Lane keeping method, apparatus, device, medium, and system

文档序号：1882885 发布日期：2021-11-26 浏览：30次中文

阅读说明：本技术 车道保持方法、装置、设备、介质及系统 (Lane keeping method, apparatus, device, medium, and system ) 是由李伟男刘斌吴杭哲刘枫于 2021-09-23 设计创作，主要内容包括：本发明实施例公开了一种车道保持方法、装置、设备、介质及系统。该方法包括：获取目标车辆的车辆行驶数据,将包括转向盘转角、车辆横向位置以及车辆横向加速度的车辆行驶数据输入至预先训练的驾驶风格辨识模型中,得到驾驶风格辨识模型输出的目标车辆对应的驾驶风格类别,确定出用户的驾驶风格,进而根据驾驶风格类别和目标车辆的车辆状态数据,确定目标车辆的目标转向盘转角,基于该目标转向盘转角调整目标车辆的转向盘转角,以控制车辆回归至车道中心线位置,实现了基于用户驾驶风格的车道保持控制,满足了不同驾驶习惯的驾驶人的差异化需求,极大地提高了用户的驾驶体验,并且,提高了车辆的系统适用性和行驶安全性。(The embodiment of the invention discloses a lane keeping method, a lane keeping device, lane keeping equipment, lane keeping media and a lane keeping system. The method comprises the following steps: the method comprises the steps of obtaining vehicle driving data of a target vehicle, inputting the vehicle driving data comprising a steering wheel corner, a vehicle transverse position and vehicle transverse acceleration into a pre-trained driving style identification model, obtaining a driving style type corresponding to the target vehicle output by the driving style identification model, determining a driving style of a user, further determining the target steering wheel corner of the target vehicle according to the driving style type and vehicle state data of the target vehicle, adjusting the steering wheel corner of the target vehicle based on the target steering wheel corner to control the vehicle to return to a lane central line position, achieving lane keeping control based on the driving style of the user, meeting the differentiated requirements of drivers with different driving habits, greatly improving the driving experience of the user, and improving the system applicability and driving safety of the vehicle.)

1. A lane keeping method, characterized in that the method comprises:

acquiring vehicle running data of a target vehicle, wherein the vehicle running data comprises a steering wheel angle, a vehicle transverse position and a vehicle transverse acceleration;

determining a driving style category corresponding to the target vehicle based on the vehicle driving data and a pre-trained driving style identification model;

determining a target steering wheel angle of the target vehicle based on the driving style category and the vehicle state data of the target vehicle, and adjusting the steering wheel angle of the target vehicle based on the target steering wheel angle.

2. The method of claim 1, further comprising:

acquiring a sample input vector and a driving style label corresponding to the sample input vector, and inputting the sample input vector to an input layer of a learning vector quantization neural network;

calculating the distance between the sample input vector and each competition layer neuron of the learning vector quantization neural network, and determining a first target neuron and a second target neuron which are closest to the sample input vector in each competition layer neuron;

updating a weight value corresponding to the first target neuron and/or the second target neuron in the learning vector quantization neural network based on a distance between the first target neuron and the sample input vector, a distance between the second target neuron and the sample input vector, a driving style predicted value corresponding to the first target neuron, a driving style predicted value corresponding to the second target neuron and a driving style label corresponding to the sample input vector;

and determining the learning vector quantization neural network as a driving style identification model.

3. The method of claim 2, wherein updating the weight corresponding to the first target neuron and/or the second target neuron in the learning vector quantization neural network based on the distance of the first target neuron from the sample input vector, the distance of the second target neuron from the sample input vector, the driving style prediction value corresponding to the first target neuron, the driving style prediction value corresponding to the second target neuron, and the driving style label corresponding to the sample input vector comprises:

if yes, updating the weight values corresponding to the first target neuron and the second target neuron in the learning vector quantization neural network based on a preset learning rate, the sample input vector and the driving style label.

4. The method of claim 1, wherein the vehicle state data includes a current vehicle speed, a lane centerline, a current position lateral coordinate, a vehicle steering system gear ratio, and a vehicle fore-aft wheelbase, and wherein the target steering wheel angle of the target vehicle is determined based on the driving style category and the vehicle state data of the target vehicle, satisfying the following equation:

5. The method of claim 1, wherein after the adjusting the steering wheel angle of the target vehicle based on the target steering wheel angle, the method further comprises:

acquiring a vehicle transverse position and a lane center line transverse position corresponding to each acquisition position;

determining whether the target vehicle regresses a lane center line based on each of the vehicle lateral positions, each of the lane center line lateral positions, and the driving style category.

6. The method of claim 5, wherein determining whether the target vehicle regresses a lane centerline based on each of the vehicle lateral positions, each of the lane centerline lateral positions, and the driving style category comprises:

and determining whether the target vehicle regresses the lane center line or not based on the polynomial coefficient and the driving style category.

7. A lane keeping apparatus, characterized in that the apparatus comprises:

the system comprises a running data acquisition module, a data processing module and a data processing module, wherein the running data acquisition module is used for acquiring vehicle running data of a target vehicle, and the vehicle running data comprises a steering wheel angle, a vehicle transverse position and a vehicle transverse acceleration;

the driving style identification module is used for determining the driving style category corresponding to the target vehicle based on the vehicle driving data and a driving style identification model trained in advance;

and the steering wheel adjusting module is used for determining a target steering wheel angle of the target vehicle based on the driving style category and the vehicle state data of the target vehicle, and adjusting the steering wheel angle of the target vehicle based on the target steering wheel angle.

8. A lane keeping system, characterized in that the system comprises an industrial personal computer and a steering wheel assembly, wherein,

the steering wheel assembly is used for acquiring a steering wheel corner of a target vehicle and sending the steering wheel corner to the industrial personal computer;

the industrial personal computer is used for adjusting the steering wheel angle of the target vehicle based on the lane keeping method of any one of claims 1 to 6.

9. An electronic device, characterized in that the electronic device comprises:

one or more processors;

a storage device for storing one or more programs,

when executed by the one or more processors, cause the one or more processors to implement the lane keeping method of any of claims 1-6.

10. A computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, carries out the lane keeping method according to any one of claims 1-6.

Technical Field

The embodiment of the invention relates to the technical field of vehicles, in particular to a lane keeping method, a lane keeping device, lane keeping equipment, lane keeping media and a lane keeping system.

Background

In recent years, with the development of smart cars, more and more researchers have started to focus on lane keeping control of vehicles. Here, the lane keeping control may be understood as keeping the vehicle running in the middle of the lane.

However, in the prior art, the measured real-time lateral offset of the vehicle relative to the middle position of the lane is usually adopted to control the vehicle to keep running in the middle of the lane, the driving style of a driver is not considered, and the differentiated driving requirements of different driving styles cannot be met.

Disclosure of Invention

Embodiments of the present invention provide a lane keeping method, apparatus, device, medium, and system, so as to implement lane keeping control based on a driving style of a user, and improve driving experience of the user.

In a first aspect, an embodiment of the present invention provides a lane keeping method, including:

acquiring vehicle running data of a target vehicle, wherein the vehicle running data comprises a steering wheel angle, a vehicle transverse position and a vehicle transverse acceleration;

determining a driving style category corresponding to the target vehicle based on the vehicle driving data and a pre-trained driving style identification model;

Optionally, the method further includes:

and determining the learning vector quantization neural network as a driving style identification model.

Optionally, the updating, based on the distance between the first target neuron and the sample input vector, the distance between the second target neuron and the sample input vector, the driving style prediction value corresponding to the first target neuron, the driving style prediction value corresponding to the second target neuron, and the driving style label corresponding to the sample input vector, the weight value corresponding to the first target neuron and/or the second target neuron in the learning vector quantization neural network includes:

Optionally, the vehicle state data includes a current vehicle speed, a lane center line, a current position lateral coordinate, a vehicle steering system transmission ratio, and a vehicle front-rear axle distance, and the target steering wheel angle of the target vehicle is determined based on the driving style category and the vehicle state data of the target vehicle, and satisfies the following formula:

Optionally, after the adjusting the steering wheel angle of the target vehicle based on the target steering wheel angle, the method further includes:

acquiring a vehicle transverse position and a lane center line transverse position corresponding to each acquisition position;

determining whether the target vehicle regresses a lane center line based on each of the vehicle lateral positions, each of the lane center line lateral positions, and the driving style category.

Optionally, the determining whether the target vehicle returns to the lane center line based on each of the vehicle lateral positions, each of the lane center line lateral positions, and the driving style category includes:

and determining whether the target vehicle regresses the lane center line or not based on the polynomial coefficient and the driving style category.

In a second aspect, an embodiment of the present invention further provides a lane keeping apparatus, including:

In a third aspect, an embodiment of the present invention further provides a lane keeping system, where the system includes an industrial personal computer and a steering wheel assembly,

the steering wheel assembly is used for acquiring a steering wheel corner of a target vehicle and sending the steering wheel corner to the industrial personal computer;

the industrial personal computer is used for adjusting the steering wheel angle of the target vehicle based on the lane keeping method provided by any embodiment of the invention.

In a fourth aspect, an embodiment of the present invention further provides an electronic device, where the electronic device includes:

one or more processors;

a storage device for storing one or more programs,

when executed by the one or more processors, cause the one or more processors to implement a lane keeping method as provided by any of the embodiments of the invention.

In a fifth aspect, embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements a lane keeping method as provided in any of the embodiments of the present invention.

The embodiment of the invention has the following advantages or beneficial effects:

the method comprises the steps of obtaining vehicle driving data of a target vehicle, inputting the vehicle driving data comprising a steering wheel corner, a vehicle transverse position and vehicle transverse acceleration into a pre-trained driving style identification model, obtaining a driving style type corresponding to the target vehicle output by the driving style identification model, determining the driving style of a user, further determining the target steering wheel corner of the target vehicle according to the driving style type and vehicle state data of the target vehicle, adjusting the steering wheel corner of the target vehicle based on the target steering wheel corner to control the vehicle to return to a lane central line position, realizing lane keeping control based on the driving style of the user, meeting the differentiated requirements of drivers with different driving habits, greatly improving the driving experience of the user, and improving the system applicability and driving safety of the vehicle.

Drawings

In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, a brief description is given below of the drawings used in describing the embodiments. It should be clear that the described figures are only views of some of the embodiments of the invention to be described, not all, and that for a person skilled in the art, other figures can be derived from these figures without inventive effort.

Fig. 1A is a schematic flowchart of a lane keeping method according to an embodiment of the present invention;

FIG. 1B is a diagram illustrating a target trajectory of a vehicle according to a first embodiment of the present invention;

fig. 2 is a flowchart illustrating a lane keeping method according to a second embodiment of the present invention;

fig. 3 is a flowchart illustrating a lane keeping method according to a third embodiment of the present invention;

fig. 4A is a schematic structural diagram of a lane keeping system according to a fourth embodiment of the present invention;

fig. 4B is a schematic diagram of components of a lane keeping system according to a fourth embodiment of the present invention;

fig. 4C is a schematic diagram of a seat and an electric cylinder in a lane keeping system according to a fourth embodiment of the present invention;

FIG. 4D is a schematic diagram of the internal connection of the lane keeping system according to the fourth embodiment of the present invention;

fig. 5 is a schematic structural diagram of a lane keeping apparatus according to a fifth embodiment of the present invention;

fig. 6 is a schematic structural diagram of an electronic device according to a sixth embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.

Example one

Fig. 1A is a schematic flowchart of a lane keeping method according to an embodiment of the present invention, where the embodiment is applicable to automatically control a vehicle to travel in a lane center position according to a driving style of a user driving the vehicle when the vehicle travels, and the method may be executed by a lane keeping device, where the lane keeping device may be implemented by hardware and/or software, and the method specifically includes the following steps:

and S110, acquiring vehicle running data of the target vehicle, wherein the vehicle running data comprises a steering wheel angle, a vehicle transverse position and a vehicle transverse acceleration.

The target vehicle may be a running vehicle, among others. Optionally, in this embodiment, when it is detected that the target vehicle deviates from the lane center line, vehicle driving data of the target vehicle may be acquired. For example, when it is detected that the distance between the vehicle center position of the target vehicle and the lane center line exceeds a preset threshold, vehicle driving data of the target vehicle is acquired to further control the target vehicle to keep driving in the middle of the lane.

In the present embodiment, the vehicle running data includes a steering wheel angle, a vehicle lateral position, and a vehicle lateral acceleration. Wherein the steering wheel angle may be a steering wheel angle value of a steering wheel of the target vehicle; the vehicle lateral position may be a lateral position of the target vehicle relative to a lane centerline; the vehicle lateral acceleration may be an acceleration of the target vehicle in a direction perpendicular to the lane center line.

Specifically, the steering wheel angle can be obtained through a steering wheel sensor, the transverse position of the vehicle can be obtained through a transverse position sensor, and the transverse acceleration of the vehicle can be obtained through an acceleration sensor.

And S120, determining the driving style type corresponding to the target vehicle based on the vehicle driving data and the driving style identification model trained in advance.

Wherein the pre-trained driving style recognition model may be a model for recognizing a driving style category of the vehicle. The input vector can be formed based on the steering wheel angle, the vehicle transverse position and the vehicle transverse acceleration of the vehicle driving data and input into the driving style identification model to obtain the driving style type output by the driving style identification model. Illustratively, the driving style categories include conservative, normal, aggressive. Of course, the driving style categories may be further divided, such as very conservative, generally conservative, normal, generally aggressive, and very aggressive.

Optionally, the driving style identification model may be a learning vector quantization neural network model, a convolutional neural network model, a decision tree model, a support vector machine model, a naive bayes model, or the like. Specifically, taking the convolutional neural network model as an example, the training process of the driving style identification model may be: and constructing a sample set, wherein the sample set comprises sample running data and sample class labels corresponding to the sample running data, inputting the sample set into the convolutional neural network, calculating a loss function based on the prediction class labels and the sample class labels output by the convolutional neural network, and reversely adjusting parameters of the convolutional neural network according to a calculation result of the loss function until a convergence condition is met.

And S130, determining a target steering wheel angle of the target vehicle based on the driving style category and the vehicle state data of the target vehicle, and adjusting the steering wheel angle of the target vehicle based on the target steering wheel angle.

The vehicle state data may be information that characterizes a current vehicle travel state of the target vehicle, among others. For example, the vehicle state data may be information such as current vehicle speed, current position lateral coordinates, vehicle fore-aft wheelbase, vehicle steering gear ratio, and the like.

Specifically, in this embodiment, the driving style category may be considered, and the target steering wheel angle corresponding to the driving style category is determined by combining the vehicle state data, so as to meet the driving requirement of the driving style category. For example, for aggressive driving style categories, the corresponding target steering wheel angle may be greater than the target steering wheel angle for conservative driving style categories.

For example, the present embodiment may determine the target steering wheel angle of the target vehicle based on the driving style category, the vehicle state data of the target vehicle, and a pre-established mapping table. The mapping table comprises various driving style categories and corresponding steering wheel rotation angles under various vehicle state data. Alternatively, a feature vector may be constructed based on the driving style class and the vehicle running state data of the target vehicle, the feature vector may be input to a steering angle determination model trained in advance, and the target steering wheel angle of the target vehicle may be determined based on the output of the steering angle determination model.

Alternatively still, in one specific embodiment, the vehicle state data includes a current vehicle speed, a lane center line, a current position lateral coordinate, a vehicle steering system transmission ratio and a vehicle front-rear axle distance, and the target steering wheel angle of the target vehicle is determined based on the driving style category and the vehicle state data of the target vehicle, and the following formula is satisfied:

wherein, theta_optTarget steering wheel angle, L vehicle fore-and-aft wheelbase, i vehicle steering ratio, C_yRepresenting the driving style class, v is the current vehicle speed, y (T) is the lateral coordinate of the current position, T is the preview time, f (T) is the lane central line at the time T, f (T + T) represents the lateral coordinate of the lane central line at the time T + T, d is the preview distance,indicating the vehicle lateral velocity. Illustratively, when the driving style category is a conservative class, C_y1 is ═ 1; when the driving style class is normal class, C_y2; when the driving style class is an aggressive class, C_y＝3。

Specifically, the derivation of the above formula can be exemplarily described with reference to a target trajectory of a vehicle shown in fig. 1B. As shown in fig. 1B, f (T) represents the center line of the lane, i.e., the target track of the vehicle, y (T) represents the current position lateral coordinate of the vehicle, and T represents the preview time. Assuming that the preview distance in the driver preview model is d, the relationship between the preview time T and the preview distance d is:

according to the automobile kinematic relationship, the vehicle lateral speed can be obtained by differentiating the current position lateral coordinate y (T) of the vehicle, the vehicle lateral acceleration can be obtained by performing second-order differentiation on the current position lateral coordinate y (T) of the vehicle, and the lateral coordinate y (T + T) of the vehicle position at the time T + T can be further predicted based on the vehicle lateral speed and the vehicle lateral acceleration obtained by differentiation:

according to the principle of minimum error, in order to enable the vehicle to achieve the optimal tracking effect on the target track, i.e. to minimize the error, the lateral coordinate y (T + T) of the vehicle position at the time T + T should be consistent with the lateral coordinate f (T + T) of the target track (i.e. the lane center line) at the time T + T, that is:

f(t+T)＝y(t+T)

further, the optimal target trajectory tracking effect of the vehicle is calculated, namely the optimal lateral acceleration capable of enabling the vehicle position to return to the positive state is calculatedComprises the following steps:

the relationship between the lateral acceleration of the vehicle and the steering wheel angle is calculated as follows:

wherein R identifies the steering radius of the vehicle, v is the current vehicle speed, theta is the current steering angle of the vehicle, L is the wheelbase of the front and rear axles of the vehicle, i is the vehicle steering system transmission ratio, Y is the steering angle of the vehicle_XRepresenting the personalized corner coefficient. Specifically, Y_XThe calculation formula of (c) may be as follows:

the optimal steering wheel rotation angle theta required by the vehicle to track the target track, namely the lane center line in the lane keeping process can be calculated by simultaneously solving the formula_optExpression of (target steering wheel angle):

in the optional implementation mode, the steering wheel turning angle required by the vehicle returning to the lane central line under the driving style category of the current user can be calculated through the minimum error principle between the predicted lateral coordinate of the vehicle position at the time T + T and the predicted target track at the time T + T, so that the accurate determination of the optimal steering wheel turning angle under the driving style of the user is considered, the differentiated requirements of drivers with different driving habits are met, and the driving experience of the user is greatly improved.

Specifically, after the target steering wheel angle of the target vehicle is determined, the steering wheel angle of the target vehicle may be adjusted to the target steering wheel angle, so as to control the target vehicle to return to the lane center line position.

The technical scheme of the embodiment is that vehicle running data of a target vehicle is acquired, the vehicle running data comprising a steering wheel angle, a vehicle transverse position and a vehicle transverse acceleration is input into a driving style identification model trained in advance, obtaining the driving style class corresponding to the target vehicle output by the driving style identification model, determining the driving style of the user, further, a target steering wheel angle of the target vehicle is determined based on the driving style category and the vehicle state data of the target vehicle, the steering wheel angle of the target vehicle is adjusted based on the target steering wheel angle, the lane keeping control based on the driving style of the user is realized by controlling the vehicle to return to the center line position of the lane, the differentiated requirements of drivers with different driving habits are met, the driving experience of the user is greatly improved, and the system applicability and the driving safety of the vehicle are improved.

Example two

Fig. 2 is a schematic flow chart of a lane keeping method according to a second embodiment of the present invention, where on the basis of the foregoing embodiments, optionally, the lane keeping method further includes: acquiring a sample input vector and a driving style label corresponding to the sample input vector, and inputting the sample input vector to an input layer of a learning vector quantization neural network; calculating the distance between the sample input vector and each competition layer neuron of the learning vector quantization neural network, and determining a first target neuron and a second target neuron which are closest to the sample input vector in each competition layer neuron; updating a weight value corresponding to the first target neuron and/or the second target neuron in the learning vector quantization neural network based on a distance between the first target neuron and the sample input vector, a distance between the second target neuron and the sample input vector, a driving style predicted value corresponding to the first target neuron, a driving style predicted value corresponding to the second target neuron and a driving style label corresponding to the sample input vector; and determining the learning vector quantization neural network as a driving style identification model.

Wherein explanations of the same or corresponding terms as those of the above embodiments are omitted. Referring to fig. 2, the lane keeping method provided by the present embodiment includes the steps of:

s210, obtaining a sample input vector and a driving style label corresponding to the sample input vector, and inputting the sample input vector to an input layer of the learning vector quantization neural network.

The sample input vector may be a sample vector formed by collecting vehicle driving data (including a steering wheel angle, a vehicle lateral position, and a vehicle lateral acceleration) in advance. The driving style label may be a predetermined driving style category corresponding to the sample input vector, such as: conservative, normal, aggressive. The sample input vector and the driving style label corresponding to the sample input vector may constitute one sample data.

In the present embodiment, the learning vector quantization neural network includes an input layer, a competition layer, and a linear output layer. The number of the output layer neurons is the same as the number of the output driving style categories, and the number of the competition layer neurons can be larger than the number of the output layer neurons. Each neuron of the input layer is connected with each neuron of the competition layer; the neurons of the competition layer are connected with the corresponding neurons of the output layer.

The weight value between each input layer neuron and each competition layer neuron in the learning vector quantization neural network is defaulted to 1. Therefore, the process of training the learning vector quantization neural network in this embodiment is mainly a process of reversely adjusting the weights between the input layer neurons and the competition layer neurons.

In this embodiment, the sample input vector may be input to an input layer of the learning vector quantization neural network to further calculate the distance between the sample input vector and each neuron in the competition layer. Specifically, the data in the sample input vector may be input to each neuron of the input layer, respectively. Illustratively, the sample input vector is [ a 1a 2 a3], then a1 may be assigned to the first neuron in the input layer, a2 may be assigned to the second neuron in the input layer, and a3 may be assigned to the third neuron in the input layer.

S220, calculating the distance between the sample input vector and each competition layer neuron of the learning vector quantization neural network, and determining a first target neuron and a second target neuron which are closest to the sample input vector in each competition layer neuron.

Wherein, for each sample input vector, the distance between the sample input vector and each neuron in the competition layer is calculated. For example, the distance between the sample input vector and each competition layer neuron of the learning vector quantization neural network can be calculated according to the following formula:

in the formula (d)_iRepresents the distance between the sample input vector and the ith neuron in the competition layer, S represents the number of neurons in the competition layer, x represents the sample input vector, and x ═ x (x)₁,x₂,…,x_R)^TR represents the number of data in the sample input vector, ω_ijRepresenting the weights between the jth input layer neuron and the ith contention layer neuron.

In this embodiment, the distances between the sample input vector and each of the competitive layer neurons may be sequentially calculated based on the above formula, and then the distances are sorted from small to large based on the calculated distances, and the competitive layer neurons corresponding to the first two distances in the sorted list are taken as the first target neuron and the second target neuron. That is, a competition layer neuron that is the first closest to the sample input vector is determined as a first target neuron, and a competition layer neuron that is the second closest to the sample input vector is determined as a second target neuron.

And S230, updating the weight corresponding to the first target neuron and/or the second target neuron in the learning vector quantization neural network based on the distance between the first target neuron and the sample input vector, the distance between the second target neuron and the sample input vector, the driving style predicted value corresponding to the first target neuron, the driving style predicted value corresponding to the second target neuron and the driving style label corresponding to the sample input vector.

The driving style predicted value corresponding to the first target neuron can be determined based on the driving style category corresponding to the output layer neuron connected with the first target neuron. If the first target neuron is connected to the output layer neuron output _ C, and the driving style category corresponding to the output layer neuron output _ C is a conservative category, the driving style predicted value C corresponding to the first target neuron is determined as the conservative category, and the driving style predicted value C corresponding to the first target neuron is determined as the conservative category_y＝1(C_y1 represents a conservative class, C_y2 stands for normal, C_yEither 3 represents an aggressive class). Correspondingly, the driving style predicted value corresponding to the second target neuron may also be a driving style category corresponding to an output layer neuron to which the second target neuron is connected.

In an embodiment, if the driving style predicted value corresponding to the first target neuron is equal to the driving style label, updating the weight value corresponding to the first target neuron; if the driving style predicted value corresponding to the second target neuron is equal to the driving style label, updating the weight corresponding to the second target neuron; and if the driving style predicted value corresponding to the first target neuron and the driving style predicted value corresponding to the second target neuron are equal to the driving style label, updating the weight values corresponding to the first target neuron and the second target neuron. It should be noted that, in this embodiment, the weight corresponding to the first target neuron is a weight between the first target neuron and the input layer neuron; the weight corresponding to the second target neuron is the weight between the second target neuron and the input layer neuron.

In another embodiment, updating a weight value corresponding to a first target neuron and/or a second target neuron in a learning vector quantization neural network based on a distance between the first target neuron and a sample input vector, a distance between the second target neuron and the sample input vector, a driving style prediction value corresponding to the first target neuron, a driving style prediction value corresponding to the second target neuron, and a driving style label corresponding to the sample input vector comprises:

judging whether a preset weight value updating condition is met or not based on the distance between the first target neuron and the sample input vector, the distance between the second target neuron and the sample input vector, the driving style predicted value corresponding to the first target neuron and the driving style predicted value corresponding to the second target neuron; and if so, updating the corresponding weight values of the first target neuron and the second target neuron in the learning vector quantization neural network based on the preset learning rate, the sample input vector and the driving style label.

The preset weight value updating condition may be that a distance between the first target neuron and the sample input vector and a distance between the second target neuron and the sample input vector satisfy a preset window width condition, and the driving style predicted value corresponding to the first target neuron and the driving style predicted value corresponding to the second target neuron are different (that is, the driving style category represented by the output layer neuron connected to the first target neuron is different from the driving style category represented by the output layer neuron connected to the second target neuron).

For example, the preset window width condition may be expressed by the following formula:

wherein d is_aIs the distance of the first target neuron from the sample input vector, d_bP is the distance of the second target neuron from the sample input vector, and is the window width. Alternatively, ρ is 0.5.

Specifically, if a preset weight update condition is satisfied, the weights corresponding to the first target neuron and the second target neuron in the learning vector quantization neural network are updated based on a preset learning rate, a sample input vector, and a driving style label, which may be:

if the driving style label is equal to the driving style predicted value corresponding to the first target neuron, the weight corresponding to the first target neuron can be increased and the weight corresponding to the second target neuron can be reduced based on the preset learning rate, the sample input vector and the driving style label; for example, the following formula is adopted to correct the weight corresponding to the first target neuron and the weight corresponding to the second target neuron, so as to increase the weight corresponding to the first target neuron and decrease the weight corresponding to the second target neuron:

ω_a-neW＝ω_a-old+η(x-ω_a-old)

ω_b-neW＝ω_b-old-η(x-ω_b-old)

if the driving style label is equal to the driving style predicted value corresponding to the second target neuron, the weight corresponding to the second target neuron can be increased and the weight corresponding to the first target neuron can be reduced based on the preset learning rate, the sample input vector and the driving style label; for example, the following formula is adopted to correct the weight corresponding to the first target neuron and the weight corresponding to the second target neuron, so as to increase the weight corresponding to the second target neuron and decrease the weight corresponding to the first target neuron:

ω_a-neW＝ω_a-old-η(x-ω_a-old)

ω_b-neW＝ω_b-old+η(x-ω_b-old)

wherein, ω is_a-neWThe weight value, omega, corresponding to the first target neuron after modification_a-oldThe weight corresponding to the first target neuron before correction is obtained, eta is a preset learning rate, and x is a sample input vector; omega_b-neWThe weight value, omega, corresponding to the modified second target neuron_b-oldThe weights corresponding to the second target neurons before correction.

Of course, if the preset weight update condition is not satisfied, only the weight corresponding to the first target neuron closest to the sample input vector may be updated. Specifically, when the preset weight updating condition is not satisfied, whether the driving style prediction value corresponding to the first target neuron closest to the sample input vector distance is consistent with the preset driving style label or not may be further determined, if so, the weight corresponding to the first target neuron may be increased, and if not, the weight corresponding to the first target neuron may be decreased.

For example, the weight corresponding to the first target neuron may be updated based on the following formula to increase the weight corresponding to the first target neuron:

ω_a-neW＝ω_a-old+η(x-ω_a-ola)

updating the weight corresponding to the first target neuron based on the following formula to reduce the weight corresponding to the first target neuron:

ω_a-neW＝ω_a-old-η(x-ω_a-old)

in the optional embodiment, by judging whether the preset weight updating condition is met or not, when the preset weight updating condition is met, the weight corresponding to the first target neuron and the weight corresponding to the second target neuron in the competition layer are adjusted according to the preset learning rate, the sample input vector and the driving style label, so that the accuracy of the training process of the learning vector quantization neural network is improved, the accuracy of the driving style identification model is improved, the accuracy of the predicted driving style category is further improved, and the lane keeping control of the vehicle according to the driving style of the user is ensured.

It should be noted that, after the process is completed to update the weights of the first target neuron and/or the second target neuron in the learning vector quantization neural network, the process may be repeated, and the other sample input vectors and the driving style labels corresponding to the other sample input vectors are continuously input into the learning vector quantization neural network, so as to continuously adjust the weights of the neurons in the competition layer in the network until the training cutoff condition is satisfied. Alternatively, the training cutoff condition may be that all neurons in the competition layer are activated.

For example, after a sample input vector is fed into the learning vector quantization neural network, the distance between the sample input vector and the competition layer is activatedThe closest neuron, the state of the neuron in the competition layer changes to '1', the states of other unactivated neurons in the competition layer are still '0', the states of the linear output layer neurons connected with the activated neuron in the competition layer are also '1', the states of other linear output layer neurons are all '0', and the linear output layer neuron y connected with the activated neuron in the competition layer (each competition layer neuron is connected with only one linear output layer neuron) outputs C_y. The process of inputting each sample input vector into the learning vector quantization neural network is repeated, neurons in the competition layer can be continuously activated until the training is finished, the neurons in the competition layer are all activated, and the weight values of the neurons in the competition layer are all updated.

And S240, determining the learning vector quantization neural network as a driving style identification model.

In this embodiment, each sample input vector is input to the learning vector quantization neural network to update the weight corresponding to the neuron in the competition layer in the learning vector quantization neural network, so that after the update of the weight of each neuron in the competition layer in the learning vector quantization neural network is completed, the learning vector quantization neural network can be determined as the driving style identification model.

And S250, acquiring vehicle running data of the target vehicle, and determining the driving style type corresponding to the target vehicle based on the vehicle running data and a driving style identification model trained in advance.

The vehicle driving data includes a steering wheel angle, a vehicle lateral position, and a vehicle lateral acceleration.

And S260, determining a target steering wheel angle of the target vehicle based on the driving style category and the vehicle state data of the target vehicle, and adjusting the steering wheel angle of the target vehicle based on the target steering wheel angle.

In this embodiment, a method for training a selectable learning vector quantization neural network is further provided, where the method can further improve the training accuracy of the learning vector quantization neural network, and further improve the prediction accuracy of the learning vector quantization neural network, and the method includes the following steps:

step 1, quantizing the learning vector to obtain a weight omega between an input layer neuron j and a competition layer neuron i in a neural network_ijInitializing, and assigning values to be 1;

step 2, the sample input vector x is equal to (x)₁,x₂,…,x_R)^TInputting the distance to an input layer of the learning vector quantization neural network, and calculating the distance between a sample input vector and each competition layer neuron based on the following formula;

step 3, selecting two competition layer neurons with the minimum distance with the sample input vector, wherein the two competition layer neurons comprise the neuron a and the neuron b;

step 4, judging whether the neuron a and the neuron b simultaneously satisfy two conditions (namely, a preset weight updating condition), wherein the condition 1: the neuron a and the neuron b represent different driving style categories, namely the predicted value corresponding to the neuron a is equal to the predicted value corresponding to the neuron b, and the condition 2: distance d between neuron a and sample input vector_aAnd the distance d of neuron b from the sample input vector_bSatisfies the following conditions:

if the two conditions are met simultaneously, executing the step 5, and if the two conditions cannot be met simultaneously, executing the step 6;

step 5, if the driving style class C corresponding to the neuron a_aDriving style label C corresponding to sample input vector_XIf the weights of the neuron a and the neuron b are consistent, the weights of the neuron a and the neuron b are modified according to the following formula:

ω_a-neW＝ω_a-ols+η(x-ω_a-old)

ω_b-neW＝ω_b-old-η(x-ω_b-old)

if the driving style class C corresponding to the neuron b_bDriving style class C corresponding to input vector_XIf the weights of the neuron a and the neuron b are consistent, the weights of the neuron a and the neuron b are modified according to the following formula:

ω_a-new＝ω_a-old-η(x-ω_a-old)

ω_b-new＝ω_b-old+η(x-ω_b-old)

step 6, if the driving style class C corresponding to the neuron a_aDriving style label C corresponding to sample input vector_xIf the weights of the neurons a are consistent, the weights of the neurons a are adjusted based on the following formula:

ω_a-new＝ω_a-old+η(x-ω_a-old)

if the driving style class c corresponding to the neuron a_aDriving style label c corresponding to sample input vector_xIf the weights of the neurons are inconsistent, the weights of the neurons a are adjusted based on the following formula:

ω_a-new＝ω_a-old-η(x-ω_a-old)

for parameters in the formulas involved in the steps, reference may be made to the aforementioned explanations for the parameters.

In the technical scheme of this embodiment, the distance between the sample input vector and each competitive layer neuron of the learning vector quantization neural network is calculated to determine a first target neuron and a second target neuron in the competitive layer, which are closest to the sample input vector, and the weight corresponding to the first target neuron and/or the second target neuron in the learning vector quantization neural network is updated according to the distance between the first target neuron and the sample input vector, the distance between the second target neuron and the sample input vector, the driving style prediction value corresponding to the first target neuron, the driving style prediction value corresponding to the second target neuron, and the driving style label corresponding to the sample input vector, so as to obtain the driving style identification model, thereby implementing accurate training of the driving style identification model, improving the output accuracy of the driving style identification model, and further, the accuracy of the predicted driving style category of the target vehicle is improved, and therefore the driving experience of the user and the safety of the vehicle are further improved.

EXAMPLE III

Fig. 3 is a schematic flow chart of a lane keeping method according to a third embodiment of the present invention, where in this embodiment, optionally after the steering wheel angle of the target vehicle is adjusted based on the target steering wheel angle, the method further includes: acquiring a vehicle transverse position and a lane center line transverse position corresponding to each acquisition position; determining whether the target vehicle regresses a lane center line based on each of the vehicle lateral positions, each of the lane center line lateral positions, and the driving style category.

Wherein explanations of the same or corresponding terms as those of the above embodiments are omitted. Referring to fig. 3, the lane keeping method provided by the present embodiment includes the steps of:

s310, vehicle running data of the target vehicle is obtained, and the driving style type corresponding to the target vehicle is determined based on the vehicle running data and a driving style identification model trained in advance.

The vehicle driving data includes a steering wheel angle, a vehicle lateral position, and a vehicle lateral acceleration.

And S320, determining a target steering wheel angle of the target vehicle based on the driving style category and the vehicle state data of the target vehicle, and adjusting the steering wheel angle of the target vehicle based on the target steering wheel angle.

S330, acquiring the transverse position of the vehicle and the transverse position of the center line of the lane corresponding to each acquisition position.

Specifically, after the steering wheel angle of the target vehicle is adjusted to control the vehicle to perform lane keeping, the present embodiment may determine the position of the vehicle and analyze whether the vehicle returns to the center line of the road.

The acquisition position may be an acquisition position determined at a preset time interval, and the preset time interval may be 0.5s for example. As shown in table 1, the vehicle lateral position and the lane center line lateral position corresponding to each collection position are shown.

Table 1 vehicle lateral position and lane center line lateral position corresponding to each collection position

Acquisition position k	1	2	…	m
					Transverse position u of vehicle_k	u₁	u₂	…	u_m
Transverse position w of lane center line_k	w₁	w₂	…	w_m

And S340, determining whether the target vehicle returns to the lane central line or not based on the transverse position of each vehicle, the transverse position of the central line of each lane and the driving style category.

In one embodiment, the position difference between each vehicle transverse position and the corresponding lane central line transverse position can be determined according to each vehicle transverse position and each lane central line transverse position, and whether the target vehicle returns to the lane central line can be determined based on the variation trend of each position difference and the driving style category.

Specifically, based on the variation trend of each position gap and the driving style category, determining whether the target vehicle returns to the lane center line may be: and if the variation trend of the difference of the positions is that the difference is gradually reduced, and the speed of the reduced difference is greater than a preset speed moving threshold corresponding to the driving style category, determining that the target vehicle returns to the center line of the lane. If the driving style category is a conservative category, considering that the speed of a driver of the conservative category returning to the center line of the lane is low, the preset moving speed threshold corresponding to the conservative category can be set to be low; the driving style category is an aggressive category, and considering that the speed of a driver of the aggressive category generally returning to the center line of the lane is high, the preset moving speed threshold corresponding to the aggressive category can be set to be large. Illustratively, the driving style category is a conservative category, the position difference changes gradually in a decreasing trend, the decreasing speed is 0.6m/s, and the decreasing speed is greater than a preset moving speed threshold (such as 0.5m/s) corresponding to the conservative category, so that the target vehicle is determined to return to the lane center line.

In another alternative embodiment, determining whether the target vehicle returns to the lane center line based on the lateral position of each vehicle, the lateral position of each lane center line, and the driving style category may further be: constructing a polynomial function according to the transverse position of the vehicle corresponding to each acquisition position, and calculating a polynomial coefficient of the polynomial function based on the transverse position of the lane center line corresponding to each acquisition position; and determining whether the target vehicle returns to the lane central line or not based on the polynomial coefficient and the driving style category.

And constructing a polynomial function corresponding to each vehicle transverse position aiming at each vehicle transverse position. Illustratively, the expression of the polynomial function constructed from the respective vehicle lateral positions is as follows:

P_n(u)＝a_nuⁿ+…+a₁u+a₀

in the formula, P_n(u) is a polynomial function corresponding to the lateral position u of the vehicle, a₀、a₁、……、a_n(n < m) is the polynomial coefficient of the polynomial function and u is the vehicle lateral position.

In this alternative embodiment, the polynomial coefficient of the polynomial function is calculated based on the lateral position of the lane center line corresponding to each acquisition position, which may be: and adjusting the polynomial coefficients in the polynomial function according to the error value between the calculation result of the polynomial function of the transverse position of the vehicle and the transverse position of the lane center line corresponding to the transverse position of the vehicle until the error value between the calculation result of the polynomial function and the transverse position of the lane center line corresponding to the transverse position of the vehicle is minimum. Or, calculating a polynomial coefficient of the polynomial function based on the lateral position of the lane center line corresponding to each acquisition position, which may also be: and performing least square fitting on the polynomial function based on the transverse position of the lane center line corresponding to the acquisition position, and determining the polynomial coefficient of the polynomial function based on the fitting result.

Illustratively, calculating polynomial coefficients of a polynomial function, i.e. solving for a₀,a₁,…,a_n(n < m), it is possible to make:

take δ (a)₀，a₁，…，a_n) The polynomial function thus determined is the data (u)_k，W_k) And k is 1,2, …, m.

That is to say that the first and second electrodes,

or the above formula can be written as:

further, introducing a mark

Then there is S_ja₀+S_j+1a₁+…+S_j+na_n＝u_j(j＝1,2,…,n)；

The above formula is P_n(u) coefficient a₀，a₁，…，a_nSatisfied system of equations when the vehicle lateral position u₁，u₂，…，u_nWhen different from each other, the equation set has unique solution a₀，a₁，…，a_nSo thatTaking the minimum value. At this time, the lateral position u of the vehicle can be obtained_kTransverse position w to lane central line_kFitting function P therebetween_n(u) when a₂+a₃+…+a_n＜0.001×C_yWhen the vehicle is fully regressed to the lane center line, it may be determined. Wherein, C_yRepresenting a determined driving style category (C)_y1 represents a conservative class, C_y2 stands for normal, C_yEither 3 represents an aggressive class).

That is, based on the polynomial coefficients and the driving style category, determining whether the target vehicle regresses the lane center line may be: and if the weighted sum of the polynomial coefficients is less than the weighted value corresponding to the driving style category, determining the center line of the regression lane of the target vehicle.

In the optional implementation mode, a polynomial function is constructed according to the vehicle transverse position corresponding to each acquisition position, each polynomial coefficient of the polynomial function is calculated according to the transverse position of the lane center line corresponding to each vehicle transverse position, and then whether the target vehicle regresses the lane center line is judged based on each polynomial coefficient and the driving style category, so that the accurate judgment of whether the vehicle regresses the lane center in combination with the driving style is realized, whether the vehicle regresses the lane center can be judged individually for drivers in each driving style, and the safety of the vehicle is further improved.

According to the technical scheme, after the steering wheel turning angle of the target vehicle is adjusted, the plurality of vehicle transverse positions and the lane central line transverse positions corresponding to the vehicle transverse positions are collected, whether the target vehicle returns to the lane central line is judged based on the vehicle transverse positions, the lane central line transverse positions corresponding to the vehicle transverse positions and the driving style types, the accurate judgment of whether the vehicle combining the driving style returns to the lane center is achieved, whether the vehicle returns to the lane center can be accurately judged under different driving styles, the driving experience of a user is improved, and meanwhile the driving safety of the vehicle is improved.

Example four

Fig. 4A is a schematic structural diagram of a lane keeping system according to a fourth embodiment of the present invention, where the system includes an industrial personal computer 410 and a steering wheel assembly 420, where the steering wheel assembly 420 is configured to obtain a steering wheel angle of a target vehicle and send the steering wheel angle to the industrial personal computer 410; the industrial personal computer 410 is used for adjusting the steering wheel angle of the target vehicle based on the lane keeping method provided by any embodiment of the application.

Illustratively, a porphyry mainboard aimb781 can be adopted in the industrial personal computer 410, a display card of the industrial personal computer 410 can be a GTX1070, a CPU of the industrial personal computer 410 can be i7, the industrial personal computer 410 is further connected with a display, a driving style identification system is arranged in the industrial personal computer 410, the driving style identification system is realized based on computer software, specifically related software includes PanoSim and MATLAB/Simulink, PanoSim is an automobile virtual simulation platform provided for solving various challenges faced by modern intelligent automobile and automobile intelligentization technologies and product development, testing and verification, and can simulate the response of an automobile to the input of a driver, a road surface and aerodynamics, Simulink is a visual simulation tool in MATLAB and can realize the functions of dynamic system modeling, simulation and analysis, and the Simulink provides an integrated environment of a dynamic modeling system, simulation and comprehensive analysis.

Optionally, the PanoSim deployed in the industrial personal computer 410 may be used to collect vehicle driving data (including a steering wheel angle, a vehicle lateral position, and a vehicle lateral acceleration) of the target vehicle, and the driving style identification system may determine a driving style category and a target steering wheel angle according to the vehicle driving data collected by the PanoSim, and input the target steering wheel angle to the PanoSim, so that the PanoSim controls the position of the target vehicle to be adjusted.

PanoSim can also be used for collecting the transverse position of the vehicle and the transverse position of the lane center line corresponding to each collecting position, and the driving style identification system can judge whether the target vehicle returns to the lane center line according to the transverse position of the vehicle and the transverse position of the lane center line collected by PanoSim.

Illustratively, the present embodiment also provides a lane keeping system comprising components as shown in fig. 4B-4C, fig. 4B illustrating a schematic view of components of a lane keeping system, and fig. 4C illustrating a schematic view of a seat and an electric cylinder in a lane keeping system. Referring to fig. 4B to 4C, the lane keeping system includes a seat 1, a steering wheel assembly 2, a pedal assembly 3, a bracket 4, a screen 5, an industrial personal computer 6, a connection plate 7, a base plate 8, an electric cylinder 9, an electric cylinder control system 10, a universal wheel 11, a speaker 12, an accelerator pedal 13, a brake pedal 14, a clutch pedal 15, and a display 16.

Wherein, the steering wheel assembly 2 is fixed on a bracket 4, the pedal assembly 3 is assembled below the bracket 4, the screen 5 is arranged in front of the steering wheel assembly 2 and the bracket 4, the seat 1 is arranged corresponding to the steering wheel assembly 2, a connecting plate 7 and a bottom plate 8 are arranged below the seat 1, an electric cylinder 9 is assembled between the top surface of the connecting plate 7 and the bottom of the seat 1, an electric cylinder 9 is also assembled between the rear side of the connecting plate 7 and the rear end plate of the bottom plate 8, the electric cylinders 9 are all connected with an electric cylinder control system 10 and controlled by the electric cylinder control system 10 to work, the steering wheel assembly 2, the pedal assembly 3, the screen 5 and the electric cylinder control system 10 are connected with an industrial personal computer 6, the steering wheel assembly 2, the pedal assembly 3, the screen 5, the industrial personal computer 6 and the electric cylinder control system 10 are all powered by the same power supply, and universal wheels 11 are assembled between the bottom surface of the connecting plate 7 and the top surface of the bottom plate 8.

The top surface of the support 4 is provided with a sound box 12, the sound box 12 is connected with the industrial personal computer 6 through a USB cable, and the sound box 12 is used for simulating sound in the real driving process. The steering Wheel assembly 2 is connected with the industrial personal computer 6 through a CAN line, the steering Wheel assembly 2 adopts a SENSO-Wheel steering Wheel assembly, the steering Wheel assembly 2 provides freely programmable and adjustable rigidity, damping and torque, steering feeling experience CAN be achieved, a steering Wheel corner sensor is integrated in the steering Wheel assembly 2, the corner sensor collects steering Wheel corner signals, and the steering Wheel corner signals are sent to the industrial personal computer 6 in a CAN message mode and used for a vehicle dynamic model in the industrial personal computer 6 to operate. The industrial personal computer 6 can adjust the steering wheel angle of the target vehicle according to the lane keeping method in any embodiment. Or, according to the lane keeping method of any of the above embodiments, the steering wheel angle of the target vehicle is adjusted, and it is determined whether the target vehicle completely returns to the lane center line.

The pedal assembly 3 is connected with the industrial personal computer 6 through a USB (universal serial bus) wire, the pedal assembly 3 adopts G29 series pedal components, an accelerator pedal 13, a brake pedal 14 and a clutch pedal 15 are sequentially arranged from right to left, respective pedal displacement sensors are integrated in the accelerator pedal 13 and the brake pedal 14, and the displacement sensors acquire corresponding pedal stroke signals and send the pedal stroke signals to the industrial personal computer 6 for the running of a vehicle dynamics model in the industrial personal computer 6. The connection relationship of the above components can be seen in fig. 4D, and fig. 4D shows the internal connection diagram of the lane keeping system.

The screen 5 is an annular screen, the annular screen adopts three NEC NP4100+ mainstream engineering projectors for projection, each projector forms a channel, three channels are generated in a horizontal crossing mode to be spliced and displayed, three VGA computer signals provide display content support, hardware nonlinear geometric correction technology is adopted between the NEC NP4100+ mainstream engineering projectors of adjacent channels, and finally good projection effect on the annular screen is achieved.

The electric cylinder control system 10 comprises a PID controller, a D/A card and a servo amplifier, an industrial personal computer 6 generates a speed control instruction and transmits the speed control instruction to the PID controller through a signal line, the PID controller resolves based on a PID algorithm of the PID controller to obtain a speed signal and transmit the speed signal to the D/A card, the D/A card converts the speed signal into a voltage signal and transmits the voltage signal to the servo controller, and the servo controller finally controls the movement of the electric cylinder 9. Two groups of four electric cylinders 9 assembled between the top surface of the connecting plate 7 and the bottom of the seat 1 are symmetrically arranged, and two electric cylinders 9 assembled between the rear side of the connecting plate 7 and the rear end plate of the bottom plate 8 are arranged. Four universal wheels 11 are arranged between the bottom surface of the connecting plate 7 and the top surface of the bottom plate 8.

The lane keeping system provided by the embodiment comprises the industrial personal computer and the steering wheel assembly, wherein the industrial personal computer can determine the driving style category of the target vehicle, the target steering wheel corner and adjust the steering wheel corner, so that lane keeping control based on the driving style of a user is realized, the differentiation requirements of drivers with different driving habits are met, the driving experience of the user is greatly improved, and the system applicability and the driving safety of the vehicle are improved.

EXAMPLE five

Fig. 5 is a schematic structural diagram of a lane keeping apparatus according to a fifth embodiment of the present invention, where the present embodiment is applicable to a case where a vehicle is automatically controlled to travel in a lane center position according to a driving style of a user driving the vehicle when the vehicle travels, and the apparatus specifically includes: a driving data acquisition module 510, a driving style recognition module 520, and a steering wheel adjustment module 530.

A driving data obtaining module 510, configured to obtain vehicle driving data of a target vehicle, where the vehicle driving data includes a steering wheel angle, a vehicle lateral position, and a vehicle lateral acceleration;

a driving style identification module 520, configured to determine a driving style category corresponding to the target vehicle based on the vehicle driving data and a pre-trained driving style identification model;

a steering wheel adjustment module 530, configured to determine a target steering wheel angle of the target vehicle based on the driving style category and the vehicle state data of the target vehicle, and adjust the steering wheel angle of the target vehicle based on the target steering wheel angle.

Optionally, the lane keeping apparatus further includes a network training module, where the network training module includes a sample input unit, a distance calculation unit, a weight adjustment unit, and a model determination unit;

the sample input unit is used for acquiring a sample input vector and a driving style label corresponding to the sample input vector, and inputting the sample input vector to an input layer of a learning vector quantization neural network;

the distance calculation unit is configured to calculate a distance between the sample input vector and each competitive layer neuron of the learning vector quantization neural network, and determine a first target neuron and a second target neuron which are closest to the sample input vector in each competitive layer neuron;

the weight adjusting unit is configured to update a weight corresponding to the first target neuron and/or the second target neuron in the learning vector quantization neural network based on a distance between the first target neuron and the sample input vector, a distance between the second target neuron and the sample input vector, a driving style prediction value corresponding to the first target neuron, a driving style prediction value corresponding to the second target neuron, and a driving style label corresponding to the sample input vector;

the model determining unit is used for determining the learning vector quantization neural network as a driving style identification model.

Optionally, the weight adjusting unit is specifically configured to:

judging whether a preset weight updating condition is met or not based on the distance between the first target neuron and the sample input vector, the distance between the second target neuron and the sample input vector, the driving style predicted value corresponding to the first target neuron and the driving style predicted value corresponding to the second target neuron; if yes, updating the weight values corresponding to the first target neuron and the second target neuron in the learning vector quantization neural network based on a preset learning rate, the sample input vector and the driving style label.

Optionally, the vehicle state data includes a current vehicle speed, a lane center line, a current position lateral coordinate, a vehicle steering system transmission ratio, and a vehicle front-rear axle distance, and the steering wheel adjustment module 530 includes a target steering angle determination unit, which is configured to determine a target steering wheel angle of the target vehicle based on the driving style category and the vehicle state data of the target vehicle according to the following formula:

Optionally, the lane keeping device further includes a regression judging module, where the regression judging module includes a position collecting unit and a position judging unit;

the position acquisition unit is used for acquiring a vehicle transverse position and a lane center line transverse position corresponding to each acquisition position after the steering wheel rotating angle of the target vehicle is adjusted based on the target steering wheel rotating angle;

the position judging unit is used for determining whether the target vehicle returns to the lane central line or not based on the transverse positions of the vehicles, the transverse positions of the lane central line and the driving style category.

Optionally, the position determining unit is specifically configured to:

constructing a polynomial function according to the transverse position of the vehicle corresponding to each acquisition position, and calculating a polynomial coefficient of the polynomial function based on the transverse position of the lane center line corresponding to each acquisition position; and determining whether the target vehicle regresses the lane center line or not based on the polynomial coefficient and the driving style category.

In the embodiment, the driving data acquisition module is used for acquiring the vehicle driving data of the target vehicle, the driving style identification module is used for inputting the vehicle driving data comprising the steering wheel corner, the vehicle transverse position and the vehicle transverse acceleration into a pre-trained driving style identification model to obtain the driving style type corresponding to the target vehicle output by the driving style identification model, the driving style of the user is determined, the steering wheel adjustment module is used for determining the target steering wheel corner of the target vehicle according to the driving style type and the vehicle state data of the target vehicle, the steering wheel corner of the target vehicle is adjusted based on the target steering wheel corner to control the vehicle to return to the lane central line position, the lane keeping control based on the driving style of the user is realized, the differentiated requirements of drivers with different driving habits are met, and the driving experience of the user is greatly improved, and, the system suitability and the driving safety of the vehicle are improved.

The lane keeping device provided by the embodiment of the invention can execute the lane keeping method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.

It should be noted that, the units and modules included in the system are merely divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the embodiment of the invention.

EXAMPLE six

Fig. 6 is a schematic structural diagram of an electronic device according to a sixth embodiment of the present invention. FIG. 6 illustrates a block diagram of an exemplary electronic device 12 suitable for use in implementing embodiments of the present invention. The electronic device 12 shown in fig. 6 is only an example and should not bring any limitation to the function and the scope of use of the embodiment of the present invention. The device 12 is typically an electronic device that assumes lane keeping functions.

As shown in FIG. 6, electronic device 12 is embodied in the form of a general purpose computing device. The components of electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a memory 28, and a bus 18 that couples the various components (including the memory 28 and the processing unit 16).

Bus 18 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, such architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, an enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

Electronic device 12 typically includes a variety of computer-readable media. Such media may be any available media that is accessible by electronic device 12 and includes both volatile and nonvolatile media, removable and non-removable media.

Memory 28 may include computer device readable media in the form of volatile Memory, such as Random Access Memory (RAM) 30 and/or cache Memory 32. The electronic device 12 may further include other removable/non-removable, volatile/nonvolatile computer storage media. By way of example only, the storage device 34 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 6, commonly referred to as a "hard drive"). Although not shown in FIG. 6, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a Compact disk-Read Only Memory (CD-ROM), a Digital Video disk (DVD-ROM), or other optical media) may be provided. In these cases, each drive may be connected to bus 18 by one or more data media interfaces. Memory 28 may include at least one program product 40, with program product 40 having a set of program modules 42 configured to carry out the functions of embodiments of the invention. Program product 40 may be stored, for example, in memory 28, and such program modules 42 include, but are not limited to, one or more application programs, other program modules, and program data, each of which examples or some combination may comprise an implementation of a network environment. Program modules 42 generally carry out the functions and/or methodologies of the described embodiments of the invention.

Electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboard, mouse, camera, etc., and display), one or more devices that enable a user to interact with electronic device 12, and/or any devices (e.g., network card, modem, etc.) that enable electronic device 12 to communicate with one or more other computing devices. Such communication may be through an input/output (I/O) interface 22. Also, the electronic device 12 may communicate with one or more networks (e.g., a Local Area Network (LAN), Wide Area Network (WAN), and/or a public Network such as the internet) via the Network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via the bus 18. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with electronic device 12, including but not limited to: microcode, device drivers, Redundant processing units, external disk drive Arrays, disk array (RAID) devices, tape drives, and data backup storage devices, to name a few.

The processor 16 executes various functional applications and data processing by running a program stored in the memory 28, for example, to implement the lane keeping method provided by the above-described embodiment of the present invention, including:

acquiring vehicle running data of a target vehicle, wherein the vehicle running data comprises a steering wheel angle, a vehicle transverse position and a vehicle transverse acceleration;

determining a driving style category corresponding to the target vehicle based on the vehicle driving data and a pre-trained driving style identification model;

Of course, those skilled in the art will understand that the processor may also implement the technical solution of the lane keeping method provided by any embodiment of the present invention.

EXAMPLE seven

An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the lane keeping method provided in any embodiment of the present invention, the method including:

acquiring vehicle running data of a target vehicle, wherein the vehicle running data comprises a steering wheel angle, a vehicle transverse position and a vehicle transverse acceleration;

determining a driving style category corresponding to the target vehicle based on the vehicle driving data and a pre-trained driving style identification model;

Computer storage media for embodiments of the invention may employ any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Computer program code for carrying out operations for embodiments of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

28页详细技术资料下载

上一篇：一种医用注射器针头装配设备

下一篇：一种车辆的自适应巡航滑模控制方法

Lane keeping method, apparatus, device, medium, and system

相关技术

网友询问留言