CN201974940U - Short-time traffic state predicting device - Google Patents
Short-time traffic state predicting device Download PDFInfo
- Publication number
- CN201974940U CN201974940U CN2011200634390U CN201120063439U CN201974940U CN 201974940 U CN201974940 U CN 201974940U CN 2011200634390 U CN2011200634390 U CN 2011200634390U CN 201120063439 U CN201120063439 U CN 201120063439U CN 201974940 U CN201974940 U CN 201974940U
- Authority
- CN
- China
- Prior art keywords
- traffic
- traffic state
- state data
- unit
- 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.)
- Expired - Lifetime
Links
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The utility model discloses a short-time traffic state predicting device, which comprises a traffic state analyzing unit, a time sequence splitting unit and a traffic state predicting unit, wherein the traffic state analyzing unit is used for reading history traffic state data, and screens the history traffic state data according to a preset time interval, so as to obtain screened history traffic state data; the time sequence splitting unit is used for splitting the screened history traffic state data into traffic state data by a time sequence; and the traffic state predicting unit is used for generating prediction results of future traffic states through the traffic state data and a preset prediction model. Because of the adoption of the technical scheme, the short-time traffic state predicting device splits traffic state data from history traffic state data, so as to precisely present the traffic state characteristics of all the traffic state data. Therefore, induction information release is performed according to history changes, so as to facilitate the changes in traffic flows. In addition, different traffic signal control strategies can be chosen conveniently.
Description
Technical field
The utility model relates to the traffic control equipment technical field, relates in particular to a kind of traffic behavior prediction unit, particularly relates to a kind of sectional traffic behavior prediction unit in short-term.
Background technology
Often will adopt technological means such as traffic flow is induced, traffic signals control, event detection in intelligentized urban traffic control technology, the real-time estimate data that characterize the various parameters of traffic behavior are to realize the foundation of above-mentioned technological means necessity.Therefore the real-time estimate of each parameter of traffic behavior is that traffic administration realizes information-based, modernized, intelligent, integrated prerequisite and key, and the accuracy that predicts the outcome of traffic behavior parameter is directly connected to intelligent real-time, accuracy and the high efficiency of inducing and controlling in the traffic administration.
But in the prior art, the traffic state data that constitutes by the detected various traffic behavior parameters of traffic detecting device (flow, occupation rate, speed etc.) all belongs to historical data, can only characterize the state of the traffic flow operation that has taken place.Adopt above-mentioned historical data as traffic flow in short-term in future induce, the decision-making foundation of technological means such as traffic signals control, event detection, certainly will cause the hysteresis of operating strategy, can not accomplish to improve traffic with suiting measures to differing conditions in terms of time.Therefore, the prediction of traffic behavior in short-term is exactly the statistics according to a large amount of historical datas, adopt relevant Forecasting Methodology to future traffic flow modes in short-term estimate, thereby improve intelligent real-time, accuracy and the high efficiency of inducing and controlling in the traffic administration.
The existing Forecasting Methodology of traffic behavior in short-term is autoregressive moving average method, autoregression method, running mean and method history averaging method etc. for example, the various supplemental characteristics of foundation are all comparatively single, parameter is generally all used the least square method On-line Estimation, have calculate easy, be easy to more new data, be convenient to advantage such as large-scale application.But these models fail to reflect the uncertainty and the speciality such as non-linear of traffic behavior, can't overcome the influence of random disturbance factor to the magnitude of traffic flow, so along with the shortening of time in predicting interval, the precision of prediction of above-mentioned model will greatly reduce.In addition, traffic behavior also has periodically variable characteristics, for example have period of state such as morning peak, Ping Feng, evening peak in one day the traffic behavior, if in each period of state, use a unified forecast model, can not very accurately characterize out the variation characteristic of traffic behavior in a certain period, thereby be unfavorable for accurate estimation traffic behavior in short-term in future.
Therefore, those skilled in the art is devoted to develop the high prediction unit of traffic behavior in short-term of a kind of precision of prediction always.
The utility model content
The technical problems to be solved in the utility model is in order to overcome the shortcoming that existing Forecasting Methodology acquisition parameter comparatively simply can't accurately characterize the traffic behavior variation characteristic, a kind of traffic behavior prediction unit is provided, by cutting apart to historical traffic state data, thereby accurately characterize the traffic behavior characteristic of each time period, with traffic signal control strategy to different traffic state data selection different schemes, thus the forecasting accuracy of raising future transportation state.
The utility model solves above-mentioned technical matters by following technical proposals:
The utility model provides a kind of prediction unit of traffic behavior in short-term, comprising:
One traffic behavior analytic unit, be used to read historical traffic state data, and obtain the historical traffic state data in screening back according to screening described historical traffic state data a predefined interval time;
One time sequence cutting unit, be used for the historical traffic state data in described screening back is partitioned into some traffic state datas by the time sequence; And
One traffic behavior predicting unit, be used for producing predicting the outcome of future transportation state by described traffic state data and default forecast model.
Preferably, described traffic behavior analytic unit also is used to preestablish a time Cycle Length, cuts apart with the time series to historical traffic state data.
Preferably, the forecast model of described traffic behavior predicting unit is fuzzy consensus forecast model, changes with the non-stationary that reduces the time series traffic behavior that the subjective factor that goes out the pedestrian in the traffic causes.
Preferably, described time series cutting unit also is used for respectively the described traffic state data that is partitioned into being carried out the trend smoothing processing, changes the influence of fluctuation to error with the big traffic that reduces indivedual appearance.
Application of the present utility model may further comprise the steps:
S
101, the user sets the interval time of traffic behavior analytic unit and the forecast model of traffic behavior predicting unit;
S
102, described traffic behavior analytic unit reads in historical traffic state data and obtains the historical traffic state data in screening back according to screening historical traffic state data described interval time;
S
103, described time series cutting unit is partitioned into some traffic state datas with the historical traffic state data in described screening back by the time sequence;
S
104, the forecast model set according to the user of described traffic behavior predicting unit and the described some traffic state datas that are partitioned into produce predicting the outcome of future transportation states.
Preferably, described traffic behavior analytic unit also is used to preestablish time cycle length, and at step S
101In also comprise: the user sets the time cycle length of traffic behavior analytic unit.
Preferably, at step S
102In further comprising the steps of: described traffic behavior analytic unit also screens historical traffic state data according to described time cycle length.
Preferably, the forecast model of described traffic behavior predicting unit is fuzzy consensus forecast model, and at step S
104In further comprising the steps of: described traffic behavior predicting unit produces predicting the outcome of future transportation state according to described simulation consensus forecast model and traffic state data.
Preferably, described time series cutting unit also is used for respectively described traffic state data being carried out the trend smoothing processing, and at step S
103In further comprising the steps of: described time series cutting unit carries out the trend smoothing processing to the described traffic state data that is partitioned into respectively.
Positive progressive effect of the present utility model is:
Traffic behavior prediction unit of the present utility model, owing to adopted above-mentioned technological means, by from historical traffic state data, being partitioned into some traffic state datas, thereby accurately characterize the traffic behavior characteristic of each traffic state data, to carry out the induction information issue according to historical variations, be convenient to the variation of macro-control traffic flow, but also be convenient to select different traffic signal control strategy.
In addition, device of the present utility model is also by adopting the method for fuzzy consensus forecast, reduced the influence of traffic behavior non-stationary variation, reduced the complexity of Forecasting Methodology and the requirement of applied environment, thereby can realize large-scale application the prediction of future transportation state.
Description of drawings
Fig. 1 is the theory structure synoptic diagram of the prediction unit of traffic behavior in short-term of the present utility model preferred embodiment.
Fig. 2 is the principle flow chart of traffic behavior Forecasting Methodology of the present utility model preferred embodiment.
Fig. 3 is the traffic state data sectionally smooth figure as a result in the utility model.
Embodiment
Provide the utility model preferred embodiment below in conjunction with accompanying drawing, to describe the technical solution of the utility model in detail.
Figure 1 shows that the theory structure synoptic diagram of the prediction unit of traffic behavior in short-term of the present utility model, the prediction unit of traffic behavior in short-term wherein of the present utility model comprises a traffic behavior analytic unit 1, a time sequence cutting unit 2 and a traffic behavior predicting unit 3 in one embodiment.
Wherein, traffic behavior analytic unit 1 is used to read historical traffic state data, and according to a predefined interval time and a time Cycle Length screening historical traffic state data.
For example, existing historical traffic state data is 30 seconds serving as to gather once at interval, but so the data in the short time interval not have too big meaning to analyzing the traffic Changing Pattern, and can increase complexity and the operation time of predicting computing.In the present embodiment, by setting interval time is to screen historical traffic state data in 5 minutes to obtain the historical traffic state data in screening back, thereby under the prerequisite that does not reduce the traffic behavior characteristic, reduce the quantity of the data that need computing as best one can, to reduce the complexity and the operation time of prediction computing effectively.
Also can freely set according to the situation of time traffic and data acquisition and not only be confined to above-mentioned length interval time above-mentioned interval time.
In addition, because road traffic is to have periodically, so traffic state data is a kind of data with cyclical variation rule.Thus, can also set the time cycle length of the historical traffic data of screening according to the actual state of traffic.
For example, be 1 day with the time cycle length setting, thereby can screen needed each cycle in the historical traffic data as required.And choosing of above-mentioned time cycle length can freely set according to the situation of traffic and data acquisition, and not only be confined to above-mentioned time cycle length.For example, the setting-up time Cycle Length is 1 time-of-week or 1 month or the like.
Above-mentioned time series cutting unit 2 is used for the historical traffic state data in screening back is partitioned into some traffic state datas by the time sequence.
For example, use
Representing the time series of historical traffic state data after the screening, wherein is time variable,
Be real number.
Get limited point
, in the prediction of traffic behavior, in order to serve actual control mode, time series all is time uniformly-spaced basically, note
Or brief note
,
Be
Constantly or the
The observed reading in cycle,
Certainly, if in the prediction of traffic behavior, need the non-equal intervals time, can also adopt other partitioning scheme that the time series of screening the historical traffic state data in back is cut apart.
By above-mentioned time series is divided into n time period, thereby can from historical traffic state data, be partitioned into traffic state data corresponding to each time period.For example, the traffic state data in 1 day 24 hours can be marked off 7 point-10 point, evening peak 16-19 point and 10 point-16, the flat peak point of morning peak according to time series.
In addition, the user can also be cut apart historical traffic state data according to other rules as required, for example cuts apart etc. by 1 time-of-week, and is not limited only to above-mentioned partitioning scheme.
Above-mentioned time series cutting unit 2 also is used for respectively each traffic state data that is partitioned into being carried out the trend smoothing processing.
Because traffic changes fluctuation greatly in the actual historical traffic state data, bigger to the influence of the error that predicts the outcome, so reducing indivedual big traffic that occur, the trend of employing smoothing processing changes the influence of fluctuation to error.
For example, respectively above-mentioned morning peak, evening peak and the flat peak historical traffic state data in the period is carried out match, thereby set up 3 function models that correspond respectively to the historical traffic state data in above-mentioned morning peak, evening peak and flat peak period, and by aforementioned 3 function models the historical traffic state data of reality is proofreaied and correct, change fluctuation thereby reduce traffic.
And, owing to respectively the historical traffic state data in the different time sections has been carried out match, so with respect to traditional technical approach that historical traffic state data in the whole time series is carried out match, the utility model has improved the accuracy of match, reduce simultaneously the operation time of match again, simplified the complexity of match computing.
Above-mentioned traffic behavior predicting unit 3 is used for producing predicting the outcome of future transportation state by traffic state data and default forecast model.
In the present embodiment, the forecast model of traffic behavior predicting unit 3 is fuzzy consensus forecast model.
Be provided with aforesaid time series in the above-mentioned consensus forecast model
, note
,
Be
On a fuzzy subset, subordinate function is
, be time series in the following formula (1) then
In fuzzy constraint
On fuzzy mean value.
The non-stationary that can reduce the time series traffic behavior that the subjective factor that goes out the pedestrian in the traffic causes by above-mentioned fuzzy consensus forecast model changes.
In addition, traffic behavior predicting unit 3 in the present embodiment also can adopt traditional average forecast model according to the reality needs of traffic behavior prediction unit or user's actual demand in short-term, when the time sequence table now is random fluctuation, and when fluctuation ratio was steady, the average forecast model was more effective.
For example, for given time observation sequence
, in order to predict next constantly or the data in cycle,
, adopt
The mean value approximate evaluation, shown in (2):
In addition because in actual traffic stream situation, recent data compared with early data for prediction
Even more important, so this gives bigger weight coefficient with regard to requiring to recent data.For this reason, can be at above-mentioned average forecast model
The middle weight coefficient that adds improves average forecast model prediction accuracy further.
Figure 2 shows that the prediction unit of traffic behavior in short-term of the present utility model principle flow chart in use, in one embodiment, may further comprise the steps at least:
Step S
101, the user sets the forecast model of interval time, time cycle length and the traffic behavior predicting unit of traffic behavior analytic unit.
For example, in the present embodiment, it is 5 minutes that the user is provided with interval time, and time cycle length is 1 day.
Step S
102, the traffic behavior analytic unit reads in historical traffic state data and according to above-mentioned interval time and time cycle length, screens historical traffic state data and obtain the historical traffic state data in screening back.
For example, with filtering out 336 groups of traffic state datas altogether with the traffic state data collected at interval in 1 minute according to the time cycle length of 5 minutes interval time and 1 day in 6 continuous two days point-20,168 groups of traffic state datas are arranged in every day wherein.
Step S
103, the time series cutting unit will screen the historical traffic state data in back and be partitioned into some traffic state datas by the time sequence.The time series cutting unit carries out the trend smoothing processing to the traffic state data that is partitioned into respectively.
As shown in Figure 3, in the present embodiment, 168 groups of traffic state datas in above-mentioned 1 day are partitioned into 4 sections according to peace peak, peak, solid line wherein be shown in the curve of 168 groups of traffic state datas.
, respectively above-mentioned 4 section traffic state datas set up 4 function models thereafter, shown in (3):
（3）
Then, 4 sections traffic state datas are carried out smoothing processing, thereby obtain the traffic state data curve after the smoothing processing of dotted line as shown in Figure 3.
Step S
104, the traffic behavior predicting unit produces predicting the outcome of future transportation state according to user simulation consensus forecast model of setting and the traffic state data that is partitioned into.
For example, adopted 2 days traffic state data in the above-described embodiments, so
Be the time domain, and the corresponding fuzzy set of setting is
, be example with the 20th group here, calculate the 20th group of traffic state data by following formula (4) corresponding to the corresponding time period in 1 day future.
According to above-mentioned using method the totally 336 groups of traffic state datas in 2 days are blured consensus forecast, thereby predicted the outcome.
By to the above-mentioned analysis that predicts the outcome, maximum error is 60%, and minimum error is about 4%, and average error was 7.4% in one day.
Using method in present embodiment comparative neural network, kalman filtering etc. aspect operability is all more simple.Though error ratio of the present utility model is big, the complexity of its operability is low, and is little to environmental requirement, be convenient on a large scale, apply on a large scale, be particularly suited for the traffic control device not high to accuracy requirement, for example different conditions induces decision-making, control strategy etc.
Though more than described embodiment of the present utility model, it will be understood by those of skill in the art that these only illustrate, protection domain of the present utility model is limited by appended claims.Those skilled in the art can make numerous variations or modification to these embodiments under the prerequisite that does not deviate from principle of the present utility model and essence, but these changes and modification all fall into protection domain of the present utility model.
Claims (4)
1. traffic behavior prediction unit in short-term is characterized in that the described prediction unit of traffic behavior in short-term comprises:
One traffic behavior analytic unit is used to read historical traffic state data, and obtains the historical traffic state data in screening back according to screening described historical traffic state data a predefined interval time;
One time sequence cutting unit is used for the historical traffic state data in described screening back is partitioned into some traffic state datas by the time sequence; And
One traffic behavior predicting unit is used for producing predicting the outcome of future transportation state by a described traffic state data and a default forecast model.
2. the prediction unit of traffic behavior in short-term as claimed in claim 1 is characterized in that described traffic behavior analytic unit also is used to preestablish a time Cycle Length.
3. the prediction unit of traffic behavior in short-term as claimed in claim 1 is characterized in that, the forecast model of described traffic behavior predicting unit is fuzzy consensus forecast model.
4. the prediction unit of traffic behavior in short-term as claimed in claim 1 is characterized in that, described time series cutting unit also is used for respectively the described traffic state data that is partitioned into being carried out the trend smoothing processing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011200634390U CN201974940U (en) | 2011-03-11 | 2011-03-11 | Short-time traffic state predicting device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011200634390U CN201974940U (en) | 2011-03-11 | 2011-03-11 | Short-time traffic state predicting device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN201974940U true CN201974940U (en) | 2011-09-14 |
Family
ID=44580037
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011200634390U Expired - Lifetime CN201974940U (en) | 2011-03-11 | 2011-03-11 | Short-time traffic state predicting device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN201974940U (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102087787A (en) * | 2011-03-11 | 2011-06-08 | 上海千年工程建设咨询有限公司 | Prediction device and prediction method for short time traffic conditions |
| CN103995808A (en) * | 2013-02-17 | 2014-08-20 | 中国电信股份有限公司 | Method and device for detecting events on time series |
| CN106779222A (en) * | 2016-12-20 | 2017-05-31 | 中国人民解放军空军装备研究院雷达与电子对抗研究所 | Airport ground stand-by period Forecasting Methodology and device |
| 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 |
-
2011
- 2011-03-11 CN CN2011200634390U patent/CN201974940U/en not_active Expired - Lifetime
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102087787A (en) * | 2011-03-11 | 2011-06-08 | 上海千年工程建设咨询有限公司 | Prediction device and prediction method for short time traffic conditions |
| CN102087787B (en) * | 2011-03-11 | 2013-06-12 | 上海千年城市规划工程设计股份有限公司 | Prediction device and prediction method for short time traffic conditions |
| CN103995808A (en) * | 2013-02-17 | 2014-08-20 | 中国电信股份有限公司 | Method and device for detecting events on time series |
| CN103995808B (en) * | 2013-02-17 | 2018-02-02 | 中国电信股份有限公司 | Event detecting method and device in time series |
| CN106779222A (en) * | 2016-12-20 | 2017-05-31 | 中国人民解放军空军装备研究院雷达与电子对抗研究所 | Airport ground stand-by period Forecasting Methodology and device |
| CN106779222B (en) * | 2016-12-20 | 2020-11-24 | 中国人民解放军空军装备研究院雷达与电子对抗研究所 | Airport ground waiting time prediction method and device |
| 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 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102087787B (en) | Prediction device and prediction method for short time traffic conditions | |
| CN105702029B (en) | A kind of Expressway Traffic trend prediction method for considering space-time relationship at times | |
| CN107085943B (en) | Short-term prediction method and system for road travel time | |
| CN102831325B (en) | A kind of bearing fault Forecasting Methodology returned based on Gaussian process | |
| CN104715292A (en) | City short-term water consumption prediction method based on least square support vector machine model | |
| CN201974940U (en) | Short-time traffic state predicting device | |
| CN112417788A (en) | Water environment pollution analysis system and method based on big data | |
| CN103903452A (en) | Traffic flow short time predicting method | |
| CN104748839A (en) | Hydroelectric generating unit vibration state region monitoring method based on real-time online monitoring | |
| CN105305426B (en) | Mapreduce two-step short-period load prediction method based on deviation control mechanism | |
| CN113222296A (en) | Flood control scheduling method based on digital twin | |
| CN102402726A (en) | Method for predicting electric quantity of large-scale distribution network based on regional load analysis | |
| CN102937534B (en) | Method for predicting fault of electromechanical device based on combined prediction model | |
| CN103679288B (en) | Method for short-term prediction of generated output of runoff small hydropower station cluster and prediction system | |
| CN108415884A (en) | A kind of modal parameters real-time tracing method | |
| CN109116444B (en) | PCA-kNN-based air quality model PM2.5Forecasting method | |
| CN102109619A (en) | System and method for predicting typhoon surge based on artificial intelligence | |
| CN104407688A (en) | Virtualized cloud platform energy consumption measurement method and system based on tree regression | |
| CN106682784A (en) | Analysis method and system of smart power grid big data | |
| CN104281776A (en) | Method for judging remarkable influence period of human activities on river flow | |
| CN111626006A (en) | Quantitative identification method for multivariate evaluation on runoff change sensitivity | |
| CN1361480A (en) | Vibration throw parameter monitoring and analyzing method for water power set | |
| CN106382102A (en) | Overflow early warning method based on clustering algorithm | |
| CN116402390A (en) | Road construction area traffic influence evaluation method and system based on primitive model | |
| CN107135125A (en) | Video IDC predicting bandwidth flow method and devices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C56 | Change in the name or address of the patentee |
Owner name: SHANGHAI THOUSAND YEAR CONSTRUCTION DESIGN CO., LT Free format text: FORMER NAME: SHANGHAI THOUSAND YEAR CONSTRUCTION CONSULTANTS CO., LTD. |
|
| CP01 | Change in the name or title of a patent holder |
Address after: 201100 Shanghai city Minhang District Zhuan Hing Road No. 688 Patentee after: Shanghai Thousandyear Urban Planning Engineering Design Co., Ltd. Address before: 201100 Shanghai city Minhang District Zhuan Hing Road No. 688 Patentee before: Shanghai Thousand Year Consultants Ltd. |
|
| AV01 | Patent right actively abandoned |
Granted publication date: 20110914 Effective date of abandoning: 20130612 |
|
| RGAV | Abandon patent right to avoid regrant |