CN114701509B - Driving intention recognition method and system for mixed traffic flow - Google Patents

Abstract

The invention discloses a driving intention recognition method and a system for mixed traffic flow, wherein the method comprises the following steps: acquiring traffic information; wherein the traffic information includes state information of the target vehicle and surrounding vehicles, and road information of the target vehicle; the status information includes a position of the vehicle and a speed of the vehicle; preprocessing the acquired traffic information; the preprocessing comprises data cleaning, data extraction, feature expansion and data standardization; predicting the driving intention of the target vehicle by using a preset driving intention recognition model based on the preprocessed traffic information; the driving intention recognition model is a deep neural network model; outputting the driving intention category with the maximum probability according to the probability of each driving intention; wherein the driving intention category includes lane change to the left, lane keeping, and lane change to the right. The invention can realize the classification of the driving intention of the human driver.

Description

A driving intention recognition method and system for mixed traffic flow

Technical Field

The present invention relates to the technical field of intelligent connected vehicles, and in particular to a driving intention recognition method and system for mixed traffic flows.

Background Art

With the vigorous development of driverless and intelligent connected vehicle technologies, mixed traffic flows consisting of different proportions of manually driven vehicles, intelligent vehicles and connected vehicles will exist for a long time in the foreseeable future traffic scenarios. Due to the poor information accessibility and strong randomness of driving behavior of manual drivers, the driving safety and traffic efficiency of intelligent vehicles and connected vehicles are seriously restricted, which greatly affects the further development of intelligent transportation. Identifying the driving intention of manual drivers is an effective way to solve the above problems.

In general, the research on driving intention recognition develops models based on vehicle operating parameters (such as accelerator pedal position, brake pedal opening and driving speed), environmental parameters (such as road curvature, lane position and road signs) and driver visual behavior characteristics (such as facial direction and line of sight). Depending on the available feature parameters, the recognition effect varies to a certain extent, and the recognition accuracy of existing recognition models is not ideal enough.

Summary of the invention

The present invention provides a driving intention recognition method and system for mixed traffic flow, so as to solve the technical problem that the recognition accuracy of the existing recognition model is not ideal enough.

In order to solve the above technical problems, the present invention provides the following technical solutions:

In one aspect, the present invention provides a driving intention recognition method for mixed traffic flow, comprising:

Acquire traffic information; wherein the traffic information includes status information of the target vehicle and surrounding vehicles, and road information of the target vehicle; the surrounding vehicles refer to vehicles within a preset range of the target vehicle; the status information includes the position of the vehicle and the speed of the vehicle;

Preprocessing the acquired traffic information; wherein the preprocessing includes data cleaning, data extraction, feature expansion and data standardization;

Based on the preprocessed traffic information, a preset driving intention recognition model is used to predict the driving intention of the target vehicle; wherein the driving intention recognition model is a deep neural network model;

According to the probability of each driving intention, the driving intention category with the greatest probability is output; wherein the driving intention category includes left lane change, lane keeping and right lane change.

Furthermore, the preprocessing of the acquired traffic information includes:

Perform data cleaning on the acquired traffic information, delete unnecessary data dimensions and remove noise;

Extract the vehicle information and historical status around the target vehicle based on the cleaned data;

Mining deep features for feature expansion based on the extracted current information;

Perform data standardization on the data that has completed feature expansion.

Furthermore, the data input of the driving intention recognition model incorporates information interaction between the target vehicle and surrounding vehicles, road information, and current and historical vehicle status information of the target vehicle;

Specifically, the input data is as follows:

I＝[ _VE , _VS ,R]

Among them, _VE is the state information of the target vehicle, _VS is the state information of the surrounding vehicles, and R is the road information; among them, the state information of the target vehicle is as follows:

_VE ＝[ _S1 , _S2 ,…, _ST ]

S _i =[x _i , y _i , _vi ], i = 1, 2,...,T

Wherein, T is the preset historical time domain, S _i is the state information of the target vehicle at time i, x _i , y _i , _vi are the horizontal coordinate, vertical coordinate and absolute speed of the target vehicle at time i respectively;

The status information of surrounding vehicles is as follows:

V _S = [V _S1 , V _S2 , V _S3 , V _S4 , V _S5 , V _S6 ]

Among them, V _S1 , V _S2 , V _S3 , V _S4 , V _S5 , and V _S6 are the status information of the left front, left rear, front, rear, right front, and right rear vehicles of the target vehicle respectively; is the status information of vehicle i at time j, are the lateral and longitudinal coordinates of vehicle i relative to the target vehicle, is the absolute speed of vehicle i;

The road information R is as follows:

R＝[L _l ,L _r ]

Among them, L _l and L _r represent the left and right lane marking positions of the target vehicle respectively.

Furthermore, the preprocessing of the acquired traffic information is to process the data input into the driving intention recognition model into a 107-dimensional sequence; wherein the 1st to 15th dimensions are the current and historical status information of the target vehicle, the 16th to 105th dimensions are the current and historical status information of the surrounding vehicles, and the 106th to 107th dimensions are the left and right lane information flags.

Furthermore, the driving intention recognition model consists of 1 input layer, 4 hidden layers and 1 output layer, the activation function is elu, the loss function is cross entropy, the optimizer is Adam, and the learning rate is 0.001.

On the other hand, the present invention also provides a driving intention recognition system for mixed traffic flow, comprising:

A data receiving module, used to obtain traffic information; wherein the traffic information includes status information of the target vehicle and surrounding vehicles, and road information of the target vehicle; the surrounding vehicles refer to vehicles within a preset range of the target vehicle; the status information includes the position of the vehicle and the speed of the vehicle;

A data preprocessing module, used for preprocessing the traffic information acquired by the data receiving module; wherein the preprocessing includes data cleaning, data extraction, feature expansion and data standardization;

A driving intention recognition module, used to predict the driving intention of the target vehicle based on the traffic information preprocessed by the data preprocessing module using a preset driving intention recognition model; wherein the driving intention recognition model is a deep neural network model;

The output module is used to output the driving intention category with the highest probability according to the probability of each driving intention; wherein the driving intention category includes changing lanes to the left, lane keeping and changing lanes to the right.

Furthermore, the data preprocessing module is specifically used for:

Perform data cleaning on the acquired traffic information, delete unnecessary data dimensions and remove noise;

Extract the vehicle information and historical status around the target vehicle based on the cleaned data;

Mining deep features for feature expansion based on the extracted current information;

Perform data standardization on the data that has completed feature expansion.

Furthermore, the data input of the driving intention recognition model incorporates information interaction between the target vehicle and surrounding vehicles, road information, and current and historical vehicle status information of the target vehicle;

Specifically, the input data is as follows:

I＝[ _VE , _VS ,R]

Among them, _VE is the state information of the target vehicle, _VS is the state information of the surrounding vehicles, and R is the road information; among them, the state information of the target vehicle is as follows:

_VE ＝[ _S1 , _S2 ,…, _ST ]

S _i =[x _i , y _i , _vi ], i = 1, 2,...,T

Wherein, T is the preset historical time domain, _Si is the state information of the target vehicle at time i, _xi , _yi , _vi are the horizontal coordinate, longitudinal coordinate and absolute speed of the target vehicle at time i respectively;

The status information of surrounding vehicles is as follows:

V _S = [V _S1 , V _S2 , V _S3 , V _S4 , V _S5 , V _S6 ]

Among them, V _S1 , V _S2 , V _S3 , V _S4 , V _S5 , and V _S6 are the status information of the left front, left rear, front, rear, right front, and right rear vehicles of the target vehicle respectively; is the status information of vehicle i at time j, are the lateral and longitudinal coordinates of vehicle i relative to the target vehicle, is the absolute speed of vehicle i;

The road information R is as follows:

R＝[L _l ,L _r ]

Among them, L _l and L _r represent the left and right lane marking positions of the target vehicle respectively.

Furthermore, the data preprocessing module processes the data input into the driving intention recognition model into a 107-dimensional sequence; wherein the 1st to 15th dimensions are the current and historical status information of the target vehicle, the 16th to 105th dimensions are the current and historical status information of the surrounding vehicles, and the 106th to 107th dimensions are the left and right lane information flags.

Furthermore, the driving intention recognition model consists of 1 input layer, 4 hidden layers and 1 output layer, the activation function is elu, the loss function is cross entropy, the optimizer is Adam, and the learning rate is 0.001.

On the other hand, the present invention further provides an electronic device, comprising a processor and a memory; wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the above method.

In yet another aspect, the present invention further provides a computer-readable storage medium, wherein the storage medium stores at least one instruction, and the instruction is loaded and executed by a processor to implement the above method.

The beneficial effects brought about by the technical solution provided by the present invention include at least:

The technical solution provided by the present invention obtains traffic information; pre-processes the obtained traffic information; predicts the driving intention of the target vehicle based on the pre-processed traffic information using a preset driving intention recognition model; and outputs the driving intention category with the highest probability (left lane change, lane keeping, right lane change) according to the probability of each driving intention. Thus, the classification of the driving intention of human drivers can be realized.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

FIG1 is a schematic diagram of an execution flow of a driving intention recognition method for mixed traffic flow provided by an embodiment of the present invention;

FIG2 is a schematic diagram of surrounding vehicle information provided by an embodiment of the present invention;

FIG3 is a schematic diagram of a driving intention recognition framework provided by an embodiment of the present invention.

DETAILED DESCRIPTION

In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

First embodiment

This embodiment provides a driving intention recognition method for mixed traffic flow, which can be implemented by an electronic device. The execution flow of the method is shown in Figure 1, and includes the following steps:

S1, obtain traffic information;

Among them, the traffic information includes status information of the target vehicle and surrounding vehicles, and environmental information (road information) of the target vehicle; the surrounding vehicles refer to vehicles within a preset range of the target vehicle; the status information includes the location of the vehicle and the speed of the vehicle.

S2, preprocessing the acquired traffic information;

Wherein, the preprocessing includes data cleaning, data extraction, feature expansion and data standardization;

Specifically, in this embodiment, the preprocessing process of S2 is as follows:

First, the acquired traffic information is cleaned to delete unnecessary data dimensions and remove noise points;

Then, based on the cleaned data, the vehicle information and historical status around the target vehicle are extracted;

Then, based on the extracted current information, deep features are mined for feature expansion;

Finally, the data with completed feature expansion is standardized.

S3, based on the preprocessed traffic information, using a preset driving intention recognition model to predict the driving intention of the target vehicle; wherein the driving intention recognition model is a deep neural network model DNN;

It should be noted that in actual traffic scenarios, driving intention is the result of the combined effects of environmental information, surrounding vehicle information, and the state of the own vehicle (target vehicle).

Therefore, the data input of the driving intention recognition model proposed in this embodiment integrates the information interaction between the vehicle and surrounding vehicles, road information, and the current and historical status of the vehicle. Specifically, the input data is as follows:

I＝[ _VE , _VS ,R]

Among them, _VE is the state information of the target vehicle, _VS is the state information of the surrounding vehicles, and R is the road information; among them, the state information of the target vehicle is as follows:

_VE ＝[ _S1 , _S2 ,…, _ST ]

S _i =[x _i ,y _i ,v _i ](i=1,2,…,T)

Wherein, T is the preset historical time domain, _Si is the state information of the target vehicle at time i, _xi , _yi , _vi are the horizontal coordinate, longitudinal coordinate and absolute speed of the target vehicle at time i respectively;

The status information of surrounding vehicles is as follows:

V _S = [V _S1 , V _S2 , V _S3 , V _S4 , V _S5 , V _S6 ]

As shown in FIG2 , in this embodiment, _VS1 , _VS2 , _VS3 , _VS4 , _VS5 , and _VS6 are the status information of the left front, left rear, front, rear, right front, and right rear vehicles of the target vehicle respectively; is the status information of vehicle i at time j, are the lateral and longitudinal coordinates of vehicle i relative to the target vehicle, is the absolute speed of vehicle i;

The road information R is as follows:

R＝[L _l ,L _r ]

Among them, L _l and L _r represent the left and right lane markings of the target vehicle respectively.

Based on the above, the driving intention recognition model of this embodiment takes the historical information of the target vehicle and the interactive information of the environment vehicles as input; and predicts the driving intention of the target vehicle based on the input environment and historical information.

The above S2 processes the input data of the model into a 107-dimensional sequence, and the selected historical time domain is 5, among which the 1st to 15th dimensions are the current and historical status information of the target vehicle, the 16th to 105th dimensions are the current and historical status information of the vehicles around the target vehicle, and the 106th to 107th dimensions are the left and right lane information flags.

Furthermore, as shown in FIG3 , the driving intention recognition model proposed in this embodiment consists of 1 input layer, 4 hidden layers and 1 output layer, the activation function is elu, the loss function is cross entropy, the optimizer is Adam, and the learning rate is 0.001.

S4, outputting a driving intention category with the greatest probability according to the probability of each driving intention; wherein the driving intention category includes changing lanes to the left, lane keeping, and changing lanes to the right.

In summary, the method provided in this embodiment obtains traffic information; pre-processes the obtained traffic information; predicts the driving intention of the target vehicle based on the pre-processed traffic information using a preset driving intention recognition model; and outputs the driving intention category with the highest probability (left lane change, lane keeping, right lane change) according to the probability of each driving intention. Thus, the classification of the driving intention of human drivers can be achieved.

Second embodiment

This embodiment provides a driving intention recognition system for mixed traffic flow, which includes:

A data receiving module, used to obtain traffic information; wherein the traffic information includes status information of the target vehicle and surrounding vehicles, and road information of the target vehicle; the surrounding vehicles refer to vehicles within a preset range of the target vehicle; the status information includes the position of the vehicle and the speed of the vehicle;

A data preprocessing module, used for preprocessing the traffic information acquired by the data receiving module; wherein the preprocessing includes data cleaning, data extraction, feature expansion and data standardization;

A driving intention recognition module, used to predict the driving intention of the target vehicle based on the traffic information preprocessed by the data preprocessing module using a preset driving intention recognition model; wherein the driving intention recognition model is a deep neural network model;

The output module is used to output the driving intention category with the highest probability according to the probability of each driving intention; wherein the driving intention category includes changing lanes to the left, lane keeping and changing lanes to the right.

The driving intention recognition system for mixed traffic flow of the present embodiment corresponds to the driving intention recognition method for mixed traffic flow of the above-mentioned first embodiment; wherein, the functions implemented by each functional module in the driving intention recognition system for mixed traffic flow of the present embodiment correspond one-to-one to each process step in the driving intention recognition method for mixed traffic flow of the above-mentioned first embodiment; therefore, they will not be repeated here.

Third embodiment

This embodiment provides an electronic device, which includes a processor and a memory; wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the method of the first embodiment.

The electronic device may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) and one or more memories, wherein the memory stores at least one instruction, and the instruction is loaded by the processor to execute the above method.

Fourth embodiment

This embodiment provides a computer-readable storage medium, which stores at least one instruction, and the instruction is loaded and executed by a processor to implement the method of the first embodiment. The computer-readable storage medium may be a ROM, a random access memory, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc. The instructions stored therein may be loaded by a processor in a terminal to execute the method.

In addition, it should be noted that the present invention can be provided as a method, an apparatus or a computer program product. Therefore, the embodiments of the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code.

The embodiments of the present invention are described with reference to the flowcharts and/or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and/or box in the flowchart and/or block diagram, as well as the combination of the processes and/or boxes in the flowchart and/or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate a device for implementing the functions specified in one process or multiple processes in the flowchart and/or one box or multiple boxes in the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes of the flowchart and/or one or more blocks of the block diagram. These computer program instructions may also be loaded onto a computer or other programmable data processing terminal device, so that a series of operation steps are performed on the computer or other programmable terminal device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable terminal device provide steps for implementing the functions specified in one or more processes of the flowchart and/or one or more blocks of the block diagram.

It should also be noted that, in this article, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

Finally, it should be noted that the above is a preferred embodiment of the present invention. It should be pointed out that although the preferred embodiment of the present invention has been described, for those skilled in the art, once the basic creative concept of the present invention is known, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention. Therefore, the attached claims are intended to be interpreted as including the preferred embodiment and all changes and modifications that fall within the scope of the embodiments of the present invention.

Claims (2)

1. A driving intention recognition method for a mixed traffic flow, comprising:

acquiring traffic information; wherein the traffic information includes state information of a target vehicle and surrounding vehicles, and road information of the target vehicle; the surrounding vehicles refer to vehicles within a preset range of the target vehicle; the status information includes a position of the vehicle and a speed of the vehicle;

preprocessing the acquired traffic information; the preprocessing comprises data cleaning, data extraction, feature expansion and data standardization;

Predicting the driving intention of the target vehicle by using a preset driving intention recognition model based on the preprocessed traffic information; the driving intention recognition model is a deep neural network model;

Outputting the driving intention category with the maximum probability according to the probability of each driving intention; wherein the driving intention category includes lane changing to the left, lane keeping, and lane changing to the right;

the preprocessing of the acquired traffic information comprises:

data cleaning is carried out on the acquired traffic information, the dimension of unnecessary data is deleted, and noise points are removed;

Extracting vehicle information and a history state around the target vehicle based on the cleaned data;

Mining deep features according to the extracted current information to perform feature expansion;

carrying out data standardization processing on the data with the characteristics expanded;

the data input of the driving intention recognition model fuses information interaction of the target vehicle and surrounding vehicles, road information and current and historical vehicle state information of the target vehicle; specifically, the input data is as follows:

I＝[V_E,V_S,R]

Wherein V _E is the state information of the target vehicle, V _S is the state information of surrounding vehicles, and R is road information; wherein, the state information of the target vehicle is as follows:

V_E＝[S₁,S₂,…,S_T]

S_i＝[x_i,y_i,v_i]，i＝1,2,…,T

Wherein T is a preset historical time domain, S _i is state information of the target vehicle at the moment i, and x _i,y_i,v_i is a transverse coordinate, a longitudinal coordinate and an absolute speed of the target vehicle at the moment i respectively;

The status information of the surrounding vehicles is as follows:

V_S＝[V_S1,V_S2,V_S3,V_S4,V_S5,V_S6]

Wherein V _S1、V_S2、V_S3、V_S4、V_S5、V_S6 is the state information of the left front, left rear, front, rear, front right and rear right vehicles of the target vehicle, respectively; As the status information of the vehicle i at the moment j, The lateral and longitudinal coordinates of the vehicle i relative to the target vehicle,Is the absolute speed of vehicle i;

the road information R is as follows:

R＝[L_l,L_r]

Wherein, L _l and L _r respectively represent left and right lane marker bits of the target vehicle;

the preprocessing of the acquired traffic information is to process the data input into the driving intention recognition model into a 107-dimensional sequence; the 1 st to 15 th dimensions are the current and historical state information of the target vehicle, the 16 th to 105 th dimensions are the current and historical state information of surrounding vehicles, and the 106 th to 107 th dimensions are the left and right lane marker bits;

the driving intention recognition model consists of 1 input layer, 4 hidden layers and 1 output layer, the activation function is elu, the loss function is cross entropy, the optimizer is Adam, and the learning rate is 0.001.

2. A driving intention recognition system for a mixed traffic stream, comprising:

The data receiving module is used for acquiring traffic information; wherein the traffic information includes state information of a target vehicle and surrounding vehicles, and road information of the target vehicle; the surrounding vehicles refer to vehicles within a preset range of the target vehicle; the status information includes a position of the vehicle and a speed of the vehicle;

The data preprocessing module is used for preprocessing the traffic information acquired by the data receiving module; the preprocessing comprises data cleaning, data extraction, feature expansion and data standardization;

The driving intention recognition module is used for predicting the driving intention of the target vehicle by using a preset driving intention recognition model based on the traffic information preprocessed by the data preprocessing module; the driving intention recognition model is a deep neural network model;

the output module is used for outputting the driving intention category with the maximum probability according to the probability of each driving intention; wherein the driving intention category includes lane changing to the left, lane keeping, and lane changing to the right;

The data preprocessing module is specifically used for:

data cleaning is carried out on the acquired traffic information, the dimension of unnecessary data is deleted, and noise points are removed;

Extracting vehicle information and a history state around the target vehicle based on the cleaned data;

Mining deep features according to the extracted current information to perform feature expansion;

carrying out data standardization processing on the data with the characteristics expanded;

the data input of the driving intention recognition model fuses information interaction of the target vehicle and surrounding vehicles, road information and current and historical vehicle state information of the target vehicle; specifically, the input data is as follows:

I＝[V_E,V_S,R]

Wherein V _E is the state information of the target vehicle, V _S is the state information of surrounding vehicles, and R is road information; wherein, the state information of the target vehicle is as follows:

V_E＝[S₁,S₂,…,S_T]

S_i＝[x_i,y_i,v_i]，i＝1,2,…,T

Wherein T is a preset historical time domain, S _i is state information of the target vehicle at the moment i, and x _i,y_i,v_i is a transverse coordinate, a longitudinal coordinate and an absolute speed of the target vehicle at the moment i respectively;

The status information of the surrounding vehicles is as follows:

V_S＝[V_S1,V_S2,V_S3,V_S4,V_S5,V_S6]

Wherein V _S1、V_S2、V_S3、V_S4、V_S5、V_S6 is the state information of the left front, left rear, front, rear, front right and rear right vehicles of the target vehicle, respectively; As the status information of the vehicle i at the moment j, The lateral and longitudinal coordinates of the vehicle i relative to the target vehicle,Is the absolute speed of vehicle i;

the road information R is as follows:

R＝[L_l,L_r]

Wherein, L _l and L _r respectively represent left and right lane marker bits of the target vehicle;

The data preprocessing module processes the data input into the driving intention recognition model into a 107-dimensional sequence; the 1 st to 15 th dimensions are the current and historical state information of the target vehicle, the 16 th to 105 th dimensions are the current and historical state information of surrounding vehicles, and the 106 th to 107 th dimensions are the left and right lane marker bits;

the driving intention recognition model consists of 1 input layer, 4 hidden layers and 1 output layer, the activation function is elu, the loss function is cross entropy, the optimizer is Adam, and the learning rate is 0.001.