CN108280540B - Method and device for predicting short-time passenger flow state of rail transit station - Google Patents

Method and device for predicting short-time passenger flow state of rail transit station Download PDF

Info

Publication number
CN108280540B
CN108280540B CN201810024755.3A CN201810024755A CN108280540B CN 108280540 B CN108280540 B CN 108280540B CN 201810024755 A CN201810024755 A CN 201810024755A CN 108280540 B CN108280540 B CN 108280540B
Authority
CN
China
Prior art keywords
passenger flow
station
speed
error correction
target station
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.)
Active
Application number
CN201810024755.3A
Other languages
Chinese (zh)
Other versions
CN108280540A (en
Inventor
张宁
张炳森
尹嵘
袁春强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CN201810024755.3A priority Critical patent/CN108280540B/en
Publication of CN108280540A publication Critical patent/CN108280540A/en
Application granted granted Critical
Publication of CN108280540B publication Critical patent/CN108280540B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • Tourism & Hospitality (AREA)
  • General Physics & Mathematics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A method and a device for predicting the short-time passenger flow state of a rail transit station are disclosed, wherein the method comprises the following steps: generating passenger flow and speed time sequences according to the acquired passenger flow volume and passenger flow speed of the rail transit target station and the adjacent stations at the upstream and downstream of the target station; performing first-order difference operation on the passenger flow and speed time sequence to obtain a stable time sequence; adopting historical sample data of a stationary time sequence to construct a vector autoregressive model; obtaining an error correction term according to historical sample data of the original passenger flow volume, the original passenger flow volume and the speed time sequence in the same time period; establishing a vector error correction model of the passenger flow volume and the passenger flow speed of the target station; and calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station in the time interval t. The method and the device for predicting the short-time passenger flow state of the rail transit station use the related station passenger flow and speed parameter data to establish a vector error correction model to predict the passenger flow volume and the passenger flow speed of the target station, so that the accuracy and the reliability of the short-time passenger flow prediction are improved.

Description

Method and device for predicting short-time passenger flow state of rail transit station
Technical Field
The invention relates to the technical field of urban rail transit management, in particular to a method and a device for predicting a short-time passenger flow state of a rail transit station.
Background
The urban rail transit has the advantages of rapidness, comfort, tidiness and the like, attracts more and more residents to choose rail transit for going out, causes great change of passenger flow space-time distribution, and further provides higher requirements for operation management. The rail transit passenger flow state prediction is one of key technologies of rail transit operation, management and control, accurate and reliable prediction can reflect the real-time change rule of passenger flow and provide data support for transport energy analysis and transport volume matching, and the prediction is also an important decision index for service level and system operation state evaluation, and has important significance for deep research on the prediction.
At present, some researches are also carried out on the short-time passenger flow prediction of rail transit, such as prediction methods of a neural network model, a time series model and the like. However, most of the existing rail transit station passenger flow state prediction is based on the historical passenger flow of the station, a single-input single-output prediction model is mainly adopted, the internal relation among parameters and effective information such as space-time correlation among stations under networked operation are ignored, and the passenger flow state is difficult to predict accurately. Therefore, the time-space correlation among the rail transit passenger flow parameters is deeply excavated, and a multivariable time series model of the rail transit station passenger flow state is constructed on the basis, so that the accuracy of passenger flow state prediction can be further improved.
Disclosure of Invention
The invention aims to provide a method and a device for predicting a short-time passenger flow state of a rail transit station, so as to improve the accuracy and reliability of short-time passenger flow prediction.
In order to achieve the purpose, the invention adopts the following technical scheme:
a method for predicting a short-time passenger flow state of a rail transit station comprises the following steps:
according to the obtained passenger flow volume and passenger flow speed of the rail transit target station and the adjacent stations at the upstream and the downstream of the target station, a passenger flow and speed time sequence { S ] is generateditI ═ 1,2,3}, where S isit＝(qit，vit)TI is 1,2,3, which respectively represents the target station upstream neighboring station, the target station and the target station downstream neighboring station, q isitAs volume of passenger, vitAs passenger flow velocity:
performing first-order difference operation on the passenger flow and speed time sequence to obtain a stationary time sequence s'it＝(q′it，v′it)TWherein, s'itThe first order difference value of the ith station in the time interval t;
adopting historical sample data of the stationary time sequence to construct a vector autoregressive model;
according to the original passenger flow volume in the same time period and the historical sample data of the passenger flow and speed time sequence, checking the passenger flow and speed time sequence{SitI is a synergistic relationship between 1,2,3, and an error correction term λ ecm is obtainedt-1；
Establishing a vector error correction model of the passenger flow volume and the passenger flow speed of the target station according to the vector autoregressive model and the error correction term;
and calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t according to the vector error correction model and the video data.
In the above scheme, the vector autoregressive model is:
Figure BDA0001543692220000021
Figure BDA0001543692220000022
wherein p and q are the hysteresis order of the vector autoregressive process, and delta q2tIs the first order difference value, delta v, of the passenger flow of the target station in the time interval t2tIs the first order differential value, delta q, of the target station passenger flow velocity over the time interval t1(t-m)、Δq3(t-m)First order difference value, Deltav, for upstream and downstream adjacent station traffic within time interval (t-m)1(t-m)、Δv3(t-m)First order differential values of the passenger flow velocities of the upstream and downstream adjacent stations within a time interval (t-m); alpha is alphaxm、βxm、γxm、δxm、εxm、∈xm、αym、βym、γym、δym、εym、∈ym、cx、cyThe parameters to be estimated of the vector autoregressive model are obtained; e is the same asxt、∈ytIs an error term of the vector autoregressive model.
In the above scheme, the vector error correction model is:
Figure BDA0001543692220000031
Figure BDA0001543692220000032
wherein λ is1、λ2Error correction factor, lambda, corresponding to the bit1ecmt-1、λ2ecmt-1Is an error correction term.
In the above scheme, calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t according to the vector error correction model and the video data includes:
obtaining actual observations from the video at time intervals (t-1), (t-2), (t-3) … … (t-p + 1);
calculating a predicted value delta q of a first-order difference time sequence of the passenger flow volume and the passenger flow speed of the target station in a time interval t according to the actual observed value and the vector error correction model2t、Δv2t；
According to the predicted value delta q of the first-order difference time sequence of the station passenger flow volume and the passenger flow speed2t、Δv2tAnd calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t.
In the above solution, before generating the passenger flow and speed time series, the method further includes: and acquiring the passenger flow volume and the passenger flow speed according to the video data acquired by the video detector in the rail transit station passage.
In the scheme, the hysteresis orders p and q in the vector autoregressive model are determined by a Bayesian criterion.
In the above scheme, the original passenger flow volume and the passenger flow and speed time series { S }itThe coordination relationship between i ═ 1,2,3} was examined by Johansen coordination examination method.
In the above scheme, the parameter to be estimated: alpha is alphaxm、βxm、γxm、δxm、εxm、∈xm、αym、βym、γym、δym、εym、∈ym、cx、cyAnd the error correction coefficient lambda1、λ2Are obtained by least square estimation.
A rail transit station short-time passenger flow state prediction device, the device comprising:
the passenger flow and speed time sequence generating unit is used for generating a passenger flow and speed time sequence { S) according to the acquired passenger flow and passenger flow speed of the rail transit target station and the adjacent stations at the upstream and downstream of the target stationitI ═ 1,2,3}, where S isit＝(qit，vit)TI is 1,2,3, respectively, representing a target station upstream neighbor station, the target station and a target station downstream neighbor station, q isitAs volume of passenger, vitIs the passenger flow velocity;
a difference calculation unit for performing a first order difference calculation on the passenger flow and speed time series to obtain a stationary time series S'it＝(q′it，v′it)TWherein, S'itThe first order difference value of the ith station in the time interval t;
the first modeling unit is used for constructing a vector autoregressive model by adopting historical sample data comprising the stationary time sequence;
a co-integration relation checking unit for checking the passenger flow and speed time sequence { S) according to the original passenger flow volume in the same time period and the historical sample data of the passenger flow and speed time sequenceitI is a synergistic relationship between 1,2,3, and an error correction term λ ecm is obtainedt-1；
The second modeling unit is used for establishing a vector error correction model of the passenger flow volume and the passenger flow speed of the target station according to the vector autoregressive model and the error correction term;
and the predicted value calculating unit is used for calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t according to the vector error correction model and the video data.
In the above scheme, the device further comprises a passenger flow information obtaining unit, which is used for obtaining the passenger flow volume and the passenger flow speed according to the video data collected by the video detector in the rail transit station passage.
According to the method and the device for predicting the short-time passenger flow state of the rail transit station, provided by the invention, the related station passenger flow and speed parameter data are used for establishing a vector error correction model to predict the passenger flow volume and the passenger flow speed of the target station, so that the accuracy and the reliability of the short-time passenger flow prediction are improved.
Drawings
FIG. 1 is a flow chart of an implementation of a method for predicting a short-term passenger flow state of a rail transit station according to an embodiment of the present invention;
FIG. 2 is a diagram showing the fitting effect of the short-term predicted value and the actual observed value of the station early peak passenger flow in the embodiment of the invention;
FIG. 3 is a diagram illustrating the fitting effect of the short-term predicted value and the actual observed value of the station early peak passenger flow speed in the embodiment of the present invention;
fig. 4 is a schematic structural diagram of a short-time passenger flow state prediction device for a rail transit station according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
As shown in fig. 1, a method for predicting a short-term passenger flow state of a rail transit station according to an embodiment of the present invention includes:
step 110, generating passenger flow and speed time sequence { S) according to the acquired passenger flow volume and passenger flow speed of the rail transit target station and the adjacent stations at the upstream and downstream of the target stationitI ═ 1,2,3}, where S isit＝(qit，vit)TI is 1,2,3, which respectively denote a target station upstream neighbor station, a target station, and a target station downstream neighbor station, q isitAs volume of passenger, vitIs the passenger flow velocity.
Step 120, performing first order difference operation on the passenger flow and speed time sequence to obtain a stationary time sequence S'it＝(q′it，v′it)TWherein, SitThe first order difference value for the ith station within time interval t.
And step 130, constructing a vector autoregressive model by using the historical sample data of the stationary time sequence.
Step 140, checking the passenger flow and speed time sequence { S ] according to the historical sample data of the original passenger flow, passenger flow and speed time sequence in the same time perioditI is a synergistic relationship between 1,2,3, and an error correction term λ ecm is obtainedt-1。
And 150, establishing a vector error correction model of the passenger flow volume and the passenger flow speed of the target station according to the vector autoregressive model and the error correction term.
And 160, calculating predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t according to the vector error correction model and the video data.
The technical scheme provided by the embodiment of the invention considers that the coordination relation exists between the station passenger flow and the speed parameter, and the passenger flow and the speed parameter are taken as the input of prediction, so that the neglect of the equilibrium relation between the parameters by univariate prediction is compensated; meanwhile, the spatial correlation of passenger flow among stations is considered, passenger flow data of the stations with the correlation are introduced, a vector error correction model is established, and the accuracy and reliability of short-time passenger flow prediction are further improved.
Before step 110, the passenger flow volume and the passenger flow speed are obtained according to the video data collected by the video detector in the railway traffic station passage.
Thereafter, in step 110, the passenger flow and speed time series generated from the passenger flow volume and the passenger flow speed are continuous time series data at intervals of 5 minutes.
In step 120, the first order difference operation formula is:
S′it＝Si(t+1)-Sit
in the formula: sitFor the first order difference value, S, of the ith station in the time interval titIs the time sequence of the ith station within the time interval t.
In step 130, the vector autoregressive model is:
Figure BDA0001543692220000071
Figure BDA0001543692220000072
wherein p and q are the hysteresis order of the vector autoregressive process, and delta q2tIs the first order difference value, delta v, of the passenger flow of the target station in the time interval t2tIs the first order differential value, delta q, of the target station passenger flow velocity over the time interval t1(t-m)、Δq3(t-m)First order difference value, Deltav, for upstream and downstream adjacent station traffic within time interval (t-m)1(t-m)、Δv3(t-m)First order differential values of the passenger flow velocities of the upstream and downstream adjacent stations within a time interval (t-m); alpha is alphaxm、βxm、γxm、δxm、εxm、∈xm、αym、βym、γym、δym、εym、∈ym、cx、cyParameters to be estimated of the vector autoregressive model; e is the same asxt、∈ytIs the error term of the vector autoregressive model.
Wherein the parameters to be estimated are: alpha is alphaxm、βxm、γxm、δxm、εxm、∈xm、αym、βym、γym、δym、εym、∈ym、cx、cyAnd estimating by using a least square method.
And determining the hysteresis orders p and q in the vector autoregressive model by a Bayesian criterion.
In step 140, raw traffic volume and traffic and speed time series SitThe general relation of coordination between i ═ 1,2, 3-The test was performed by the Johansen cooperative test method.
In step 150, the vector error correction model is:
Figure BDA0001543692220000073
Figure BDA0001543692220000074
wherein λ is1、λ2Error correction factor, lambda, corresponding to the bit1ecmt-1、λ2ecmt-1Is an error correction term. Wherein the error correction coefficient lambda1、λ2And estimating by using a least square method.
In step 160, actual observations at time intervals (t-1), (t-2), (t-3) … … (t-p +1) are obtained from the video;
calculating a predicted value delta q of a first-order difference time sequence of the passenger flow volume and the passenger flow speed of the target station in a time interval t according to an actual and vector error correction model2t、Δv2t；
According to the predicted value delta q of the first-order difference time sequence of the station passenger flow volume and the passenger flow speed2t、Δv2tAnd calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t.
Wherein,
Figure BDA0001543692220000081
in an example of the embodiment of the invention, a first station of urban rail transit is selected as a target station, and a second station and a third station of adjacent stations of the urban rail transit are respectively an upstream station and a downstream station of the first station.
Firstly, according to a picture shot by monitoring a working day early peak video within the period of 2016, 8 months, 1 day to 8 months, 26 days, a pedestrian target in a subway monitoring video is identified by a directional gradient histogram feature descriptor and a support vector machine classifier, and a target window is tracked by adopting a continuous self-adaptive Mean SHIFT algorithm (Camshift algorithm for short), so that the statistics of passenger flow and speed parameters is realized. And taking 5 minutes as a time interval, taking the early-peak data of the working days of the first three weeks as sample data for calibrating the model parameters, and taking the data of the last week for evaluating the predictive performance of the model.
And then, acquiring an original sequence according to the acquired passenger flow volume and passenger flow speed, and performing first-order difference operation to convert the original sequence into a stable time sequence.
Then, a vector autoregressive model is established for the stationary time sequence, and the expression is as follows:
Figure BDA0001543692220000091
Figure BDA0001543692220000092
the lag order p, q of the target site vector autoregressive model can be determined by bayesian information criterion, the results of which are given in table 1.
TABLE 1 Drum building station vector autoregression model lag order
Figure BDA0001543692220000093
And on the basis of determining the hysteresis order, carrying out the co-integration relation test on the original parameter sequences of the target station and the adjacent stations by adopting a Johansen co-integration test method based on the regression coefficient. The embodiment provides the results of checking the coordination relationship between the first station and the upstream and downstream stations of the adjacent station as shown in table 2.
TABLE 2 inspection results of the co-integration relationship between the target site and the neighboring sites
Figure BDA0001543692220000094
The principle of the Johansen co-integration test is to judge the co-integration relation among variables by using a maximum likelihood estimation method, wherein feature root statistics and critical values are statistical values under 95% confidence level, and when the feature root statistics is larger than the critical values, the original hypothesis is rejected, otherwise, the acceptance is accepted. The result shows that a co-integration relationship exists between the target site and the upstream and downstream site parameter sequences, and a Vector Error Correction (VEC) model can be established.
Then, a vector error correction model is constructed, and the vector error correction model with the passenger flow variable lag order of 3 and the speed variable lag order of 2 is established by combining the vector autoregressive model and the co-integration relation as follows:
Figure BDA0001543692220000101
Figure BDA0001543692220000102
all the parameters to be estimated are estimated by adopting a common least square method by using sample data, and the result is shown in table 3 because the system cx、cyIs statistically insignificant, so in the VEC model of the first vehicle station, cx、cyAre all 0.
TABLE 3VEC model parameter estimation results
Figure BDA0001543692220000103
Finally, calculating a first-order difference time sequence predicted value delta q of the passenger flow volume of the first vehicle station, the passenger flow volume and the speed time sequence2t、Δv2tAnd further calculating the predicted value of the horizontal sequence within the time interval t as q2t＝q2(t-1)+Δq2t，v2t＝v2(t-1)+Δv2tTraffic volume and flowThe fitting effect of the predicted velocity value and the actual observed value is shown in fig. 2 and fig. 3. As shown in fig. 2, a curve 201 is a passenger flow observation value, and a curve 202 is a passenger flow prediction value. As shown in fig. 3, a curve 301 is a passenger flow velocity observed value, and a curve 302 is a passenger flow velocity predicted value.
The present embodiment uses Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percent Error (MAPE) to evaluate the predicted performance of target station passenger flow and velocity. In order to compare with the traditional prediction method, the univariate time series model ARIMA (0, 1, 1) is constructed by adopting the same sample data, the model does not consider the space-time correlation and the balance relation between parameters, and the prediction results of the two models are shown in a table 4.
TABLE 4 comparison of VEC and ARIMA model predicted Performance
Figure BDA0001543692220000111
As can be seen from the results given in Table 4, the VEC model is obviously superior to the ARIMA model in the passenger flow and speed prediction, and the VEC model constructed by the method has better prediction performance.
According to the method for predicting the short-time passenger flow state of the rail transit station, the related station passenger flow and speed parameter data are used, a vector error correction model is built to predict the passenger flow volume and the passenger flow speed of the target station, and the accuracy and the reliability of the short-time passenger flow prediction are improved.
Example two
The embodiment of the invention provides a device for predicting the short-time passenger flow state of a rail transit station, which comprises:
a passenger flow and speed time sequence generating unit 410, configured to generate a passenger flow and speed time sequence { S) according to the acquired passenger flow volume and passenger flow speed of the target station of rail transit and the stations adjacent to the upstream and downstream of the target stationitI ═ 1,2,3}, where S isit＝(qit，vit )TI is 1,2,3, respectively representing a target station upstream neighbor station, a target station and a target stationDownstream neighbor site, qitAs volume of passenger, vitIs the passenger flow velocity.
A difference operation unit 420 for performing a first order difference operation on the passenger flow and the velocity time series to obtain a stationary time series S ″it＝(q′it，v′it)TWherein, S'itThe first order difference value for the ith station within time interval t.
A first modeling unit 430, configured to construct a vector autoregressive model using historical sample data including stationary time series.
A co-integration relation checking unit 440 for checking the passenger flow and the speed time series { S }according to the original passenger flow, the historical sample data of the passenger flow and the speed time series in the same time perioditI is a synergistic relationship between 1,2,3, and an error correction term λ ecm is obtainedt-1。
The second modeling unit 450 is configured to establish a vector error correction model of the passenger flow volume and the passenger flow speed of the target station according to the vector autoregressive model and the error correction term.
And a predicted value calculating unit 460, configured to calculate predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t according to the vector error correction model and the video data.
The technical scheme provided by the embodiment of the invention considers that the coordination relation exists between the station passenger flow and the speed parameter, and the passenger flow and the speed parameter are taken as the input of prediction, so that the neglect of the equilibrium relation between the parameters by univariate prediction is compensated; meanwhile, the spatial correlation of passenger flow among stations is considered, passenger flow data of the stations with the correlation are introduced, a vector error correction model is established, and the accuracy and reliability of short-time passenger flow prediction are further improved.
The device for predicting the short-time passenger flow state of the rail transit station further comprises a passenger flow information acquisition unit, wherein the passenger flow information acquisition unit is used for acquiring the passenger flow volume and the passenger flow speed according to the video data acquired by the video detector in the rail transit station passage.
The device for predicting the short-time passenger flow state of the rail transit station uses the related station passenger flow and speed parameter data to establish a vector error correction model to predict the passenger flow volume and the passenger flow speed of the target station, so that the accuracy and the reliability of the short-time passenger flow prediction are improved.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention.

Claims (8)

1. A method for predicting a short-time passenger flow state of a rail transit station is characterized by comprising the following steps:
according to the obtained passenger flow volume and passenger flow speed of the rail transit target station and the adjacent stations at the upstream and the downstream of the target station, a passenger flow and speed time sequence { S ] is generatedit1,2,3, wherein the passenger flow volume and the passenger flow speed are obtained from video data in the channel; sit＝(qit，vit)TI is 1,2,3, which respectively represents the target station upstream neighboring station, the target station and the target station downstream neighboring station, q isitAs volume of passenger, vitIs the passenger flow velocity;
performing first-order difference operation on the passenger flow and speed time sequence to obtain a stationary time sequence S'it＝(q′it，v′it)TWherein, S'itThe first order difference value of the ith station in the time interval t;
adopting the historical sample data of the stationary time sequence to construct a vector autoregressive model, wherein the vector autoregressive model is as follows:
Figure FDA0003107874060000011
Figure FDA0003107874060000012
wherein p, q is the hysteresis order of the vector autoregressive process, Δ q2tIs the first order difference value of the passenger flow of the target station in the time interval t, delta v2tIs the first order difference value of the passenger flow speed of the target station in the time interval t, delta q1(t-m)、△q3(t-m)For the first order difference value of the upstream and downstream adjacent station passenger flow within the time interval (t-m), Deltav1(t-m)、△v3(t-m)First order differential values of the passenger flow velocities of the upstream and downstream adjacent stations within a time interval (t-m); alpha is alphaxm、βxm、γxm、δxm、εxm、∈xm、αym、βym、γym、δym、εym、∈ym、cx、cyThe parameters to be estimated of the vector autoregressive model are obtained; e is the same asxt、∈ytAn error term for the vector autoregressive model;
according to the original passenger flow volume in the same time period and historical sample data of the passenger flow and speed time sequence, checking the passenger flow and speed time sequence { S }itI is a synergistic relationship between 1,2,3, and an error correction term λ ecm is obtainedt-1；
According to the vector autoregressive model and the error correction term, a vector error correction model of the passenger flow volume and the passenger flow speed of the target station is established, wherein the vector error correction model is as follows:
Figure FDA0003107874060000021
Figure FDA0003107874060000022
wherein λ is1、λ2Error correction factor, lambda, corresponding to the bit1ecmt-1、λ2ecmt-1Is an error correction term;
and calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t according to the vector error correction model and the video data.
2. The method of claim 1, wherein said calculating predicted values of target station traffic volume and traffic velocity over time interval t from said vector error correction model and video data comprises:
obtaining actual observations from the video at time intervals (t-1), (t-2), (t-3) … … (t-p + 1);
calculating a predicted value delta q of a first-order difference time sequence of the passenger flow volume and the passenger flow speed of the target station in a time interval t according to the actual observed value and the vector error correction model2t、△v2t；
According to the predicted value delta q of the first-order difference time sequence of the station passenger flow volume and the passenger flow speed2t、△v2tAnd calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t.
3. The method of claim 1, wherein prior to generating the time series of passenger flows and velocities, the method further comprises: and acquiring the passenger flow volume and the passenger flow speed according to the video data acquired by the video detector in the rail transit station passage.
4. The method of claim 1, wherein the hysteresis order p, q in the vector autoregressive model is determined by a bayesian criterion.
5. Method according to claim 1, characterized in that said original traffic volume and said traffic and speed time series { S }itThe coordination relationship between i ═ 1,2,3} was examined by Johansen coordination examination method.
6. The method according to any of claims 1 to 5, characterized in that the parameter to be estimated: alpha is alphaxm、βxm、γxm、δxm、εxm、∈xm、αym、βym、γym、δym、εym、∈ym、cx、cyAnd anThe error correction coefficient lambda1、λ2Are obtained by least square estimation.
7. A rail transit station short-time passenger flow state prediction device is characterized by comprising:
the passenger flow and speed time sequence generating unit is used for generating a passenger flow and speed time sequence { S) according to the acquired passenger flow and passenger flow speed of the rail transit target station and the adjacent stations at the upstream and downstream of the target stationit1,2,3, wherein the passenger flow volume and the passenger flow speed are obtained from video data in the channel; sit＝(qit，vit)TI is 1,2,3, respectively, representing a target station upstream neighbor station, the target station and a target station downstream neighbor station, q isitAs volume of passenger, vitIs the passenger flow velocity;
a difference calculation unit for performing a first order difference calculation on the passenger flow and speed time series to obtain a stationary time series S'it＝(q′it，v′it) TWherein, S'itThe first order difference value of the ith station in the time interval t;
a first modeling unit, configured to construct a vector autoregressive model using the historical sample data of the stationary time sequence, where the vector autoregressive model is:
Figure FDA0003107874060000031
Figure FDA0003107874060000032
wherein p, q is the hysteresis order of the vector autoregressive process, Δ q2tIs the first order difference value of the passenger flow of the target station in the time interval t, delta v2tIs the first order difference value of the passenger flow speed of the target station in the time interval t, delta q1(t-m)、△q3(t-m)For the first order difference value of the upstream and downstream adjacent station passenger flow within the time interval (t-m), Deltav1(t-m)、△v3(t-m)First order differential values of the passenger flow velocities of the upstream and downstream adjacent stations within a time interval (t-m); alpha is alphaxm、βxm、γxm、δxm、εxm、∈xm、αym、βym、γym、δym、εym、∈ym、cx、cyThe parameters to be estimated of the vector autoregressive model are obtained; e is the same asxt、∈ytAn error term for the vector autoregressive model;
a co-integration relation checking unit for checking the passenger flow and speed time sequence { S) according to the original passenger flow volume in the same time period and the historical sample data of the passenger flow and speed time sequenceitI is a synergistic relationship between 1,2,3, and an error correction term λ ecm is obtainedt-1；
The second modeling unit is used for establishing a vector error correction model of the passenger flow volume and the passenger flow speed of the target station according to the vector autoregressive model and the error correction term, and the vector error correction model is as follows:
Figure FDA0003107874060000041
Figure FDA0003107874060000042
wherein λ is1、λ2Error correction factor, lambda, corresponding to the bit1ecmt-1、λ2ecmt-1Is an error correction term;
and the predicted value calculating unit is used for calculating the predicted values of the passenger flow volume and the passenger flow speed of the target station within the time interval t according to the vector error correction model and the video data.
8. The device as claimed in claim 7, further comprising a passenger flow information obtaining unit for obtaining the passenger flow volume and the passenger flow speed according to the video data collected by the video detector in the station passage of the rail transit.
CN201810024755.3A 2018-01-10 2018-01-10 Method and device for predicting short-time passenger flow state of rail transit station Active CN108280540B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810024755.3A CN108280540B (en) 2018-01-10 2018-01-10 Method and device for predicting short-time passenger flow state of rail transit station

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810024755.3A CN108280540B (en) 2018-01-10 2018-01-10 Method and device for predicting short-time passenger flow state of rail transit station

Publications (2)

Publication Number Publication Date
CN108280540A CN108280540A (en) 2018-07-13
CN108280540B true CN108280540B (en) 2021-09-28

Family

ID=62803389

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810024755.3A Active CN108280540B (en) 2018-01-10 2018-01-10 Method and device for predicting short-time passenger flow state of rail transit station

Country Status (1)

Country Link
CN (1) CN108280540B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108877272B (en) * 2018-08-02 2020-12-22 哈尔滨工程大学 Vehicle navigation system and method based on destination state
CN110119845A (en) * 2019-05-11 2019-08-13 北京京投亿雅捷交通科技有限公司 A kind of application method of track traffic for passenger flow prediction
CN111401643B (en) * 2020-03-19 2022-10-04 卡斯柯信号有限公司 Urban rail transit passenger flow loop self-adaptive intelligent train scheduling method
CN112418518A (en) * 2020-11-20 2021-02-26 佳都新太科技股份有限公司 Passenger flow prediction method and device based on time characteristic weight and network topology
CN113159408B (en) * 2021-04-14 2023-11-21 交控科技股份有限公司 Rail transit station passenger flow prediction method and device
CN113537596B (en) * 2021-07-16 2024-08-06 南京理工大学 Short-time passenger flow prediction method for new line station of urban rail transit

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103730006A (en) * 2014-01-26 2014-04-16 吉林大学 Short-time traffic flow combined forecasting method
CN107067076A (en) * 2017-05-27 2017-08-18 重庆大学 A kind of passenger flow forecasting based on time lag NARX neutral nets

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103730006A (en) * 2014-01-26 2014-04-16 吉林大学 Short-time traffic flow combined forecasting method
CN107067076A (en) * 2017-05-27 2017-08-18 重庆大学 A kind of passenger flow forecasting based on time lag NARX neutral nets

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
城市轨道交通客流增长滞后性与预测方法研究;陈小鸿等;《城市轨道交通研究》;20141110(第2014期);p22-27 *
基于向量误差修正模型的短时交通参数预测;邴其春等;《吉林大学学报》;20150731;第45卷(第4期);p1076-1081 *

Also Published As

Publication number Publication date
CN108280540A (en) 2018-07-13

Similar Documents

Publication Publication Date Title
CN108280540B (en) Method and device for predicting short-time passenger flow state of rail transit station
Lu et al. A combined method for short-term traffic flow prediction based on recurrent neural network
CN107610464B (en) A kind of trajectory predictions method based on Gaussian Mixture time series models
CN104408913B (en) A kind of traffic flow three parameter real-time predicting method considering temporal correlation
CN103730006B (en) A kind of combination forecasting method of Short-Term Traffic Flow
CN103903430B (en) Dynamic fusion type travel time predicting method with multi-source and isomorphic data adopted
Lindemann et al. Anomaly detection and prediction in discrete manufacturing based on cooperative LSTM networks
CN114495507B (en) Traffic flow prediction method integrating space-time attention neural network and traffic model
CN107153874A (en) Water quality prediction method and system
CN114325395B (en) Battery state determining method and device
CN110991776A (en) Method and system for realizing water level prediction based on GRU network
CN115223365B (en) Road network speed prediction and anomaly identification method based on damping Holt model
CN104599500A (en) Grey entropy analysis and Bayes fusion improvement based traffic flow prediction method
CN109508788A (en) A kind of SDN method for predicting based on arma modeling
CN113468720B (en) Service life prediction method for digital-analog linked random degradation equipment
CN108549955A (en) A kind of charging pile abnormal rate determines method and device
Luger et al. Identification of representative operating conditions of HVAC systems in passenger rail vehicles based on sampling virtual train trips
CN116305985A (en) Local intelligent ventilation method based on multi-sensor data fusion
CN115907154A (en) Industrial internet prediction method and system based on frequency domain and long-term and short-term feature fusion
CN117370771A (en) Knowledge embedding filling soft measurement method based on conditional fractional diffusion
Wang et al. Research on construction cost estimation based on artificial intelligence technology
CN107688556A (en) A kind of real-time travel time computation method based on function type principal component analysis
CN116915122B (en) Self-adaptive control method and system for coal mine frequency conversion equipment
CN118035670A (en) Typhoon wind speed prediction method and system based on Deep-Pred framework
CN112927507A (en) Traffic flow prediction method based on LSTM-Attention

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