CN112085077B - Method and device for determining lane change of vehicle, storage medium and electronic equipment - Google Patents
Method and device for determining lane change of vehicle, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN112085077B CN112085077B CN202010889856.4A CN202010889856A CN112085077B CN 112085077 B CN112085077 B CN 112085077B CN 202010889856 A CN202010889856 A CN 202010889856A CN 112085077 B CN112085077 B CN 112085077B
- Authority
- CN
- China
- Prior art keywords
- relative motion
- track
- vehicle
- lane change
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000008859 change Effects 0.000 title claims abstract description 387
- 238000000034 method Methods 0.000 title claims abstract description 67
- 230000015654 memory Effects 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 19
- 230000001133 acceleration Effects 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 230000007704 transition Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 17
- 238000013528 artificial neural network Methods 0.000 description 14
- 238000012549 training Methods 0.000 description 11
- 238000004891 communication Methods 0.000 description 8
- 239000004973 liquid crystal related substance Substances 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 238000010606 normalization Methods 0.000 description 6
- 238000006073 displacement reaction Methods 0.000 description 4
- 230000005236 sound signal Effects 0.000 description 4
- 238000012417 linear regression Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- KLDZYURQCUYZBL-UHFFFAOYSA-N 2-[3-[(2-hydroxyphenyl)methylideneamino]propyliminomethyl]phenol Chemical compound OC1=CC=CC=C1C=NCCCN=CC1=CC=CC=C1O KLDZYURQCUYZBL-UHFFFAOYSA-N 0.000 description 1
- 229930091051 Arenine Natural products 0.000 description 1
- 241000288105 Grus Species 0.000 description 1
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 239000011668 ascorbic acid Substances 0.000 description 1
- 230000003915 cell function Effects 0.000 description 1
- 201000001098 delayed sleep phase syndrome Diseases 0.000 description 1
- 208000033921 delayed sleep phase type circadian rhythm sleep disease Diseases 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000000153 supplemental effect Effects 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- 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
-
- 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
Abstract
The disclosure relates to a method, a device, a storage medium and an electronic device for determining lane change of a vehicle, wherein the method is applied to a first vehicle and comprises the following steps: the track data set and the relative motion data set are acquired, the track data set is respectively input into a pre-trained track recognition model and a pre-trained track clustering model to determine track lane change probability according to a track recognition result output by the track recognition model and a track clustering result output by the track clustering model, the relative motion data set is respectively input into the pre-trained relative motion recognition model and the pre-trained relative motion clustering model to determine relative motion lane change probability according to the relative motion recognition result output by the relative motion recognition model and the relative motion clustering result output by the relative motion clustering model, the lane change probability of the second vehicle is determined according to the track lane change probability and the relative motion lane change probability, and whether the second vehicle is about to change the lane is determined according to the lane change probability and a preset lane change threshold.
Description
Technical Field
The disclosure relates to the technical field of vehicle control, in particular to a method and a device for determining lane change of a vehicle, a storage medium and electronic equipment.
Background
With the increasing of the holding quantity of automobiles, the running safety of the automobiles is increasingly paid attention to. At present, traffic accidents are more frequent because drivers cannot find that vehicles in front change lanes in time, and life and property safety of the drivers and pedestrians are seriously affected. Therefore, to ensure safe running of the vehicle, it is necessary to accurately judge whether the preceding vehicle is lane-changing.
In general, the way of determining a lane change of a vehicle may be divided into two ways, one is to determine whether the lane change of the vehicle ahead by the forward direction of the vehicle ahead and the curvature of the lane line direction, and this way is easily affected by misoperation of the driver, instantaneous change of the vehicle motion state, and the like, and the accuracy of the determination is unstable. The other is to judge whether the front vehicle is changed according to the operation data (such as steering wheel angle, steering lamp state, clutch brake state, driver state, etc.) of the front vehicle based on the Markov model, for the vehicle, it is difficult to accurately acquire the operation data of the front vehicle, so that the accuracy of the judgment mode is not high, and the judgment result is mainly dependent on the data of the current time point because the Markov model depends on the homogeneous Markov assumption and the observation independence assumption, and the accuracy of the judgment result is further reduced.
Disclosure of Invention
The disclosure aims to provide a method and a device for determining lane change of a vehicle, a storage medium and electronic equipment, which are used for solving the problem of low lane change accuracy of the vehicle in the prior art.
To achieve the above object, according to a first aspect of embodiments of the present disclosure, there is provided a method for determining lane change of a vehicle, applied to a first vehicle, the method including:
acquiring a track data set and a relative motion data set, wherein the track data set comprises track data of a second vehicle acquired at a plurality of acquisition moments within a preset duration, the relative motion data set comprises relative motion data between the second vehicle and the first vehicle acquired at the plurality of acquisition moments within the preset duration, the plurality of acquisition moments comprise current acquisition moments, and the second vehicle is a vehicle which runs on an adjacent lane of the lane where the first vehicle is located and is positioned in front of the first vehicle;
respectively inputting the track data set into a pre-trained track recognition model and a pre-trained track clustering model to determine track change probability according to a track recognition result output by the track recognition model and a track clustering result output by the track clustering model;
Respectively inputting the relative motion data set into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model to determine the relative motion lane change probability according to the relative motion recognition result output by the relative motion recognition model and the relative motion clustering result output by the relative motion clustering model;
determining the lane change probability of the second vehicle according to the track lane change probability and the relative motion lane change probability;
and determining whether the second vehicle is about to change the lane according to the lane change probability and a preset lane change threshold value.
Optionally, the acquiring the trajectory dataset and the relative motion dataset includes:
acquiring the track data and the relative motion data acquired at each acquisition time within the preset time, wherein the track data comprises: the lateral and longitudinal positions of the second vehicle, the relative motion data comprising: a relative longitudinal speed and a relative longitudinal distance of the first vehicle and the second vehicle;
determining supplementary track data corresponding to each acquisition time according to the track data acquired at the acquisition time, wherein the supplementary track data comprises: a lateral speed and a lateral acceleration of the second vehicle;
Taking the track data acquired at each acquisition moment and the corresponding supplementary track data as the track data set;
and processing the relative motion data acquired at each acquisition time according to a preset rule to obtain the relative motion data set.
Optionally, the inputting the track data set into a pre-trained track recognition model and a pre-trained track clustering model respectively, so as to determine track lane change probability according to a track recognition result output by the track recognition model and a track clustering result output by the track clustering model, including:
inputting the track data set into the track recognition model to obtain the track recognition result, and inputting the track data set into the track clustering model to obtain a track clustering result, wherein the track recognition result is used for indicating straight running or lane changing, and the track clustering result is used for indicating the category of the track data collected at each collection moment;
if the track identification result indicates straight going, determining that the track lane change probability is zero;
and if the track identification result indicates track change, determining the track change probability according to the track clustering result.
Optionally, the track recognition model is an Attention-LSTM model, and the inputting the track dataset into the track recognition model to obtain the track recognition result includes:
inputting the track data set into the Attention-LSTM model to acquire the track recognition result output by the Attention-LSTM model and the Attention value of each acquisition moment;
the determining the track lane-changing probability according to the track clustering result comprises the following steps:
determining the initial track lane change probability of the acquisition time according to the category of the track data acquired at each acquisition time and the corresponding relation between the preset category and the initial track lane change probability, wherein the category is included in the track clustering result;
and determining the track lane change probability according to the initial track lane change probability of each acquisition time and the attention value of each acquisition time.
Optionally, the inputting the relative motion data set into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model respectively, so as to determine a relative motion lane-changing probability according to a relative motion recognition result output by the relative motion recognition model and a relative motion clustering result output by the relative motion clustering model, including:
Inputting the relative motion data set into the relative motion recognition model to obtain the relative motion recognition result, and inputting the relative motion data set into the relative motion clustering model to obtain a relative motion clustering result, wherein the relative motion recognition result is used for indicating straight running or lane changing, and the relative motion clustering result is used for indicating the category to which the relative motion data acquired at each acquisition moment belong;
if the relative motion recognition result indicates straight going, determining that the relative motion lane change probability is zero;
and if the relative motion recognition result indicates lane change, determining the relative motion lane change probability according to the relative motion clustering result.
Optionally, the relative motion recognition model is an Attention-GRU model, and the inputting the relative motion data set into the relative motion recognition model to obtain the relative motion recognition result includes:
inputting the relative motion data set into the Attention-GRU model to acquire the relative motion recognition result output by the Attention-GRU model and the Attention value of each acquisition time;
the determining the relative motion lane-change probability according to the relative motion clustering result comprises the following steps:
Determining initial relative motion lane change probability at the acquisition time according to the category to which the relative motion data acquired at each acquisition time included in the relative motion clustering result belong and the corresponding relation between a preset category and the initial relative motion lane change probability;
determining the relative motion lane change probability according to the initial relative motion lane change probability of each acquisition time and the attention value of each acquisition time.
Optionally, the acquiring the trajectory dataset and the relative motion dataset includes:
acquiring the track data and the relative motion data acquired at each acquisition time within a specified duration;
dividing the appointed time length into appointed number of sliding windows, wherein the length of each sliding window is the preset time length;
the track data and the relative motion data acquired in each sliding window are used as the track data set and the relative motion data set corresponding to the sliding window;
the determining whether the second vehicle is about to change lane according to the lane change probability and a preset lane change threshold value comprises:
the lane change probability determined according to the track data set and the relative motion data set corresponding to each sliding window is weighted and summed to determine total lane change probability, and the weight corresponding to each sliding window is inversely proportional to the distance between the sliding window and the current moment;
If the total lane change probability is greater than or equal to the lane change threshold, determining that the second vehicle is about to change lanes;
and if the total lane change probability is smaller than the lane change threshold, determining that the second vehicle is going straight.
Optionally, the track recognition result is used for indicating straight going, left lane change or right lane change; after determining whether the second vehicle is about to change lane according to the lane change probability and a preset lane change threshold, the method further includes:
if the second vehicle is determined to be changed, counting a first number corresponding to a left lane change, a second number corresponding to a right lane change and a third number corresponding to a straight line in the track identification result determined according to the track data set corresponding to each sliding window, wherein the sum of the first number, the second number and the third number is equal to the appointed number;
if the first number is greater than the second number, determining that the second vehicle is about to lane-change left;
and if the first quantity is smaller than the second quantity, determining that the second vehicle is about to change lanes right.
Optionally, after determining whether the second vehicle is about to change lanes according to the lane change probability and a preset lane change threshold, the method further includes:
If the second vehicle is determined to be lane-changing, determining lane-changing time of the second vehicle according to the transverse position, the transverse speed and the course angle of the second vehicle, which are included in the track data acquired at the current acquisition time;
determining termination track data of the second vehicle at the end of lane change according to the lane change time and the longitudinal speed included in the track data acquired at the current acquisition time;
and fitting according to the termination track data and the track data acquired at the current acquisition moment and a Bessel function to obtain the track changing track of the second vehicle.
Optionally, the trajectory recognition model is trained by:
acquiring a first sample input set and a first sample output set, wherein each first sample input in the first sample input set comprises a plurality of sample track data, the first sample output set comprises first sample output corresponding to each first sample input, each first sample output comprises a sample track identification result marked by the corresponding plurality of sample track data, and the sample track identification result is used for indicating straight line or lane change;
Taking the first sample input set as the input of the track recognition model, and taking the first sample output set as the output of the track recognition model so as to train the track recognition model;
the relative motion recognition model is trained by the following steps:
acquiring a second sample input set and a second sample output set, wherein each second sample input in the second sample input set comprises a plurality of sample relative motion data, each second sample output set comprises a second sample output corresponding to each second sample input, each second sample output comprises a sample relative motion recognition result marked by the corresponding plurality of sample relative motion data, and the sample relative motion recognition result is used for indicating straight going or lane changing;
and taking the second sample input set as the input of the relative motion recognition model, and taking the second sample output set as the output of the relative motion recognition model so as to train the relative motion recognition model.
Optionally, the track recognition model is an LSTM model, and the relative motion recognition model is a GRU model;
the taking the first sample input set as the input of the track recognition model and the first sample output set as the output of the track recognition model to train the track recognition model comprises:
Inputting the first sample input set into an initial LSTM model according to a first input weight, and taking the first sample output set as the output of the initial LSTM model to train the initial LSTM model;
updating the first input weight according to the trained initial LSTM model;
repeatedly executing the step of inputting the first sample input set, inputting an initial LSTM model according to a first input weight, taking the first sample output set as the output of the initial LSTM model to train the initial LSTM model, and updating the first input weight according to the trained initial LSTM model, wherein the initial LSTM model and the first input weight obtained by executing a preset number of iterations are taken as the input weights corresponding to the track recognition model and the track recognition model;
the taking the second sample input set as the input of the relative motion recognition model and the second sample output set as the output of the relative motion recognition model to train the relative motion recognition model includes:
inputting the second sample input set into the initial GRU model according to a second input weight, and taking the second sample output set as the output of the initial GRU model to train the initial GRU model;
Updating the second input weight according to the trained initial GRU model;
and repeatedly executing the step of inputting the second sample input set, inputting the initial GRU model according to a second input weight, taking the second sample output set as the output of the initial GRU model to train the initial GRU model until the step of updating the second input weight according to the trained initial GRU model, and taking the initial GRU model and the second input weight obtained by executing the iteration number as the input weights corresponding to the relative motion recognition model and the relative motion recognition model.
According to a second aspect of embodiments of the present disclosure, there is provided a vehicle lane change determining apparatus applied to a first vehicle, the apparatus including:
the track data set comprises track data of a second vehicle acquired at a plurality of acquisition moments within a preset duration, the relative motion data set comprises relative motion data between the second vehicle and the first vehicle acquired at the plurality of acquisition moments within the preset duration, the plurality of acquisition moments comprise current acquisition moments, and the second vehicle is a vehicle which runs on an adjacent lane of the lane where the first vehicle is located and is positioned in front of the first vehicle;
The first processing module is used for inputting the track data set into a pre-trained track recognition model and a pre-trained track clustering model respectively so as to determine track lane change probability according to a track recognition result output by the track recognition model and a track clustering result output by the track clustering model;
the second processing module is used for inputting the relative motion data set into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model respectively so as to determine the relative motion lane change probability according to the relative motion recognition result output by the relative motion recognition model and the relative motion clustering result output by the relative motion clustering model;
the first determining module is used for determining the lane change probability of the second vehicle according to the track lane change probability and the relative motion lane change probability;
and the second determining module is used for determining whether the second vehicle is about to change the lane according to the lane change probability and a preset lane change threshold value.
According to a third aspect of the disclosed embodiments, there is provided a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the steps of the method of the first aspect of the disclosed embodiments.
According to a fourth aspect of embodiments of the present disclosure, there is provided an electronic device, comprising:
a memory having a computer program stored thereon;
a processor for executing the computer program in the memory to implement the steps of the method of the first aspect of the embodiments of the present disclosure.
Through the technical scheme, the first vehicle firstly acquires the track data set and the relative motion data set which are acquired at a plurality of acquisition moments within the preset duration. And inputting the track data set into a track recognition model and a track clustering model to determine track lane change probability according to the track recognition result and the track clustering result, and inputting the relative motion data set into a relative motion recognition model and a relative motion clustering model to determine the relative motion lane change probability according to the relative motion recognition result and the relative motion clustering result. And determining the lane change probability of a second vehicle according to the track lane change probability and the relative motion lane change probability, and finally determining whether the second vehicle is about to change lanes according to the lane change probability and a preset lane change threshold value, wherein the second vehicle is a vehicle in front of the first vehicle and positioned in an adjacent lane. According to the method and the device, the track data set and the relative motion data set are acquired by the first vehicle, the operation data in the second vehicle do not need to be acquired, the accuracy of the data is improved, whether the second vehicle is about to change the track is determined by the track data set and the relative motion data set, the continuity of the second vehicle in space and time in the running process is comprehensively considered, and the accuracy of vehicle track change judgment is improved.
Additional features and advantages of the present disclosure will be set forth in the detailed description which follows.
Drawings
The accompanying drawings are included to provide a further understanding of the disclosure, and are incorporated in and constitute a part of this specification, illustrate the disclosure and together with the description serve to explain, but do not limit the disclosure. In the drawings:
FIG. 1 is a flow chart illustrating a method of determining a lane change of a vehicle according to an exemplary embodiment;
FIG. 2 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment;
FIG. 3 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment;
FIG. 4 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment;
FIG. 5 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment;
FIG. 6 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment;
FIG. 7 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment;
FIG. 8 is a scatter plot of a linear regression shown according to an exemplary embodiment;
FIG. 9 is a schematic diagram of a lane change trajectory shown in accordance with an exemplary embodiment;
FIG. 10 is a block diagram illustrating a vehicle lane change determination apparatus according to an exemplary embodiment;
FIG. 11 is a block diagram illustrating another vehicle lane change determination apparatus according to an exemplary embodiment;
FIG. 12 is a block diagram illustrating another vehicle lane change determination apparatus according to an example embodiment;
FIG. 13 is a block diagram illustrating another vehicle lane change determination apparatus according to an example embodiment;
FIG. 14 is a block diagram illustrating another vehicle lane change determination apparatus according to an example embodiment;
FIG. 15 is a block diagram illustrating another vehicle lane change determination apparatus according to an example embodiment;
FIG. 16 is a block diagram illustrating another vehicle lane change determination apparatus according to an example embodiment;
fig. 17 is a block diagram of an electronic device, according to an example embodiment.
Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as detailed in the accompanying claims.
Before describing the method, the device, the storage medium and the electronic equipment for determining the lane change of the vehicle, the application scenario related to each embodiment of the disclosure is first described, where the application scenario includes a first vehicle and a second vehicle running on a road, the first vehicle runs on a first lane, the second vehicle runs on a second lane, the second lane is an adjacent lane of the first lane, and the second vehicle is located in front of the first vehicle. Further, the road may further include a third vehicle that is located in front of the first vehicle and travels on the first lane, and may further include a fourth vehicle that is located in front of the second vehicle and travels on the second lane. The execution subject of the embodiments provided by the present disclosure is a first vehicle, on which a data acquisition device may be provided to acquire trajectory data and relative motion data mentioned below. The data acquisition device may include, for example: GNSS (English: global Navigation Satellite System, chinese: global navigation satellite System) devices, inertial navigation devices, cameras, infrared devices, radars, etc., which are not particularly limited in this disclosure.
FIG. 1 is a flowchart illustrating a method of determining lane change of a vehicle, as shown in FIG. 1, applied to a first vehicle, according to an exemplary embodiment, comprising the steps of:
step 101, acquiring a track data set and a relative motion data set, wherein the track data set comprises track data of a second vehicle acquired at a plurality of acquisition moments within a preset duration, the relative motion data set comprises relative motion data between the second vehicle and a first vehicle acquired at the plurality of acquisition moments within the preset duration, the plurality of acquisition moments comprise current acquisition moments, and the second vehicle is a vehicle which runs on an adjacent lane of the lane where the first vehicle is located and is positioned in front of the first vehicle.
For example, during the running of the first vehicle, the track data P of the second vehicle may be acquired at each acquisition time according to a preset acquisition period T i And relative movement data E between the second vehicle and the first vehicle i Wherein i represents the sequence number of the acquisition time, namely, the time difference between the ith acquisition time and the (i+1) th acquisition time is T. P (P) i It is understood that data reflecting the second vehicle travel trajectory may include a plurality of trajectory features, such as: lateral position x of second vehicle i Longitudinal position y i Velocity v i Acceleration a i 。E i It is understood that data reflecting the relative movement between the first vehicle and the second vehicle may include a plurality of characteristics of the relative movement, such as: relative longitudinal speed Deltav of first and second vehicles i And a relative longitudinal distance deltay i . The track data and the relative motion data can be directly acquired according to the data acquisition device arranged on the first vehicle, so that the accuracy of the track data and the relative motion data can be ensured without acquiring operation data (such as steering wheel angle, steering lamp state, clutch brake state, driver state and the like of the second vehicle) in the second vehicle. Then, the track data acquired at a plurality of acquisition moments within a preset duration can be used as a track data set,and taking the relative motion data acquired at a plurality of acquisition moments within a preset duration as a relative motion data set. The preset duration may be understood as a time range including the current acquisition time, where a plurality of acquisition times are included. For example, the preset time length is 30s, the acquisition period is 0.1s, and the track data set is { P } 1 ，P 2 ，…，P i ，…，P 300 300 pieces of trajectory data, the set of relative motion data is { E } 1 ，E 2 ，…，E i ，…，E 300 300 pieces of relative motion data are included.
Step 102, inputting the track data set into a pre-trained track recognition model and a pre-trained track clustering model respectively, so as to determine track lane change probability according to the track recognition result output by the track recognition model and the track clustering result output by the track clustering model.
Step 103, inputting the relative motion data set into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model respectively, so as to determine the relative motion lane change probability according to the relative motion recognition result output by the relative motion recognition model and the relative motion clustering result output by the relative motion clustering model.
For example, a track recognition model and a track clustering model may be trained in advance for the track data set, where the track recognition model may recognize that the track of the second vehicle reflected by the input track data set belongs to lane change or straight running, i.e. the track recognition result output by the track recognition model is used to indicate that the track data set belongs to lane change or straight running. The track clustering model can cluster an input track data set and determine which category each track data belongs to, namely a track clustering result output by the track clustering model is used for indicating the category to which each track data belongs. In this way, the track data set can be respectively input into the track recognition model and the track clustering model to obtain the track recognition result and the track clustering result, and then the track change probability is determined according to the track recognition result and the track clustering result. The track change probability may be understood as the probability that the second vehicle will change track, as determined from the track data set. In one implementation, if the track recognition result indicates straight going, the track lane change probability may be directly determined to be zero, and if the track recognition result indicates lane change, the track lane change probability may be determined according to the track clustering result. In another implementation, the first track lane-change probability may be determined according to the track recognition result, then the second track lane-change probability may be determined according to the track clustering result, and then the product of the first track lane-change probability and the second track lane-change probability may be used as the track lane-change probability.
Similarly, a relative motion recognition model and a relative motion clustering model may be trained in advance for the relative motion data set, where the relative motion recognition model is capable of recognizing that the relative motion between the first vehicle and the second vehicle reflected by the input relative motion data set belongs to lane change or straight running, that is, the relative motion recognition result output by the relative motion recognition model is used to indicate that the relative motion data set belongs to lane change or straight running. The relative motion clustering model can cluster input relative motion data sets and determine which category each relative motion data belongs to, namely, a relative motion clustering result output by the relative motion clustering model is used for indicating the category to which each relative motion data belongs. In this way, the relative motion data set can be respectively input into the relative motion recognition model and the relative motion clustering model to obtain a relative motion recognition result and a relative motion clustering result, and then the relative motion lane change probability is determined according to the relative motion recognition result and the relative motion clustering result. The relative motion lane change probability may be understood as the probability that the second vehicle will change lanes as determined from the relative motion data set. In one implementation, if the relative motion recognition result indicates straight line, the relative motion lane change probability may be directly determined to be zero, and if the relative motion recognition result indicates lane change, the relative motion lane change probability may be determined according to the relative motion clustering result. In another implementation, the first relative motion lane-change probability may be determined according to the relative motion recognition result, then the second relative motion lane-change probability may be determined according to the relative motion clustering result, and then the product of the first relative motion lane-change probability and the second relative motion lane-change probability may be used as the relative motion lane-change probability.
Step 104, determining the lane change probability of the second vehicle according to the track lane change probability and the relative motion lane change probability.
Step 105, determining whether the second vehicle is about to change track according to the track change probability and the preset track change threshold value.
For example, after determining the track change probability and the relative motion change probability, the track change probability and the relative motion change probability may be weighted and summed to obtain the track change probability of the second vehicle. For example, the track change probability is S, and the track change probability is S P The probability of the relative motion lane change is S E S=α×s P +β*S E Where α is a weight corresponding to the track lane change probability (e.g., may be set to 0.5), and β is a weight corresponding to the relative motion lane change probability (e.g., may be set to 0.5). And finally, comparing the lane change probability with a preset lane change threshold value so as to determine whether the second vehicle is about to change lanes. For example, the lane change threshold is 0.5, if the lane change probability is greater than 0.5, it is determined that the second vehicle is about to change lanes, and if the lane change probability is less than or equal to 0.5, it is determined that the second vehicle is about to keep traveling straight. After determining that the second vehicle is about to change lanes, the first vehicle may also send a prompt message to prompt the driver of the first vehicle that the second vehicle is about to change lanes, thereby making a decision in advance for the driver of the first vehicle to avoid the second vehicle. The track data set and the relative motion data set comprise a plurality of data (comprising track data and relative motion data) acquired at the acquisition time, can reflect the running track of the second vehicle and the relative motion between the first vehicle and the second vehicle, comprehensively consider the continuity of the running process of the second vehicle in space and time, and therefore combine the track lane change probability determined according to the track data set and the lane change probability obtained according to the relative motion lane change probability determined by the relative motion data set, can more accurately determine whether the second vehicle is about to change lanes or not, and improve the accuracy of lane change judgment of the vehicle.
In summary, in the present disclosure, the first vehicle first acquires a track data set and a relative motion data set acquired at a plurality of acquisition moments within a preset duration. And inputting the track data set into a track recognition model and a track clustering model to determine track lane change probability according to the track recognition result and the track clustering result, and inputting the relative motion data set into a relative motion recognition model and a relative motion clustering model to determine the relative motion lane change probability according to the relative motion recognition result and the relative motion clustering result. And determining the lane change probability of a second vehicle according to the track lane change probability and the relative motion lane change probability, and finally determining whether the second vehicle is about to change lanes according to the lane change probability and a preset lane change threshold value, wherein the second vehicle is a vehicle in front of the first vehicle and positioned in an adjacent lane. According to the method and the device, the track data set and the relative motion data set are acquired by the first vehicle, the operation data in the second vehicle do not need to be acquired, the accuracy of the data is improved, whether the second vehicle is about to change the track is determined by the track data set and the relative motion data set, the continuity of the second vehicle in space and time in the running process is comprehensively considered, and the accuracy of vehicle track change judgment is improved.
FIG. 2 is a flow chart illustrating another method of determining a lane change of a vehicle, according to an exemplary embodiment, as shown in FIG. 2, the implementation of step 101 may include:
step 1011, acquiring track data and relative motion data acquired at each acquisition time within a preset duration, where the track data includes: the lateral and longitudinal positions of the second vehicle, the relative motion data comprising: the relative longitudinal speed and the relative longitudinal distance of the first vehicle and the second vehicle.
Step 1012, determining, according to the track data collected at each collection time, supplementary track data corresponding to the collection time, where the supplementary track data includes: lateral speed and lateral acceleration of the second vehicle.
In step 1013, the track data collected at each collection time and the corresponding supplementary track data are used as a track data set.
In one application scenario, the track data set may be acquired by first acquiring track data at each acquisition time within a preset duration. Then track at each acquisition time is checked according to the track data acquired at the acquisition timeTrace data is supplemented. For example, the trajectory data acquired at the acquisition time may include the lateral position x of the second vehicle i And a longitudinal position y i Two trajectory features. On the basis, the track data may also include the speed v of the second vehicle i And acceleration a i Then the trajectory data includes x in total i ，y i ，v i ，a i Four trajectory features.
Further, the lateral velocity v of the second vehicle at the acquisition time can be determined according to the track data at the acquisition time xi ＝(x i -x i-1 ) T and lateral acceleration a xi ＝(v xi -v x(i-1) ) and/T is taken as the supplementary track data corresponding to the acquisition time. On the basis of this, the longitudinal speed v of the second vehicle at the time of acquisition can also be determined yi ＝(y i -y i-1 ) T, longitudinal acceleration a yi ＝(v yi -v y(i-1) ) /T, heading angle Ang i ＝tan -1 (y i -y i-1 )/(x i -x i-1 ) And v is set yi 、a yi 、Ang i And supplementing the supplementary track data corresponding to the acquisition time. Thus, the supplemental trajectory data collectively includes v xi 、v yi 、a xi 、a yi 、Ang i Five trajectory features. And then splicing the track data acquired at each acquisition time and the corresponding supplementary track data into total track data (including nine track features) at the acquisition time as a track data set. I.e. the trajectory dataset is { P ] 1 ’，P 2 ’，…，P i ’，…，P N ' where N is the number of acquisition times in a preset time period, and the total track data of the ith acquisition time is P i ’＝{x i ，y i ，v i ，a i ，v xi ，v yi ，a xi ，a yi ，Ang i }. Further, the track features in each total track data can be subjected to maximum and minimum normalization processing to map each track feature to [0,1 ] ]Within the interval. By trace feature x i For example, for x i The normalization process is carried out, the processing is carried out,post-treatment x i * ＝(x i -x min )/(x max -x min ) Wherein x is max And x min The largest lateral position and the smallest lateral position in the trajectory dataset.
Step 1014, processing the relative motion data acquired at each acquisition time according to a preset rule to obtain a relative motion data set.
Correspondingly, the relative motion data set may be acquired by first acquiring the relative motion data at each acquisition time within a preset duration. The relative motion data may include a relative longitudinal speed Deltav of the first vehicle and the second vehicle 1i And a relative longitudinal distance deltay 1i It may further include: relative longitudinal speed Deltav of first and third vehicles 2i And a relative longitudinal distance deltay 2i Relative longitudinal speeds Deltav of the second and fourth vehicles 3i And a relative longitudinal distance deltay 3i Six features of relative motion are provided. Then the relative motion data acquired at the ith acquisition time is E i ＝{Δy 1i ，Δy 2i ，Δy 3i ，Δv 1i ，Δv 2i ，Δv 3i }. And then processing the relative motion data acquired at each acquisition time according to a preset rule to obtain a relative motion data set. The relative motion data set is { E 1 ，E 2 ，…，E i ，…，E N And (2) N is the number of acquisition moments in a preset time length.
Wherein, the preset rule may include at least one of the following: and (3) according to the rule I, if the relative motion characteristic acquired at a certain acquisition moment exceeds the normal range corresponding to the relative motion characteristic, deleting the relative motion characteristic acquired at the acquisition moment. And a rule II, if a certain relative motion characteristic acquired at a certain acquisition time exceeds a fluctuation range corresponding to the relative motion characteristic (which can be understood as jump of the relative motion characteristic acquired at the acquisition time), deleting the relative motion characteristic acquired at the acquisition time. Rule three, if a certain relative motion feature acquired at a certain acquisition time does not exist (it can be understood that the relative motion feature is not acquired, and also can be understood that the relative motion feature is deleted in rule one or rule two), fitting the relative motion feature acquired at the acquisition time according to a gaussian distribution. Similarly, the maximum and minimum normalization processing can be performed on the relative motion characteristics in each relative motion data, so as to map each relative motion characteristic into the [0,1] interval.
FIG. 3 is a flowchart illustrating another method of determining a lane change of a vehicle, according to an example embodiment, as shown in FIG. 3, step 102 may include:
step 1021, inputting the track data set into the track recognition model to obtain a track recognition result, and inputting the track data set into the track clustering model to obtain a track clustering result, wherein the track recognition result is used for indicating straight running or track changing, and the track clustering result is used for indicating the category of the track data collected at each collection time.
For example, the track recognition model may be an Attention-LSTM (English: attention Long Short-Term Memory, chinese: attention long-Term Memory) model, which may be understood as adding Attention weights to the hidden layer of the LSTM. Then after inputting the track data set into the Attention-LSTM model, the track recognition result output by the Attention-LSTM model and the Attention value at each acquisition time can be obtained. The trace dataset is { P 1 ，P 2 ，…，P s ，…，P t And the t is the number of acquisition moments in a preset time period. The track data set is input into the Attention-LSTM model, and the obtained Attention value of each acquisition time can be understood as the Attention between the acquisition time and the current acquisition time (i.e. the t acquisition time), and can be obtained by the following formula:
e t ＝(h 1 ,h 2 ,…,h s ,…,h t ) T h t
Wherein the method comprises the steps of，e t The attention value representing the current acquisition instant and all other acquisition instants,it can be understood that the pair e is a function of softmax t Obtained by normalization, represents the Attention value, e, of the s-th acquisition moment output by the Attention-LSTM model ts Representation e t The s-th element, h s And the hidden layer output of the Attention-LSTM model at the s-th acquisition moment is shown under the Self-Attention mechanism.
The trajectory clustering model may be a Kohonen neural network, and after the trajectory data set is input into the Kohonen neural network, the Kohonen neural network can cluster the trajectory data in the trajectory data set, so that a category to which each trajectory data belongs is determined according to a clustering result. The Kohonen neural network includes an input layer and a competition layer, and the number of clustered neurons in the competition layer can be preset, for example, 10, so that the Kohonen neural network can divide the trajectory data into 10 categories.
Step 1022, if the track recognition result indicates straight going, determining that the track lane change probability is zero.
Step 1023, if the track recognition result indicates track change, determining track change probability according to the track clustering result.
For example, if the track recognition result indicates straight going, the track lane change probability may be directly determined to be zero, and if the track recognition result indicates lane change, the track lane change probability may be determined further according to the track clustering result. Specifically, the track change probability may be determined by:
Step 1) determining the initial track lane change probability of each acquisition time according to the category of the track data acquired at the acquisition time and the corresponding relation between the preset category and the initial track lane change probability, which are included in the track clustering result.
For example, the initial track lane-changing probability at each acquisition time can be determined according to the category to which the track data acquired at each acquisition time belongs and the corresponding relation between the preset category and the initial track lane-changing probability, and the initial track lane-changing probability can be understood as the probability determined according to the track clustering result. Wherein the correspondence of the category and the initial trajectory transition probability may be determined when the Kohonen neural network is established. For example, a large amount of trajectory data that has been marked as a lane change, and a large amount of trajectory data that has been marked as a straight line may be input to the Kohonen neural network, resulting in the number of trajectory data that has been marked as a lane change clustered into M categories, and the number of trajectory data that has been marked as a straight line clustered into M categories, as shown in table 1:
TABLE 1
Where nI represents the number of track data clustered under category I, marked as lane change, and mI represents the number of track data clustered under category I, marked as straight. Then, the correspondence between the category and the initial track change probability may be as shown in table 2:
TABLE 2
| Class 1 | Class 2 | … | Class I | … | Class M |
| q1 | q2 | … | qI | … | qM |
Wherein qI represents the initial track change probability corresponding to the category I, qI =ni/(ni+mi). With trace dataset as { P ] 1 ，P 2 ，…，P s ，…，P t The method comprises the steps of inputting a track data set into a Kohonen neural network to obtain a track clustering result { O }, wherein t is the number of acquisition moments in a preset time period for example 1 ，O 2 ，…，O s ，…，O t O, where s The category to which the s-th acquisition time belongs. And then determining the initial track lane change probability of each acquisition time according to the table 2: { S P1 ，S P2 ，…，S Ps ，…，S Pt S, where S Ps The initial track change probability at the s-th acquisition time is represented.
Step 2) determining the track change probability according to the initial track change probability at each acquisition time and the attention value at each acquisition time.
For example, the product of the initial track change probability at each acquisition time and the attention value at each acquisition time can be summed to obtain the track change probability, i.e
FIG. 4 is a flowchart illustrating another method of determining a lane change of a vehicle, according to an example embodiment, as shown in FIG. 4, an implementation of step 103 may include:
step 1031, inputting the relative motion data set into a relative motion recognition model to obtain a relative motion recognition result, and inputting the relative motion data set into a relative motion clustering model to obtain a relative motion clustering result, wherein the relative motion recognition result is used for indicating straight line or lane change, and the relative motion clustering result is used for indicating the category to which the relative motion data acquired at each acquisition time belongs.
For example, the relative motion recognition model may be an Attention-GRU (English: attention Gate Recurrent Unit, chinese: attention-gate cycle unit) model, which may be understood as adding Attention weights to the hidden layer of the GRU. Then after inputting the relative motion data set into the Attention-GRU model, the relative motion recognition result output by the Attention-GRU model and the Attention value at each acquisition time can be obtained. The relative motion data set is { E 1 ，E 2 ，…，E s ，…，E t And the t is the number of acquisition moments in a preset time period. The relative motion data set is input into the Attention-GRU model, and the obtained Attention value of each acquisition time can be understood as the Attention between the acquisition time and the current acquisition time (i.e. the t acquisition time), and can be obtained by the following formula:
e t ＝(h 1 ,h 2 ,…,h s ,…,h t ) T h t
wherein e t Attention value a representing current acquisition time and all other acquisition times t r s el It can be understood that the pair e is a function of softmax t Obtained by normalization, represents the Attention value, e, of the s-th acquisition moment output by the Attention-GRU model ts Representation e t The s-th element, h s And the hidden layer output of the Attention-GRU model at the s-th acquisition time under the Self-Attention mechanism is shown.
The relative motion clustering model may be a Kohonen neural network, and after the relative motion data set is input into the Kohonen neural network, the Kohonen neural network can cluster the relative motion data in the relative motion data set, so that the category to which each relative motion data belongs is determined according to the clustering result. The Kohonen neural network includes an input layer and a competition layer, and the number of clustered neurons in the competition layer may be preset, for example, 10, and then the Kohonen neural network may classify the relative motion data into 10 categories.
In step 1032, if the relative motion recognition result indicates straight line, the relative motion lane change probability is determined to be zero.
Step 1033, if the relative motion recognition result indicates lane change, determining the relative motion lane change probability according to the relative motion clustering result.
For example, if the relative motion recognition result indicates straight going, the relative motion lane change probability may be directly determined to be zero, and if the relative motion recognition result indicates lane change, the relative motion lane change probability may be determined further according to the relative motion clustering result. Specifically, the relative motion lane change probability may be determined by:
step 3) determining the initial relative motion lane change probability of each acquisition moment according to the category to which the relative motion data acquired at each acquisition moment included in the relative motion clustering result belong and the corresponding relation between the preset category and the initial relative motion lane change probability.
For example, the initial relative motion lane-change probability at each acquisition time can be determined according to the category to which the relative motion data acquired at each acquisition time belongs and the corresponding relation between the preset category and the initial relative motion lane-change probability, and the initial relative motion lane-change probability can be understood as the probability determined according to the relative motion clustering result. The corresponding relation between the category and the initial relative motion lane change probability is the same as the corresponding relation between the category and the initial track lane change probability, and is not described herein.
With relative motion data set { E ] 1 ，E 2 ，…，E s ，…，E t T is the number of acquisition moments in a preset time length for example, and the relative motion data set is input into a Kohonen neural network to obtain a relative motion clustering result { Q } 1 ，Q 2 ，…，Q s ，…，Q t }, wherein Q s The category to which the s-th acquisition time belongs. Then according to the category and the initial relative motion lane change probabilityCorresponding relation, determining initial relative motion lane change probability at each acquisition time: { S E1 ，S E2 ，…，S Es ，…，S Et S, where S Es The initial relative motion lane change probability at the s-th acquisition time is represented.
And 4) determining the relative motion lane change probability according to the initial relative motion lane change probability at each acquisition time and the attention value at each acquisition time.
For example, the product of the initial relative motion lane change probability at each acquisition time and the attention value at each acquisition time can be summed to obtain the relative motion lane change probability, i.e
Fig. 5 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment, and step 101 may be performed by:
step 1015, obtaining track data and relative motion data collected at each collection time within a specified time.
Step 1016, dividing the specified duration into a specified number of sliding windows, each sliding window having a length of a preset duration.
Step 1017, using the track data and the relative motion data collected in each sliding window as the track data set and the relative motion data set corresponding to the sliding window.
In another implementation, multiple sets of trajectory data sets and relative motion data sets may be acquired. And then, respectively executing steps 102 to 104 on each group of track data sets and the relative motion data sets to obtain the lane change probability determined according to each group of track data sets and the relative motion data sets, then, obtaining the total lane change probability, and then, comparing the total lane change probability with a lane change threshold value according to the total lane change probability to judge whether the second vehicle is to change lanes or not.
Specifically, the track data and the relative motion data collected at each collection time within a specified time period can be obtained first, wherein the specified time period is longer than a preset time period. And then dividing the appointed time length into appointed number of sliding windows according to a preset step length, wherein the length of each sliding window is the preset time length. For example, the specified duration is 2min, the preset duration is 30s, and the step size can be set to 15s, so that 7 (i.e. the specified number of) sliding windows can be obtained: [0, 30s ], [15s,45s ], [30s,60s ], [45s,75s ], [60s,90s ], [75s,105s ], [90s,120s ]. Then, the track data and the relative motion data collected in each sliding window are used as the track data set and the relative motion data set corresponding to the sliding window, so that the track data set and the relative motion data set of the designated number of groups are obtained.
Accordingly, step 105 may be implemented by:
step 1051, the track changing probability determined according to the track data set and the relative motion data set corresponding to each sliding window is weighted and summed to determine the total track changing probability, and the weight corresponding to each sliding window is inversely proportional to the distance between the sliding window and the current moment.
If the total lane change probability is greater than or equal to the lane change threshold, step 1052, it is determined that the second vehicle is about to change lanes.
If the total lane change probability is less than the lane change threshold, it is determined 1053 that the second vehicle will remain straight.
For example, steps 102 to 104 may be performed on the trajectory data set and the relative motion data set corresponding to each sliding window, respectively, to obtain the lane change probability determined according to the trajectory data set and the relative motion data set corresponding to each sliding window. And then carrying out weighted summation on the lane change probability determined according to the track data set corresponding to each sliding window and the relative motion data set so as to obtain the total lane change probability. Wherein the weight corresponding to each sliding window is inversely proportional to the distance between the sliding window and the current moment. And if the total lane change probability is greater than or equal to the lane change threshold, determining that the second vehicle is about to change lanes, and if the total lane change probability is less than the lane change threshold, determining that the second vehicle is about to keep straight.
Taking the example of determining the track data sets and the relative motion data sets corresponding to the N sliding windows in step 1017, N lane change probabilities may be obtained: s is S 1 ，S 2 ，…，S i ，…，S N Wherein S is i And determining the lane change probability for the track data set and the relative motion data set corresponding to the ith sliding window. The total lane change probability may be determined by the following formula:
Wherein S is total Represents the total lane change probability, W i The weight corresponding to the ith sliding window is represented, wherein D represents a designated duration, D i Representing the distance of the ith sliding window from the current time, e.g., d=120s, 7 sliding windows: [0, 30s]、[15s，45s]、[30s，60s]、[45s，75s]、[60s，90s]、[75s，105s]、[90s，120s]Then d 1 ＝120-30＝90，d 2 =120-45=75, and so on.
Fig. 6 is a flowchart illustrating another method of determining a lane change of a vehicle according to an exemplary embodiment, and the trajectory recognition result is used to indicate a straight line, a left lane change, or a right lane change as shown in fig. 6. After step 105, the method may further include:
step 106, if it is determined that the second vehicle is about to change lanes, counting a first number corresponding to the left lane change, a second number corresponding to the right lane change and a third number corresponding to the straight line in the track recognition result determined according to the track data set corresponding to each sliding window, where the sum of the first number, the second number and the third number is equal to the specified number.
If the first number is greater than the second number, it is determined that the second vehicle is about to lane-change left, step 107.
If the first number is less than the second number, it is determined that the second vehicle is about to change lanes right, step 108.
For example, if the track recognition result indicates that the track of the second vehicle belongs to a left lane change, a right lane change or a straight line (i.e. the scene of the above lane change is divided into a left lane change and a right lane change), it may also be determined whether the second vehicle is about to change the left lane or the right lane change after determining that the second vehicle is about to change the lane in step 105. Specifically, step 102 may be performed on the track data sets corresponding to the specified number of sliding windows, so as to obtain the specified number of track recognition results determined according to the track data set corresponding to each sliding window. And counting the first quantity corresponding to the left lane change, the second quantity corresponding to the right lane change, the third quantity corresponding to the straight line in the identification results of the specified quantity of the tracks, wherein the sum of the first quantity, the second quantity and the third quantity is equal to the specified quantity.
If the first number is greater than the second number, it may be determined that the second vehicle is about to lane left. If the first number is less than the second number, it may be determined that the second vehicle is about to lane-change. If the first number is equal to the second number, the next acquisition time can be waited, and the judgment can be carried out again after new track data are acquired.
FIG. 7 is a flowchart illustrating another method of determining a lane change of a vehicle, as shown in FIG. 7, according to an exemplary embodiment, after step 105, the method may further include:
step 109, if it is determined that the second vehicle is going to change track, determining the track changing time of the second vehicle according to the lateral position, the lateral speed and the heading angle of the second vehicle included in the track data acquired at the current acquisition time.
For example, after determining that the second vehicle is about to change track in step 105, the track change track of the second vehicle may be further determined according to the track data collected at the current collection time in the track data set. Specifically, the lane change time of the second vehicle may be determined according to the lateral position, the lateral speed, and the heading angle of the second vehicle included in the track data acquired at the current acquisition time. Taking the current acquisition time as i and the corresponding acquired track data as P i For example, the transverse position included therein is x i Transverse velocity v xi And heading angle Ang i Then the lane change time of the second vehicle may be calculated by the following formulaAnd (3) obtaining:
wherein, time is pre,i Represents the lane change time, x tar And the central position of a target lane to which the second lane is to be changed is represented, xi, omega and lambda are preset correlation coefficients, and theta is a preset constant. It should be noted that, ζ, ω, λ, and θ may be obtained by constructing a correlation coefficient matrix according to a predetermined training data set and then using a linear regression method. For example, the training data set is { T } 1 ，T 2 ，…，T j ，…，T N Wherein N is the number of training data contained in the training data set, T j ＝{x j ，v xj ，Ang j For each training data, calculate the time according to the above formula pre,i A correlation matrix was constructed as shown in table 3:
TABLE 3 Table 3
Wherein, the liquid crystal display device comprises a liquid crystal display device,cov (X, Y) represents tan (Ang) j ) And->Is represented by σX, tan (Ang j ) Sigma Y represents +.>And the other correlation coefficients are similar, and are not described in detail herein. According to the correlation matrix pair tan (Ang j )、/>And (x) j -x tar ) By performing linear regression, as shown in FIG. 8, it is possible to obtainζ, ω, λ and θ.
And 110, determining termination track data of the second vehicle at the end of the lane change according to the lane change time and the longitudinal speed included in the track data acquired at the current acquisition time.
And step 111, fitting according to Bessel functions according to the termination track data and the track data acquired at the current acquisition time to obtain a track changing track of the second vehicle.
For example, after determining the lane change time, the longitudinal displacement of the second vehicle during the lane change may be determined according to the lane change time and the longitudinal speed included in the track data collected at the current collection time, and further, the termination track data of the second vehicle at the end of the lane change may be determined according to the longitudinal displacement and the center position of the target lane to which the second vehicle is to be lane-changed. Taking the current acquisition time as i and the corresponding acquired track data as P i Including a longitudinal speed v yi For example, then longitudinal displacement y pre,i ＝v yi time pre,i . And determining the transverse position, the longitudinal position, the speed and the course angle of the second vehicle at the end of lane change according to the longitudinal displacement and the central position of the target lane, and taking the transverse position, the longitudinal position, the speed and the course angle as termination track data.
And finally, fitting according to the termination track data and the track data acquired at the current acquisition moment and the Bessel function to obtain the track changing track of the second vehicle. For example, the trajectory data acquired at the current acquisition time may be used as the start point P 0 Terminating track data as an end point P 3 Then P is taken up 0 Extending a length (d) along the current driving direction of the second vehicle as an extension point P 1 And then P is added 3 Extending a length of the vehicle in the opposite direction of the current running direction of the second vehicle as a reverse extension point P 2 For P 0 、P 1 、P 2 、P 3 Fitting according to a third-order Bessel function: b (t) =p 0 (1-t) 3 +P 1 t(1-t) 2 +P 2 t 2 (1-t)+P 3 t 3 Wherein t is in the range of [0,1]In the section, the fitted lane change track is shown in fig. 9. In determining the second vehicleAfter the track changing track, the track changing track can be displayed on a central control display screen of the first vehicle, so that a driver of the first vehicle makes a decision in advance to avoid the second vehicle.
It should be noted that, the trajectory recognition model in the above embodiment is trained by the following steps:
step A, a first sample input set and a first sample output set are obtained, each first sample input in the first sample input set comprises a plurality of sample track data, the first sample output set comprises first sample output corresponding to each first sample input, each first sample output comprises a sample track recognition result marked by the corresponding plurality of sample track data, and the sample track recognition result is used for indicating straight running or lane change.
For example, a first set of sample inputs and a first set of sample outputs for training the trajectory recognition model may be obtained prior to training the trajectory recognition model, wherein the first set of sample inputs includes a plurality of first sample inputs, each first sample input including a plurality of sample trajectory data, i.e., each first sample input is a set of sample trajectory data, the set of sample trajectory data being labeled with sample trajectory recognition results for indicating that the set of sample trajectory data is straight or lane-changing (may also be used for indicating that the set of sample trajectory data is straight, lane-changing left or lane-changing right). Thus, each first sample input corresponds to a sample trace recognition result, and a first sample output set is formed. The rule for determining the sample track recognition result may be that sample track data collected when the second vehicle passes through the lane line and sample track data collected in 4s before the second vehicle passes through the lane line are marked as lane change. Sample track data collected in the rest time is marked as straight.
Wherein each sample trace data may include: the transverse position, the longitudinal position, the speed, the acceleration, the transverse speed, the longitudinal speed, the transverse acceleration, the longitudinal acceleration and the course angle are nine track characteristics. For example, the first sample input set includes N first sample inputs, which may be { A } 1 ，A 2 ，…，A n ，…，A N (wherein A) n Includes L sample trace data, which may be { a } n1 ，a n2 ，…，a nl ，…，a nL }。
And B, taking the first sample input set as the input of the track recognition model, and taking the first sample output set as the output of the track recognition model so as to train the track recognition model.
After the first sample input set and the first sample output set are acquired, the first sample input set may be used as an input of a track recognition model, and the first sample output set may be used as an output of the track recognition model to train the track recognition model, so that when any first sample input is input, the output of the track recognition model is matched with the first sample output corresponding to the first sample input.
The following specifically describes a training process of the trajectory recognition model by taking the trajectory recognition model as an LSTM model as an example:
for example, an input weight may be added to the input layer of the LSTM model, and it may be understood that when any first sample input is taken as an input of the LSTM model, the first sample input is multiplied by the input weight, that is, a forgetting gate model of the LSTM model is:
/>
Wherein n is t Represents the t first sample input, f t Output of forgetting gate, h t-1 Represents the output of the LSTM model at the last moment, W f Weight representing forgetting gate, b f Indicating a deviation from the forgotten gate,representing an input weight;
the input gate model of the LSTM model is:
wherein i is t Representing an input doorOutput, W i Representing the weight of the input gate, b i Representing the bias of the input gate.
Candidate gate models for the LSTM model are:
wherein, the liquid crystal display device comprises a liquid crystal display device,representing candidate vectors, W C Representing the weights of candidate gates, b C Representing the bias of the candidate gate. Correspondingly, the memory cell function is: />
The output gate model of the LSTM model is:
wherein o is t Representing the output of the output gate, W o Representing the weight of the output gate, b o Representing the bias of the output gate.
And controlling the memory unit by using the tanh activation function, wherein the output of the LSTM model is as follows:
h t ＝o t ·tanh(C t )
specifically, the input weight of the LSTM may be determined by:
and B1, inputting a first sample input set into the initial LSTM model according to a first input weight, and taking the first sample output set as the output of the initial LSTM model to train the initial LSTM model.
And step B2, updating the first input weight according to the trained initial LSTM model.
And B3, repeatedly executing the steps B1 to B2, and taking the initial LSTM model and the first input weight obtained by executing the preset iteration number times as a track recognition model and the input weight corresponding to the track recognition model.
The current iteration number is represented by K max Representing a preset number of iterations for illustration. The first sample input set is { A 1 ，A 2 ，…，A n ，…，A N (wherein A) n Is { a } n1 ，a n2 ，…，a nl ，…，a nL Examples are shown. K=1, performing a first iteration, wherein the initial value of the first input weight corresponding to the first sample input set may be W K ＝{W 1 K ，W 2 K ，…，W n K ，…，W N K }，W n K Representation A n Corresponding input weight, W n K ＝{W n1 K ，W n2 K ，…，W nl K ，…，W nL K W, where nl K Representing a in the K-th iteration nl Corresponding input weight, at this time, W can be made to be nl K ＝1/L。
Let n=1, i.e. a n According to W n K Is input into the initial LSTM model, will A n The corresponding first sample output is used as the output of the initial LSTM model to train the initial LSTM model, and the trained initial LSTM model is obtained. New A can be obtained according to the trained initial LSTM model n Corresponding first input weight, i.e. W n K+1 。W n K+1 ＝{W n1 K+1 ，W n2 K +1 ，…，W nl K+1 ，…，W nL K+1 },Wherein, the liquid crystal display device comprises a liquid crystal display device,can be understood as normalization factor, alpha K G is a preset parameter K (W nl K A n ) To A n According to W n K Input weight input initial L of (2)STM model, output of initial LSTM model.
Then, n=n+1 is repeated in turn until n=n, and W is obtained K+1 ＝{W 1 K+1 ，W 2 K+1 ，…，W n K+1 ，…，W N K+1 }. Afterwards, let k=k+1 again until k=k max Through two-layer circulation, W is obtained Kmax ＝{W 1 Kmax ，W 2 Kmax ，…，W n Kmax ，…，W N Kmax W, where n Kmax ＝{W n1 Kmax ，W n2 Kmax ，…，W nl Kmax ，…，W nL Kmax }。
The relative motion recognition model is trained by the following steps:
and C, acquiring a second sample input set and a second sample output set, wherein each second sample input in the second sample input set comprises a plurality of sample relative motion data, each second sample output set comprises a second sample output corresponding to each second sample input, each second sample output comprises a sample relative motion recognition result marked by the corresponding plurality of sample relative motion data, and the sample relative motion recognition result is used for indicating straight line or lane change.
And D, taking the second sample input set as the input of the relative motion recognition model, and taking the second sample output set as the output of the relative motion recognition model so as to train the relative motion recognition model.
Likewise, a second set of sample inputs and a second set of sample outputs for training the relative motion recognition model may be obtained prior to training the relative motion recognition model, wherein the second set of sample inputs includes a plurality of second sample inputs, each second sample input including a plurality of sample relative motion data, i.e., each second sample input is a set of sample relative motion data, the set of sample relative motion data being labeled with sample relative motion recognition results for indicating that the set of sample relative motion data is straight or lane-changing (and may also be used for indicating that the set of sample relative motion data is straight, lane-changing left or lane-changing right). Thus, each second sample inputs a corresponding sample relative motion recognition result, forming a second sample output set. The rule for determining the sample relative movement recognition result can be that sample relative movement data acquired when the second vehicle passes through the lane line and sample relative movement data acquired in 4s before the second vehicle passes through the lane line are marked as lane change. And the relative motion data of the samples collected in the rest time are marked as straight.
Wherein each sample relative motion data may comprise: the relative longitudinal speed and relative longitudinal distance of the first vehicle and the second vehicle, the relative longitudinal speed and relative longitudinal distance of the first vehicle and the third vehicle, and the relative longitudinal speed and relative longitudinal distance of the second vehicle and the fourth vehicle, are six relative motion characteristics. For example, the second sample input set includes N second sample inputs, which may be { B } 1 ，B 2 ，…，B n ，…，B N }, wherein B is n Includes L samples of relative motion data, which may be { b } n1 ，b n2 ，…，b nl ，…，b nL }。
After the second sample input set and the second sample output set are obtained, the second sample input set may be used as an input of the relative motion recognition model, and the second sample output set may be used as an output of the relative motion recognition model to train the relative motion recognition model, so that when any second sample input is input, the output of the relative motion recognition model matches with the second sample output corresponding to the second sample input.
The following specifically describes a training process of the relative motion recognition model, taking the relative motion recognition model as a GRU model as an example:
for example, an input weight may be added to the input layer of the GRU model, and it may be understood that when any second sample input is used as an input of the GRU model, the second sample input is multiplied by the input weight, that is, the updated gate model of the GRU model is:
Wherein n' t Represents the t second sample input, r t Representing the output of the update gate, W r Representing the weight of the update gate, h' t-1 Representing the output of the GRU model at the previous moment,representing the input weights.
The reset gate model of the GRU model is:
wherein z is t Representing the output of the reset gate, W z Indicating the weight of the reset gate.
Candidate gate models for the GRU model are:
wherein, the liquid crystal display device comprises a liquid crystal display device,as candidate vector, W h Is the weight of the candidate gate. The output of the corresponding GRU model is:
specifically, the input weight of the GRU may be determined by:
and D1, inputting a second sample input set into the initial GRU model according to a second input weight, and taking the second sample output set as the output of the initial GRU model to train the initial GRU model.
And D2, updating the second input weight according to the trained initial GRU model.
And D3, repeatedly executing the steps D1 to D2, and taking the initial GRU model and the second input weight obtained by executing the iteration number times as the relative motion recognition model and the input weight corresponding to the relative motion recognition model.
The input weights of the GRUs in the above embodiment are the same as those of the LSTM in the above embodiment, and will not be described in detail here.
In summary, in the present disclosure, the first vehicle first acquires a track data set and a relative motion data set acquired at a plurality of acquisition moments within a preset duration. And inputting the track data set into a track recognition model and a track clustering model to determine track lane change probability according to the track recognition result and the track clustering result, and inputting the relative motion data set into a relative motion recognition model and a relative motion clustering model to determine the relative motion lane change probability according to the relative motion recognition result and the relative motion clustering result. And determining the lane change probability of a second vehicle according to the track lane change probability and the relative motion lane change probability, and finally determining whether the second vehicle is about to change lanes according to the lane change probability and a preset lane change threshold value, wherein the second vehicle is a vehicle in front of the first vehicle and positioned in an adjacent lane. According to the method and the device, the track data set and the relative motion data set are acquired by the first vehicle, the operation data in the second vehicle do not need to be acquired, the accuracy of the data is improved, whether the second vehicle is about to change the track is determined by the track data set and the relative motion data set, the continuity of the second vehicle in space and time in the running process is comprehensively considered, and the accuracy of vehicle track change judgment is improved.
Fig. 10 is a block diagram illustrating a vehicle lane change determination apparatus according to an exemplary embodiment, and as shown in fig. 10, the apparatus 200 is applied to a first vehicle, and includes:
the acquiring module 201 is configured to acquire a track data set and a relative motion data set, where the track data set includes track data of a second vehicle acquired at a plurality of acquisition moments within a preset duration, and the relative motion data set includes relative motion data between the second vehicle and a first vehicle acquired at the plurality of acquisition moments within the preset duration, and the plurality of acquisition moments includes a current acquisition moment, and the second vehicle is a vehicle that travels on an adjacent lane of the lane where the first vehicle is located and is located in front of the first vehicle.
The first processing module 202 is configured to input the track data set into a pre-trained track recognition model and a pre-trained track clustering model, so as to determine a track lane change probability according to a track recognition result output by the track recognition model and a track clustering result output by the track clustering model.
The second processing module 203 is configured to input the relative motion data set into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model, so as to determine a relative motion lane-changing probability according to a relative motion recognition result output by the relative motion recognition model and a relative motion clustering result output by the relative motion clustering model.
The first determining module 204 is configured to determine a lane-change probability of the second vehicle according to the track lane-change probability and the relative motion lane-change probability.
The second determining module 205 is configured to determine whether the second vehicle is about to change lanes according to the lane change probability and a preset lane change threshold.
Fig. 11 is a block diagram of another vehicle lane change determination apparatus according to an exemplary embodiment, and as shown in fig. 11, the acquisition module 201 includes:
the first obtaining submodule 2011 is configured to obtain track data and relative motion data collected at each collection time within a preset duration, where the track data includes: the lateral and longitudinal positions of the second vehicle, the relative motion data comprising: the relative longitudinal speed and the relative longitudinal distance of the first vehicle and the second vehicle.
The processing submodule 2012 is configured to determine, according to the track data collected at each collection time, supplementary track data corresponding to the collection time, where the supplementary track data includes: lateral speed and lateral acceleration of the second vehicle. And taking the track data acquired at each acquisition time and the corresponding supplementary track data as a track data set. And processing the relative motion data acquired at each acquisition time according to a preset rule to obtain a relative motion data set.
Fig. 12 is a block diagram of another vehicle lane change determination apparatus according to an exemplary embodiment, and as shown in fig. 12, the first processing module 202 includes:
the first input submodule 2021 is configured to input a track data set into the track recognition model to obtain a track recognition result, and input the track data set into the track clustering model to obtain a track clustering result, where the track recognition result is used to indicate straight running or lane changing, and the track clustering result is used to indicate a category to which the track data collected at each collection time belongs.
The first judging submodule 2022 is configured to determine that the track lane change probability is zero if the track recognition result indicates straight running. And if the track identification result indicates track change, determining track change probability according to the track clustering result.
In one embodiment, the trajectory recognition model is the Attention-LSTM model, and the first input sub-module 2021 is configured to:
the track data set is input into the Attention-LSTM model to acquire the track recognition result output by the Attention-LSTM model and the Attention value of each acquisition moment.
The first judgment sub-module 2022 is configured to implement the following steps:
step 1) determining the initial track lane change probability of each acquisition time according to the category of the track data acquired at the acquisition time and the corresponding relation between the preset category and the initial track lane change probability, which are included in the track clustering result.
Step 2) determining the track change probability according to the initial track change probability at each acquisition time and the attention value at each acquisition time.
Fig. 13 is a block diagram of another vehicle lane change determination apparatus according to an exemplary embodiment, and as shown in fig. 13, the second processing module 203 includes:
the second input sub-module 2031 is configured to input the relative motion data set into a relative motion recognition model to obtain a relative motion recognition result, and input the relative motion data set into a relative motion clustering model to obtain a relative motion clustering result, where the relative motion recognition result is used to indicate straight line or lane change, and the relative motion clustering result is used to indicate a category to which the relative motion data collected at each collection time belongs.
The second judging submodule 2032 is configured to determine that the relative motion lane change probability is zero if the relative motion identification result indicates straight going. And if the relative motion recognition result indicates lane change, determining the relative motion lane change probability according to the relative motion clustering result.
Optionally, the relative motion recognition model is an Attention-GRU model, and the second input submodule 2031 is configured to:
and inputting the relative motion data set into the Attention-GRU model to acquire the relative motion recognition result output by the Attention-GRU model and the Attention value of each acquisition time.
The second judging submodule 2032 is configured to implement the following steps:
step 3) determining the initial relative motion lane change probability of each acquisition moment according to the category to which the relative motion data acquired at each acquisition moment included in the relative motion clustering result belong and the corresponding relation between the preset category and the initial relative motion lane change probability.
And 4) determining the relative motion lane change probability according to the initial relative motion lane change probability at each acquisition time and the attention value at each acquisition time.
Fig. 14 is a block diagram of another vehicle lane change determination apparatus according to an exemplary embodiment, and as shown in fig. 14, the acquisition module 201 includes:
a second acquiring sub-module 2013, configured to acquire track data and relative motion data acquired at each acquisition time within a specified duration.
A dividing submodule 2014, configured to divide the specified duration into a specified number of sliding windows, where a length of each sliding window is a preset duration. And taking the track data and the relative motion data acquired in each sliding window as a track data set and a relative motion data set corresponding to the sliding window.
The second determining module 205 includes:
the first determining submodule 2051 is configured to weight and sum the lane change probabilities determined according to the track data set and the relative motion data set corresponding to each sliding window to determine a total lane change probability, where the weight corresponding to each sliding window is inversely proportional to the distance between the sliding window and the current moment.
The second determining submodule 2052 is configured to determine that the second vehicle is about to change lanes if the total lane change probability is greater than or equal to the lane change threshold. If the total lane change probability is less than the lane change threshold, it is determined that the second vehicle will remain straight.
Fig. 15 is a block diagram of another vehicle lane change determination apparatus according to an exemplary embodiment, as shown in fig. 15, with the track recognition result indicating straight, left lane change, or right lane change. The apparatus 200 further comprises:
and a third determining module 206, configured to, after determining whether the second vehicle is about to change the track according to the track changing probability and the preset track changing threshold, count a first number corresponding to the left track, a second number corresponding to the right track, and a third number corresponding to the straight track in the track recognition result determined according to the track data set corresponding to each sliding window if the second vehicle is determined to change the track, where a sum of the first number, the second number, and the third number is equal to the specified number. If the first number is greater than the second number, it is determined that the second vehicle is about to lane-change left. If the first number is less than the second number, it is determined that the second vehicle is about to lane-change.
Fig. 16 is a block diagram of another vehicle lane change determination apparatus according to an exemplary embodiment, and as shown in fig. 16, the apparatus 200 further includes:
The track generation module 207 is configured to determine, after determining whether the second vehicle is about to change the track according to the track change probability and the preset track change threshold, if it is determined that the second vehicle is about to change the track, determine a track change time of the second vehicle according to a lateral position, a lateral speed and a heading angle of the second vehicle included in the track data acquired at the current acquisition time. And determining termination track data of the second vehicle at the end of the lane change according to the lane change time and the longitudinal speed included in the track data acquired at the current acquisition time. And fitting according to the ending track data and the track data acquired at the current acquisition time and the Bessel function to obtain the track changing track of the second vehicle.
It should be noted that, the trajectory recognition model in the above embodiment is trained by the following steps:
step A, a first sample input set and a first sample output set are obtained, each first sample input in the first sample input set comprises a plurality of sample track data, the first sample output set comprises first sample output corresponding to each first sample input, each first sample output comprises a sample track recognition result marked by the corresponding plurality of sample track data, and the sample track recognition result is used for indicating straight running or lane change.
And B, taking the first sample input set as the input of the track recognition model, and taking the first sample output set as the output of the track recognition model so as to train the track recognition model.
The relative motion recognition model is trained by the following steps:
and C, acquiring a second sample input set and a second sample output set, wherein each second sample input in the second sample input set comprises a plurality of sample relative motion data, each second sample output set comprises a second sample output corresponding to each second sample input, each second sample output comprises a sample relative motion recognition result marked by the corresponding plurality of sample relative motion data, and the sample relative motion recognition result is used for indicating straight line or lane change.
And D, taking the second sample input set as the input of the relative motion recognition model, and taking the second sample output set as the output of the relative motion recognition model so as to train the relative motion recognition model.
In one application scenario, the trajectory recognition model is an LSTM model and the relative motion recognition model is a GRU model.
Step B may be implemented by:
and B1, inputting a first sample input set into the initial LSTM model according to a first input weight, and taking the first sample output set as the output of the initial LSTM model to train the initial LSTM model.
And step B2, updating the first input weight according to the trained initial LSTM model.
And B3, repeatedly executing the steps B1 to B2, and taking the initial LSTM model and the first input weight obtained by executing the preset iteration number times as a track recognition model and the input weight corresponding to the track recognition model.
Step D may be achieved by:
and D1, inputting a second sample input set into the initial GRU model according to a second input weight, and taking the second sample output set as the output of the initial GRU model to train the initial GRU model.
And D2, updating the second input weight according to the trained initial GRU model.
And D3, repeatedly executing the steps D1 to D2, and taking the initial GRU model and the second input weight obtained by executing the iteration number times as the relative motion recognition model and the input weight corresponding to the relative motion recognition model.
The specific manner in which the various modules perform the operations in the apparatus of the above embodiments have been described in detail in connection with the embodiments of the method, and will not be described in detail herein.
In summary, in the present disclosure, the first vehicle first acquires a track data set and a relative motion data set acquired at a plurality of acquisition moments within a preset duration. And inputting the track data set into a track recognition model and a track clustering model to determine track lane change probability according to the track recognition result and the track clustering result, and inputting the relative motion data set into a relative motion recognition model and a relative motion clustering model to determine the relative motion lane change probability according to the relative motion recognition result and the relative motion clustering result. And determining the lane change probability of a second vehicle according to the track lane change probability and the relative motion lane change probability, and finally determining whether the second vehicle is about to change lanes according to the lane change probability and a preset lane change threshold value, wherein the second vehicle is a vehicle in front of the first vehicle and positioned in an adjacent lane. According to the method and the device, the track data set and the relative motion data set are acquired by the first vehicle, the operation data in the second vehicle do not need to be acquired, the accuracy of the data is improved, whether the second vehicle is about to change the track is determined by the track data set and the relative motion data set, the continuity of the second vehicle in space and time in the running process is comprehensively considered, and the accuracy of vehicle track change judgment is improved.
Fig. 17 is a block diagram of an electronic device 300, according to an example embodiment. As shown in fig. 17, the electronic device 300 may include: a processor 301, a memory 302. The electronic device 300 may also include one or more of a multimedia component 303, an input/output (I/O) interface 304, and a communication component 305.
The processor 301 is configured to control the overall operation of the electronic device 300 to perform all or part of the steps in the above-mentioned method for determining lane change of a vehicle. The memory 302 is used to store various types of data to support operation at the electronic device 300, which may include, for example, instructions for any application or method operating on the electronic device 300, as well as application-related data, such as contact data, transceived messages, pictures, audio, video, and the like. The Memory 302 may be implemented by any type or combination of volatile or non-volatile Memory devices, such as static random access Memory (Static Random Access Memory, SRAM for short), electrically erasable programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM for short), erasable programmable Read-Only Memory (Erasable Programmable Read-Only Memory, EPROM for short), programmable Read-Only Memory (Programmable Read-Only Memory, PROM for short), read-Only Memory (ROM for short), magnetic Memory, flash Memory, magnetic disk, or optical disk. The multimedia component 303 may include a screen and an audio component. Wherein the screen may be, for example, a touch screen, the audio component being for outputting and/or inputting audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 302 or transmitted through the communication component 305. The audio assembly further comprises at least one speaker for outputting audio signals. The I/O interface 304 provides an interface between the processor 301 and other interface modules, which may be a keyboard, mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 305 is used for wired or wireless communication between the electronic device 300 and other devices. Wireless communication, such as Wi-Fi, bluetooth, near field communication (Near Field Communication, NFC for short), 2G, 3G, 4G, NB-IOT, eMTC, or other 5G, etc., or one or a combination of more of them, is not limited herein. The corresponding communication component 305 may thus comprise: wi-Fi module, bluetooth module, NFC module, etc.
In an exemplary embodiment, the electronic device 300 may be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, abbreviated as ASIC), digital signal processors (Digital Signal Processor, abbreviated as DSP), digital signal processing devices (Digital Signal Processing Device, abbreviated as DSPD), programmable logic devices (Programmable Logic Device, abbreviated as PLD), field programmable gate arrays (Field Programmable Gate Array, abbreviated as FPGA), controllers, microcontrollers, microprocessors, or other electronic components for performing the above-described method of determining lane changes in a vehicle.
In another exemplary embodiment, a computer readable storage medium is also provided, comprising program instructions which, when executed by a processor, implement the steps of the above-described method of determining a lane change of a vehicle. For example, the computer readable storage medium may be the memory 302 including program instructions described above, which are executable by the processor 301 of the electronic device 300 to perform the method of determining lane changes of a vehicle described above.
In another exemplary embodiment, a computer program product is also provided, which comprises a computer program executable by a programmable apparatus, the computer program having code portions for performing the above-mentioned method of determining a lane change of a vehicle when being executed by the programmable apparatus.
The preferred embodiments of the present disclosure have been described in detail above with reference to the accompanying drawings, but the present disclosure is not limited to the specific details of the above embodiments, and various simple modifications may be made to the technical solutions of the present disclosure within the scope of the technical concept of the present disclosure, and all the simple modifications belong to the protection scope of the present disclosure.
In addition, the specific features described in the foregoing embodiments may be combined in any suitable manner, and in order to avoid unnecessary repetition, the present disclosure does not further describe various possible combinations.
Moreover, any combination between the various embodiments of the present disclosure is possible as long as it does not depart from the spirit of the present disclosure, which should also be construed as the disclosure of the present disclosure.
Claims (14)
1. A method for determining lane change of a vehicle, applied to a first vehicle, the method comprising:
acquiring a track data set and a relative motion data set, wherein the track data set comprises track data of a second vehicle acquired at a plurality of acquisition moments within a preset duration, the relative motion data set comprises relative motion data between the second vehicle and the first vehicle acquired at the plurality of acquisition moments within the preset duration, the plurality of acquisition moments comprise current acquisition moments, and the second vehicle is a vehicle which runs on an adjacent lane of the lane where the first vehicle is located and is positioned in front of the first vehicle;
Respectively inputting the track data set into a pre-trained track recognition model and a pre-trained track clustering model to determine track change probability according to a track recognition result output by the track recognition model and a track clustering result output by the track clustering model;
respectively inputting the relative motion data set into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model to determine the relative motion lane change probability according to the relative motion recognition result output by the relative motion recognition model and the relative motion clustering result output by the relative motion clustering model;
determining the lane change probability of the second vehicle according to the track lane change probability and the relative motion lane change probability;
and determining whether the second vehicle is about to change the lane according to the lane change probability and a preset lane change threshold value.
2. The method of claim 1, wherein the acquiring the trajectory dataset and the relative motion dataset comprises:
acquiring the track data and the relative motion data acquired at each acquisition time within the preset time, wherein the track data comprises: the lateral and longitudinal positions of the second vehicle, the relative motion data comprising: a relative longitudinal speed and a relative longitudinal distance of the first vehicle and the second vehicle;
Determining supplementary track data corresponding to each acquisition time according to the track data acquired at the acquisition time, wherein the supplementary track data comprises: a lateral speed and a lateral acceleration of the second vehicle;
taking the track data acquired at each acquisition moment and the corresponding supplementary track data as the track data set;
and processing the relative motion data acquired at each acquisition time according to a preset rule to obtain the relative motion data set.
3. The method of claim 1, wherein inputting the trajectory dataset into a pre-trained trajectory recognition model and a pre-trained trajectory clustering model, respectively, to determine a trajectory transition probability based on a trajectory recognition result output by the trajectory recognition model and a trajectory clustering result output by the trajectory clustering model, comprises:
inputting the track data set into the track recognition model to obtain the track recognition result, and inputting the track data set into the track clustering model to obtain a track clustering result, wherein the track recognition result is used for indicating straight running or lane changing, and the track clustering result is used for indicating the category of the track data collected at each collection moment;
If the track identification result indicates straight going, determining that the track lane change probability is zero;
and if the track identification result indicates track change, determining the track change probability according to the track clustering result.
4. The method of claim 3, wherein the trajectory recognition model is an Attention-LSTM model, and the inputting the trajectory dataset into the trajectory recognition model to obtain the trajectory recognition result comprises:
inputting the track data set into the Attention-LSTM model to acquire the track recognition result output by the Attention-LSTM model and the Attention value of each acquisition moment;
the determining the track lane-changing probability according to the track clustering result comprises the following steps:
determining the initial track lane change probability of the acquisition time according to the category of the track data acquired at each acquisition time and the corresponding relation between the preset category and the initial track lane change probability, wherein the category is included in the track clustering result;
and determining the track lane change probability according to the initial track lane change probability of each acquisition time and the attention value of each acquisition time.
5. The method of claim 1, wherein inputting the relative motion dataset into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model, respectively, to determine a relative motion lane-change probability based on the relative motion recognition result output by the relative motion recognition model and the relative motion clustering result output by the relative motion clustering model, comprises:
inputting the relative motion data set into the relative motion recognition model to obtain the relative motion recognition result, and inputting the relative motion data set into the relative motion clustering model to obtain a relative motion clustering result, wherein the relative motion recognition result is used for indicating straight running or lane changing, and the relative motion clustering result is used for indicating the category to which the relative motion data acquired at each acquisition moment belong;
if the relative motion recognition result indicates straight going, determining that the relative motion lane change probability is zero;
and if the relative motion recognition result indicates lane change, determining the relative motion lane change probability according to the relative motion clustering result.
6. The method of claim 5, wherein the relative motion recognition model is an Attention-GRU model, and wherein the inputting the relative motion dataset into the relative motion recognition model to obtain the relative motion recognition result comprises:
Inputting the relative motion data set into the Attention-GRU model to acquire the relative motion recognition result output by the Attention-GRU model and the Attention value of each acquisition time;
the determining the relative motion lane-change probability according to the relative motion clustering result comprises the following steps:
determining initial relative motion lane change probability at the acquisition time according to the category to which the relative motion data acquired at each acquisition time included in the relative motion clustering result belong and the corresponding relation between a preset category and the initial relative motion lane change probability;
determining the relative motion lane change probability according to the initial relative motion lane change probability of each acquisition time and the attention value of each acquisition time.
7. The method of any of claims 1-6, wherein the acquiring a trajectory dataset and a relative motion dataset comprises:
acquiring the track data and the relative motion data acquired at each acquisition time within a specified duration;
dividing the appointed time length into appointed number of sliding windows, wherein the length of each sliding window is the preset time length;
The track data and the relative motion data acquired in each sliding window are used as the track data set and the relative motion data set corresponding to the sliding window;
the determining whether the second vehicle is about to change lane according to the lane change probability and a preset lane change threshold value comprises:
the lane change probability determined according to the track data set and the relative motion data set corresponding to each sliding window is weighted and summed to determine total lane change probability, and the weight corresponding to each sliding window is inversely proportional to the distance between the sliding window and the current moment;
if the total lane change probability is greater than or equal to the lane change threshold, determining that the second vehicle is about to change lanes;
and if the total lane change probability is smaller than the lane change threshold, determining that the second vehicle is going straight.
8. The method of claim 7, wherein the trajectory recognition result is used to indicate straight, left lane change, or right lane change; after determining whether the second vehicle is about to change lane according to the lane change probability and a preset lane change threshold, the method further includes:
if the second vehicle is determined to be changed, counting a first number corresponding to a left lane change, a second number corresponding to a right lane change and a third number corresponding to a straight line in the track identification result determined according to the track data set corresponding to each sliding window, wherein the sum of the first number, the second number and the third number is equal to the appointed number;
If the first number is greater than the second number, determining that the second vehicle is about to lane-change left;
and if the first quantity is smaller than the second quantity, determining that the second vehicle is about to change lanes right.
9. The method according to any one of claims 1-6, wherein after the determining whether the second vehicle is about to lane change based on the lane change probability and a preset lane change threshold, the method further comprises:
if the second vehicle is determined to be lane-changing, determining lane-changing time of the second vehicle according to the transverse position, the transverse speed and the course angle of the second vehicle, which are included in the track data acquired at the current acquisition time;
determining termination track data of the second vehicle at the end of lane change according to the lane change time and the longitudinal speed included in the track data acquired at the current acquisition time;
and fitting according to the termination track data and the track data acquired at the current acquisition moment and a Bessel function to obtain the track changing track of the second vehicle.
10. The method of claim 1, wherein the trajectory recognition model is trained by:
Acquiring a first sample input set and a first sample output set, wherein each first sample input in the first sample input set comprises a plurality of sample track data, the first sample output set comprises first sample output corresponding to each first sample input, each first sample output comprises a sample track identification result marked by the corresponding plurality of sample track data, and the sample track identification result is used for indicating straight line or lane change;
taking the first sample input set as the input of the track recognition model, and taking the first sample output set as the output of the track recognition model so as to train the track recognition model;
the relative motion recognition model is trained by the following steps:
acquiring a second sample input set and a second sample output set, wherein each second sample input in the second sample input set comprises a plurality of sample relative motion data, each second sample output set comprises a second sample output corresponding to each second sample input, each second sample output comprises a sample relative motion recognition result marked by the corresponding plurality of sample relative motion data, and the sample relative motion recognition result is used for indicating straight going or lane changing;
And taking the second sample input set as the input of the relative motion recognition model, and taking the second sample output set as the output of the relative motion recognition model so as to train the relative motion recognition model.
11. The method of claim 10, wherein the trajectory recognition model is an LSTM model and the relative motion recognition model is a GRU model;
the taking the first sample input set as the input of the track recognition model and the first sample output set as the output of the track recognition model to train the track recognition model comprises:
inputting the first sample input set into an initial LSTM model according to a first input weight, and taking the first sample output set as the output of the initial LSTM model to train the initial LSTM model;
updating the first input weight according to the trained initial LSTM model;
repeatedly executing the step of inputting the first sample input set, inputting an initial LSTM model according to a first input weight, taking the first sample output set as the output of the initial LSTM model to train the initial LSTM model, and updating the first input weight according to the trained initial LSTM model, wherein the initial LSTM model and the first input weight obtained by executing a preset number of iterations are taken as the input weights corresponding to the track recognition model and the track recognition model;
The taking the second sample input set as the input of the relative motion recognition model and the second sample output set as the output of the relative motion recognition model to train the relative motion recognition model includes:
inputting the second sample input set into the initial GRU model according to a second input weight, and taking the second sample output set as the output of the initial GRU model to train the initial GRU model;
updating the second input weight according to the trained initial GRU model;
and repeatedly executing the step of inputting the second sample input set, inputting the initial GRU model according to a second input weight, taking the second sample output set as the output of the initial GRU model to train the initial GRU model until the step of updating the second input weight according to the trained initial GRU model, and taking the initial GRU model and the second input weight obtained by executing the iteration number as the input weights corresponding to the relative motion recognition model and the relative motion recognition model.
12. A lane change determination apparatus for a first vehicle, the apparatus comprising:
The track data set comprises track data of a second vehicle acquired at a plurality of acquisition moments within a preset duration, the relative motion data set comprises relative motion data between the second vehicle and the first vehicle acquired at the plurality of acquisition moments within the preset duration, the plurality of acquisition moments comprise current acquisition moments, and the second vehicle is a vehicle which runs on an adjacent lane of the lane where the first vehicle is located and is positioned in front of the first vehicle;
the first processing module is used for inputting the track data set into a pre-trained track recognition model and a pre-trained track clustering model respectively so as to determine track lane change probability according to a track recognition result output by the track recognition model and a track clustering result output by the track clustering model;
the second processing module is used for inputting the relative motion data set into a pre-trained relative motion recognition model and a pre-trained relative motion clustering model respectively so as to determine the relative motion lane change probability according to the relative motion recognition result output by the relative motion recognition model and the relative motion clustering result output by the relative motion clustering model;
The first determining module is used for determining the lane change probability of the second vehicle according to the track lane change probability and the relative motion lane change probability;
and the second determining module is used for determining whether the second vehicle is about to change the lane according to the lane change probability and a preset lane change threshold value.
13. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any one of claims 1-11.
14. An electronic device, comprising:
a memory having a computer program stored thereon;
a processor for executing the computer program in the memory to implement the steps of the method of any one of claims 1-11.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010889856.4A CN112085077B (en) | 2020-08-28 | 2020-08-28 | Method and device for determining lane change of vehicle, storage medium and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010889856.4A CN112085077B (en) | 2020-08-28 | 2020-08-28 | Method and device for determining lane change of vehicle, storage medium and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112085077A CN112085077A (en) | 2020-12-15 |
| CN112085077B true CN112085077B (en) | 2023-10-31 |
Family
ID=73729247
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010889856.4A Active CN112085077B (en) | 2020-08-28 | 2020-08-28 | Method and device for determining lane change of vehicle, storage medium and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112085077B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113044042B (en) * | 2021-06-01 | 2021-09-21 | 禾多科技（北京）有限公司 | Vehicle predicted lane change image display method and device, electronic equipment and readable medium |
| CN113386775B (en) * | 2021-06-16 | 2022-06-17 | 杭州电子科技大学 | Driver intention identification method considering human-vehicle-road characteristics |
| CN114019497B (en) * | 2022-01-05 | 2022-03-18 | 南京楚航科技有限公司 | Target lane change identification method based on millimeter wave radar variance statistics |
| CN114771539B (en) * | 2022-06-16 | 2023-02-28 | 小米汽车科技有限公司 | Vehicle lane change decision method and device, storage medium and vehicle |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102855638A (en) * | 2012-08-13 | 2013-01-02 | 苏州大学 | Detection method for abnormal behavior of vehicle based on spectrum clustering |
| CN103605362A (en) * | 2013-09-11 | 2014-02-26 | 天津工业大学 | Learning and anomaly detection method based on multi-feature motion modes of vehicle traces |
| CN111079590A (en) * | 2019-12-04 | 2020-04-28 | 东北大学 | Peripheral vehicle behavior pre-judging method of unmanned vehicle |
| CN111104969A (en) * | 2019-12-04 | 2020-05-05 | 东北大学 | Method for pre-judging collision possibility between unmanned vehicle and surrounding vehicle |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100023265A1 (en) * | 2008-07-24 | 2010-01-28 | Gm Global Technology Operations, Inc. | Adaptive vehicle control system with integrated driving style recognition |
| US8260515B2 (en) * | 2008-07-24 | 2012-09-04 | GM Global Technology Operations LLC | Adaptive vehicle control system with driving style recognition |
-
2020
- 2020-08-28 CN CN202010889856.4A patent/CN112085077B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102855638A (en) * | 2012-08-13 | 2013-01-02 | 苏州大学 | Detection method for abnormal behavior of vehicle based on spectrum clustering |
| CN103605362A (en) * | 2013-09-11 | 2014-02-26 | 天津工业大学 | Learning and anomaly detection method based on multi-feature motion modes of vehicle traces |
| CN111079590A (en) * | 2019-12-04 | 2020-04-28 | 东北大学 | Peripheral vehicle behavior pre-judging method of unmanned vehicle |
| CN111104969A (en) * | 2019-12-04 | 2020-05-05 | 东北大学 | Method for pre-judging collision possibility between unmanned vehicle and surrounding vehicle |
Non-Patent Citations (3)
| Title |
|---|
| 结构化道路中动态车辆的轨迹预测;谢辉;高斌;熊硕;王悦;;汽车安全与节能学报(04);全文 * |
| 车辆再识别技术综述;刘凯;李浥东;林伟鹏;;智能科学与技术学报(01);全文 * |
| 驾驶员避撞转向行为的改进K-means聚类与识别;赵治国;冯建翔;周良杰;王凯;胡昊锐;张海山;宁忠麟;;汽车工程(01);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112085077A (en) | 2020-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112085077B (en) | Method and device for determining lane change of vehicle, storage medium and electronic equipment | |
| Hou et al. | Interactive trajectory prediction of surrounding road users for autonomous driving using structural-LSTM network | |
| US11480972B2 (en) | Hybrid reinforcement learning for autonomous driving | |
| Zhang et al. | Vehicle motion prediction at intersections based on the turning intention and prior trajectories model | |
| CN111459168B (en) | Fused automatic-driving automobile pedestrian crossing track prediction method and system | |
| Altché et al. | An LSTM network for highway trajectory prediction | |
| Jeong et al. | Surround vehicle motion prediction using LSTM-RNN for motion planning of autonomous vehicles at multi-lane turn intersections | |
| US9784592B2 (en) | Turn predictions | |
| KR20180068511A (en) | Apparatus and method for generating training data for training neural network determining information related to road included in an image | |
| US11782158B2 (en) | Multi-stage object heading estimation | |
| CN111284485B (en) | Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium | |
| US11501449B2 (en) | Method for the assessment of possible trajectories | |
| Schulz et al. | Learning interaction-aware probabilistic driver behavior models from urban scenarios | |
| Song et al. | Multi-vehicle tracking using microscopic traffic models | |
| Jeong et al. | Bidirectional long shot-term memory-based interactive motion prediction of cut-in vehicles in urban environments | |
| Jin et al. | Driver intention recognition based on continuous hidden Markov model | |
| Wirthmüller et al. | Predicting the time until a vehicle changes the lane using LSTM-based recurrent neural networks | |
| Li et al. | Development and evaluation of two learning-based personalized driver models for pure pursuit path-tracking behaviors | |
| CN112833903B (en) | Track prediction method, device, equipment and computer readable storage medium | |
| Amsalu et al. | Driver behavior modeling near intersections using hidden Markov model based on genetic algorithm | |
| CN114127810A (en) | Vehicle autonomous level function | |
| Toledo-Moreo et al. | Maneuver prediction for road vehicles based on a neuro-fuzzy architecture with a low-cost navigation unit | |
| WO2022035602A2 (en) | Agent trajectory prediction using target locations | |
| Hori et al. | Driver confusion status detection using recurrent neural networks | |
| Sharma et al. | Highway lane-changing prediction using a hierarchical software architecture based on support vector machine and continuous hidden markov model |
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 |