CN111276008A - Device and method for individualized guidance of vehicle speed in acceleration lane - Google Patents

Abstract

The invention discloses a device for personalizing vehicle speed guidance in an acceleration lane, comprising a visual camera, a digital image processor, a millimeter-wave radar, a data processor and a display screen. A guidance method is also disclosed. The visual camera collects the vehicle information entering the acceleration lane, and the digital image processor processes the vehicle information to obtain the vehicle type information; the millimeter wave radar collects the speed of the incoming vehicle, the distance between the millimeter wave radar and the incoming vehicle, and the driving distance. The speed of the rear vehicle on the road, the distance between the rear vehicle and the millimeter-wave radar; the data processor processes the information collected by the millimeter-wave radar to obtain the remaining distance of the incoming vehicle and the relative distance between the incoming vehicle and the rear vehicle; the data processor also receives The vehicle type information sent by the digital image processor is compared with the optimal speed curve in the database to determine the difference level, and output a reminder signal through the display screen according to the difference level. The present invention conducts personalized guidance for importing vehicles, thereby reducing the risk of importing.

Description

Device and method for individualized guidance of vehicle speed in acceleration lane

technical field

The invention relates to the technical field of traffic safety facilities, in particular to a device and method for individualized guidance of vehicle speed in an acceleration lane.

Background technique

With the development of the national economy, the number of car ownership has also risen rapidly. While road traffic safety is being paid attention to, the degree of individualized demands of drivers on vehicles and roads is also rising rapidly. Because different drivers have different driving habits and driving styles, traffic safety facilities sometimes cannot meet the needs of different drivers, which in turn affects the driving rhythm of drivers. Drivers will require the road traffic safety reminder device to not only meet the safety requirements, but also meet the driving habits and styles of different drivers in the process of guiding the driver. With the continuous development of intelligent transportation systems, the safety reminder devices for future road traffic must become more intelligent and personalized.

The merging area where vehicles enter the main lane from the acceleration lane has always been a key area that affects the efficiency of traffic flow. The vehicles in the acceleration lane will affect the passage of vehicles in the main lane during the merging process, especially in the case of forced merging. This leads to the failure of traffic flow on the main lane, which in turn leads to congestion and possibly even traffic risks. In the traditional environment, experienced drivers need to observe the vehicles in the main lane through the rearview mirror, obtain the traffic information on the main lane, and judge the risk of inflow based on the current traffic environment, and start the inflow after confirming that it is safe. But this process is often difficult for novice drivers. In addition, at this stage, the quality of drivers in our country is uneven. Some drivers do not accelerate or observe the forced entry when entering, which may also endanger the driving safety of vehicles on the main lane.

SUMMARY OF THE INVENTION

In view of the problems existing in the prior art, the purpose of the present invention is to provide a device and method for individualized guidance of vehicle speed in the acceleration lane, so as to provide intelligent and individualized guidance for vehicles entering the acceleration lane, and to improve the safety of the merging area. performance and traffic efficiency.

In order to achieve the above object, the present invention adopts the following technical solutions to achieve.

Technical solution one

A device for personalizing vehicle speed guidance in an acceleration lane, comprising: a visual camera, a digital image processor, a millimeter-wave radar, a data processor and a display screen;

The output end of the visual camera is connected with the input end of the digital image processor, and the output end of the digital image processor and the output end of the millimeter wave radar are respectively connected with the input end of the data processor , the output end of the data processor is connected with the input end of the display screen;

The visual camera is used to collect the image information of the vehicle entering the acceleration lane, and transmit the collected image information of the vehicle to the digital image processor;

The digital image processor is used for processing the received image information of the vehicle to obtain the vehicle type information;

The millimeter wave radar is used to collect the motion parameters of the vehicles in the acceleration lane and the driving lane and the distance parameters between the vehicles;

The data processor is used to receive the information transmitted by the digital image processor and the millimeter-wave radar, and determine the difference level by comparing it with the speed curve of the corresponding vehicle model in the database when the import risk is the lowest;

The display screen is used for receiving the information transmitted by the data processor and outputting a reminder signal.

The characteristic and further improvement of technical scheme one of the present invention are:

The vision camera is installed at the entrance of the acceleration lane.

The millimeter wave radar is installed at the end of the acceleration lane.

Technical solution two

A method for individualized guidance of vehicle speed in the acceleration lane, based on the above-mentioned device for individualized guidance of vehicle speed in the acceleration lane, comprising the following steps:

Step 1, the visual camera collects the image information of the vehicle entering the acceleration lane;

Step 2, the digital image processor processes the image information of the imported vehicle collected by the visual camera, and obtains the model information of the imported vehicle;

Step 3, the millimeter-wave radar collects the speed v ₀ of the incoming vehicle on the acceleration lane, the distance L ₀ between the incoming vehicle and the millimeter-wave radar, the speed v ₁ of the rear vehicle on the driving lane and behind the incoming vehicle, and the The distance L ₁ between the rear vehicle and the millimeter-wave radar;

Step 4, the data processor calculates the relative distance d ₁ between the incoming vehicle and the following vehicle, and the remaining length Dr on the acceleration lane of the incoming vehicle according to the information collected by the millimeter-wave radar in step 3;

Step 5, the data processor determines the difference level according to the information collected in step 3 and the relative distance d ₁ calculated in step 4, combined with the optimal speed curve of the corresponding vehicle model in the database, and finally outputs a reminder according to the difference level. Signal.

The characteristic and further improvement of technical scheme two of the present invention are:

In step 2, the method for obtaining the model information of the imported vehicle adopts a three-branch convolutional neural network based on AlexNET.

Further, the method for obtaining the model information of the imported vehicle is specifically:

First, normalize the size of the collected image information imported into the vehicle, then cut it, and then convert it into a tensor data structure. Forward the traditional path, and finally get the image classification score, and judge the image type according to the image classification score.

In step 4, the specific method of calculation is: the data processor establishes a rectangular coordinate system, the abscissa axis of the rectangular coordinate system is parallel to the main road, the ordinate axis of the rectangular coordinate system and the abscissa axis are in the same horizontal plane, and perpendicular to the abscissa axis; the angles between the millimeter-wave radar and the incoming vehicle and the rear car behind the incoming vehicle are α ₀ and α ₁ respectively, and the distances are L ₀ and L ₁ respectively. According to the cosine theorem, the incoming vehicle is calculated by calculating The relative distance d ₁ to the vehicle behind, and the remaining length Dr of the merging vehicle on the acceleration lane.

In step 5, the optimal speed curve is specifically: collecting and recording each import action of different models entering the acceleration lane in advance, and judging the risk level of the import action, corresponding to the speed curve of the import vehicle when the risk level is the smallest. , which is the optimal speed curve.

The risk level of the judgment import action is:

where round is the rounding function.

In step 5, the difference level is specifically:

The data processor queries the database according to the model information of the imported vehicle, the relative speed and distance between the imported vehicle and the rear vehicle, and matches an optimal speed curve; Poor, divide the difference level into N grades.

Compared with the prior art, the beneficial effects of the present invention are:

The device for personalizing vehicle speed guidance in the acceleration lane provided by the present invention processes the vehicle information collected by the visual camera through the digital image processor to obtain the vehicle type information; the data processor processes the information sent by the visual image processor and the millimeter wave radar. After processing, the driving environment of the imported vehicle is obtained, and the difference level is determined in combination with the optimal speed curve of the corresponding vehicle model in the database, and a reminder signal is output to provide personalized guidance to the imported vehicle.

The method for personalized guidance of vehicle speed in the acceleration lane provided by the present invention can provide personalized guidance for different vehicle models. During the guidance process, only the incoming vehicle needs to be observed through a visual camera, and the incoming vehicle is identified through a deep learning algorithm. The model of the vehicle, combined with the optimal speed curve stored in the database according to the risk level of the import action of different models, and compare the speed of the imported vehicle to get the difference level, send the reminder information to the display screen, and the display screen will remind through the screen image The import vehicle adjusts the speed to achieve personalized guidance for different models in different environments, thereby reducing the risk of import.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

1 is a schematic structural diagram of an embodiment of a device for personalizing vehicle speed guidance in an acceleration lane provided by the present invention;

2 is a schematic flowchart of an embodiment of a method for personalizing vehicle speed guidance in an acceleration lane provided by the present invention;

FIG. 3 is a schematic diagram of a scene for personalized guidance of vehicle speed in the acceleration lane provided by the present invention; in the figure, 0 represents an incoming vehicle, and 1 represents a rear vehicle.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Referring to FIG. 1 , an embodiment of the present invention provides a device for personalizing vehicle speed guidance in an acceleration lane, including a visual camera, a digital image processor, a millimeter-wave radar, a data processor, and a display screen. The output end of the visual camera is connected with the input end of the digital image processor, the output end of the digital image processor and the output end of the millimeter wave radar are respectively connected with the input end of the data processor, and the output end of the data processor is connected to the input end of the data processor. The input terminals of the display are connected.

Referring to FIG. 3 , a visual camera is installed at the entrance of the acceleration lane, used to collect image information of vehicles entering the acceleration lane, and transmit the collected image information of the vehicle to a digital image processor.

The digital image processor is used for processing the received image information of the vehicle to obtain the vehicle type information.

Referring to Figure 3, the millimeter-wave radar is installed at the end of the acceleration lane to collect the motion parameters of the vehicles in the acceleration lane and the driving lane and the distance parameters between the vehicles. The motion parameters include speed, and the distance parameters include distance and angle.

The data processor is used to receive the information transmitted by the digital image processor and the millimeter wave radar, and by comparing it with the speed curve of the corresponding vehicle model in the database when the import risk is the lowest, to determine the difference level, and output a reminder signal through the difference level.

The display screen is used to receive the reminder signal transmitted by the data processor and output the reminder information.

Referring to FIG. 2 and FIG. 3 , the present invention also provides a method for individualized guidance of vehicle speed in the acceleration lane, based on the above-mentioned device for individualized guidance of vehicle speed in the acceleration lane, including the following steps:

Step 1, the visual camera collects image information of the vehicle entering the acceleration lane.

In step 2, the digital image processor processes the image information of the imported vehicle collected by the visual camera to obtain the vehicle type information of the imported vehicle.

The deep learning method is used to obtain the model information of the imported vehicle, specifically a three-branch convolutional neural network based on AlexNET.

Specifically, the size of the collected image information imported into the vehicle is first normalized, then cut, and then converted into a tensor data structure. The forward traditional path formed by the connection layer finally obtains the image classification score, and the image type is determined according to the image classification score.

其中K(m,n)为卷积核各点数值，A(x,y)为输入图像各点的数值，f(i,j)为输入的图像(i,j)点的值；采用最大池化方式，计算方式为f(i,j)＝max_0≤n,n<k{A(i×s+m,i×s+n)}，其中s为池化窗口移动的步长，m,n分别为池化窗口的宽和高，A(x,y)为输入图像各点的数值，f(i,j)为输入的图像(i,j)点的值。Specifically, the convolution kernel is

where K(m,n) is the value of each point of the convolution kernel, A(x,y) is the value of each point of the input image, and f(i,j) is the value of the input image (i,j); Pooling method, the calculation method is f(i,j)=max _0≤n,n<k {A(i×s+m,i×s+n)}, where s is the step size of the pooling window movement, m, n are the width and height of the pooling window respectively, A(x, y) is the value of each point of the input image, and f(i, j) is the value of the input image (i, j) point.

Specifically, in the three-branch convolutional neural network based on AlexNET, the three-branch structure is embodied in the first five convolutional layers, each branch contains five convolutional layers, the sixth layer is a feature fusion layer, and the seventh layer is a feature fusion layer. Layer and layer 8 are fully connected layers, refer to AlexNET, each with 4096 neurons. The number of neurons in the last fully connected layer is the same as the number of classification categories.

Specifically, the activation function in the convolutional neural network is ReLU, and the loss function is a cross-entropy function.

Specifically, the backpropagation of the convolutional neural network adopts the stochastic gradient descent method to update the weights.

Step 3, the millimeter-wave radar collects the speed v ₀ of the incoming vehicle on the acceleration lane, the distance L ₀ between the incoming vehicle and the millimeter-wave radar, the speed v ₁ of the rear vehicle on the driving lane and behind the incoming vehicle, and the The distance L ₁ between the rear vehicle and the millimeter-wave radar.

As shown in Figure 3, the millimeter-wave radar is installed at the end of the acceleration lane to scan and monitor the driving vehicles in the driving lane and the acceleration lane.

Step 4, the data processor calculates the relative distance d ₁ between the incoming vehicle and the following vehicle, and the remaining length Dr on the acceleration lane of the incoming vehicle according to the information collected by the millimeter wave radar in step 3 .

Specifically, the calculation method is as follows: the data processor establishes a rectangular coordinate system, the abscissa axis of the rectangular coordinate system is parallel to the main road, the ordinate axis and the abscissa axis of the rectangular coordinate system are in the same horizontal plane, and perpendicular to the Abscissa axis; the angles between the millimeter-wave radar and the incoming vehicle and the rear car behind the incoming vehicle are α ₀ and α ₁ respectively, and the distances are L ₀ and L ₁ respectively. According to the cosine theorem, the incoming vehicle and the rear car are calculated. The relative distance d ₁ of , and the remaining length Dr of the merging vehicle on the acceleration lane.

Step 5, the data processor determines the difference level according to the information collected in step 3 and the relative distance d ₁ calculated in step 4, combined with the optimal speed curve of the corresponding vehicle model in the database, and finally outputs a reminder according to the difference level. Signal.

In this embodiment, the optimal speed curve is: collect and record in advance each entry action of different models entering the acceleration lane within the last three months, and determine the risk level of the entry action. When the corresponding risk level is the smallest, The speed profile of the incoming vehicle is the optimal speed profile.

Specifically, the rules for determining the import risk level are:

First, after the incoming vehicle enters the acceleration lane, after the vehicle enters the acceleration lane, the data processor determines the incoming risk level Risk _level according to the current environment; the risk _level is composed of two parts, the first part is related to the remaining length of the acceleration lane, and the second part is related to the remaining length of the acceleration lane. Part of it is related to the time to collision (TTC, time to collision) between the incoming vehicle and the following vehicle on the main lane. The calculation formula is:

Dr为加速车道剩余长度；round为四舍五入函数；即加速车道剩余长度和碰撞时间与汇入风险成反比；上述计算公式在计算过程中，所涉及的参数均采用传感器获取到的数值，不考虑单位影响。in,

Dr is the remaining length of the acceleration lane; round is a rounding function; that is, the remaining length of the acceleration lane and the collision time are inversely proportional to the risk of merging; in the calculation process of the above calculation formula, the parameters involved are the values obtained by the sensor, regardless of the unit influences.

When working, the visual camera collects the vehicle information entering the acceleration lane, and the digital image processor processes the collected vehicle information to obtain the vehicle type information; the millimeter wave radar collects the vehicle speed v ₀ of the incoming vehicle, and the distance between the millimeter wave radar and the incoming vehicle L ₀ , the speed v ₁ of the rear vehicle on the driving lane, the distance L ₁ between the rear vehicle and the millimeter wave radar, and the angles α ₀ , α ₁ between the millimeter wave radar and the merging vehicle and the rear vehicle behind the merging vehicle; The information collected by the millimeter wave radar is processed to obtain the remaining distance Dr of the incoming vehicle and the relative distance d ₁ between the incoming vehicle and the rear vehicle; at the same time, the data processor also receives the vehicle type information sent by the digital image processor, and passes through the database. Compare the optimal speed curves in , and divide the difference level into 3 levels; when the speed difference at the same time point is less than 3km/h, the difference level is level 1; when it is greater than or equal to 3km/h and less than or equal to 5km/h, the difference level is The difference level is level 2; when it is greater than 5km/h, the difference level is level 3. When the difference level reaches level 3, the data processor judges that the current imported vehicle speed does not meet the optimal speed curve, and sends a prompt message to the display screen. When the difference between the speed of the incoming vehicle and the rear vehicle is negative, the display will display the word "accelerate"; when the difference between the speed of the incoming vehicle and the rear vehicle is positive, the display will display the word "deceleration"; thus realizing the speed of the incoming vehicle guide to reduce the risk of inflow.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. An apparatus for personalized guidance of vehicle speed in an acceleration lane, comprising: the system comprises a visual camera, a digital image processor, a millimeter wave radar, a data processor and a display screen;

the output end of the visual camera is connected with the input end of the digital image processor, the output end of the digital image processor and the output end of the millimeter wave radar are respectively connected with the input end of the data processor, and the output end of the data processor is connected with the input end of the display screen;

the visual camera is used for acquiring the image information of the vehicle entering the acceleration lane and transmitting the acquired image information of the vehicle to the digital image processor;

the digital image processor is used for processing the received image information of the vehicle to obtain the vehicle type information of the vehicle;

the millimeter wave radar is used for acquiring motion parameters of vehicles in an acceleration lane and a traffic lane and distance parameters between the vehicles;

the data processor is used for receiving information transmitted by the digital image processor and the millimeter wave radar, comparing the information with a speed curve of a corresponding vehicle type in the database when the risk of the vehicle type is the lowest, determining a difference grade and outputting a reminding signal according to the difference grade;

the display screen is used for receiving the reminding signal transmitted by the data processor and outputting reminding information.

2. The device for personalized guidance of vehicle speed in an acceleration lane according to claim 1, characterized in that the vision camera is mounted at the entrance of the acceleration lane.

3. The apparatus for guiding vehicle speed individually for an acceleration lane according to claim 1, wherein said millimeter wave radar is installed at the end position of the acceleration lane.

4. A method for personalized guidance of vehicle speed in an acceleration lane, based on the device for personalized guidance of vehicle speed in an acceleration lane of claim 1, characterized by comprising the following steps:

step 1, a visual camera collects image information of vehicles converged into an acceleration lane;

step 2, the digital image processor processes the image information of the imported vehicle collected by the visual camera to obtain the vehicle type information of the imported vehicle;

step 3, the millimeter wave radar collects the speed v of the vehicle converged on the acceleration lane₀Distance L between the incoming vehicle and the millimeter wave radar₀A rear vehicle speed v on the traffic lane and behind the merging vehicle₁And the distance L between the rear vehicle and the millimeter wave radar₁；

Step 4, the data processor calculates the relative distance d between the afflux vehicle and the rear vehicle according to the information collected by the millimeter wave radar in the step 3₁And the remaining length Dr of the merging vehicle on the acceleration lane;

step 5, the data processor calculates the relative distance d obtained in step 4 according to the information collected in step 3₁And determining a difference grade by combining the optimal speed curve of the corresponding vehicle type in the database, and finally outputting a reminding signal according to the difference grade.

5. The method for personalized guidance of vehicle speed on an acceleration lane according to claim 4, characterized in that, in step 2, the method for obtaining the vehicle type information of the imported vehicle adopts a three-branch convolutional neural network based on AlexNET.

6. The method for personalized guidance of vehicle speed in an acceleration lane according to claim 5, wherein the method for obtaining the vehicle type information of the imported vehicle is specifically:

firstly, carrying out size normalization on collected image information of an imported vehicle, then cutting, converting into a tensor data structure, carrying out regularization treatment, and finally obtaining an image classification score through a forward traditional path formed by a convolution layer, a pooling layer and a full-connection layer, and judging the image type according to the image classification score.

7. The method for personalized guidance of vehicle speed on an acceleration lane according to claim 4, wherein in step 4, the specific calculation method is that the data processor establishes a rectangular coordinate system, the abscissa axis of the rectangular coordinate system is parallel to the main lane, the ordinate axis of the rectangular coordinate system and the abscissa axis are in the same horizontal plane and perpendicular to the abscissa axis, and the angles of the millimeter wave radar and the rear vehicle converged into the vehicle and converged behind the vehicle are α respectively₀、α₁Respectively, distance is L₀、L₁According to the cosine theorem, the relative distance d between the incoming vehicle and the rear vehicle is calculated₁And the remaining length Dr of the merging vehicle on the acceleration lane.

8. The method as claimed in claim 4, wherein in step 5, the optimal speed curve is obtained by collecting and recording each importing action of different vehicle types to the acceleration lane in advance, and determining a risk level of the importing action, and the speed curve of the vehicle when the corresponding risk level is minimum is the optimal speed curve.

9. The method for personalized guidance of vehicle speed in an acceleration lane according to claim 8, wherein the determining risk level of the influx maneuver is:

where round is a rounding function.

10. The method for personalized guidance of vehicle speed in an acceleration lane according to claim 9, characterized in that in step 5, the difference level is specifically:

the data processor queries a database according to the vehicle type information of the imported vehicle, the relative speed and the relative distance between the imported vehicle and the rear vehicle, and matches an optimal speed curve; and comparing the speed of the current vehicle, and dividing the difference grade into N grades according to the speed difference at the same time point.

Images