CN106777703A - A kind of bus passenger real-time analyzer and its construction method - Google Patents
A kind of bus passenger real-time analyzer and its construction method Download PDFInfo
- Publication number
- CN106777703A CN106777703A CN201611182098.2A CN201611182098A CN106777703A CN 106777703 A CN106777703 A CN 106777703A CN 201611182098 A CN201611182098 A CN 201611182098A CN 106777703 A CN106777703 A CN 106777703A
- Authority
- CN
- China
- Prior art keywords
- bus
- passenger
- website
- data
- time
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
- G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
-
- 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
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
Abstract
The invention discloses the construction method of bus passenger real-time analyzer, based on Spark cloud computing platforms, with reference to the two-way rider history data analysis of swiping the card of public transport, Preprocessing is carried out to history public transport data according to data processing rule；Each website of passenger is obtained using the two-way behavior historical data of getting on the bus of passenger to get off behavior, so as to build each circuit bus passenger website of getting on or off the bus using statistical learning method be distributed；Generation obtains the data analysis computational methods of each interval passenger's density using passenger loading behavior；Real-time public transport data prediction is carried out using Spark, and dynamic generation bus station list is recorded by Stop of bus；Real-time Feedback of the back-end data result to foreground is realized by Spark Streaming technologies, and realizes that dynamic public transport data high-speed data base read-write is operated by Redis；The real-time public transport data analysis carried out by historical data analysis result and passenger's density computational methods.
Description
Technical field
The present invention relates to traffic data analyzing field, more particularly to a large amount of historical datas and cloud platform dynamic realtime are utilized
Ground carries out efficient analysis and data visualization to the bus card-reading real time data constantly updated.
Background technology
Urban traffic pressure rapidly increases in recent years, is made to urban economy development, talent introduction, environment sustainable development etc.
Into huge pressure.Therefore, first develop city bus and improve passenger and the satisfaction of public transit system is had become instantly
Study hotspot both domestic and external.
Intelligent bus are the important sons of intelligent transportation system (Intelligent Transportation System, ITS)
System.The system is intended to using intelligent dispatching system, automatic counter system, passenger information system, automatic vehicle location system etc.
Fully improve the purposes such as intuitive, mobility and the objectivity of utilization and the scheduling of existing Public Resource.
All rested on due to lacking the systems such as the support of high performance platform, the scheduling of many correlations, Bus information confluence analysis
To in single bus, the analysis of single time period, it is impossible to whole public transit systems are provided with high reliability analysis and to passenger
The fact carry out comprehensive information feedback.Simultaneously as lack the effective analysis to a large amount of public transport historical datas, the result of its analysis
Lack confidence level and have that accessibility is poor, time-consuming high drawback.
Traditional data mining algorithm based on machine learning is due to by coupling between each characteristic value of public transport, information
The influence of the problems such as dynamic, larger temporal correlation causes that machine learning algorithm structure design difficulty is larger.Simultaneously as public
Data analysis is handed over, is related to the professional knowledges such as bus dispatching evaluation, public transit system management, bus dispatching system planning, therefore make
Realize that public transport data analysis difficulty is larger with conventional machines learning method.
The content of the invention
In order to solve above-mentioned technical problem present in prior art, analyzed in real time the invention provides a kind of bus passenger
The construction method of system, comprises the following steps：
Step one, bus station automatically generate；
Step 2, site match of getting on the bus；
Step 3, bus passenger get-off stop are calculated；
Wherein step 2 is specially：Public transit vehicle arrival time and passenger getting on/off time match based on MapReduce,
The passenger getting on/off situation of swiping the card is clustered using K-Means, is found out passenger getting on/off centrostigma and is arrived with each website of public transport
The difference of every public transit vehicle time recorder and public transport Non-contact Media Reader time is matched and calculated up to the time；Root
Clustered according to the time difference opposite sex and K-Means passenger loadings, according to getting on the bus, the website time criteria for classifying is carried out to passenger loading website
Divide.
Further, step 3 is complete from two-way passenger data using the model automatization selection algorithm based on crosscheck
Concentrating iteratively sample drawn data, as test data set, algorithm model selection is carried out to every public bus network.
Further, the algorithm is：When bus passenger is gone on a journey, trip website approximately obeys Poisson distribution, therefore passenger
Got on the bus from i stations and take probability calculation formula n stations got off as shown in formula 4.1：
λ is current passenger's average riding station number；
The website is equal to using λ to the half of the website number of terminus to calculate, specific formula is as shown in formula 4.2：
M-CurrN is when the station number from terminus of setting out in advance to make arrangements；
J-1 stations are crossed in public transport, to j websites before the ridership that still remains on car have relation relation shown in formula 4.3,4.4：
yij=yi(j-1)-xi(j-1)(formula 4.3)
Therefore, the passenger's transition probability between each website can be derived and obtained by equation below, as shown in formula 4.5：
Thus website transition probability model, by xijAnd be saved in the naive Bayesian probability form of each website so that
Each form deposits each preamble website to the transition probability of this website, therefore need to only obtain certain period each website passenger loading number feelings
Condition, you can draw correspondingly get-off stop situation, so as to build OD matrixes.
Further, the algorithm is：On the premise of bus station temperature change degree is little, getting off for each website is drawn
Distribution is unrelated with preamble website number of getting on the bus, i.e. after public transport operation to i stations, the j website numbers of getting off are：Xj=Tj*Ai, wherein T
It is website temperature, A is number on car.
Further, the algorithm is：Read the two-way pick-up time of each user and with reference to the method for backward inference, pass through
The website that passenger reversely gets on the bus determines each website forward direction temperature in certain period, follows the trail of the record of swiping the card of each user, protects
Deposit frequency of occurrences highest positive and negative two-way getting on the bus website and to count data set and carry out calculating assessment, the data in the data set are pressed
Multiple data sets are divided into according to the period and temperature is calculated to each data set；
It is formula 4.8 by temperature simplified formula：
Wherein：
1.A circuits i stations forward direction t1Moment gets on the bus ridership：Ai；
2.A circuits i reversely gets on the bus and in t at station1The Shi Keyou riderships that car is recorded forward：A2ti；
3.A circuits t1Moment, forward direction i stations in passenger of getting on the bus got off ridership：Bti；
A circuits forward direction i stations t1Moment temperature：Tti。
On the basis of above-mentioned various Data Modeling Methods, the present invention contain it is a kind of for each bar public bus network use
Automatic algorithms model selection algorithm.Each algorithm model set up by the above method can effectively predict the upper of each website of public transport
Get off number, have preferable effect in the accuracy and practicality of model.However, each model has it during prediction
Model characteristics and result tendentiousness, the characteristic attribute of each bar public bus network are also different.For this characteristic of public transit system,
The present invention contains a model of mind selection algorithm.Model of mind selection algorithm is multiplied using the thought of crosscheck by two-way
The data of visitor are divided into multiple tables by public bus network, and the crosscheck of 10 packets is carried out to each table, wherein, choose one of them
Packet is used as test verification data.During crosscheck, choose each modeling method is carried out according to data therein
Statistical learning, so as to set up model, and carries out test verification using last test data group.Most have for each public bus network
Model will be chosen for final forecast model.
Beneficial effect：Using cloud high-performance treatments historical data and by statistical learning method carry out data scrubbing with
Integrate, and feature extraction is carried out to each public transport data problem and calculate analysis with MapReduce；Three kinds of model selection storehouses are built,
Using crosscheck, the model for automated to each bar public bus network, customizing is selected and parameter adaptation；Use public bus network
Crowding method of determining and calculating, optimal circuit by bus is prompted the user with by marking congestion route on map.Use Spark
The public transport data that Streaming real-time processings are received by data-interface, and by valuable treatment after supercomputing treatment
Result output is in database.Quickly read using Redis memory databases, write data.
The system utilizes cloud computing correlation technique, realizes the quick treatment to big data using Spark platforms, and use
Spark Streaming read, process, integrating, exporting a large amount of related datas in real time.In terms of data processing, present case is used
Data digging method, arrives at a station the data such as information, passenger loading information according to public transport, and analysis obtains optimal scheduling strategy, map heat
The function such as route choosing reference that point speculates, passenger rides.Meanwhile, in terms of the information representation to passenger is implemented, this literary grace
With Redis key assignments memory databases, realize that the high speed of inquiry reads and shows.
Brief description of the drawings
Fig. 1 is bus passenger real-time analyzer Organization Chart of the invention；
Fig. 2 .1 are bus station automatic generating calculation flow charts；
Fig. 2 .2 are the model automatization selection algorithm flow charts based on crosscheck.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.
The system efforts be made so as to be integrated and divided with big data technology of Internet of things by the mass data processing that public transit system is generated
Each client is dealt into, making full use of for data is realized.Its major function includes：Public transport position data is distributed to mobile phone visitor in real time
Family end, facilitates popular bus trip demand；The real time status of public transport operation are fed back into bus dispatching and driver and conductor, is helped
Helping it carries out Real-Time Scheduling；Mass historical data to accumulating is analyzed integration, finds the bottleneck of road traffic system, is system
Determine relevant policies and decision support is provided, and for following Urban Traffic Planning provides theoretical foundation；The entirety of public transit system is transported
Row situation Real-time Feedback helps it preferably to control the situation that public transit system is run to the personnel of clothes are driven.
The data source of the system mainly has：Driving GPS is recorded, bus station is arrived at a station, and record, public transit system IC-card are swiped the card
Record, bus station geographical location information, city traffic road condition information etc..
The real-time initial data of public transport operation is due to above-mentioned data, with data volume is big, requirement of real-time strong, data
Source dispersion and correlation is strong, the low feature of the quality of data, larger difficulty is caused to data processing.
The application is related to transport information and control, intelligent public transportation system, traffic programme and management, Computer Simulation, cooperates with
The multidisciplinary field such as control, decision theory, the crossing research of category traffic engineering, control science and information science etc..Therefore, for intelligence
The particular demands in energy public transport field, it is necessary to using new data fusion, decision support and analysis means, comprehensive artificial intelligence, machine
The achievement in research of the aspect such as device study and pattern-recognition carries out data mining, the domain knowledge of combined with intelligent public transit system, research
The concrete model for being adapted to intelligent public transportation system information excavating is set up, effective mining algorithm is designed, is planning, the battalion of public transit system
Fortune management etc. provides the accurate decision-making foundation of science and feasible scheme.By integrated use Field Research, qualitative and quantitative analysis,
Simulation modeling and experiment, plan strategies for and optimize etc. ways and means.Shown in technology path Fig. 1 that plan is taken.
In terms of system overall architecture, the system is using Spark Streaming real-time processing datas and uses interior poke
Rapidly and efficiently access strategy is realized according to storehouse Redis.System architecture diagram is as shown in Figure 1.
Bus station automatic generating calculation
The situation for jumping station often occurs due to public transit vehicle, therefore simple certain bus station arrival time can not be used
Order carries out website generation, and needs to carry out website order point respectively in situation of arriving at a station within the next few days in all buses of each circuit
Analysis and the generation of sequence.The method can substantially avoid the data statistics on business handed over system website entry time delayed and bring
Problem.Algorithm particular flow sheet such as Fig. 2 .1.
Bus loading zone point matching algorithm
The time record of the IC-card card reader used by the time of each public transit vehicle and passenger swipe the card, therefore need
The two systems time is corrected, and site match is carried out to the record after correction.K- is carried out to passenger loading behavior
Means cluster analyses, so as to obtain passenger loading time block, by calculating gap of the time point away from other times point in each block
So that it is determined that the center time point of each time block.The center time point is matched with the record time in public transit vehicle, from
And be directed to the time correction of each public transit vehicle.On this basis, run time between bus stop is done into decile, so that
Determine passenger loading website.
Bus passenger get-off stop prediction model
It is that public transport operation situation is accurately estimated, the get-off stop of passenger is particularly important.But, due to existing
The design feature of bus IC card card-punching system, the get-off stop of user is difficult to be estimated.Thus, for bus passenger get-off stop
Prediction just turn into intelligent bus key topics.At home and abroad association area has been obtained for sufficiently paying attention to and grinding the problem
Study carefully.The system is analyzed using the passenger's individuality with two-way brushing card data, and computation model is extended to all passengers
On the model construction of colony.
It is directed to model selection and the parameter optimization of each circuit
Because the property differences such as each bus route, run time, the frequency are larger, therefore only design a unified system
Meter study analysis model is difficult to carry out good fitting to whole public transport data entirely without exception.It is traditional based on artificial parameter
Adjustment can be manually selected with model selection according to the accuracy of forecast of each model of each circuit.However, this traditional side
Formula is difficult to the renewal that the bus routes larger to data movement are carried out in real time algorithm model selection.If additionally, by two-way brush
Card passenger complete or collected works will certainly cause result to be difficult to the problem tested as training data.
The present invention proposes the model automatization selection algorithm based on crosscheck, the iteration from two-way passenger data complete or collected works
Ground sample drawn data as test data set, so as to realize the function.Detailed process is shown in Fig. 2 .2.
Calculating process is：
When bus passenger is gone on a journey, trip website approximately obeys Poisson distribution, therefore passenger gets on the bus from i stations and takes n stations and get off
Probability calculation formula as shown in formula 4.1：
λ is current passenger's average riding station number, and survey shows, city bus averagely take website number for 8 stations.And
And, for last 16 station of vehicle, it is desirable to reduce the value of λ is avoiding terminus number of getting off excessive.Specific method of adjustment
For：For last 16 websites, it is equal to the website to the half of the website number of terminus using λ to calculate, specific formula such as formula
Shown in 4.2：
M-CurrN is that, when the station number from terminus of setting out in advance to make arrangements, AVG is city bus average trip distance.
Again because j-1 stations are crossed in public transport, to j websites before the ridership that still remains on car have shown in formula 4.3,4.4 and close
It is relation：
yij=yi(j-1)-xi(j-1)(formula 4.3)
Therefore, the passenger's transition probability between each website can be derived and obtained by equation below, as shown in formula 4.5：
Thus website transition probability model, by xijAnd be saved in the naive Bayesian probability form of each website so that
Each form deposits each preamble website to the transition probability of this website.Therefore certain period each website passenger loading number feelings need to only be obtained
Condition, you can draw correspondingly get-off stop situation, so as to build OD matrixes.
Website prediction based on Markov property is then on the premise of it is assumed that bus station temperature change degree is little, to draw
The distribution of getting off of each website is unrelated with preamble website number of getting on the bus, so as to carry out statistical learning analysis.
That is, after public transport operation is stood to i, the j website numbers of getting off are：Xj=Tj*Ai, wherein T is website temperature, and A is people on car
Number.
And website Forecasting Methodology at times this be the calculating website temperature for counting at times.
The algorithm needs to read the two-way pick-up time of each user and with reference to the method for backward inference, by passenger it is reverse on
The website of car determines each website forward direction temperature in certain period.Because website of reversely being got on the bus by bus by passenger can draw just
Got off number ratio to website, so, can be compared by record of swiping the card and show that the website of each period is got off patronage ratio
Example.
In specific calculating, because the number ratio of getting off of each website of the positive day part of website ratio reflection of reversely getting on the bus
Example, it is therefore desirable to follow the trail of the record of swiping the card of each user, preservation frequency of occurrences highest positive and negative two-way getting on the bus and counts website
Data set carries out calculating assessment.Afterwards, it is necessary to according to the period data in the data set are divided into multiple data sets and to every
Individual data set calculates temperature.
Can be formula 4.8 by temperature simplified formula for convenience of calculating：
Wherein：
4.A circuits i stations forward direction t1Moment gets on the bus ridership：Ai；
5.A circuits i reversely gets on the bus and in t at station1The Shi Keyou riderships that car is recorded forward：A2ti；
6.A circuits t1Moment, forward direction i stations in passenger of getting on the bus got off ridership：Bti；
7.A circuits forward direction i stations t1Moment temperature：Tti。
Public bus network crowding method of determining and calculating
Road crowding in public transport driving distance is always one of most concerned problem of passenger, and the size of crowding is very big
Time that passenger arrives at is affected in degree and bus dispatching personnel are provided and sent or subtract the data for sending bus
Foundation.In order to realize that the crowding in certain route bus traffic route is calculated, the system is by each website bus and adjacent two
Running time between standing is driven a vehicle using the difference value of itself and standard average running time as foundation as criterion
Road crowding judges in route.During implementing, it is necessary to pass through big data cloud platform calculate each website between it is each
Circuit running time, and Mllab storehouses by Spark carry out chi square distribution fitting to it.Then, at treatment real-time bus station
Between point during crowding, substituted into the chi square distribution for having completed parameter fitting running time between nearest one class of bus two station
In, half point is calculated than position, then on this basis, the route crowding between website is shown by color.
Center processing process
The data comprising Spark streaming and Redis memory databases for being adopted as transit hub and optimizing
Framework, the characteristic for the high noisy information in public transit system employs Kafka as public transport data transmission scheme to improve appearance
Mistake；Stream process are carried out to public transport data using Spark streaming.Meanwhile, according to public transport data variation is very fast, data category
Property less characteristic the concurrent characteristic of read-write high is reached using Redis memory databases, the public transport data for being readily able to change are present
Internal storage data is synchronized to hard disk in every 20 minutes by read or write speed faster on internal memory rather than in the hard disk of traditional data library storage
Or in backup cloud platform.Secondary method simplify traditional Relational DataBase storage change degree it is higher, attribute is less, the scale of construction is larger
The storage redundancy of public transport data, so as to improve public transport data read-write efficiency.
The application is used in discrete data unified integration a to cloud platform, and line number is entered by big data technology unification
According to the method such as connection and statistical learning, analysis between cleaning, tables of data.To realize the quick read functions of data, the system
Using the HDFS storage strategies of cloud platform, a large amount of public transport historical data piecemeals are stored in disk array, so as to make full use of
System disk storage performance.Meanwhile, the system also utilizes Map Reduce characteristics, realizes historical data in cloud platform
High speed processing, and using the Spark Steaming of Spark platforms, Stream Processing is carried out to real time data, it is achieved thereby that many
The center of source historical data and multi-source real-time data is uniformly processed.
Analyze and process in real time
By Spark Steaming, the system realizes the Stream Processing function to real time data.The system is by being
System interface, obtains public transport real time data from each data source in real time, and carry out quick merger, except making an uproar, analyze and process work.In profit
With high-performance cloud platform to acquired real time data, after differentiating that scheduling algorithm carries out data analysis with OD matrixes, crowding,
The data result of acquisition is flowed to export in database by Spark and is called for foreground.
Collect the complexity of a large amount of bus real time datas in order to overcome, the application using kafka carry out Data Collection with
Distribution.
In order to tackle the real time data processing in the case of big data, Stream Processing clothes quasi real time are provided using Spark
Business.Spark Streaming are the map-reduce modes based on spark realizes the quick treatment to data.Spark be by
The map-reduce data processing shelfs based on internal memory of University of California Berkeley AMPLab exploitations, its bullet for passing through definition
Property distributed data collection (RDD) can realize to the treatment of the high performance parallel of mass data.And spark streaming conducts
One expansion of spark, can effectively complete the data processing task in the case of quasi real time.
Claims (7)
1. a kind of construction method of bus passenger real-time analyzer, comprises the following steps：
Step one, bus station automatically generate；
Step 2, site match of getting on the bus；
Step 3, bus passenger get-off stop are calculated；
Wherein step 2 is specially：Public transit vehicle arrival time and passenger getting on/off time match based on Map Reduce, profit
The passenger getting on/off situation of swiping the card is clustered with K-Means, is found out passenger getting on/off centrostigma and is reached with each website of public transport
Time is matched and is calculated the difference of every public transit vehicle time recorder and public transport Non-contact Media Reader time；According to
The time difference opposite sex and K-Means passenger loadings cluster, according to getting on the bus, the website time criteria for classifying is drawn to passenger loading website
Point.
2. the construction method of bus passenger real-time analyzer as claimed in claim 1, it is characterised in that：Step 3 uses base
In crosscheck model automatization selection algorithm, from two-way passenger data complete or collected works iteratively sample drawn data as test
Data set, algorithm model selection is carried out to every public bus network.
3. the construction method of bus passenger real-time analyzer as claimed in claim 2, it is characterised in that：
The algorithm is：When bus passenger is gone on a journey, trip website approximately obeys Poisson distribution, therefore passenger gets on the bus from i stations and takes n
The probability calculation formula that station is got off is as shown in formula 4.1：
λ is current passenger's average riding station number；
The website is equal to using λ to the half of the website number of terminus to calculate, specific formula is as shown in formula 4.2：
M-CurrN is that, when the station number from terminus of setting out in advance to make arrangements, Avg is city bus average trip distance；
J-1 stations are crossed in public transport, to j websites before the ridership that still remains on car have relation relation shown in formula 4.3,4.4：
yij=yi(j-1)-xi(j-1) (formula 4.3)
Therefore, the passenger's transition probability between each website can be derived and obtained by equation below, as shown in formula 4.5：
Thus website transition probability model, by xijAnd be saved in the naive Bayesian probability form of each website so that each table
Lattice deposit each preamble website to the transition probability of this website, therefore need to only obtain certain period each website passenger loading number situation,
Correspondingly get-off stop situation can be drawn, so as to build OD matrixes.
4. the construction method of bus passenger real-time analyzer as claimed in claim 2, it is characterised in that：The algorithm is：
On the premise of bus station temperature change degree is little, show that the distribution of getting off of each website is unrelated with preamble website number of getting on the bus,
That is, after public transport operation is stood to i, the j website numbers of getting off are：Xj=Tj*Ai, wherein T is website temperature, and A is number on car.
5. the construction method of bus passenger real-time analyzer as claimed in claim 2, it is characterised in that：The algorithm is：
The two-way pick-up time of each user is read and with reference to the method for backward inference, when the website reversely got on the bus by passenger determines certain
Each website forward direction temperature in section, follows the trail of the record of swiping the card of each user, and preservation frequency of occurrences highest is positive and negative two-way to get on the bus
Website simultaneously counts data set and carries out calculating assessment, and according to the period data in the data set are divided into multiple data sets and to every
Individual data set calculates temperature；
It is formula 4.8 by temperature simplified formula：
Wherein：
1. A circuits i stations forward direction t1Moment gets on the bus ridership：Ai；
2. A circuits i stations are reversely got on the bus and in t1The Shi Keyou riderships that car is recorded forward：A2ti；
3. A circuits t1Moment, forward direction i stations in passenger of getting on the bus got off ridership：Bti；
4. A circuits forward direction i station t1Moment temperature：Tti。
6. the construction method of the bus passenger real-time analyzer as described in one of claim 2-5, it is characterised in that：Using friendship
The fork method of inspection, selects optimal model for each bus routes automatically：For each public bus network is built using three kinds of algorithms
Vertical statistical model, on Spark platforms being 10 groups by each bar public bus network Data Integration can verify that data, wherein 9 groups numbers of extraction
Statistical model is set up according to according to above-mentioned modeling method, is compared with last 1 group of data result according to each statistical model result of calculation
It is right, repeat said process 10 times, the optimal model of Average Accuracy is chosen as forecast model.
7. the construction method of bus passenger real-time analyzer as claimed in claim 6, it is characterised in that：It is adopted as public transport number
The data framework comprising Spark streaming and Redis memory databases optimized according to model, in public transit system
The characteristic of high noisy information employs Kafka as public transport data transmission scheme to improve fault-tolerance；Using Spark
Streaming carries out stream process to public transport data；Using Redis memory databases, there is reading in the public transport data for being readily able to change
Writing rate faster on internal memory and, and internal storage data is synchronized in hard disk or backup cloud platform for every 20 minutes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611182098.2A CN106777703A (en) | 2016-12-19 | 2016-12-19 | A kind of bus passenger real-time analyzer and its construction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611182098.2A CN106777703A (en) | 2016-12-19 | 2016-12-19 | A kind of bus passenger real-time analyzer and its construction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106777703A true CN106777703A (en) | 2017-05-31 |
Family
ID=58891209
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611182098.2A Pending CN106777703A (en) | 2016-12-19 | 2016-12-19 | A kind of bus passenger real-time analyzer and its construction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106777703A (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107025790A (en) * | 2017-06-08 | 2017-08-08 | 河北城兴市政设计院股份有限公司 | Urban road green trip temperature data collecting system and method |
| CN107818412A (en) * | 2017-10-18 | 2018-03-20 | 福州大学 | A kind of extensive bus passenger OD parallel calculating methods based on MapReduce |
| CN107945560A (en) * | 2017-12-21 | 2018-04-20 | 大连海事大学 | A kind of public transport smart electronics stop sign information display control method and system |
| CN108038728A (en) * | 2017-12-11 | 2018-05-15 | 北京奇虎科技有限公司 | Public transport vehicle body advertisement putting Decision Making of Line Schemes generation method, device and electronic equipment |
| CN108171971A (en) * | 2017-12-18 | 2018-06-15 | 武汉烽火众智数字技术有限责任公司 | Vehicular real time monitoring method and system based on Spark Streaming |
| CN108492608A (en) * | 2018-03-12 | 2018-09-04 | 北京航空航天大学 | A kind of analysis method and system of the bus passenger flow volume based on cloud model |
| CN108846503A (en) * | 2018-05-17 | 2018-11-20 | 电子科技大学 | A kind of respiratory disease illness person-time dynamic prediction method neural network based |
| CN109615036A (en) * | 2018-11-30 | 2019-04-12 | 深圳大学 | A kind of fine particle exposure appraisal procedure based on bus IC card-punching system |
| CN109615850A (en) * | 2018-12-27 | 2019-04-12 | 连尚（新昌）网络科技有限公司 | It is a kind of for determining the method and apparatus of the transit riding information of user |
| CN110085048A (en) * | 2019-06-04 | 2019-08-02 | 湖南智慧畅行交通科技有限公司 | A kind of bus based on GPS data arrives point calculating method leaving from station in real time |
| CN110309241A (en) * | 2018-03-15 | 2019-10-08 | 北京嘀嘀无限科技发展有限公司 | Method for digging, device, server, computer equipment and readable storage medium storing program for executing |
| CN110459056A (en) * | 2019-08-26 | 2019-11-15 | 南通大学 | A kind of public transport arrival time prediction technique based on LSTM neural network |
| CN110472813A (en) * | 2019-06-24 | 2019-11-19 | 广东浤鑫信息科技有限公司 | A kind of school bus website self-adapting regulation method and system |
| CN111401663A (en) * | 2020-04-12 | 2020-07-10 | 广州通达汽车电气股份有限公司 | Method and device for updating public transport space-time OD matrix in real time |
| CN111667087A (en) * | 2019-03-08 | 2020-09-15 | 南京农业大学 | Bus station-jumping operation method considering pollution emission |
| CN111858806A (en) * | 2020-07-09 | 2020-10-30 | 武汉译码当先科技有限公司 | Passenger travel track detection method, device, equipment and storage medium |
| CN112005562A (en) * | 2018-04-23 | 2020-11-27 | 谷歌有限责任公司 | Determining vehicle congestion using real-time location data |
| CN113053156A (en) * | 2021-03-31 | 2021-06-29 | 华录智达科技股份有限公司 | Intelligent bus station addressing method based on bus radius method |
| CN113191223A (en) * | 2021-04-15 | 2021-07-30 | 宁波市民卡运营管理有限公司 | Passenger density evaluation method and device, computer equipment and storage medium |
| CN113628474A (en) * | 2021-08-09 | 2021-11-09 | 百度在线网络技术（北京）有限公司 | Bus arrival station identification method and device, electronic equipment and readable storage medium |
| CN114724375A (en) * | 2022-05-05 | 2022-07-08 | 厦门理工学院 | Reverse passenger identification system based on Internet of things |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102097002A (en) * | 2010-11-22 | 2011-06-15 | 东南大学 | Method and system for acquiring bus stop OD based on IC card data |
| JP2011210092A (en) * | 2010-03-30 | 2011-10-20 | Nippon Conlux Co Ltd | Route bus guide system |
| 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 |
| CN105023437A (en) * | 2015-08-21 | 2015-11-04 | 苏州大学张家港工业技术研究院 | Method and system for establishing public transit OD matrix |
| CN105390013A (en) * | 2015-11-18 | 2016-03-09 | 北京工业大学 | Method for predicting bus arrival time based on bus IC card |
-
2016
- 2016-12-19 CN CN201611182098.2A patent/CN106777703A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011210092A (en) * | 2010-03-30 | 2011-10-20 | Nippon Conlux Co Ltd | Route bus guide system |
| CN102097002A (en) * | 2010-11-22 | 2011-06-15 | 东南大学 | Method and system for acquiring bus stop OD based on IC card data |
| 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 |
| CN105023437A (en) * | 2015-08-21 | 2015-11-04 | 苏州大学张家港工业技术研究院 | Method and system for establishing public transit OD matrix |
| CN105390013A (en) * | 2015-11-18 | 2016-03-09 | 北京工业大学 | Method for predicting bus arrival time based on bus IC card |
Non-Patent Citations (4)
| Title |
|---|
| 孙慈嘉，等: "基于云计算的公交OD矩阵构建方法", 《江苏大学学报（自然科学版）》 * |
| 康琦、吴启迪: "《机器学习中的不平衡分类方法》", 31 October 2017 * |
| 窦慧丽，等: "基于站点上下客人数的公交客流OD反推方法研究", 《交通与计算机》 * |
| 陈绍辉等: "基于特征站点的公交IC卡数据站点匹配方法研究", 《北京工业大学学报》 * |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107025790A (en) * | 2017-06-08 | 2017-08-08 | 河北城兴市政设计院股份有限公司 | Urban road green trip temperature data collecting system and method |
| CN107818412A (en) * | 2017-10-18 | 2018-03-20 | 福州大学 | A kind of extensive bus passenger OD parallel calculating methods based on MapReduce |
| CN108038728A (en) * | 2017-12-11 | 2018-05-15 | 北京奇虎科技有限公司 | Public transport vehicle body advertisement putting Decision Making of Line Schemes generation method, device and electronic equipment |
| CN108038728B (en) * | 2017-12-11 | 2022-03-11 | 北京奇虎科技有限公司 | Bus body advertisement delivery line scheme generation method and device and electronic equipment |
| CN108171971A (en) * | 2017-12-18 | 2018-06-15 | 武汉烽火众智数字技术有限责任公司 | Vehicular real time monitoring method and system based on Spark Streaming |
| CN107945560A (en) * | 2017-12-21 | 2018-04-20 | 大连海事大学 | A kind of public transport smart electronics stop sign information display control method and system |
| CN108492608A (en) * | 2018-03-12 | 2018-09-04 | 北京航空航天大学 | A kind of analysis method and system of the bus passenger flow volume based on cloud model |
| CN110309241A (en) * | 2018-03-15 | 2019-10-08 | 北京嘀嘀无限科技发展有限公司 | Method for digging, device, server, computer equipment and readable storage medium storing program for executing |
| CN112005562A (en) * | 2018-04-23 | 2020-11-27 | 谷歌有限责任公司 | Determining vehicle congestion using real-time location data |
| CN112005562B (en) * | 2018-04-23 | 2022-02-01 | 谷歌有限责任公司 | Method, computing system and computer readable medium for determining vehicle congestion using real-time location data |
| CN108846503A (en) * | 2018-05-17 | 2018-11-20 | 电子科技大学 | A kind of respiratory disease illness person-time dynamic prediction method neural network based |
| CN109615036A (en) * | 2018-11-30 | 2019-04-12 | 深圳大学 | A kind of fine particle exposure appraisal procedure based on bus IC card-punching system |
| CN109615850A (en) * | 2018-12-27 | 2019-04-12 | 连尚（新昌）网络科技有限公司 | It is a kind of for determining the method and apparatus of the transit riding information of user |
| CN111667087A (en) * | 2019-03-08 | 2020-09-15 | 南京农业大学 | Bus station-jumping operation method considering pollution emission |
| CN110085048A (en) * | 2019-06-04 | 2019-08-02 | 湖南智慧畅行交通科技有限公司 | A kind of bus based on GPS data arrives point calculating method leaving from station in real time |
| CN110472813A (en) * | 2019-06-24 | 2019-11-19 | 广东浤鑫信息科技有限公司 | A kind of school bus website self-adapting regulation method and system |
| CN110472813B (en) * | 2019-06-24 | 2023-12-22 | 广东浤鑫信息科技有限公司 | Self-adaptive adjustment method and system for school bus station |
| CN110459056A (en) * | 2019-08-26 | 2019-11-15 | 南通大学 | A kind of public transport arrival time prediction technique based on LSTM neural network |
| CN111401663B (en) * | 2020-04-12 | 2021-04-27 | 广州通达汽车电气股份有限公司 | Method and device for updating public transport space-time OD matrix in real time |
| CN111401663A (en) * | 2020-04-12 | 2020-07-10 | 广州通达汽车电气股份有限公司 | Method and device for updating public transport space-time OD matrix in real time |
| CN111858806A (en) * | 2020-07-09 | 2020-10-30 | 武汉译码当先科技有限公司 | Passenger travel track detection method, device, equipment and storage medium |
| CN113053156A (en) * | 2021-03-31 | 2021-06-29 | 华录智达科技股份有限公司 | Intelligent bus station addressing method based on bus radius method |
| CN113053156B (en) * | 2021-03-31 | 2022-09-02 | 华录智达科技股份有限公司 | Intelligent bus station addressing method based on bus radius method |
| CN113191223A (en) * | 2021-04-15 | 2021-07-30 | 宁波市民卡运营管理有限公司 | Passenger density evaluation method and device, computer equipment and storage medium |
| CN113628474A (en) * | 2021-08-09 | 2021-11-09 | 百度在线网络技术（北京）有限公司 | Bus arrival station identification method and device, electronic equipment and readable storage medium |
| CN114724375A (en) * | 2022-05-05 | 2022-07-08 | 厦门理工学院 | Reverse passenger identification system based on Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106777703A (en) | A kind of bus passenger real-time analyzer and its construction method | |
| Zhao et al. | Predictive task assignment in spatial crowdsourcing: a data-driven approach | |
| Zhang et al. | Understanding taxi service strategies from taxi GPS traces | |
| CN105023437B (en) | A kind of construction method and system of public transport OD matrixes | |
| Chen et al. | Modeling of emergency supply scheduling problem based on reliability and its solution algorithm under variable road network after sudden-onset disasters | |
| CN107481511A (en) | A kind of method and system for calculating candidate bus station | |
| CN108986453A (en) | A kind of traffic movement prediction method based on contextual information, system and device | |
| CN110836675B (en) | Decision tree-based automatic driving search decision method | |
| CN109376906B (en) | Travel time prediction method and system based on multi-dimensional trajectory and electronic equipment | |
| CN108737492A (en) | A method of the navigation based on big data system and location-based service | |
| CN113033110B (en) | Important area personnel emergency evacuation system and method based on traffic flow model | |
| CN105205052A (en) | Method and device for mining data | |
| Liu et al. | Modeling the interaction coupling of multi-view spatiotemporal contexts for destination prediction | |
| Ai et al. | A deep learning approach to predict the spatial and temporal distribution of flight delay in network | |
| Moharm et al. | Big data in ITS: Concept, case studies, opportunities, and challenges | |
| Xia et al. | SW-BiLSTM: a Spark-based weighted BiLSTM model for traffic flow forecasting | |
| Ramesh et al. | Station-level demand prediction for bike-sharing system | |
| CN114493535A (en) | Asset model-based data center system for data driving | |
| Wu et al. | A novel dynamically adjusted regressor chain for taxi demand prediction | |
| Wang | An intelligent passenger flow prediction method for pricing strategy and hotel operations | |
| Rodríguez-Rueda et al. | Origin–Destination matrix estimation and prediction from socioeconomic variables using automatic feature selection procedure-based machine learning model | |
| Zhu et al. | Validating rail transit assignment models with cluster analysis and automatic fare collection data | |
| CN115965163A (en) | Rail transit short-time passenger flow prediction method for generating countermeasures to loss based on space-time diagram | |
| Ruch et al. | The impact of fleet coordination on taxi operations | |
| Hu et al. | An attention-mechanism-based traffic flow prediction scheme for smart city |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170531 |
|
| RJ01 | Rejection of invention patent application after publication |