CN105059184B - The rollover early warning of passenger stock bend and actively prevention and control device and determination methods thereof - Google Patents

Abstract

The invention discloses an early warning and active prevention and control device and a judging method for passenger vehicle curve rollover, including a digital signal processor, the input end of the digital signal processor is connected to an image acquisition module and a signal acquisition module; the described The signal acquisition module is connected to the vehicle speed sensor through the data line P1; the output end of the digital signal processor is connected to the signal output module, liquid crystal display, storage hard disk and alarm module. The present invention uses a signal processor to compare the real-time vehicle speed signal with the calculated critical vehicle speed. According to the size relationship between the two, the alarm module voice prompts the driver's vehicle speed, makes a corresponding early warning of rollover risk, and takes vehicle preparation in advance. Measures to achieve the purpose of actively preventing rollovers and reduce the casualty rate of passenger vehicle accidents; achieve the combination of early warning and active prevention and control of rollover risks, which will help improve the safety assurance capabilities of my country's highway passenger transport enterprises.

Description

Early warning and active prevention and control device and judgment method for passenger vehicle curve rollover

technical field

The invention belongs to the field of active safety of passenger transport vehicles, and relates to an early warning and main prevention and control device and a judging method for passenger vehicle curve rollover.

Background technique

Passenger vehicle rollover is a relatively common traffic accident. Excessive centrifugal force causes the vehicle to lose its driving stability and topple over. The proportion of rollover in traffic accidents is not high, but the casualty rate caused by rollover is much higher than the incidence of rollover accidents. Especially once the rollover of passenger vehicles occurs, serious casualties and economic losses often occur.

Rollover usually occurs after the vehicle enters a curve, and the research and development of rollover warning or rollover warning at home and abroad often rely on the signal provided by the sensor and the preset threshold. When reaching a certain value, the anti-rollover system will work. The early warning and prevention and control measures are too late to provide enough time for the driver to react. The driver is often nervous in the face of sudden warnings, and there is a certain risk of misjudgment or misoperation, which is easy to interweave with the movement of other vehicles and cause secondary accidents. ACCIDENT. Due to the high speed of current vehicles, especially highway vehicles, the existing anti-rollover technology has the following problems: first, there is no early warning of rollover risk sections; second, anti-rollover tendencies that may be required on risky sections The necessary vehicle preparations cannot be made before the operation; the third is that the calculation amount of the controller in the short moment before the danger is too large, and there are certain missed and false alarms; the fourth is that there is no record of the road ahead and the driver's operation behavior.

Contents of the invention

In view of the above-mentioned problems and defects in the prior art, the purpose of the present invention is to provide an early warning and active prevention and control device for passenger vehicle curve rollover, which can allow passenger vehicles to identify risky curves before entering a rollover risky curve Before entering the curve, it will automatically take certain vehicle preparation measures to reduce the rollover risk of passenger vehicles driving in the curve, and achieve the combination of early warning and active prevention and control; and improve the algorithm, Reduce the amount of calculation, and save the driver's driving trajectory and other vehicle movement information in the field of view of the CCD camera on the road to the hard disk for timely reading.

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

The early warning and active prevention and control device for passenger vehicle curve rollover includes a digital signal processor, the input end of the digital signal processor is connected to the image acquisition module and the signal acquisition module; the signal acquisition module is connected through the data line P1 vehicle speed sensor;

The output end of the digital signal processor is connected to the execution signal output data line, the liquid crystal display, the storage hard disk and the alarm module, and the execution signal output data line is connected to the execution module.

The storage hard disk is connected to the image acquisition module.

Specifically, the image acquisition module includes a CCD camera and an image decoder installed inside the front windshield of the passenger vehicle, and the CCD camera is connected to the image decoder.

Further, the digital signal processor, the signal acquisition module, the liquid crystal display, the alarm module and the storage hard disk are all fixed in a metal casing through a card slot or bolts, and the metal casing is fixed on the instrument panel of the passenger vehicle through brackets and bolts above.

Specifically, the signal output data lines include three signal output data lines, which are data line Q1, data line Q2 and data line Q3, and the execution module includes a retarder switch, a semi/active suspension switch and a power assist Turning to the system, the three are respectively connected to the data line Q1, the data line Q2 and the data line Q3.

A method for judging early warning and active prevention and control of rollover of a passenger car on a curve, specifically comprising the following steps:

Step 1: Calibrate the CCD camera to obtain the internal parameters and external parameters of the CCD camera;

Step 2: Utilize the CCD camera to collect the image of the road ahead, and transmit it to the digital signal processor 1 through the image decoder;

Step 3: Preprocessing the road image collected in step 2;

Step 4: extracting the lane lines of the preprocessed road image obtained in step 3;

Step 5: Carry out curve recognition on the ratio relationship of the parameters of the lane line equation obtained in step 4, and judge whether the vehicle is about to enter the curve; if the vehicle is about to enter the curve, perform step 6; otherwise, go to step 2;

Step 6: Calculate the curve radius R of the lane line obtained in step 4;

Step 7: According to the curve radius R obtained in step 6, find the critical speed of the anti-rollover behavior when driving on the curve. If the critical speed is greater than the maximum speed of the passenger vehicle, go to step 2, otherwise go to step 8;

Step 8: Collect the actual real-time speed of passenger vehicles, and judge whether there is a rollover risk in the curve ahead based on the comparison with the critical speed, and implement graded early warning and rollover prevention and control measures.

Specifically, the specific implementation method of the step 4 is as follows:

Step 4.1: The 1/4 area below the road image is set as the first detection image segment, which is divided into ROIA and ROIB. During the extraction process of the ROI left and right lane lines of the first detection image, search for the left and right lane lines within a certain range; Extract the left and right lane lines in ROIA and ROIB, the middle point of the line connecting the endpoints of the two lanes is a, and the slope mean value K of the left and right lanes of ROIA and _ROIB ;

Step 4.2: Start from a and use K ₁ as the slope to make the dividing line of the region of interest in the next image. The two sides of the dividing line are ROIC and ROID respectively. The left and right lane lines are detected in these two areas. Two lanes The middle point of the connection between the endpoints on the line is b, and the slope mean value of the left and right lanes of ROIC and ROID is K ₂ ;

Step 4.3: Start from b and use K ₂ as the slope to divide the region of interest in the next segment of the image, and so on until the lane lines of all ROIs in the six segment images of the road image are detected.

Specifically, the concrete realization method of described step 8 is as follows:

By collecting the actual real-time speed V of passenger vehicles, and judging by comparing it with the critical speed, an early warning is implemented in grades; the specific method is as follows:

V≤0.8V _T , the voice reminds the driver that the vehicle is about to enter a curve;

0.8V _T < V ≤ V _T , the driver will be prompted by voice that the vehicle is about to enter a curve, and the current speed and critical speed will be displayed on the display screen, and the risk of rollover will be notified, and the three-level risk automatic response operation will be carried out automatically, that is, Start the on-board retarder and remind the driver that the retarder has been automatically turned on;

1.1 V _T > V > V _T , the voice prompts the driver that the vehicle speed is too high, and displays the current vehicle speed and critical vehicle speed on the display screen, and automatically performs the second-level risk automatic response operation, that is, on the basis of the third-level response operation, Remind the driver to adjust the speed of the vehicle and it is recommended to mount it to resist, such as a passenger vehicle equipped with a semi/active suspension;

V≥1.1V _T , voice prompts the driver that the vehicle speed is too high, and displays the current vehicle speed and critical vehicle speed on the display screen, and will automatically perform the first-level risk automatic response operation, that is, on the basis of the second and third-level response operations, assist The steering provides 5 percent negative assist.

Compared with the prior art, the present invention has the following technical effects:

1. The present invention uses a signal processor to compare the real-time vehicle speed signal with the calculated critical vehicle speed. According to the size relationship between the two, the alarm module voice prompts the driver's vehicle speed, makes a corresponding early warning of rollover risk, and passes the signal The output module turns on the relevant equipment of passenger vehicles in advance, so as to prepare in advance for the possible anti-rollover behavior after entering the curve, and reserve time for the possible anti-rollover operation of passenger vehicles on the curve, which can give the driver certain The psychological preparation of the vehicle and the completion of the vehicle preparation, the combination of early warning and active prevention and control, to achieve the purpose of prevention and control of rollover, reduce the casualty rate of passenger vehicle accidents, and help improve the safety guarantee capabilities of my country's road passenger transport enterprises.

2. The present invention adopts the method of combining subdivided areas to extract lane lines, so as to extract lane lines subdivided into regions, and the amount of calculation is relatively reduced, and the method of dynamically dividing ROI is also beneficial to reduce the amount of calculation, and to a certain extent improve The real-time and stability of the algorithm.

3. The present invention sets the storage hard disk for storing the vehicle trajectory of the driver's emergency operation and other vehicle movement information in the field of view of the CCD camera. Once a traffic accident occurs, it can provide a basis for judgment by the traffic control department.

Description of drawings

Figure 1 is a schematic diagram of the installation of a CCD camera;

Fig. 2 is the hardware structure schematic diagram of the present invention;

Fig. 3 is a combined ROI dynamic partition diagram;

Fig. 4 is method flowchart of the present invention;

The symbols in the figure represent: 1—digital signal processor, 2—image acquisition module, 3—signal acquisition module, 4—liquid crystal display, 5—alarm module, 6—storage hard disk, 7—execution signal output data line, 8—vehicle speed Sensor, 9—metal shell, 10—power switch.

The solution of the present invention will be further explained and described in detail in conjunction with the accompanying drawings and embodiments.

detailed description

According to the above-mentioned technical scheme, referring to Fig. 1 and 2, the early warning and active prevention and control device for passenger vehicle curve rollover of the present invention includes a digital signal processor 1, and the input end of the digital signal processor 1 is connected to an image acquisition module 2 and signal acquisition module 3; described signal acquisition module 3 is connected vehicle speed sensor 8 by data line P1; The output terminal of described digital signal processor 1 is connected execution signal output data line 7, liquid crystal display 4, storage hard disk 6 and The alarm module 5, the execution signal output data line 7 is connected to the execution module.

Specifically, the image acquisition module includes a CCD camera and an image decoder. The CCD camera is installed inside the front windshield of the passenger vehicle, obliquely pointing to the lane ahead, and the CCD camera is connected to the image decoder. The installation height is h, the inclination angle is α, and the lens faces forward slightly downward.

The present invention utilizes the CCD camera to shoot the instantaneous image of the road ahead of the vehicle, and transmits the image signal to the digital signal processor 1 after passing through the image decoder; the digital signal processor 1 judges whether there is a curve in the road ahead according to its built-in program. When there is a curve on the road ahead, the critical speed at which the passenger vehicle rolls over is calculated.

The signal acquisition module 3 collects the real-time vehicle speed signal measured by the vehicle speed sensor 8 and transmits it to the digital signal processor 1 .

The digital signal processor 1 compares the real-time vehicle speed signal with the calculated critical vehicle speed, and uses the alarm module 5 to voice prompt the driver's vehicle speed according to the size relationship between the two, make corresponding early warning of rollover risk, and output data through the signal Harness 7 opens passenger vehicle-related equipment in advance, such as retarder, semi/active suspension, and power steering system, so as to prepare for the possible anti-rollover behavior after entering a curve, and prepares for passenger vehicles that may The time reserved for the anti-rollover operation can give the driver a certain amount of psychological preparation and complete vehicle preparation, so as to prevent rollover and reduce the casualty rate of passenger vehicle accidents. It is helpful to improve the safety and security capabilities of my country's highway passenger transport enterprises.

The storage hard disk 6 is used to store the driver's driving trajectory and other vehicle movement information in the field of view of the CCD camera on the road. Once a traffic accident occurs, it can provide a basis for judgment by the traffic control department.

The liquid crystal display is used to display the danger level and the critical speed value to remind the driver to drive safely, and the liquid crystal display is synchronized with the alarm module in terms of time and content.

Optionally, the CCD camera adopts a SONY 1/4 "machine vision industrial camera, the model is WAT-232, and the model of the image decoder is TVP5150. The output of the image decoder is connected to the digital The signal processor 1 is connected, and the image decoder adopts TVP5150.

Specifically, the alarm module 5 includes a voice chip and a loudspeaker.

Further, the digital signal processor 1, the signal acquisition module 3, the liquid crystal display 4, the alarm module 5 and the storage hard disk 6 are all fixed in the metal casing 9 by a card slot or bolts, and the metal casing 9 is fixed by a bracket and bolted to the top of the dashboard of the passenger vehicle.

A power switch 10 is installed on the metal shell 9, and the power switch 10 is connected with the digital signal processor 1, the signal acquisition module 3, the liquid crystal display 4, the alarm module 5 and the storage hard disk 6 through respective power conversion circuits , the power conversion circuit is used to regulate the voltage to supply power for each module.

Specifically, the signal output data line 7 includes three signal output data lines, which are data line Q1, data line Q2 and data line Q3, and the execution module includes a retarder switch, a semi/active suspension switch and the power steering system, the three are respectively connected to the data line Q1, the data line Q2 and the data line Q3.

Optionally, the digital signal processor 1 adopts TMS320DM6437, the model adopted by the CCD camera is SONY WAT-232, the signal acquisition module 3 adopts MC9S12DG128B, the liquid crystal display 4 adopts TVG-703, and the early warning module 5 The speech chip in adopts ISD4004-08, and loudspeaker adopts 2 inch loudspeaker, and storage hard disk 6 adopts Seagate STEA2000400; Described vehicle speed sensor 8 adopts KF-01004 Hall vehicle speed sensor.

The early warning and active prevention and control judging method of passenger car vehicle curve rollover according to the present invention, as shown in Fig. 4, specifically includes the following steps:

Step 1: Calibrate the CCD camera

Before the device is installed, the CCD camera needs to be calibrated to obtain the internal parameters and external parameters of the CCD camera.

Firstly, a calibration board is made, on which a calibration figure is set, and the calibration figure is a black and white square frame with a side length of 10 cm. CCD cameras are used to collect images from different angles of the calibration plate, and the collected images are calibrated using matlab software to obtain the internal parameters and external parameters of the CCD camera. The internal parameters mainly include: effective focal length f, distortion coefficient fc and distortion coefficient kc, these parameters reflect the distortion of the image by the camera itself. The external parameters obtained by camera calibration include the height h of the camera from the ground, the distance d from the side of the vehicle, the rotation angle γ, and the pitch angle α. After the CCD camera is calibrated, fix the device, and turn on the power switch to start the system during actual work.

Step 2: Acquire road images

Utilize the CCD camera to collect the image of the road ahead, and send it to the digital signal processor 1 through the image decoder.

Step 3: Preprocessing the road image

Because the actual road conditions are generally more complicated, the road image captured by the CCD camera in step 2 has oil stains and other disturbances, so the digital signal processor 1 needs to preprocess the road image to remove the interference points and useless points on the image. point, and increase image contrast.

Step 3.1: Use Wiener filtering to perform filtering processing on the road image to filter out some random noises. The Wiener filter is a modern filter, which is still applicable when the signal and interference frequency bands overlap, and has an obvious filtering effect on white noise. At the same time, Wiener filtering has a relatively small amount of calculation and storage space among various modern filters, which is beneficial to the real-time calculation of the system.

Step 3.2: In the road image captured by the CCD camera, the gray level difference between objects may be small, which makes the contrast of the image poor. The invention utilizes the histogram to correct the distance between the gray scales of the image, which is beneficial to enhancing the overall contrast of the road image. The entire gray value of the image is adjusted within the interval (0,255).

Step 4: Combination of subdivided areas to extract lane lines

The processed lane image obtained in step 3 is combined by region to extract lane lines. Usually the top 1/4 area of the road image is a natural landscape. In order to reduce the amount of calculation, the present invention uses the bottom 3/4 area of the image as the initial ROI (Region Of Interest) candidate area, and divides the image into multiple areas, as shown in Figure 3 shown. The lower 3/4 area of the road image is divided into four areas according to the slope characteristics of the lane lines. In order to further reduce the amount of calculation, the straight line detection of the left and right lane lines is processed separately, and each area is subdivided into multiple areas according to their respective characteristics. In the ROI area, the lane lines in each ROI are regarded as straight lines due to the large number of partitions. Area I is located in the bottom 1/4 of the image, and is equally divided into left and right ROI areas. The lane line in this area is generally a straight line; area II occupies 1/4 of the image width. If there is a curve in the image in this area, the slope does not change. It will be too large, divide this area into two parts with the same width, and set two left and right ROI areas for each block; Area III occupies 1/8 of the width of the road image, because the lane line at this position generally has a higher slope and the lane The length of the line is short, and it is better to further subdivide it, so divide this area into equal width, and set ROI on the left and right; the IV area occupies 1/8 of the width of the image, because the probability of lane lines appearing in this area is small And the slope of the line segment is relatively short. In order to avoid the amount of computation, it is no longer divided into equal widths, and only two left and right ROIs are drawn in this area.

The present invention adopts the method for dynamically dividing the ROI area, and the lane line slope mean value and the left and right lane line endpoints detected in the two ROIs of the previous area are used as the basis for dividing the ROI areas on both sides of the next area, which is beneficial to the ROI division on the basis of Quickly detect lane lines. The method of extracting lane lines by combining subregions is beneficial to reduce the amount of calculation, and improve the real-time and stability of the algorithm to a certain extent. The specific implementation steps are as follows:

Step 4.1: The 1/4 area below the road image is set as the first ROI (ie ROIA and ROIB). In the process of detecting the left and right lane lines of the first ROI, set the initial left lane line search angle range from 40 degrees to 70 degrees, the right lane line search angle range is determined to be 110 degrees to 160 degrees, which helps to narrow the search range and reduce the calculation amount of the digital signal processor; extract the left and right lane lines in ROIA and ROIB, and the two lane lines The middle point of the connecting line between the endpoints is a, and the slope mean K ₁ of the left and right lanes of ROIA and ROIB;

Step 4.2: Since the lane line is continuous and there is no distortion, start from a and use K ₁ as the slope to divide the region of interest in the next image. The two sides of the division line are ROIC and ROID respectively. The left and right lane lines are detected in the area, the middle point of the line connecting the end points of the two lane lines is b, and the slope mean value K ₂ of the left and right lanes of ROIC and ROID.

Step 4.3: _Start from b and use K2 as the slope to divide the region of interest in the next segment of the image, and so on until the lane lines of all ROIs in the six segments of the road image are detected.

Using the above method, multiple areas are combined and divided, which not only reduces the amount of calculation but also ensures the fineness of lane line extraction; in the process of lane line extraction in each area, the average value of the slope of the left and right lane lines is used as the next area of interest. The division basis, the above design fully considers the changing characteristics of the lane line when the vehicle turns, so it is more suitable for the actual application, making the extracted lane line more accurate, ensuring the accuracy of the calculation of the lane radius and critical speed; at the same time, gradually narrowing the search range In this way, the accuracy and real-time performance of the algorithm are increased.

Wherein, the specific method for extracting the left and right lane lines is as follows:

(a): Use the Sobel operator to detect the edge of the lane line, and obtain the lane line image after the edge line of the lane is enhanced. The image has pixel gray level discontinuity, especially at the edge where the pixel gray level changes greatly. Sobel operator can be used to detect the sudden edge line of the image pixel and strengthen the edge line. The edge detection effect of the Sobel operator is obvious, and it is less affected by noise. It can quickly detect the edge line segments in the image, including the left and right lane lines. The Sobel operator is determined as follows, SR is the edge detection operator of the right lane line, and SL is the edge detection operator of the left lane line:

$\begin{array}{cc}{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; =</span>\left[\begin{array}{ccc}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; 1</span>}\\ <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; 2</span>}\\ <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right]& {\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}<span\; class="notranslate">\; =</span>\left[\begin{array}{ccc}\mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}\\ \mathrm{<span\; class="notranslate">\; 2</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}\\ <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right]\end{array}$

(b): Image binarization. Carry out binarization processing on the enhanced lane line image of the road edge line obtained in step (a), and set the gray value of the pixel on the image to 0 or 255. Images can appear distinctly black and white, with white edge lines and an entirely black background. The specific method is as follows: use the dynamic threshold segmentation method to segment the enhanced lane line image of the road edge line obtained in step (a), use a dynamic threshold to classify the gray histogram of the image, and set the gray level below the threshold to 0, the gray level above this threshold is set to 255, so that the edge lines appear in the image.

The method of determining the dynamic threshold is as follows: through the first 3 frames of images, find the maximum value of the pixels in the three frames of road images, and use the built-in program to obtain the average value Pm of the maximum gray value of the three frames of images, and calculate it in the 1/6 area below the image. Road average gray value Pd, dynamic threshold Pt=1/5|Pm-Pd|. The system automatically updates the dynamic threshold every 6 hours.

(c): Probabilistic Hough transform is performed on the binarized image in step (b) to extract lane lines. The Hough transform can transform the entire curve or straight line in the original road image into a point in the parameter space. The method of detecting curves or straight lines in the original road image can be replaced by the method of identifying peak points in the parameter space. Convert the image straight line into a polar coordinate equation: ρ=xcosθ+ysinθ. The specific steps of the probability Hough transform are as follows:

The parameter space of each ROI is divided into several small areas, and an accumulator is set corresponding to each area, and the initial value is 0.

Map a point ( _xi , y ₎ on the edge of the image to the polar coordinate formula, calculate the corresponding ρ, add 1 to the accumulator, and then delete the point from the point set.

Process all (x _i , y _i ) in each area until the image edge point set is empty, and judge whether the value of each accumulator is greater than the threshold thr, if greater, it means that there is a meaningful line segment.

The line segment parameters in the image are jointly determined by ( _xi , y _i ) and (ρ, θ), and the broken parts are connected.

Step 5: Curve Judgment

Use the ratio relationship of the parameters of the lane line equation obtained in step 4 to determine whether the vehicle is about to enter the curve.

The slopes of the equations of the left and right straight line segments identified in each ROI area are respectively denoted as K _L1 , K _L2 , K _L3 , K _L4 , K _L5 and K _R1 , K _R2 , K _R3 , K _R4 , K _R5 from bottom to top, and denoted as P _L1i =|K _L1 /K _Li |, P _R1i =|K _R1 /K _Ri |, i≥2.

Considering the real-time requirements of the system, the algorithm must have a very low time complexity. The present invention uses the method of finding the maximum value of the slope ratio to preliminarily identify whether there is a curve ahead.

Max[P _L1i ]≤1.5 and Max[P _R1i ]≤1.5 _{, the} system defaults that the lane line in the distant view is a straight road, the system does not perform the subsequent steps, and goes to step 2.

Max[P _L1i ]＞1.5 or Max[P _R1i ]＞1.5, the system defaults that the lane line in the distant view has a curve, go to step 6.

Step 6: Calculate the curve radius of the lane ahead

Extract three pixel point groups on the lane line segment obtained in step 4, the specific operation method is as follows: as shown in Figure 3, starting from the top, the intersection of the (right) lane line in the adjacent image segment is J, K, L, M. Extract the pixel point group of the lane line image at the two points J and K and the pixel point group at the position of the midpoint between J and K, and calculate the coordinate mean values of the above three pixel point groups in the image coordinates, respectively (x ₁ ,y ₁ ),(x ₂ ,y ₂ ),(x ₃ ,y ₃ .) If there is no lane segment between J and K or the lane segment is too short, at two points K, L and the midpoint between them Extract pixel point groups, and so on.

According to the CCD sensor parameters and the installation height and pitch angle of the sensor, the conversion formula between the image coordinates and the world coordinates is obtained. The coordinates on the world coordinates obtained after transforming the three points on the image coordinates are (X1, Y1), (X2, Y2) ,(X3,Y3).

Use the general formula of the circle equation X ² +Y ² +DX+EY+F=0, put the coordinates of three points into the above formula to get the values of D, E, F, and thus get the radius of the curve

Due to the vibration of the vehicle during actual driving and certain accidental error factors, the average value of the radius values R ₁ , R ₂ , and R ₃ obtained from three consecutive frames of images is taken as the value of the radius R of the curve. The selected three frames of images are: the image of the curve detected for the first time and the next two frames of images.

Step 7: Critical Vehicle Speed Calculation

According to the curve radius R obtained in step 6, calculate the critical speed of the anti-rollover behavior in the curve, and display it on the LCD.

use the formula Obtain the critical vehicle speed VT, b is the wheelbase of the passenger vehicle, and h is the height of the center of mass of the passenger vehicle. Vmax is the maximum speed of the passenger vehicle, if V _T ≥ V _max , go to step 2; if V _T < V _max , go to step 8.

Step 8: Risk level early warning and rollover prevention and control measures

By collecting the actual real-time speed V of passenger vehicles and comparing it with the critical speed, it is judged whether there is a risk of rollover in the curve ahead, and early warning and early prevention and control measures for rollover are implemented.

In order to avoid the occurrence of rollover events, the relevant devices can be activated before entering the curve, which can shorten the operation time of the possible anti-rollover behavior. By collecting the actual real-time speed V of passenger vehicles, and judging by comparing with the critical speed, an early warning is implemented by grade. The specific method is as follows:

V≤0.8V _T , the voice reminds the driver that the vehicle is about to enter a curve.

0.8V _T < V ≤ V _T , the driver will be prompted by voice that the vehicle is about to enter a curve, and the current speed and critical speed will be displayed on the display screen, and the risk of rollover will be notified, and the three-level risk automatic response operation will be carried out automatically, that is, Start the on-board retarder and remind the driver that the retarder has been automatically turned on, which can avoid the excessive speed of passenger vehicles in the curve and allow the driver to control the speed within a reasonable range.

1.1 V _T > V > V _T , the voice prompts the driver that the vehicle speed is too high, and displays the current vehicle speed and critical vehicle speed on the display screen, and automatically performs the second-level risk automatic response operation, that is, on the basis of the third-level response operation, To remind the driver to adjust the vehicle speed, it is recommended to put it on the block. For example, if the passenger vehicle is equipped with a semi/active suspension, activate the semi/active suspension to improve the anti-roll performance of the passenger vehicle when driving in a curve.

V≥1.1V _T , voice prompts the driver that the vehicle speed is too high, and displays the current vehicle speed and critical vehicle speed on the display screen, and will automatically perform the first-level risk automatic response operation, that is, on the basis of the second and third-level response operations, assist The steering system provides 5% negative power assist to delay the driver's steering behavior in time and prevent excessive instantaneous steering angle.

Claims (2)

1. a Passenger Vehicle bend rollover early warning and actively prevention and control determination methods, it is characterised in that specifically include following Step:

Step 1: demarcate ccd video camera, obtains inner parameter and the external parameter of ccd video camera；

Step 2: utilize ccd video camera to gather road ahead image, and be sent to digital signal processor by image decoder In；

Step 3: the road image collecting step 2 carries out pretreatment；

Step 4: the lane line of the pretreated road image that extraction step 3 obtains；

Step 5: the lane line linear equation aspect ratio relation obtaining step 4 carries out bend identification, it is judged that whether vehicle is Bend will be sailed into；If vehicle will sail bend into, then perform step 6, otherwise, go to step 2；

Step 6: the turning radius R of the lane line that calculation procedure 4 obtains；

Step 7: the turning radius R obtained according to step 6, seeks the critical speed in negotiation of bends anti-rollover behavior, such as critical car Speed more than passenger stock max. speed, then performs step 2, otherwise performs step 8；

Step 8: gather the actual speed in real time of passenger stock, judge with the contrast of critical speed according to it, it is judged that in curve ahead Travel and whether there is rollover risk, implement graduation early warning and prevention and control rollover measure；

The concrete methods of realizing of described step 4 is as follows:

Step 4.1: below road image, 1/4 region is set as that first paragraph detects image sections, is divided into ROIA and ROIB, in detection During about the ROI lane line of first paragraph image extracts, search left and right lane line within the specific limits；Extract ROIA and Left and right lane line in ROIB, the slope in the track, left and right that line midpoint is a, ROIA and ROIB of two lane line upper extreme points is equal Value K₁；

Step 4.2: from a and with K₁The dividing line of area-of-interest in next section of image, the both sides of this dividing line are made for slope Being respectively ROIC and ROID, go out left and right lane line at the two region detection, the line midpoint of two lane line upper extreme points is b, Slope average K in the track, left and right of ROIC and ROID₂；

Step 4.3: from b and with K₂Do the dividing line of area-of-interest in next section of image for slope, the rest may be inferred until profit Till the lane line detecting six sections of all ROI in region of road image.

2. Passenger Vehicle bend rollover early warning as claimed in claim 1 and actively prevention and control determination methods, it is characterised in that The concrete methods of realizing of described step 8 is as follows:

By gathering the actual vehicle velocity V in real time of passenger stock, judge according to the contrast with critical speed, implement graduation the most pre- Alert；Concrete grammar is as follows:

V≤0.8V_T, voice reminder driver vehicle will enter bend；

0.8V_T＜ V≤V_T, voice message driver vehicle will enter bend, and current vehicle speed and critical speed are shown aobvious In display screen, and inform existence rollover risk, automatically tackle operation by automatically carrying out tertiary risk, i.e. start vehicle-mounted retarder, and Driver's retarder is reminded to automatically turn on；

1.1V_T＞ V ＞ V_T, voice message driver's speed is too high, and current vehicle speed and critical speed are shown on a display screen, Automatically tackle operation by automatically carrying out light breeze danger, i.e. on the basis of three grades of reply operations, remind driver to regulate speed and build View is linked into keeps out；

V≥1.1V_T, voice message driver's speed is the highest, and current vehicle speed and critical speed are shown on a display screen, will be from The dynamic prime risk that carries out tackles operation automatically, and i.e. on the basis of two, three grades of reply operations, servo steering system provides 5% Negative power-assisted.