CN110516866A - A kind of real-time estimation method for handing over subway crowding for city rail - Google Patents
A kind of real-time estimation method for handing over subway crowding for city rail Download PDFInfo
- Publication number
- CN110516866A CN110516866A CN201910774323.9A CN201910774323A CN110516866A CN 110516866 A CN110516866 A CN 110516866A CN 201910774323 A CN201910774323 A CN 201910774323A CN 110516866 A CN110516866 A CN 110516866A
- Authority
- CN
- China
- Prior art keywords
- flow
- passengers
- period
- crowding
- volume
- 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
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000005259 measurement Methods 0.000 claims abstract description 37
- 238000013528 artificial neural network Methods 0.000 claims abstract description 12
- 239000000523 sample Substances 0.000 claims description 32
- 238000001514 detection method Methods 0.000 claims description 7
- 238000013527 convolutional neural network Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 3
- 230000000717 retained effect Effects 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000000630 rising effect Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000033001 locomotion Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 235000017060 Arachis glabrata Nutrition 0.000 description 1
- 244000105624 Arachis hypogaea Species 0.000 description 1
- 235000010777 Arachis hypogaea Nutrition 0.000 description 1
- 235000018262 Arachis monticola Nutrition 0.000 description 1
- 241000208340 Araliaceae Species 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000004218 nerve net Anatomy 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 235000020232 peanut Nutrition 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
- H04W4/42—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for mass transport vehicles, e.g. buses, trains or aircraft
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Pure & Applied Mathematics (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Operations Research (AREA)
- Evolutionary Computation (AREA)
- Marketing (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Aviation & Aerospace Engineering (AREA)
- Entrepreneurship & Innovation (AREA)
- Signal Processing (AREA)
- Multimedia (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Development Economics (AREA)
- Evolutionary Biology (AREA)
- Game Theory and Decision Science (AREA)
- Computer Networks & Wireless Communication (AREA)
Abstract
The invention belongs to the technical fields that city rail hands over intelligent management, disclose a kind of real-time estimation method that subway crowding is handed over for city rail, for same train by the crowding prediction of operation sections different on route, including using certain number alignment vehicle by certain operation section originate the moment after period before T1 time and arrival time between the T2 time as a measurement period, rush hour current signature day section is divided into N number of measurement period;The WiFi volume of the flow of passengers, the video passenger flow amount of vehicle in current signature day each measurement period are calculated, a upper characteristic day corresponds to the AFC section volume of the flow of passengers in measurement period;Step 3: utilizing neural network, with the WiFi volume of the flow of passengers, the video passenger flow amount in current statistic period current signature day, a upper characteristic day corresponds to the AFC section volume of the flow of passengers in measurement period as input, the practical volume of the flow of passengers of vehicle in current statistic period real-time estimation current signature day, to complete the real-time estimation to train crowding.
Description
Technical field
The invention belongs to the technical fields that city rail hands over intelligent management, and in particular to one kind hands over subway crowded for city rail
The real-time estimation method of degree.
Background technique
With the continuous development of Chinese national economy, life of urban resident is horizontal constantly to be promoted, as the big function in city four
One of urban transportation increasingly by everybody attention, wherein train by its trip feature safely, conveniently, on schedule by
The favor of numerous city dwellers, but since passenger flow has time, distribution character spatially, the passenger flow on subway train will appear
Local congestion phenomenon, driving low efficiency difficult so as to cause passenger loading, the problems such as passenger traffic risk is big.Therefore, accurately to column
Car team vehicle crowding carries out real-time judgment, arranges the adjustment of subway train passenger traffic scheme, guarantee subway train traffic safety, raising
Vehicle capacity utilization rate, raising train service level all have significance.
In recent years, train information and intelligent construction are being carried forward vigorously in each city, to obtain the real-time of subway train
Crowding, the automatic passenger flow monitoring technology such as AFC, video, WiFi, bluetooth come into operation successively on subway train.Through investigation point
Analysis, existing train detection of passenger flow technology and its disadvantage are as follows:
(1) wobble effects video identification precision intelligent video analysis technology: is generated in travelling process of train;And due to video
Monitoring range repeats or there are blind area, it is caused to be not suitable for passenger flow identification or estimation applied to permutation vehicle global scope;
(2) WiFi probe technique: sample rate is higher, is easy to be suitable under certain sample rate by unique mac address duplicate removal
Permutation vehicle passenger flow estimation;But WiFi sample rate may have fluctuation in the case that passenger flow concentration is high;
(3) it AFC passenger flow acquisition technique: swipes the card or the data such as two dimensional code using passenger AFC is out of the station, is obtained by sorting model
To 15min (minimum time granularity can also be 1 hour etc.) the section volume of the flow of passengers, belong to bulk sample data, sampled data can be risen
To supplementary function;But real-time section passenger flow data can not be obtained, and can not accurately obtain the passenger flow data of a certain train, it needs
Carry out statistical disposition;
(4) bluetooth location technology: sample rate is extremely low, is unable to satisfy the Least sampling rate requirement of passenger flow statistics analysis, at present
The real-time passenger flow of train can not be also applied to determine.
Summary of the invention
The present invention provides a kind of real-time estimation methods that subway crowding is handed over for city rail, solve existing crowding
Calculation method it is effective poor, the problems such as accuracy is low.
The present invention can be achieved through the following technical solutions:
A kind of real-time estimation method for handing over subway crowding for city rail, for same train by fortune different on route
The crowding of row section is predicted, comprising the following steps:
Step 1: with the train of certain number line traffic direction by it is a certain operation section originate the moment after T1 time and end
P rush hour current signature day section is divided into N number of by the period before to the moment between the T2 time as a measurement period
Measurement period;
Step 2: calculating the WiFi volume of the flow of passengers of vehicle in each measurement period of current signature day pVideo passenger flow amountA upper characteristic day p corresponds to the AFC section volume of the flow of passengers in measurement period
Step 3: using neural network, with the WiFi volume of the flow of passengers in p current statistic period current signature dayVideo
The volume of the flow of passengersA upper characteristic day p corresponds to the AFC section volume of the flow of passengers in measurement periodAs input, real-time estimation is worked as
The practical volume of the flow of passengers Y of vehicle in the preceding characteristic day p current statistic period(p), to complete to p current statistic period current signature day
The real-time estimation of interior train crowding;
Step 4: repeating step 2 to three, completes the real-time of train crowding in certain peak period N number of measurement period and estimate
Meter.
Further, each compartment on train is divided into several grids, is arranged one in each net center of a lattice
WIFI probe is calculated whole in p current statistic period current signature day using the MAC Address of passenger's mobile terminal as test object
The WiFi volume of the flow of passengers of vehicle
Further, the detection data of WIFI probes all to each compartment inside is screened, and is first lower than look-in frequency
MAC Address data removal twice, then identical MAC Address data are only retained one, by all interior residue MAC
The WiFi volume of the flow of passengers that the quantity of location is summed as vehicle in p current statistic period current signature day
Further, the T1 time corresponding period is less than the T2 time corresponding period.
Further, the AFC section that a upper characteristic day p corresponds to entire train in measurement period is calculated using following equation
The volume of the flow of passengers
Further, one or more cameras are set inside each compartment, and the camera is for acquiring interior
Passenger's video, the coverage of the camera inside adjacent compartment is not overlapped, and utilizes convolutional neural networks and ridge regression knot
The video people counting algorithm of conjunction calculates the volume of the flow of passengers inside each compartment, takes its summation as current signature day p current statistic
Video passenger flow amount in period
Further, using following equation, train crowding D in p current statistic period current signature day is calculated
The present invention is beneficial to be had the technical effect that
Current signature day current statistic week is calculated as measurement period by the period of operation section using certain number alignment vehicle
The WiFi volume of the flow of passengers, video passenger flow amount in phase, the AFC section volume of the flow of passengers in the identical measurement period of a upper characteristic day, and with this
It is inputted as neural network, the practical volume of the flow of passengers of train in current statistic period real-time estimation current signature day, thus completion pair
The real-time estimation of train crowding in current statistic period current signature day, real-time estimation method of the invention combine WiFi spy
Needle detection, the analysis of video number and AFC sorting data system carry out fusion estimation to the volume of the flow of passengers, are reduced as far as single detection
The limitation of method expands the acquisition range of data, realizes and has complementary advantages, and improves volume of the flow of passengers detection accuracy, and then improve
The real-time estimation precision crowded to train, while a large amount of human and material resources are saved, alleviate peak for platform staff and gathers around
It is stifled to provide data foundation, for the when space division for parsing the traffic characteristic of passenger in peak period station, emulation and passenger in pre- survey station
Cloth, optimization train station passenger transportation management scheme, starting station large passenger flow prediction scheme etc. provide parameter foundation, for improving train station
Large passenger flow management level ensures that passenger's safety plays a significant role.
Detailed description of the invention
Fig. 1 is overall procedure schematic diagram of the invention;
Fig. 2 is the WIFI probe distribution schematic diagram of entrance of the invention;
Fig. 3 is the flow diagram of video people counting algorithm of the invention;
Fig. 4 is the structural schematic diagram of neural network of the invention;
Fig. 5 is the schematic diagram data of training sample of the invention;
Fig. 6 is the volume of the flow of passengers predicted using method of the invention and the contrast schematic diagram of the practical volume of the flow of passengers.
Specific embodiment
With reference to the accompanying drawing and the preferred embodiment specific embodiment that the present invention will be described in detail.
As shown in Figure 1, the present invention proposes a kind of real-time estimation method for handing over subway crowding for city rail, for same
Train by operation sections different on route crowdings prediction, using certain number alignment vehicle by the period of operation section as uniting
The period is counted, the WiFi volume of the flow of passengers of vehicle in p current statistic period current signature day is calculatedVideo passenger flow amountUpper one
A characteristic day p corresponds to the AFC section volume of the flow of passengers in measurement periodAnd inputted in this, as neural network, real-time estimation is worked as
The practical volume of the flow of passengers Y of vehicle in the preceding characteristic day p current statistic period(p), to complete to p current statistic period current signature day
The real-time estimation of interior train crowding, in this way, when the train crowding situation for learning a certain operation section in the current statistic period,
Staff can suitably plan the volume of the flow of passengers of next operation section, or suitably be advised to the arrival situation of train
It draws, to alleviate the crowded state as far as possible, improves the intelligent level of city rail traffic control reason.
Specifically includes the following steps:
Step 1: with certain number alignment vehicle by a certain operation section originate the moment after T2 before T1 time and arrival time when
Between between period as a measurement period, p rush hour current signature day section is divided into N number of measurement period.For
In the case of avoiding train dwelling, deceleration of entering the station, outbound acceleration, shadow that WiFi probe on train and camera data are acquired
Ring, the present invention chooses train by during a certain operation section, from its originate after to this period before Zhongdao as a system
The period is counted, there is no on-board and off-board on train during this, the volume of the flow of passengers is stablized constant, carries out data convenient for camera and WiFi probe
Acquisition also accurately provides good environment for acquisition data, it is corresponding less than the T2 time to can use the T1 time corresponding period
Period, leave more times as far as possible for data processing and transmission, so as in train by whole service section
Real-time estimation is completed in time, the crowding for improving follow-up operation section earlier for staff provides foundation.
Step 2: calculating the WiFi volume of the flow of passengers of vehicle in each measurement period of current signature day pVideo passenger flow amountA upper characteristic day p corresponds to the AFC section volume of the flow of passengers in measurement period
For the WiFi volume of the flow of passengers
Each compartment on train is divided into several grids, one WIFI probe is set in each net center of a lattice,
As shown in Fig. 2, calculating and entirely being arranged in p current statistic period current signature day using the corresponding MAC Address of passenger as test object
The WiFi volume of the flow of passengers of vehicleThe quantity of its grid can determine according to the actual volume of the actual performance of WIFI probe and train
It is fixed, it is ensured that the investigative range of WIFI probe can cover entire train.
In order to improve the accuracy of subsequent calculating, the detection data of WIFI probes all to each compartment inside is needed to carry out
Screening.
(1) in order to avoid on ground and elevated line in non-train WiFi equipment data influence, screen out in measurement period
By the spy frequency lower than MAC Address twice;
(2) data deduplication: screening out and repeat collected data in measurement period, i.e., identical MAC Address only retains one.
Finally, calculating the sum of quantity of all interior residue MAC Address is used as p current statistic period current signature day
The WiFi volume of the flow of passengers of interior entire train
For video passenger flow amount
Camera is set inside each compartment, which is used to acquire passenger's video of interior, adjacent compartment
The coverage of internal camera is not overlapped.For a certain frame video image in measurement period, convolutional Neural net is utilized
Video people counting algorithm of the network in conjunction with ridge regression, as shown in figure 3, returning the frame video image by convolutional neural networks
Middle number of people central point obtains crowd density distribution characteristics figure, then uses ridge regression model analysis crowd density distribution characteristics figure
Obtain the corresponding number of the frame video image, finally take the train each compartment select camera analyze to obtain number of people number the sum of
Video passenger flow amount as in p current statistic period current signature day
For the AFC section volume of the flow of passengers
Since AFC sorting data have time delay, the whole network data out of the station pass through in the just unified sorting of morning next day
AFC sorting data can only obtain the section volume of the flow of passengers for corresponding to statistical time range history feature day, therefore the present invention uses predicted characteristics day
A upper characteristic day correspondence 15min, the average value of the AFC sorting section volume of the flow of passengers on corresponding operation section it is disconnected as AFC
The face volume of the flow of passengersCalculation formula is as follows:
Step 3: using neural network, as shown in figure 4, with the WiFi volume of the flow of passengers in p current statistic period current signature dayVideo passenger flow amountA upper characteristic day p corresponds to the AFC section volume of the flow of passengers in measurement periodAs input,
The practical volume of the flow of passengers Y of train in p current statistic period real-time estimation current signature day(p), to complete to work as current signature day p
The real-time estimation of train crowding in preceding measurement period.
Before being predicted, it is necessary first to be trained to neural network, training data uses meter described above
Calculation method obtains, and same train is corresponded in the measurement period of prediction in the period in such as peak period, off-peak period, by difference
Run the WiFi volume of the flow of passengers of sectionVideo passenger flow amountThe AFC section volume of the flow of passengersAs input, manually to regard
Even if the mode that frequency counting mode manually counts originates the people that gets on or off the bus of each car door of respectively standing by video statistics that stands from train
Number, and then train is calculated according to the following formula by the practical volume of the flow of passengers Y of each operation sectiontAs output, to nerve net
Network is trained.In order to ensure the precision of prediction of neural network, the sample number for training data is no less than 120, if when current
The data deficiencies of section is enough, can take on same train a characteristic day or again up number of the corresponding characteristic day in the identical period
According to.
Yt=∑ the number of getting on the bus-∑ is got off number
Using following equation, train crowding D in p current statistic period current signature day is calculated
When train crowding is in 100% < D≤130% by the present invention, it is considered as serious crowded;In 80% < D≤
When 100%, it is considered as crowded;When in 60% < D≤80%, it is considered as general crowded;In D≤60% when, be considered as not crowded.
Step 4: repeating step 2 to three, completes the real-time of train crowding in some peak period N number of measurement period and estimate
Meter.
By taking No. 9 alignment vehicles of Shanghai Underground as an example, detailed description of the present invention method.
Step 1: the Time-distribution of analysis No. 9 alignment vehicle passenger flow datas of Shanghai Underground, by predicted characteristics day, prediction
Time segments division is weekday rush hours, off-peak period on working day, three kinds of periods of nonworkdays period.The predicted characteristics day is taken to be
Friday on December 28th, 2018, predicted characteristics period are 7:00-7:02 in morning peak period 7:00-9:00, and number is XXX at this time
Train is in by running the section Songjiang College Town river rising in Ningxia and flowing into central Shaanxi Zhan Yudong station.
Step 2: using existing peanut WiFi probe collection data in No. 9 alignment vehicle each compartment, and in the every section of train
The camera that a fixed congestion position is chosen in compartment is analyzed for video data, it is ensured that non-overlapping monitoring region.
Step 3: choosing No. 9 alignment Che Song according to the train operation ATS data comprising No. 9 line train arrival and leaving station moment
30s and the period before arrival time between 50s after originating the moment on operation section between river university city station and hole river rising in Ningxia and flowing into central Shaanxi station
For the whole story moment of measurement period, i.e. the sample data statistic period T as WiFi probe, video and AFC data.In this example
T=1min.
Step 4: sample size n=120 is taken, according to the measurement period of each sample data, to the data of WiFi probe collection
It is screened according to following rule, to obtain effective WiFi probe data collection:
(1) in order to avoid on ground and elevated line in non-train WiFi equipment data influence, screen out sample data system
It counts and detects the frequency in the period lower than MAC Address twice;
(2) data deduplication.It screens out and repeats collected data in measurement period, i.e., identical MAC Address only retains one.
According to the uniqueness of device mac address, No. 9 alignment vehicles on December 7th, 2018,14 days, (Friday) 7 on the 21st are obtained:
00-9:00, number are the WiFi volume of the flow of passengers of the XXX train by different operation sectionsThe partial data of acquisition is shown in Fig. 5, sampling
This quantity is 120.
Step 5: choosing train according to each sample data measurement period and passing through between the station of the river rising in Ningxia and flowing into central Shaanxi Songjiang College Town Zhan Yudong
Run during section that a certain frame of camera video is selected in each compartment in the sample statistics period, using convolutional neural networks with
The video demographics that ridge regression combines calculate method, i.e., are returned in train supervision video image in the number of people by convolutional neural networks
Heart point obtains crowd density distribution characteristics figure, then obtains the frame using ridge regression model analysis crowd density distribution characteristics figure
The corresponding number of video image, to obtain No. 9 alignment vehicles on the December 7th, 2018,14 days, (Friday) on the 21st that number is XXX
7:00-9:00, it selectes camera by different operation sections, each compartment and analyzes to obtain the average value i.e. video passenger flow amount of number of people numberThe partial data of acquisition is shown in Fig. 5.
Step 6: extracting and calculating No. 9 lines of Shanghai Underground in 2018 2018 11 according to each sample data measurement period
The moon 30, December 14, December 7 (Friday), identical 15min, number were that XXX train passes through different fortune between 7:00-9:00
The average value of AFC sorting train passenger number on row section is as the AFC section volume of the flow of passengersThe partial data of acquisition is shown in Fig. 5.
Wherein, the calculation formula of the AFC section volume of the flow of passengers is as follows:
Step 7: taking sample size n=120, in conjunction with train operation ATS data, originated using artificial mode from train
It stands through video counts 9 alignment vehicles 2018 on December 7,14 days, the 7:00-9:00 of (Friday) on the 21st at each station, each car door
It gets on or off the bus number, and it is that XXX train is disconnected by the actual motions on different operation sections that number is calculated according to the following formula
Face train volume of the flow of passengers Yt, the partial data of acquisition is shown in Fig. 5:
Yt=∑ the number of getting on the bus-∑ is got off number
Step 8: establishing No. 9 line WiFi volumes of the flow of passengers of Shanghai Underground using BP neural networkVideo passenger flow amount
The AFC section volume of the flow of passengersTo actual motion section train volume of the flow of passengers YtPrediction model, with reference to Fig. 4.And utilize the 120 of acquisition
Group sample data is learnt and is trained to BP neural network, and the X for meeting training precision requirement is establishedt→YtPrediction model, ginseng
Examine Fig. 4.
Step 9: inputting the 7:00-7 of December 28 (Friday) in 2018 using trained BP neural network prediction model:
02, train originates 30s and the operation section before arrival time between 50s after the moment by the Songjiang College Town river rising in Ningxia and flowing into central Shaanxi Zhan Yudong station
The WiFi volume of the flow of passengersVideo passenger flow amountThe AFC section volume of the flow of passengersPrediction obtains in current statistic period 7:00-7:02
True train section ridership Yp.Prediction passenger people of No. 9 alignment vehicles of Shanghai Underground in predicted characteristics day, predicted characteristics period
Number is as shown in Figure 6 with actual passenger number comparison diagram.It can be seen that the registration of predicted value curve and actual value curve compared with
The maximum absolute deviation of mean value of height, train passenger number predicted value and train actual passenger number is 178 people, and average absolute is opposite accidentally
Rate is 5.1%.
Break step 10: the true train section ridership that prediction model is predicted is converted to true train using following formula
Face crowding D:
Train section crowding grade is judged as follows:
When true train section crowding is in 100% < D≤130%, it is considered as serious crowded;In 80% < D≤
When 100%, it is considered as crowded;When in 60% < D≤80%, it is considered as general crowded;In D≤60% when, be considered as not crowded.
Although specific embodiments of the present invention have been described above, it will be appreciated by those of skill in the art that these
Be merely illustrative of, under the premise of without departing substantially from of the invention and essence, these embodiments can be made numerous variations or
Modification, therefore, protection scope of the present invention is defined by the appended claims.
Claims (7)
1. a kind of real-time estimation method for handing over subway crowding for city rail, which is characterized in that pass through line for same train
The crowding prediction of the different operation sections in road, comprising the following steps:
Step 1: with the train of certain number line traffic direction by a certain operation section originate the moment after T1 time and when Zhongdao
P rush hour current signature day section is divided into N number of statistics as a measurement period by the period before carving between the T2 time
Period;
Step 2: calculating the WiFi volume of the flow of passengers of vehicle in each measurement period of current signature day pVideo passenger flow amount
A upper characteristic day p corresponds to the AFC section volume of the flow of passengers in measurement period
Step 3: using neural network, with the WiFi volume of the flow of passengers in p current statistic period current signature dayVideo passenger flow
AmountA upper characteristic day p corresponds to the AFC section volume of the flow of passengers in measurement periodAs input, real-time estimation is currently special
Levy the practical volume of the flow of passengers Y of vehicle in the day p current statistic period(p), to complete to being arranged in p current statistic period current signature day
The real-time estimation of vehicle crowding;
Step 4: repeating step 2 to three, the real-time estimation of train crowding in certain peak period N number of measurement period is completed.
2. the real-time estimation method according to claim 1 for handing over subway crowding for city rail, it is characterised in that: will arrange
Each compartment is divided into several grids on vehicle, and a WIFI probe is arranged in each net center of a lattice, mobile eventually with passenger
The MAC Address at end calculates the WiFi volume of the flow of passengers of vehicle in p current statistic period current signature day as test object
3. the real-time estimation method according to claim 2 for handing over subway crowding for city rail, it is characterised in that: to every
The detection data of section all WIFI probes of interior is screened, and first goes look-in frequency lower than MAC Address data twice
It removes, then identical MAC Address data is only retained one, the quantity summation of all interior residue MAC Address is used as and is worked as
The WiFi volume of the flow of passengers of vehicle in the preceding characteristic day p current statistic period
4. the real-time estimation method according to claim 1 for handing over subway crowding for city rail, it is characterised in that: described
The T1 time corresponding period is less than the T2 time corresponding period.
5. the real-time estimation method according to claim 3 for handing over subway crowding for city rail, it is characterised in that: utilize
Following equation calculates the AFC section volume of the flow of passengers that a characteristic day p corresponds to entire train in measurement period
6. the real-time estimation method according to claim 3 for handing over subway crowding for city rail, it is characterised in that: every
It saves interior and one or more cameras is set, the camera is used to acquire passenger's video of interior, in adjacent compartment
The coverage of the camera in portion is not overlapped, and utilizes video people counting algorithm of the convolutional neural networks in conjunction with ridge regression, meter
The volume of the flow of passengers inside each compartment is calculated, takes its summation as the video passenger flow amount in p current statistic period current signature day
7. the real-time estimation method according to claim 1 for handing over subway crowding for city rail, it is characterised in that: utilize
Following equation calculates train crowding D in p current statistic period current signature day
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910774323.9A CN110516866B (en) | 2019-08-21 | 2019-08-21 | Real-time estimation method for urban rail transit train crowding degree |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910774323.9A CN110516866B (en) | 2019-08-21 | 2019-08-21 | Real-time estimation method for urban rail transit train crowding degree |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110516866A true CN110516866A (en) | 2019-11-29 |
CN110516866B CN110516866B (en) | 2020-06-26 |
Family
ID=68626105
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910774323.9A Active CN110516866B (en) | 2019-08-21 | 2019-08-21 | Real-time estimation method for urban rail transit train crowding degree |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110516866B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111372196A (en) * | 2020-02-20 | 2020-07-03 | 杭州海康威视***技术有限公司 | Data processing method and device, electronic equipment and machine-readable storage medium |
CN112215409A (en) * | 2020-09-24 | 2021-01-12 | 交控科技股份有限公司 | Rail transit station passenger flow prediction method and system |
CN112231640A (en) * | 2020-10-14 | 2021-01-15 | 中国民航信息网络股份有限公司 | Civil aviation passenger quantity prediction method, device and equipment and readable storage medium |
CN112766950A (en) * | 2020-12-31 | 2021-05-07 | 广州广电运通智能科技有限公司 | Dynamic path cost determination method, device, equipment and medium |
CN113068215A (en) * | 2021-03-12 | 2021-07-02 | 华扬联众数字技术(深圳)有限公司 | Weight and filtering-based multi-WiFi probe MAC address dynamic line algorithm |
CN114785943A (en) * | 2022-03-31 | 2022-07-22 | 联想(北京)有限公司 | Data determination method, equipment and computer readable storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060198302A1 (en) * | 2005-03-03 | 2006-09-07 | Sofman Lev B | Traffic dimensioning in a metro area with IPTV architecture |
CN105279572A (en) * | 2015-09-16 | 2016-01-27 | 北京城建设计发展集团股份有限公司 | City track traffic passenger flow density index calculating and releasing system |
CN107360590A (en) * | 2017-06-29 | 2017-11-17 | 上海工程技术大学 | Track station part congestion points passenger flow condition judgement method |
-
2019
- 2019-08-21 CN CN201910774323.9A patent/CN110516866B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060198302A1 (en) * | 2005-03-03 | 2006-09-07 | Sofman Lev B | Traffic dimensioning in a metro area with IPTV architecture |
CN105279572A (en) * | 2015-09-16 | 2016-01-27 | 北京城建设计发展集团股份有限公司 | City track traffic passenger flow density index calculating and releasing system |
CN107360590A (en) * | 2017-06-29 | 2017-11-17 | 上海工程技术大学 | Track station part congestion points passenger flow condition judgement method |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111372196A (en) * | 2020-02-20 | 2020-07-03 | 杭州海康威视***技术有限公司 | Data processing method and device, electronic equipment and machine-readable storage medium |
CN112215409A (en) * | 2020-09-24 | 2021-01-12 | 交控科技股份有限公司 | Rail transit station passenger flow prediction method and system |
CN112215409B (en) * | 2020-09-24 | 2024-01-30 | 交控科技股份有限公司 | Rail transit station passenger flow prediction method and system |
CN112231640A (en) * | 2020-10-14 | 2021-01-15 | 中国民航信息网络股份有限公司 | Civil aviation passenger quantity prediction method, device and equipment and readable storage medium |
CN112766950A (en) * | 2020-12-31 | 2021-05-07 | 广州广电运通智能科技有限公司 | Dynamic path cost determination method, device, equipment and medium |
CN113068215A (en) * | 2021-03-12 | 2021-07-02 | 华扬联众数字技术(深圳)有限公司 | Weight and filtering-based multi-WiFi probe MAC address dynamic line algorithm |
CN113068215B (en) * | 2021-03-12 | 2022-01-11 | 华扬联众数字技术(深圳)有限公司 | Weight and filtering-based multi-WiFi probe MAC address dynamic line algorithm |
CN114785943A (en) * | 2022-03-31 | 2022-07-22 | 联想(北京)有限公司 | Data determination method, equipment and computer readable storage medium |
CN114785943B (en) * | 2022-03-31 | 2024-03-05 | 联想(北京)有限公司 | Data determination method, device and computer readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN110516866B (en) | 2020-06-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110516866A (en) | A kind of real-time estimation method for handing over subway crowding for city rail | |
CN110111575B (en) | Urban traffic flow network analysis method based on complex network theory | |
CN110866649B (en) | Method and system for predicting short-term subway passenger flow and electronic equipment | |
CN107845259B (en) | Bus running condition real-time feedback system and bus real-time running data processing method | |
CN105513351A (en) | Traffic travel characteristic data extraction method based on big data | |
CN108322891B (en) | Traffic area congestion identification method based on user mobile phone signaling | |
Petrovska et al. | Traffic congestion analysis visualisation tool | |
CN104424812B (en) | A kind of public transport arrival time forecasting system and method | |
CN103593974B (en) | A kind of public transport passenger capacity collection method based on locating information | |
CN103544837B (en) | Traffic information estimation method and system combining cross-regional position updating and communication | |
CN108171993B (en) | Highway vehicle speed calculation method based on mobile phone signaling big data | |
CN106503843A (en) | A kind of regular public traffic line network optimization and method of adjustment based on mobile phone signaling data | |
CN110493816A (en) | A kind of real-time predicting method for handing over the subway station volume of the flow of passengers for rail | |
CN106297288A (en) | A kind of bus passenger passenger flow data gathers and the method for analysis | |
CN104200667A (en) | Traffic jam grade detection method based on cellphone signal data | |
EP2608181B1 (en) | Method for detecting traffic | |
CN111882867A (en) | City wisdom traffic early warning system based on thing networking | |
CN105702041A (en) | Highway multisource data fusion state estimation system based on neural network and method thereof | |
CN105070057A (en) | Real-time road traffic monitoring method | |
CN104636611A (en) | Urban road/ road segment vehicle speed evaluation method | |
CN106920388A (en) | A kind of highway monitoring system and control method | |
CN103906086B (en) | A kind of personnel based on mobile network flow to monitoring method and device | |
CN106920417A (en) | A kind of vehicle path planning system and method | |
CN105096584A (en) | Traffic decision support method, device, and system | |
CN108665178A (en) | A kind of building staircase passenger flow forecast method in the subway station based on AFC |
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 | ||
CB03 | Change of inventor or designer information |
Inventor after: Hu Hua Inventor after: Zhao Yuan Inventor after: Deng Zihuan Inventor after: Liu Zhigang Inventor after: Hao Yanxi Inventor after: Liu Xiulian Inventor before: Hu Hua Inventor before: Deng Zihuan Inventor before: Liu Zhigang Inventor before: Hao Yanxi Inventor before: Liu Xiulian |
|
CB03 | Change of inventor or designer information | ||
GR01 | Patent grant | ||
GR01 | Patent grant |