CN114446046A - LSTM model-based weak traffic participant track prediction method - Google Patents
LSTM model-based weak traffic participant track prediction method Download PDFInfo
- Publication number
- CN114446046A CN114446046A CN202111566618.0A CN202111566618A CN114446046A CN 114446046 A CN114446046 A CN 114446046A CN 202111566618 A CN202111566618 A CN 202111566618A CN 114446046 A CN114446046 A CN 114446046A
- Authority
- CN
- China
- Prior art keywords
- street
- crossing
- lstm model
- vulnerable
- vulnerable traffic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000002452 interceptive effect Effects 0.000 claims abstract description 10
- 241000283070 Equus zebra Species 0.000 claims abstract description 4
- 238000011835 investigation Methods 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 12
- 230000004913 activation Effects 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 4
- 230000004927 fusion Effects 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims 1
- 238000001514 detection method Methods 0.000 description 3
- 206010039203 Road traffic accident Diseases 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- 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
-
- 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
- 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
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
-
- 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
- G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed
Abstract
The invention relates to a method for predicting a vulnerable traffic participant track based on an LSTM model, which is characterized by comprising the following steps of: step 1: selecting zebra crossing areas of the vulnerable traffic participants and vehicles under the condition of pedestrian and vehicle mixing for early investigation; step 2: acquiring the motion state information of the street-crossing vulnerable traffic participants, the individual characteristic information of the street-crossing vulnerable traffic participants and the interactive scene information; and 3, step 3: establishing an LSTM model and training the LSTM model; and 4, step 4: the method has the advantages that the track prediction is carried out on the street-crossing vulnerable traffic participants through the trained LSTM model, the predicted track of the street-crossing vulnerable traffic participants within the first preset time length in the future is obtained, and compared with the prior art, the method has the advantages that the street-crossing safety of the vulnerable traffic participants is improved, the road traffic capacity is improved, and the like.
Description
Technical Field
The invention relates to the field of automatic driving decision-making algorithms, in particular to a method for predicting a vulnerable traffic participant track based on an LSTM model.
Background
In recent years, as the number of global cars is increasing, Traffic accidents between cars and Vulnerable Traffic Participants (VTPs) frequently occur. The key problems of ensuring the traveling safety of the vulnerable traffic participants, reducing the rate of automobile traffic accidents and the like need to be solved urgently.
The existing protection system for pedestrians, non-motor vehicles and motor vehicles by automatically driving automobiles is mainly based on a target detection algorithm, and through detection and identification of traffic participants, if collision risks exist, the automatically driving automobiles can carry out early warning and collision avoidance control. Especially for the vulnerable traffic participants such as pedestrians and non-motor vehicles, the track is flexible and changeable, the uncertainty is strong, and the small change in the dynamic state may also result in completely different motion tracks. At present, a security protection system for vulnerable traffic participants only relying on target detection lacks the prejudgment on future behaviors of the vulnerable traffic participants, and information which can be provided for an automatic driving automobile for intelligent decision is very limited, so that the automatic driving automobile does not have the capability of coping with complex traffic scenes completely, the protection capability for the vulnerable traffic participants is not strong, and efficient and safe traffic cannot be realized. Trajectory prediction for vulnerable traffic participants is a difficult point for autonomous vehicles.
The existing method for predicting the track of the vulnerable traffic participants can predict the future positions of the vulnerable traffic participants according to physical parameters and link the acceleration, the speed and the positions of the vulnerable traffic participants with the external environment. The simplest models comprise a Constant Velocity (CV) model and a Constant Acceleration (CA) model, which are both assumed that a vulnerable traffic participant makes linear motion, only consider the kinematic model of the vulnerable traffic participant, and are only suitable for track prediction in a short period, but when an external environment changes (such as an obstacle appears in the front, the vehicle in front decelerates, and the like), the method cannot accurately predict the future track of the vulnerable traffic participant. The method also has the advantages that the social force model is used for predicting the track of the vulnerable traffic participants, although the prediction effect is good, only the instant movement reaction of the vulnerable traffic participants can be simulated, long-term dependence information and adaptation to complex movement scenes cannot be considered like a neural network, and the precision of the track prediction of the vulnerable traffic participants can be reduced.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a method for predicting the track of a vulnerable traffic participant based on an LSTM model.
The purpose of the invention can be realized by the following technical scheme:
a method for predicting the track of a vulnerable traffic participant based on an LSTM model comprises the following steps:
step 1: selecting zebra crossing areas of the vulnerable traffic participants and vehicles under the condition of pedestrian and vehicle mixing for early investigation;
step 2: acquiring the motion state information of the street-crossing vulnerable traffic participants, the individual characteristic information of the street-crossing vulnerable traffic participants and the interactive scene information;
and 3, step 3: establishing an LSTM model and training the LSTM model;
and 4, step 4: and predicting the track of the street-crossing weak traffic participants through the trained LSTM model to obtain the predicted track of the street-crossing weak traffic participants within the first preset time in the future.
In the step 2, the process of acquiring data specifically comprises:
the traffic flow video in the area is shot through a sensor carried by an automatic driving automobile, the traffic flow video is preprocessed, and the motion state information of the street-crossing weak traffic participants, the individual characteristic information of the street-crossing weak traffic participants and the interactive scene information in the range of the vehicle safety envelope are obtained based on a multi-sensor information fusion algorithm.
The individual characteristic information of the street-crossing vulnerable traffic participants comprises ages and sexes.
The motion state information of the street-crossing vulnerable traffic participant comprises the current position and the current speed of the vulnerable traffic participant.
The interactive scene information comprises the current position, the speed and the type of the vehicle.
In step 3, the structure of the LSTM model specifically is as follows:
hiding the layer: the dimension of the hidden layer is set to 256 and consists of LSTM units;
an input layer: the input unit comprises a plurality of input units, wherein each input unit respectively adopts a tanh activation function, and each input unit respectively corresponds to an input characteristic;
an output layer: the system comprises a plurality of output units, wherein each input unit respectively adopts a tanh activation function, and the corresponding outputs are respectively the X-direction movement position and the Y-direction movement position of the street-crossing weak traffic participant within a first preset time length.
The LSTM unit comprises 3 control gates, namely an input gate, a forgetting gate and an output gate, and is used for controlling the relation among input, output and internal states.
The input characteristics are the speed of the street weak traffic participant in the X direction, the speed of the street weak traffic participant in the Y direction, the position of the street weak traffic participant in the X direction, the position of the street weak traffic participant in the Y direction, the age of the street weak traffic participant, the sex of the street weak traffic participant, the vehicle speed in the X direction, the vehicle speed in the Y direction, the vehicle position in the X direction, the vehicle position in the Y direction and the vehicle type.
In step 3, the process of training the LSTM model specifically includes the following steps:
step 301: the expression of the LSTM model obtained based on the relationship between the input, output, and internal states is:
it=σ(Wi·[ht-1,Xt]+bi)
ft=σ(Wf·[ht-1,Xt]+bf)
Ot=σ(Wo·[ht-1,Xt]+bo)
Ct=ft*Ct-1+it*tanh(Wc·[ht-1,Xt]+bc)
ht=Ot*tanh(Ct)
wherein itDenotes an input gate, ftIndicating forgetting gate, OtIndicating output gate, CtIndicating the state of the cell at the current time t, htIndicating a hidden state at the current time t, Ct-1Indicates the state of the cell at the previous time, ht-1Indicating the hidden state at the previous moment, XtRepresenting the input vector at the present time t, WiWeight matrix, W, representing input gatesfWeight matrix representing forgetting gate, WoWeight matrix, W, representing output gatescWeight matrix representing the state of the cell, biRepresenting the offset term of the input gate, bfBias term representing forgetting gate, boRepresenting the offset term of the output gate, bcA bias term representing a cell state, σ represents a sigmoid function; tanh represents a tanh function;
step 302: and importing the motion state information, the individual characteristic information and the interaction information of the street-crossing weak traffic participants, which are acquired by the vehicle-mounted sensor of the automatic driving automobile, into the LSTM model, and training the structure weight and the bias parameters of the LSTM model.
In the step 4, the process of predicting the track of the street-crossing vulnerable traffic participants based on the LSTM model specifically comprises the following steps:
step 401: acquiring data information of a current vehicle and a street-crossing weak traffic participant through a sensor carried by an automatic driving automobile, namely acquiring 11 input characteristics;
step 402: inputting the input characteristics into the trained LSTM model, outputting prediction data, and performing inverse normalization on the relevant data to obtain the movement positions of the street-crossing vulnerable traffic participants within the first preset time length in the future, namely the predicted tracks of the street-crossing vulnerable traffic participants within the first preset time length in the future.
Compared with the prior art, the invention has the following advantages:
1. the invention fully considers the heterogeneity of the vulnerable traffic participants, selects the LSTM neural network model capable of reflecting the time sequence to predict the track, applies the predicted track to the intelligent decision field of the automatic driving automobile, and can improve the street crossing safety of the vulnerable traffic participants and the road traffic capacity.
2. The method can take factors such as the surrounding traffic environment, the individual difference of the vulnerable traffic participants and the like as the factors to be merged into the LSTM model, accurately predict the track of the street-crossing vulnerable traffic participants, effectively reduce the error of the actual track of the vulnerable traffic participants, and further meet the prediction requirement required by automatic driving of the automobile.
Drawings
FIG. 1 is a flow chart of the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments.
Examples
As shown in fig. 1, the present invention provides a method for predicting a trajectory of a vulnerable traffic participant based on an LSTM model, which comprises the following steps:
step 1: acquiring the motion state information, the individual characteristic information and the interactive scene information of the street-crossing vulnerable traffic participants:
selecting zebra crossing areas of the vulnerable traffic participants and vehicles under the condition of pedestrian-vehicle mixture for early investigation, and acquiring the motion state information of the street-crossing vulnerable traffic participants, the individual characteristic information of the street-crossing vulnerable traffic participants and the interactive scene information within the range of a vehicle safety envelope by utilizing sensors such as laser, a camera and a millimeter wave radar which are carried by an automatic driving vehicle through a multi-sensor information fusion algorithm;
the individual characteristic information of the street-crossing vulnerable traffic participants comprises ages and sexes;
the motion state information of the street-crossing vulnerable traffic participant comprises the current position and the current speed of the vulnerable traffic participant;
the interactive scene information comprises the current position, the speed and the type of the vehicle.
The structure of the LSTM model is specifically as follows:
hidden layer: the dimension of the hidden layer is set to be 256, the hidden layer comprises a gate control unit, the gate control unit adopts a sigmoid activation function, and the LSTM unit comprises 3 control gates which are an input gate, a forgetting gate and an output gate respectively and are used for controlling the relation among input, output and internal states;
an input layer: the system comprises a plurality of input units, wherein each input unit respectively adopts a tanh activation function, and each input unit respectively corresponds to input characteristics, and the input characteristics are the speed of a street-crossing weak traffic participant in the X direction, the speed of the street-crossing weak traffic participant in the Y direction, the position of the street-crossing weak traffic participant in the X direction, the position of the street-crossing weak traffic participant in the Y direction, the age of the street-crossing weak traffic participant, the sex of the street-crossing weak traffic participant, the vehicle speed in the X direction, the vehicle speed in the Y direction, the vehicle position in the X direction, the vehicle position in the Y direction and the vehicle type;
and (3) an output layer: the system comprises a plurality of output units, wherein each input unit respectively adopts a tanh activation function, and the corresponding outputs are respectively the X-direction movement position and the Y-direction movement position of the street-crossing weak traffic participant within a first preset time length.
In step 3, the process of training the LSTM model specifically includes the following steps:
step 301: the expression of the LSTM model obtained based on the relationship between the input, output, and internal states is:
it=σ(Wi·[ht-1,Xt]+bi)
ft=σ(Wf·[ht-1,Xt]+bf)
Ot=σ(Wo·[ht-1,Xt]+bo)
Ct=ft*Ct-1+it*tanh(Wc·[ht-1,Xt]+bc)
ht=Ot*tanh(Ct)
wherein itDenotes an input gate, ftIndicating forgetting gate, OtIndicating output gate, CtIndicating the state of the cell at the current time t, htIndicating a hidden state at the current time t, Ct-1Indicates the state of the cell at the previous time, ht-1Indicating the hidden state at the previous moment, XtRepresenting the input vector at the present time t, WiWeight matrix, W, representing input gatesfWeight matrix representing forgetting gate, WoWeight matrix, W, representing output gatescWeight matrix representing the state of the cell, biRepresenting the offset term of the input gate, bfBias term representing forgetting gate, boRepresenting the offset term of the output gate, bcA bias term representing a cell state, σ represents a sigmoid function; tanh represents a tanh function.
Step 302: and importing the motion state information, the individual characteristic information and the interaction information of the street-crossing weak traffic participants, which are acquired by the vehicle-mounted sensor of the automatic driving automobile, into the LSTM model, and training the structure weight and the bias parameters of the LSTM model.
In step 4, the process of predicting the track of the street-crossing vulnerable traffic participants based on the LSTM model specifically comprises the following steps:
step 401: acquiring data information of a current vehicle and a street-crossing weak traffic participant through a vehicle-mounted sensor, namely acquiring 11 input characteristics;
step 402: and carrying out data normalization operation on the acquired input features, inputting the input features into the trained LSTM model, outputting prediction data, and carrying out reverse normalization on the related data to obtain the movement positions of the street-crossing weak traffic participants within the first preset time length in the future, namely the predicted tracks of the street-crossing weak traffic participants within the first preset time length in the future.
While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and those skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A method for predicting the track of a vulnerable traffic participant based on an LSTM model is characterized by comprising the following steps:
step 1: selecting zebra crossing areas of the vulnerable traffic participants and vehicles under the condition of pedestrian and vehicle mixing for early investigation;
step 2: acquiring the motion state information of the street-crossing vulnerable traffic participants, the individual characteristic information of the street-crossing vulnerable traffic participants and the interactive scene information;
and 3, step 3: establishing an LSTM model and training the LSTM model;
and 4, step 4: and predicting the track of the street-crossing weak traffic participants through the trained LSTM model to obtain the predicted track of the street-crossing weak traffic participants within the first preset time in the future.
2. The LSTM model-based vulnerable traffic participant trajectory prediction method according to claim 1, wherein in the step 2, the process of collecting data specifically comprises:
the method comprises the steps of shooting a traffic flow video in the area through a sensor carried by an automatic driving automobile, preprocessing the traffic flow video, and acquiring the motion state information of the street-crossing vulnerable traffic participants, the individual characteristic information of the street-crossing vulnerable traffic participants and the interactive scene information in the range of a vehicle safety envelope line based on a multi-sensor information fusion algorithm.
3. The LSTM model-based vulnerable traffic participant trajectory prediction method of claim 1, wherein the street-crossing vulnerable traffic participant individual characteristic information includes age and gender.
4. The LSTM model-based vulnerable traffic participant trajectory prediction method of claim 1, wherein the street-crossing vulnerable traffic participant movement state information comprises the current location and speed of the vulnerable traffic participant.
5. The LSTM model-based method for predicting the trajectory of a vulnerable traffic participant of claim 1, wherein the interactive scene information includes the current position of the vehicle, the speed of the vehicle, and the type of the vehicle.
6. The LSTM model-based vulnerable traffic participant trajectory prediction method according to claim 2, wherein in the step 3, the structure of the LSTM model specifically is:
hiding the layer: the dimension of the hidden layer is set to 256 and consists of LSTM units;
an input layer: the input unit comprises a plurality of input units, wherein each input unit respectively adopts a tanh activation function, and each input unit respectively corresponds to an input characteristic;
an output layer: the system comprises a plurality of output units, wherein each input unit respectively adopts a tanh activation function, and the corresponding outputs are respectively the X-direction movement position and the Y-direction movement position of the street-crossing weak traffic participant within a first preset time length.
7. The LSTM model-based vulnerable traffic participant trajectory prediction method of claim 6, wherein the LSTM unit comprises 3 control gates, i.e. an input gate, a forgetting gate and an output gate, respectively, for controlling the relationship among input, output and internal states.
8. The LSTM model-based trajectory prediction method for vulnerable traffic participants of claim 7, wherein the input features are the speed of the X-direction of the street-crossing vulnerable traffic participant, the speed of the Y-direction of the street-crossing vulnerable traffic participant, the position of the X-direction of the street-crossing vulnerable traffic participant, the position of the Y-direction of the street-crossing vulnerable traffic participant, the age of the street-crossing vulnerable traffic participant, the sex of the street-crossing vulnerable traffic participant, the vehicle speed of the X-direction, the vehicle speed of the Y-direction, the vehicle position of the X-direction, the vehicle position of the Y-direction, and the vehicle type, respectively.
9. The LSTM model-based vulnerable traffic participant trajectory prediction method according to claim 8, wherein the step 3 of training the LSTM model specifically comprises the following steps:
step 301: the expression of the LSTM model obtained based on the relationship between the input, output, and internal states is:
it=σ(Wi·[ht-1,Xt]+bi)
ft=σ(Wf·[ht-1,Xt]+bf)
Ot=σ(Wo·[ht-1,Xt]+bo)
Ct=ft*Ct-1+it*tanh(Wc·[ht-1,Xt]+bc)
ht=Ot*tanh(Ct)
wherein itDenotes an input gate, ftIndicating forgetting gate, OtIndicating output gate, CtIndicating the state of the cell at the current time t, htIndicating a hidden state at the current time t, Ct-1Indicates the state of the cell at the previous time, ht-1Indicating the hidden state at the previous moment, XtRepresenting the input vector at the present time t, WiWeight matrix, W, representing input gatesfWeight matrix representing forgetting gate, WoWeight matrix, W, representing output gatescWeight matrix representing the state of the cell, biRepresenting the offset term of the input gate, bfBias term representing forgetting gate, boRepresenting the offset term of the output gate, bcIndicating cell stateσ represents a sigmoid function; tanh represents a tanh function;
step 302: and importing the motion state information, the individual characteristic information and the interaction information of the street-crossing weak traffic participants, which are acquired by the vehicle-mounted sensor of the automatic driving automobile, into the LSTM model, and training the structure weight and the bias parameters of the LSTM model.
10. The LSTM model-based vulnerable traffic participant trajectory prediction method according to claim 9, wherein in the step 4, the process of predicting the trajectories of the street-crossing vulnerable traffic participants based on the LSTM model specifically comprises the following steps:
step 401: acquiring data information of a current vehicle and a street-crossing weak traffic participant through a sensor carried by an automatic driving automobile, namely acquiring 11 input characteristics;
step 402: inputting the input characteristics into the trained LSTM model, outputting prediction data, and performing inverse normalization on the related data to obtain the motion positions of the street-crossing vulnerable traffic participants in the first preset time length in the future, namely the predicted tracks of the street-crossing vulnerable traffic participants in the first preset time length in the future.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111566618.0A CN114446046A (en) | 2021-12-20 | 2021-12-20 | LSTM model-based weak traffic participant track prediction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111566618.0A CN114446046A (en) | 2021-12-20 | 2021-12-20 | LSTM model-based weak traffic participant track prediction method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114446046A true CN114446046A (en) | 2022-05-06 |
Family
ID=81363818
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111566618.0A Pending CN114446046A (en) | 2021-12-20 | 2021-12-20 | LSTM model-based weak traffic participant track prediction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114446046A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114912719A (en) * | 2022-07-15 | 2022-08-16 | 北京航空航天大学 | Heterogeneous traffic individual trajectory collaborative prediction method based on graph neural network |
CN115662166A (en) * | 2022-09-19 | 2023-01-31 | 长安大学 | Automatic driving data processing method and automatic driving traffic system |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109910909A (en) * | 2019-02-25 | 2019-06-21 | 清华大学 | A kind of interactive prediction technique of vehicle track net connection of more vehicle motion states |
CN111161322A (en) * | 2019-12-31 | 2020-05-15 | 大连理工大学 | LSTM neural network pedestrian trajectory prediction method based on human-vehicle interaction |
CN111459168A (en) * | 2020-04-23 | 2020-07-28 | 上海交通大学 | Fused automatic-driving automobile pedestrian crossing track prediction method and system |
CN111553232A (en) * | 2020-04-22 | 2020-08-18 | 陕西师范大学 | Gate loop unit network pedestrian trajectory prediction method based on scene state iteration |
CN111695737A (en) * | 2020-06-15 | 2020-09-22 | 中国电子科技集团公司第二十八研究所 | Group target advancing trend prediction method based on LSTM neural network |
CN112418421A (en) * | 2020-11-06 | 2021-02-26 | 常州大学 | Roadside end pedestrian trajectory prediction algorithm based on graph attention self-coding model |
-
2021
- 2021-12-20 CN CN202111566618.0A patent/CN114446046A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109910909A (en) * | 2019-02-25 | 2019-06-21 | 清华大学 | A kind of interactive prediction technique of vehicle track net connection of more vehicle motion states |
CN111161322A (en) * | 2019-12-31 | 2020-05-15 | 大连理工大学 | LSTM neural network pedestrian trajectory prediction method based on human-vehicle interaction |
CN111553232A (en) * | 2020-04-22 | 2020-08-18 | 陕西师范大学 | Gate loop unit network pedestrian trajectory prediction method based on scene state iteration |
CN111459168A (en) * | 2020-04-23 | 2020-07-28 | 上海交通大学 | Fused automatic-driving automobile pedestrian crossing track prediction method and system |
CN111695737A (en) * | 2020-06-15 | 2020-09-22 | 中国电子科技集团公司第二十八研究所 | Group target advancing trend prediction method based on LSTM neural network |
CN112418421A (en) * | 2020-11-06 | 2021-02-26 | 常州大学 | Roadside end pedestrian trajectory prediction algorithm based on graph attention self-coding model |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114912719A (en) * | 2022-07-15 | 2022-08-16 | 北京航空航天大学 | Heterogeneous traffic individual trajectory collaborative prediction method based on graph neural network |
CN115662166A (en) * | 2022-09-19 | 2023-01-31 | 长安大学 | Automatic driving data processing method and automatic driving traffic system |
CN115662166B (en) * | 2022-09-19 | 2024-04-09 | 长安大学 | Automatic driving data processing method and automatic driving traffic system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111459168B (en) | Fused automatic-driving automobile pedestrian crossing track prediction method and system | |
EP3678911B1 (en) | Pedestrian behavior predictions for autonomous vehicles | |
US10235881B2 (en) | Autonomous operation capability configuration for a vehicle | |
CN108919795B (en) | Automatic driving automobile lane change decision method and device | |
CN112435503B (en) | Intelligent automobile active collision avoidance method for identifying intention of high-risk pedestrians | |
CN110843789B (en) | Vehicle lane change intention prediction method based on time sequence convolution network | |
CN111079590A (en) | Peripheral vehicle behavior pre-judging method of unmanned vehicle | |
CN111930110A (en) | Intent track prediction method for generating confrontation network by combining society | |
CN110077398B (en) | Risk handling method for intelligent driving | |
CN114446046A (en) | LSTM model-based weak traffic participant track prediction method | |
CN112614373B (en) | BiLSTM-based weekly vehicle lane change intention prediction method | |
CN111145552A (en) | Planning method for vehicle dynamic lane changing track based on 5G network | |
Yu et al. | Machine Learning-Based Vehicle Intention Trajectory Recognition and Prediction for Autonomous Driving | |
CN114462667A (en) | SFM-LSTM neural network model-based street pedestrian track prediction method | |
CN113424209B (en) | Trajectory prediction using deep learning multi-predictor fusion and Bayesian optimization | |
Jeong | Predictive lane change decision making using bidirectional long shot-term memory for autonomous driving on highways | |
Mänttäri et al. | Learning to predict lane changes in highway scenarios using dynamic filters on a generic traffic representation | |
CN116501820A (en) | Vehicle track prediction method, device, equipment and storage medium | |
CN114889608A (en) | Attention mechanism-based vehicle lane change prediction method | |
Guo et al. | Forwarding collision assessment with the localization information using the machine learning method | |
Miani et al. | Young drivers’ pedestrian anti-collision braking operation data modelling for ADAS development | |
Raj et al. | Lane Prediction by Autonomous Vehicle in Highway Traffic using Artificial Neural Networks | |
Zyner | Naturalistic driver intention and path prediction using machine learning | |
Leelavathy et al. | Effective traffic model for intelligent traffic monitoring enabled deep RNN algorithm for autonomous vehicles surveillance systems | |
Lv et al. | A lane-changing decision-making model of bus entering considering bus priority based on GRU neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220506 |
|
RJ01 | Rejection of invention patent application after publication |