112288131A (en) * | 2020-09-24 | 2021-01-29 | 和智信(山东)大数据科技有限公司 | Bus stop optimization method, electronic device and computer-readable storage medium |
CN112288131B (en) * | 2020-09-24 | 2021-06-11 | 和智信(山东)大数据科技有限公司 | Bus stop optimization method, electronic device and computer-readable storage medium |
CN112991722A (en) * | 2021-02-03 | 2021-06-18 | 浙江浙大中控信息技术有限公司 | Method and system for predicting real-time intersection of bus at high-frequency gps point |
CN112991722B (en) * | 2021-02-03 | 2022-07-19 | 浙江中控信息产业股份有限公司 | High-frequency gps (gps) point bus real-time intersection prediction method and system |
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 |
CN112966218B (en) * | 2021-02-26 | 2023-06-16 | 佳都科技集团股份有限公司 | Real-time calculation method and device for passenger carrying number of carriage |
CN112949939A (en) * | 2021-03-30 | 2021-06-11 | 福州市电子信息集团有限公司 | Taxi passenger carrying hotspot prediction method based on random forest model |
CN113658433A (en) * | 2021-08-18 | 2021-11-16 | 苏州工业园区测绘地理信息有限公司 | Method for extracting passenger flow characteristics based on bus card swiping and code scanning data |
CN113780851A (en) * | 2021-09-16 | 2021-12-10 | 上海世脉信息科技有限公司 | Expected driving cost-based bus parking lot site selection method |
CN113780851B (en) * | 2021-09-16 | 2024-04-26 | 上海世脉信息科技有限公司 | Bus parking lot address selection method based on expected driving cost |
CN113781822A (en) * | 2021-09-24 | 2021-12-10 | 湖北惠诚共创科技有限公司 | Bus dispatching system based on big data |
CN114241770A (en) * | 2021-12-21 | 2022-03-25 | 杭州图软科技有限公司 | Bus scheduling method and system based on accurate real-time information |
CN115186049B (en) * | 2022-09-06 | 2023-02-03 | 深圳市城市交通规划设计研究中心股份有限公司 | Intelligent bus alternative station site selection method, electronic equipment and storage medium |
CN115186049A (en) * | 2022-09-06 | 2022-10-14 | 深圳市城市交通规划设计研究中心股份有限公司 | Intelligent bus alternative station site selection method, electronic equipment and storage medium |
CN117542181A (en) * | 2024-01-10 | 2024-02-09 | 四川三思德科技有限公司 | Real-time abnormality early warning method and system for multi-mode deep learning |
CN117542181B (en) * | 2024-01-10 | 2024-04-30 | 四川三思德科技有限公司 | Real-time abnormality early warning method and system for multi-mode deep learning |
Also Published As
Publication number | Publication date |
---|---|
CN109035770B (en) | 2022-01-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109035770A (en) | The real-time analyzing and predicting method of public transport passenger capacity under a kind of big data environment | |
Kondor et al. | Estimating savings in parking demand using shared vehicles for home–work commuting | |
Zhong et al. | Inferring building functions from a probabilistic model using public transportation data | |
Gundlegård et al. | Travel demand estimation and network assignment based on cellular network data | |
Phithakkitnukoon et al. | Taxi-aware map: Identifying and predicting vacant taxis in the city | |
CN105142106A (en) | Traveler home-work location identification and trip chain depicting method based on mobile phone signaling data | |
CN105809962A (en) | Traffic trip mode splitting method based on mobile phone data | |
He et al. | Evaluation of city-scale built environment policies in New York City with an emerging-mobility-accessible synthetic population | |
Coffey et al. | Time of arrival predictability horizons for public bus routes | |
Wu et al. | Human-centric urban transit evaluation and planning | |
CN108733774A (en) | A kind of unemployment dynamic monitoring method based on big data | |
CN110110243A (en) | A kind of historical track destination prediction technique based on echo state network | |
Breyer et al. | Comparative analysis of travel patterns from cellular network data and an urban travel demand model | |
Fourie et al. | Using smartcard data for agent-based transport simulation | |
Di Lorenzo et al. | Identifying human spatio-temporal activity patterns from mobile-phone traces | |
Liu et al. | Destination estimation for bus passengers based on data fusion | |
Duan et al. | MobilePulse: Dynamic profiling of land use pattern and OD matrix estimation from 10 million individual cell phone records in Shanghai | |
Straub et al. | Forecasting the spatial and temporal charging demand of fully electrified urban private car transportation based on large-scale traffic simulation | |
Mepparambath et al. | A novel modelling approach of integrated taxi and transit mode and route choice using city-scale emerging mobility data | |
Ross-Perez et al. | Identifying trip purpose from a dockless bike-sharing system in Manchester | |
Cui et al. | Usage demand forecast and quantity recommendation for urban shared bicycles | |
Beaulieu et al. | A dynamic mixed logit model with agent effect for pedestrian next location choice using ubiquitous Wi-Fi network data | |
CN110263783A (en) | Multiple features charging addressing analysis of Influential Factors method and system based on deep learning | |
Batran et al. | Urban travel time estimation in greater maputo using mobile phone big data | |
CN114169632A (en) | Passenger flow distribution prediction method and device based on multi-view passenger flow change trend modeling |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |