CN114549930A - Rapid road short-time vehicle head interval prediction method based on trajectory data - Google Patents
Rapid road short-time vehicle head interval prediction method based on trajectory data Download PDFInfo
- Publication number
- CN114549930A CN114549930A CN202210157748.7A CN202210157748A CN114549930A CN 114549930 A CN114549930 A CN 114549930A CN 202210157748 A CN202210157748 A CN 202210157748A CN 114549930 A CN114549930 A CN 114549930A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- data
- lstm
- gan
- time sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 230000003137 locomotive effect Effects 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 14
- 230000009466 transformation Effects 0.000 claims abstract description 13
- 238000010606 normalization Methods 0.000 claims abstract description 8
- 230000008859 change Effects 0.000 claims abstract description 5
- 238000013256 Gubra-Amylin NASH model Methods 0.000 claims description 18
- 239000013598 vector Substances 0.000 claims description 16
- 230000006870 function Effects 0.000 claims description 14
- 238000012549 training Methods 0.000 claims description 8
- 150000001875 compounds Chemical class 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000003042 antagnostic effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Probability & Statistics with Applications (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Analytical Chemistry (AREA)
- Bioinformatics & Computational Biology (AREA)
- Chemical & Material Sciences (AREA)
- Primary Health Care (AREA)
- Educational Administration (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Train Traffic Observation, Control, And Security (AREA)
Abstract
The invention discloses a method for predicting the short-time locomotive spacing of a fast way based on trajectory data, which belongs to the technical field of locomotive spacing prediction and comprises the following steps: the method comprises the following steps: acquiring vehicle track data of the express way and carrying out normalization processing; step two: converting the normalized vehicle i time sequence data into a polar coordinate system time sequence by using a coordinate change method; step three: carrying out trigonometric function transformation of angles and/or differences; step four: inputting a vehicle head distance two-dimensional image obtained by utilizing the GADF to a GAN-LSTM model, and outputting a vehicle head distance predicted value; the time sequence of the vehicle head distance is extracted based on the vehicle track data, and the GAF-GAN-LSTM model is utilized to predict the vehicle head distance between the express way vehicles, so that the position distribution characteristics of the vehicles on the express way in the future period are obtained, and support is provided for traffic state prediction, vehicle guidance and intelligent management and control of the express way.
Description
Technical Field
The invention belongs to the technical field of vehicle head distance prediction, and particularly relates to a method for predicting a short-time vehicle head distance of a fast path based on track data.
Background
With the increase of the quantity of motor vehicles kept, the problem of expressway traffic jam is increasingly highlighted. In order to alleviate the problem of traffic jam of the express way, the future traffic flow running state of the express way needs to be predicted, so that a real-time traffic management scheme is formulated. The existing method for predicting short-term traffic flow of the express way mainly focuses on predicting macroscopic traffic flow parameters, such as: traffic flow, interval average speed, interval travel time, etc. With the popularization and application of radar monitoring technology and vehicle-mounted GPS equipment, the acquisition of microscopic vehicle track data becomes possible. The distance between the two car heads is used as an important parameter in the field of microscopic traffic flow, and reflects the position relation between the two front and rear cars. In addition, the development of the car networking technology provides technical support for real-time guidance of the vehicle. By predicting the distribution characteristic of the distance between the vehicle heads of the express way at the future time, the ramp networked vehicles are guided to merge into the main line by using the main line gap, so that the main line gap is utilized to the maximum extent, the influence of continuous merging of the ramp vehicles on the operation of the main line traffic flow is reduced, the fine management and control of the express way are realized, and the integral stable operation of the express way is ensured.
However, the existing short-time locomotive interval prediction method cannot fully mine the correlation between the historical locomotive interval sequence and the future locomotive interval sequence, and meanwhile, the prediction precision is limited. Therefore, the method combines the strong advantages of deep learning in the field of feature extraction and image prediction, converts one-dimensional headway distance time sequence data into a two-dimensional picture by using GAF, and extracts the correlation between the historical headway distance and the future headway distance based on LSTM to obtain the headway distance feature vector with stronger representation capability. And the strong advantages of the generated countermeasure network GAN in the field of image prediction are utilized to predict the distance between the car heads at the future moment.
Disclosure of Invention
In order to solve the problems existing in the scheme, the invention provides a method for predicting the short-time locomotive distance of the express way based on track data.
The purpose of the invention can be realized by the following technical scheme:
a method for predicting a short-time locomotive interval of a fast way based on track data comprises the following specific steps:
the method comprises the following steps: acquiring vehicle track data of the express way and carrying out normalization processing;
step two: converting the normalized vehicle i time sequence data into a polar coordinate system time sequence by using a coordinate change method;
step three: carrying out trigonometric function transformation of angles and/or differences;
step four: and inputting the two-dimensional image of the vehicle head space obtained by utilizing the transformation of the gram difference angular field GADF into the GAN-LSTM model, and outputting a predicted value of the vehicle head space.
Further, the method for performing normalization processing comprises:
the individual vehicles are labeled as i,obtaining the time sequence data of the head interval of a single vehicle iScaling the time sequence of the intervals between the car heads to [0, 1%]Or [ -1,1 [)]An interval; according to the formula:
in [0,1 ]]Or [ -1,1 [)]Normalizing the interval of (A); in the formula (I), the compound is shown in the specification,in order to normalize the distance between the car heads,for the raw headway data, min (d) and max (d) are the minimum and maximum values, respectively, in the vehicle i headway timing data set.
Further, the method for converting the normalized vehicle i time sequence data into the polar coordinate system time sequence comprises the following steps:
in the formula (I), the compound is shown in the specification,is the polar angle corresponding to the distance between the heads of the vehicle i at the time t,for the normalized headway distance of the vehicle i at the time t,when the distance between the heads of the vehicles iAnd in the ordinal data set, r is the polar radius corresponding to the headway distance of the vehicle i at the time t, M is the total number of the headway distance timestamps of the vehicle i, and N is a timestamp set.
Further, the method of performing trigonometric function transformation of angles and/or differences comprises:
using a cos function of the two angle sums to obtain a gram sum angle field GASF, and using a sin function of the two angle differences to obtain a gram difference angle field GADF; the GASF and GADF calculation methods are as follows:
further, the method for establishing the GAN model comprises the following steps:
step SA 1: acquiring and generating a confrontation type network GAN, and setting initial parameters of a GAN model, wherein the initial parameters of the GAN model comprise: learning rate eta, batch processing quantity m and generator initial parameter thetag0And an initial parameter theta of the discriminatord0Number of generator iterations nGAnd the number of iterations n of the arbiter in each generator iterationD(ii) a Inputting the GAF converted image into a generation countermeasure network GAN;
step SA 2: randomly select m sample points { d } from the dataset1,d2,...,dm-randomly selecting m vectors z from the noisy data distribution1,z2,...,zm};
Step SA 3: inputting Z into the generator to obtain m generated data d1,d2,...,dm};
Step SA 5: judging whether the iteration number is more than nDWhen the number of iterations is greater than nDIf so, stopping iteration; the number of iterations is not more than nDIf so, continuing to step SA 1-step SA4 until an iteration stop condition is met;
step SA 6: again, m vectors z are randomly sampled from the noise distribution1,z2,...,zm};
Step SA 8: judging whether the maximum iteration number n of the generator is reachedGWhen the maximum number of iterations n of the generator is reachedGIf so, stopping iterative training and outputting the trained GAN model; when the maximum iteration number n of the generator is not reachedGAnd if so, continuing to step SA 2-step SA7 until an iteration stop condition is met.
Further, in step SA 4;
further, in step SA 7:
further, the method for establishing the GAN-LSTM model comprises the following steps:
step SB 1: connecting a generator in the trained GAN model with the LSTM;
step SB 2: setting initial parameters of the LSTM modelThe number of the components comprises: learning rate eta, batch processing quantity m, and LSTM initial parameter thetaf0Maximum number of iterations n of LSTMfAnd vehicle i locomotive spacing time sequence data; inputting the GAF-converted image into the LSTM;
step SB 3: randomly selecting m vectors { v ] from real data1,ν2,...,νm};
Step SB 4: updating parameter θ of LSTMf;
Step SB 5: judging whether the maximum iteration number n of the LSTM is reachedfWhen the maximum number of iterations n of LSTM is reachedfStopping iterative training and outputting the trained GAN-LSTM model; when the maximum number of iterations n of the LSTM is not reachedfAnd if so, continuing to step SB 2-step SB4 until the iteration stop condition is met.
Compared with the prior art, the invention has the beneficial effects that: by extracting a time sequence of the vehicle head distance based on vehicle track data and predicting the vehicle head distance between the vehicles on the express way by using a GAN-LSTM model, the position distribution characteristics of the vehicles on the express way in a future period are obtained, and support is provided for traffic state prediction, vehicle guidance and intelligent management and control of the express way.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of the present invention for vehicle headway distance prediction;
FIG. 2 is a schematic representation of the GADF of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
As shown in fig. 1 to 2, a method for predicting a short-time locomotive distance of a fast way based on trajectory data specifically includes:
the method comprises the following steps: acquiring vehicle track data of the express way and carrying out normalization processing;
acquiring vehicle track data of the express way by methods including but not limited to video acquisition, radar equipment acquisition, vehicle-mounted GPS or Beidou navigation equipment acquisition and the like, and acquiring corresponding vehicle track data by a corresponding vehicle track data acquisition method;
illustratively, a navigation map is acquired by collecting through a vehicle-mounted GPS or Beidou navigation equipment, and the position of a vehicle in the navigation map is acquired in real time according to the vehicle-mounted GPS or Beidou navigation equipment to acquire a vehicle track, so as to acquire time sequence data corresponding to the vehicle track.
The method for carrying out normalization processing comprises the following steps:
label individual vehicle as i; obtaining the time sequence data of the head interval of a single vehicle iScaling the time sequence of the intervals between the car heads to [0, 1%]Or [ -1,1 [)]An interval; according to the formula:
in [0,1 ]]Or [ -1,1 [)]Normalizing the interval of (A); in the formula (I), the compound is shown in the specification,in order to normalize the distance between the car heads,and (c) taking the original headway data, namely the headway value of the ith vehicle at the time t, wherein min (D) and max (D) are respectively the minimum value and the maximum value in the headway time sequence data set of the vehicle i.
Step two: converting the normalized vehicle i time sequence data into a polar coordinate system time sequence by using a coordinate change method;
the method for converting the normalized vehicle i time sequence data into the polar coordinate system time sequence comprises the following steps:
in the formula (I), the compound is shown in the specification,is the polar angle corresponding to the distance between the heads of the vehicle i at the time t,for the normalized headway distance of the vehicle i at the time t,the method comprises the steps of collecting time sequence data of the vehicle head interval of a vehicle i, wherein r is a polar radius corresponding to the vehicle head interval of the vehicle i at the moment t, M is the total number of timestamps of the vehicle head interval of the vehicle i, and N is a timestamp collection.
Step three: carrying out trigonometric function transformation of angles and/or differences;
the method for performing trigonometric function transformation of angles and/or differences comprises:
using a cos function of the two angle sums to obtain a gram sum angle field GASF, and using a sin function of the two angle differences to obtain a gram difference angle field GADF; the calculation methods for GASF and GADF are as follows:
as shown in fig. 2, the present application selects a gram difference angular field GADF to convert the one-dimensional time series data of the headway distance of the vehicle into a two-dimensional image.
Step four: and inputting the two-dimensional image of the vehicle head space obtained by utilizing the transformation of the gram difference angular field GADF into the GAN-LSTM model, and outputting a predicted value of the vehicle head space.
The method for establishing the GAN model comprises the following steps:
step SA 1: acquiring and generating a confrontation type network GAN, and setting initial parameters of a GAN model, wherein the initial parameters of the GAN model comprise: learning rate eta, batch processing amount m, generator initial parameter thetag0And an initial parameter theta of a discriminatord0Number of generator iterations nGAnd the number of iterations n of the arbiter in each generator iterationD(ii) a Inputting the GAF converted image into a generation countermeasure network GAN; generating an antagonistic network GAN as an existing deep learning model;
step SA 2: randomly choosing m sample points { d } from the dataset1,d2,...,dm-randomly selecting m vectors z from the noisy data distribution1,z2,...,zm}; the noise data distribution comprises a Gaussian distribution mode, a normal distribution mode and other distribution modes;
step SA 3: inputting Z into the generator to obtain m generated data d1,d2,...,dm}; z is the vector in step SA 2;
step SA 4: updating the parameter θ of the discriminatordTo maximizeWherein the content of the first and second substances,is as follows;
step SA 5: judging whether the iteration number is more than nDWhen the number of iterations is greater than nDIf so, stopping iteration; the number of iterations is not more than nDIf so, continuing to step SA 1-step SA4 until an iteration stop condition is met;
step SA 6: again, m vectors z are randomly sampled from the noise distribution1,z2,...,zmThe vector does not need to be consistent with the vector in step SA 2;
Step SA 8: judging whether the maximum iteration number n of the generator is reachedGWhen the maximum number of iterations n of the generator is reachedGIf so, stopping iterative training and outputting the trained GAN model; when the maximum iteration number n of the generator is not reachedGAnd if so, continuing to step SA 2-step SA7 until an iteration stop condition is met.
The method for establishing the GAN-LSTM model comprises the following steps:
step SB 1: connecting a generator in the trained GAN model with the LSTM;
step SB 2: setting initial parameters of the LSTM model, wherein the initial parameters of the LSTM model comprise: learning rate eta, batch processing quantity m, and LSTM initial parameter thetaf0Maximum number of iterations n of LSTMfAnd vehicle i locomotive spacing time sequence data; inputting the GAF-converted image into the LSTM;
step SB 3: randomly selecting m vectors { v ] from real data1,ν2,...,νm};
Step SB 4: updating parameter θ of LSTMf;
Step SB 5: judging whether the maximum iteration number n of the LSTM is reachedfWhen the maximum number of iterations n of the LSTM is reachedfStopping iterative training and outputting the trained GAN-LSTM model; when the maximum number of iterations n of the LSTM is not reachedfAnd if so, continuing to step SB 2-step SB4 until the iteration stop condition is met.
The time sequence of the vehicle head distance is extracted based on the vehicle track data, and the GAF-GAN-LSTM model is utilized to predict the vehicle head distance between the express way vehicles, so that the position distribution characteristics of the vehicles on the express way in the future period are obtained, and support is provided for traffic state prediction, vehicle guidance and intelligent management and control of the express way.
The above formulas are all calculated by removing dimensions and taking numerical values thereof, the formula is a formula which is obtained by acquiring a large amount of data and performing software simulation to obtain the closest real situation, and the preset parameters and the preset threshold value in the formula are set by the technical personnel in the field according to the actual situation or obtained by simulating a large amount of data.
Working principle of the invention: acquiring vehicle track data of the express way and carrying out normalization processing; converting the normalized vehicle i time sequence data into a polar coordinate system time sequence by using a coordinate change method; using a cos function of the two angle sums to obtain a gram sum angle field GASF, and using a sin function of the two angle differences to obtain a gram difference angle field GADF; acquiring a generated countermeasure network GAN, setting initial parameters of a GAN model, and inputting an image converted by GAF into the generated countermeasure network GAN; randomly selecting m sample points from the data set, and randomly selecting m vectors from the noise data distribution; inputting Z into a generator to obtain m generated data; updating the parameter θ of the discriminatordTo maximizeJudging whether the iteration number is more than nDWhen the number of iterations is greater than nDIf so, stopping iteration; the number of iterations is not more than nDIf so, continuing to step SA 1-step SA4 until an iteration stop condition is met; randomly sampling m vectors from the noise distribution again; updating the parameter θ of the generatordTo minimizeJudging whether the maximum iteration number n of the generator is reachedGWhen the maximum number of iterations n of the generator is reachedGIf so, stopping iterative training and outputting the trained GAN model; when the maximum iteration number n of the generator is not reachedGIf so, continuing to step SA 2-step SA7 until an iteration stop condition is met;
connecting a generator in the trained GAN model with the LSTM; setting initial parameters of an LSTM model, and inputting the GAF converted image into the LSTM; randomly selecting m vectors from real data; updating parameter θ of LSTMf(ii) a Judging whether the maximum iteration number n of the LSTM is reachedfWhen the maximum number of iterations n of the LSTM is reachedfStopping iterative training and outputting the trained GAN-LSTM model; when the maximum number of iterations n of the LSTM is not reachedfIf so, continuing the step SB 2-the step SB4 until the iteration stop condition is met; will utilize gramAnd inputting the two-dimensional image of the vehicle head space obtained by converting the difference angle field GADF into the GAN-LSTM model, and outputting a predicted value of the vehicle head space.
Although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the spirit and scope of the present invention.
Claims (8)
1. A method for predicting the short-time locomotive interval of a fast way based on track data is characterized by comprising the following steps:
the method comprises the following steps: acquiring vehicle track data of the express way and carrying out normalization processing;
step two: converting the normalized vehicle i time sequence data into a polar coordinate system time sequence by using a coordinate change method;
step three: carrying out trigonometric function transformation of angles and/or differences;
step four: and inputting the two-dimensional image of the vehicle head space obtained by utilizing the transformation of the gram difference angular field GADF into the GAN-LSTM model, and outputting a predicted value of the vehicle head space.
2. The method for predicting the short-time locomotive distance of the express way based on the trajectory data as claimed in claim 1, wherein the method for carrying out normalization processing comprises the following steps:
marking a single vehicle as i, and acquiring the time sequence data of the distance between the heads of the single vehicle iScaling the time sequence of the intervals between the car heads to [0, 1%]Or [ -1,1 [)]An interval; according to the formula:
in [0,1 ]]Or [ -1,1 [)]Normalizing the interval of (A); in the formula (I), the compound is shown in the specification,in order to normalize the distance between the car heads,for the raw headway data, min (d) and max (d) are the minimum and maximum values, respectively, in the vehicle i headway timing data set.
3. The method for predicting the short-term headway of the express way based on the track data as claimed in claim 1, wherein the method for converting the normalized vehicle i time sequence data into a polar coordinate system time sequence comprises the following steps:
in the formula (I), the compound is shown in the specification,the polar angle corresponding to the distance between the heads of the vehicle i at the time t,for the normalized headway distance of the vehicle i at the time t,the method comprises the steps of collecting time sequence data of the vehicle head interval of a vehicle i, wherein r is a polar radius corresponding to the vehicle head interval of the vehicle i at the moment t, M is the total number of timestamps of the vehicle head interval of the vehicle i, and N is a timestamp collection.
4. The method for predicting the short-term headway of the express way based on the track data as claimed in claim 1, wherein the method for performing trigonometric function transformation on angles and/or differences comprises the following steps:
using a cos function of the two angle sums to obtain a gram sum angle field GASF, and using a sin function of the two angle differences to obtain a gram difference angle field GADF; the calculation methods for GASF and GADF are as follows:
5. the method for predicting the short-term headway of the express way based on the trajectory data as claimed in claim 1, wherein the method for establishing the GAN model comprises the following steps:
step SA 1: acquiring and generating a confrontation type network GAN, and setting initial parameters of a GAN model, wherein the initial parameters of the GAN model comprise: learning rate eta, batch processing amount m, generator initial parameter thetag0And an initial parameter theta of the discriminatord0Number of generator iterations nGAnd the number of iterations n of the discriminators in each generator iterationD(ii) a Inputting the GAF converted image into a generation countermeasure network GAN;
step SA 2: randomly choosing m sample points { d } from the dataset1,d2,...,dm-randomly selecting m vectors z from the noisy data distribution1,z2,...,zm};
Step SA 3: inputting Z into the generator to obtain m generated data d1,d2,...,dm};
Step SA 5: judging whether the iteration number is more than nDWhen the number of iterations is greater than nDIf so, stopping iteration; the number of iterations is not more than nDIf so, continuing to step SA 1-step SA4 until an iteration stop condition is met;
step SA 6: the m vectors z are again randomly sampled from the noise distribution1,z2,...,zm};
Step SA 8: judging whether the maximum iteration number n of the generator is reachedGWhen the maximum number of iterations n of the generator is reachedGIf so, stopping iterative training and outputting the trained GAN model; when the maximum iteration number n of the generator is not reachedGAnd if so, continuing to step SA 2-step SA7 until an iteration stop condition is met.
8. the method for predicting the short-time locomotive distance of the express way based on the trajectory data as claimed in claim 5, wherein the method for establishing the GAN-LSTM model comprises the following steps:
step SB 1: connecting a generator in the trained GAN model with the LSTM;
step SB 2: setting initial parameters of the LSTM model, wherein the initial parameters of the LSTM model comprise: learning rate eta, batch processing quantity m, and LSTM initial parameter thetaf0Maximum number of iterations n of LSTMfAnd vehicle i locomotive spacing time sequence data; inputting the GAF-converted image into the LSTM;
step SB 3: randomly selecting m vectors { v ] from real data1,ν2,...,νm};
Step SB 4: updating parameter θ of LSTMf;
Step SB 5: judging whether the maximum iteration number n of the LSTM is reachedfWhen the maximum number of iterations n of the LSTM is reachedfStopping iterative training and outputting the trained GAN-LSTM model; when the maximum iteration number n of the LSTM is not reachedfAnd if so, continuing to step SB 2-step SB4 until the iteration stop condition is met.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210157748.7A CN114549930B (en) | 2022-02-21 | 2022-02-21 | Rapid road short-time vehicle head interval prediction method based on trajectory data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210157748.7A CN114549930B (en) | 2022-02-21 | 2022-02-21 | Rapid road short-time vehicle head interval prediction method based on trajectory data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114549930A true CN114549930A (en) | 2022-05-27 |
CN114549930B CN114549930B (en) | 2023-01-10 |
Family
ID=81675553
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210157748.7A Active CN114549930B (en) | 2022-02-21 | 2022-02-21 | Rapid road short-time vehicle head interval prediction method based on trajectory data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114549930B (en) |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101739814A (en) * | 2009-11-06 | 2010-06-16 | 吉林大学 | SCATS coil data-based traffic state online quantitative evaluation and prediction method |
CN103116808A (en) * | 2013-01-18 | 2013-05-22 | 同济大学 | Method of real-timely predicting short time traffic flow of express way |
CN109671433A (en) * | 2019-01-10 | 2019-04-23 | 腾讯科技(深圳)有限公司 | A kind of detection method and relevant apparatus of keyword |
CN109948117A (en) * | 2019-03-13 | 2019-06-28 | 南京航空航天大学 | A kind of satellite method for detecting abnormality fighting network self-encoding encoder |
CN110414719A (en) * | 2019-07-05 | 2019-11-05 | 电子科技大学 | A kind of vehicle flowrate prediction technique based on Multi-variable Grey Model time series |
CN110598851A (en) * | 2019-08-29 | 2019-12-20 | 北京航空航天大学合肥创新研究院 | Time series data abnormity detection method fusing LSTM and GAN |
CN111931902A (en) * | 2020-07-03 | 2020-11-13 | 江苏大学 | Countermeasure network generation model and vehicle track prediction method using the same |
CN112000830A (en) * | 2020-08-26 | 2020-11-27 | 中国科学技术大学 | Time sequence data detection method and device |
CN112257850A (en) * | 2020-10-26 | 2021-01-22 | 河南大学 | Vehicle track prediction method based on generation countermeasure network |
CN112330952A (en) * | 2020-09-14 | 2021-02-05 | 浙江工业大学 | Traffic flow prediction method based on generating type countermeasure network |
CN113033619A (en) * | 2021-03-04 | 2021-06-25 | 浙江工业大学 | DVGAE-GAN-based traffic network data restoration method |
CN113469253A (en) * | 2021-07-02 | 2021-10-01 | 河海大学 | Electricity stealing detection method based on triple twin network |
CN113487855A (en) * | 2021-05-25 | 2021-10-08 | 浙江工业大学 | Traffic flow prediction method based on EMD-GAN neural network structure |
-
2022
- 2022-02-21 CN CN202210157748.7A patent/CN114549930B/en active Active
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101739814A (en) * | 2009-11-06 | 2010-06-16 | 吉林大学 | SCATS coil data-based traffic state online quantitative evaluation and prediction method |
CN103116808A (en) * | 2013-01-18 | 2013-05-22 | 同济大学 | Method of real-timely predicting short time traffic flow of express way |
CN109671433A (en) * | 2019-01-10 | 2019-04-23 | 腾讯科技(深圳)有限公司 | A kind of detection method and relevant apparatus of keyword |
CN109948117A (en) * | 2019-03-13 | 2019-06-28 | 南京航空航天大学 | A kind of satellite method for detecting abnormality fighting network self-encoding encoder |
CN110414719A (en) * | 2019-07-05 | 2019-11-05 | 电子科技大学 | A kind of vehicle flowrate prediction technique based on Multi-variable Grey Model time series |
CN110598851A (en) * | 2019-08-29 | 2019-12-20 | 北京航空航天大学合肥创新研究院 | Time series data abnormity detection method fusing LSTM and GAN |
CN111931902A (en) * | 2020-07-03 | 2020-11-13 | 江苏大学 | Countermeasure network generation model and vehicle track prediction method using the same |
CN112000830A (en) * | 2020-08-26 | 2020-11-27 | 中国科学技术大学 | Time sequence data detection method and device |
CN112330952A (en) * | 2020-09-14 | 2021-02-05 | 浙江工业大学 | Traffic flow prediction method based on generating type countermeasure network |
CN112257850A (en) * | 2020-10-26 | 2021-01-22 | 河南大学 | Vehicle track prediction method based on generation countermeasure network |
CN113033619A (en) * | 2021-03-04 | 2021-06-25 | 浙江工业大学 | DVGAE-GAN-based traffic network data restoration method |
CN113487855A (en) * | 2021-05-25 | 2021-10-08 | 浙江工业大学 | Traffic flow prediction method based on EMD-GAN neural network structure |
CN113469253A (en) * | 2021-07-02 | 2021-10-01 | 河海大学 | Electricity stealing detection method based on triple twin network |
Non-Patent Citations (1)
Title |
---|
ZHAN XU ET AL: ""Satellite Image Prediction Relying on GAN and LSTM Neural Networks"", 《IEEE》 * |
Also Published As
Publication number | Publication date |
---|---|
CN114549930B (en) | 2023-01-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111540201B (en) | Vehicle queuing length real-time estimation method and system based on roadside laser radar | |
CN113401143B (en) | Individualized self-adaptive trajectory prediction method based on driving style and intention | |
US5884212A (en) | Process for monitoring traffic for automatic vehicle incident detection | |
CN104200657A (en) | Traffic flow parameter acquisition method based on video and sensor | |
CN103236166A (en) | Method for recognizing vehicle violation behaviors with satellite positioning technology | |
CN109145982A (en) | The personal identification method and device of driver, storage medium, terminal | |
CN113511204A (en) | Vehicle lane changing behavior identification method and related equipment | |
CN115657002A (en) | Vehicle motion state estimation method based on traffic millimeter wave radar | |
CN113674525B (en) | Signalized intersection vehicle queuing length prediction method based on sparse data | |
CN116704750B (en) | Traffic state identification method based on clustering algorithm, electronic equipment and medium | |
CN114549930B (en) | Rapid road short-time vehicle head interval prediction method based on trajectory data | |
Luo et al. | A statistical method for parking spaces occupancy detection via automotive radars | |
CN115761551A (en) | Automatic driving simulation scene generation method, device, equipment and medium | |
CN115223144A (en) | Unmanned mine car sensor data screening method and device based on cloud data | |
CN115629385A (en) | Vehicle queuing length real-time detection method based on correlation of millimeter wave radar and camera | |
CN111860613B (en) | Multi-target tracking and state predicting method based on multi-source heterogeneous signals | |
CN113674329A (en) | Vehicle driving behavior detection method and system | |
KR101346220B1 (en) | Apparatus and method for providing traffic information | |
CN112489423A (en) | Vision-based urban road traffic police command method | |
Wen et al. | Analysis of vehicle driving styles at freeway merging areas using trajectory data | |
Sun et al. | A novel method of symbolic representation in diving data mining: A case study of highways in China | |
CN115691164B (en) | Intelligent traffic management method and system based on big data | |
CN116913096B (en) | Traffic situation investigation equipment and method based on Beidou short message communication technology | |
CN117610971B (en) | Highway electromechanical system health index evaluation method | |
Yang et al. | Dynamic safety estimation of airport pick-up area based on video trajectory data |
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 |