CN106114422B - Autonomous car-following system and its control method for the minimum safe inter-vehicle distance - Google Patents

Abstract

Present invention is disclosed a kind of autonomous with car system, system is provided with the front anticollision millimetre-wave radar installed in vehicle front, the front anticollision millimetre-wave radar outputs signal to main control unit through CAN, the main control unit output alarm signal is to warning reminding unit, the main control unit exports brake signal to brake-by-wire unit, the front anticollision millimetre-wave radar both sides are respectively provided with a ultrasonic radar, the ultrasonic radar outputs signal to ultrasonic radar control unit, the ultrasonic radar control unit outputs signal to main control unit through CAN.The advantage of the invention is that the present invention combines inexpensive ultrasonic radar carries out accurate measurement to the relative distance of close-in target, millimetre-wave radar closely range error deficiency bigger than normal is compensate for.

Description

Autonomous car-following system and its control method for the minimum safe inter-vehicle distance

technical field

The invention relates to the technical field of smart cars, in particular to an autonomous car-following system for a smart car and a minimum safety distance control method thereof.

Background technique

With the continuous development of vehicle intelligent technology and users' demand for vehicle intelligent driving, advanced assisted driving of vehicles and intelligent vehicles will be popularized rapidly in the future, bringing a lot of convenience to people's daily life. Among them, the autonomous car-following system can partially replace humans to control the vehicle under high-speed or urban conditions. The current adaptive cruise control system (Adaptive Cruise Control, ACC) can realize autonomous car-following control, and uses 77GHz millimeter-wave radar to detect targets in front in real time. Carry out alarm reminder and brake intervention when danger occurs. The radar detection accuracy is only ±0.5 meters, which makes it impossible to achieve close-range accurate detection. In addition, the current model estimates the warning distance and braking intervention distance generally have large errors, and cannot be accurately estimated in real time based on the driver, vehicle braking performance, and driving road conditions. , leading to an increase in the risk of a collision if the safety distance is too small or a reduction in road driving efficiency if the safety distance is too large.

Contents of the invention

The technical problem to be solved by the present invention is to realize a system that reduces the following distance as much as possible and can ensure driving safety.

In order to achieve the above object, the technical solution adopted in the present invention is: an autonomous car-following system, the system is provided with a front collision avoidance millimeter-wave radar installed in front of the vehicle, and the front collision prevention millimeter-wave radar outputs signals to the main control unit through the CAN bus , the main control unit outputs an alarm signal to the alarm reminder unit, the main control unit outputs a braking signal to the brake-by-wire unit, and an ultrasonic radar is installed on both sides of the front anti-collision millimeter wave radar, and the ultrasonic radar The signal is output to the ultrasonic radar control unit, and the ultrasonic radar control unit outputs the signal to the main control unit through the CAN bus.

A vehicle front-view camera is arranged in front of the vehicle, and the vehicle front-view camera outputs signals to the main control unit through the CAN bus.

A vehicle following detection radar is provided at the rear of the vehicle, and the vehicle following detection radar detects the information behind the vehicle and outputs a signal to the main control unit, and the main control unit outputs a control signal to the display and prompt unit at the rear of the vehicle.

Side detection radars are provided on both sides of the vehicle, and the side detection radars output signals to the main control unit, and the main control unit outputs control signals to the direction control unit.

The control method based on the minimum safe inter-vehicle distance of the autonomous car-following system:

1) Power on the system;

2) Fault self-inspection;

3) If there is a fault, a fault alarm will be issued and the system will be shut down; if there is no fault, the system will start;

4) Detect the status of the vehicle ahead in real time, detect and identify the vehicle ahead through the millimeter wave radar sensor, ultrasonic radar sensor and visual perception system, if the vehicle in front decelerates, perform step 5), otherwise continue to detect the vehicle ahead in real time;

5) Estimate the warning distance _Dw , and then judge, if the measured distance D is greater than or equal to _Dw , repeat step 5);

6) Collision warning, while estimating the braking intervention distance D _z , and then judging, if the measured distance D is greater than or equal to D _z , then repeat step 6, otherwise perform step 7;

7) If the vehicle in front is stationary, the speed of the vehicle is greater than 0 or the speed of the vehicle is greater than the speed of the vehicle in front, perform braking intervention, and then perform step 8), otherwise perform step 6);

8) Estimate the safety distance D _h , if the measured distance DD _h is greater than or equal to 1 meter, the distance is too large, increase the braking deceleration of the vehicle, and perform step 7; if the measured distance D _h -D is greater than or equal to 1 meter, then the distance If it is too small, reduce the braking deceleration of the vehicle and go to step 7; otherwise, keep the braking deceleration of the vehicle and go to step 7.

In the formula, v _b is the current speed of the vehicle (m s ^-1 ); T _d is the driver’s response time (s) and coefficient, T _z is the braking coordination time (s); a _b and k are the vehicle’s Estimated value of braking deceleration (m·s ^-2 ) and coefficient; v _f is the current speed of the vehicle in front (m·s ^-1 ); a _f is the estimated value of braking deceleration of the vehicle in front (m·s ^-2 ); d is the minimum safe distance (m) between the vehicle in front and the vehicle in front after the vehicle in front and the vehicle in front stop.

Assuming that D _h is the estimated value of the safe distance between the vehicle and the vehicle in front during the braking process of the vehicle, and D is the relative distance between the vehicle and the vehicle in front measured periodically by the radar, the minimum safe vehicle distance control model is:

In the formula, t is the radar ranging period s, and d is the minimum safe distance between the vehicle and the vehicle in front after the relative speed between the vehicle and the vehicle in front is eliminated.

In each radar distance measurement period, the vehicle braking deceleration can be adjusted by adjusting the coefficient k so that the measured value D is close to the estimated value D _h .

Collect the distance and speed of the vehicle in front, as well as the speed of the vehicle. If the deceleration of the vehicle in front is too large to make the distance between the vehicle in front and the vehicle in front difficult to meet the braking distance, then detect whether there are vehicles or obstacles on both sides of the vehicle. If there are obstacles or vehicles on both sides , the direction is not actively adjusted. If there is no obstacle or vehicle on one side, the direction of the vehicle is adjusted to that side. If there are no obstacles or vehicles on both sides, the direction of the vehicle is adjusted to the right side of the vehicle.

When actively adjusting the direction of the vehicle, turn on the turn signal on the side of the adjusted direction and sound the horn at the same time.

The advantage of the present invention is that the present invention combines the low-cost ultrasonic radar to accurately measure the relative distance of the short-distance target, and makes up for the shortcoming that the millimeter-wave radar short-range distance measurement error is relatively large. The safety distance estimation model adopted in the present invention adds a self-learning module, and considers the braking performance of different vehicles and the personality differences between drivers, as well as the driving road surface, the braking performance of the front vehicle and the driving state and other influencing factors, and has better adaptability. Safe distance estimation is more accurate. The present invention adopts the minimum safe vehicle distance closed-loop control model, overcomes the hysteresis and failure of the distance estimation method in which the measured value is used as the model input, and the problem that the safe distance estimation generally has a large deviation, and ensures the stability of the minimum safe distance control and robustness, it makes up for the deficiency that the safety distance estimation model is affected by various factors.

At the same time, the invention proposes a closed-loop control method for the minimum safe vehicle distance, which makes up for the shortcoming that the minimum safe distance is too small or too large due to large estimation deviations of the early warning distance and the braking intervention distance.

Description of drawings

The following is a brief description of the content expressed by each piece of accompanying drawing in the description of the present invention:

Figure 1 is a block diagram of the autonomous car-following system;

Figure 2 is a flow chart of distance control.

detailed description

As shown in Figure 1, the system includes a front collision avoidance millimeter-wave radar; a front ultrasonic radar sensor; a front target and environment perception camera; unit and CAN bus. The forward collision avoidance millimeter-wave radar is used to detect the relative distance, speed and azimuth angle of moving and stationary targets in real time, further judge whether the vehicle is dangerous to go straight and the degree of danger, and whether to collect early warning or brake intervention.

Two low-cost ultrasonic sensors are installed in front of the vehicle to cooperate with the forward-looking radar and the forward-looking camera to perceive the road environment ahead. At the same time, based on the existing safety distance model, a safety warning distance estimation model and a braking intervention distance are proposed. The estimation model and the closed-loop control model of the minimum safe vehicle distance after braking intervention enable the automatic following system to ensure driving safety while improving road driving efficiency.

The range of 100 meters in front of the vehicle is detected by the forward-looking millimeter-wave radar, and the front camera is used to detect and identify the targets or obstacles in the front area. Obstacle detection, especially the precise measurement of safety distances. Judging the degree of danger and making decisions, and decelerating and avoiding collisions through smart car braking by wire. Ultrasonic radar is used to detect the relative distance between the target within one meter in front of the vehicle and the vehicle, to make up for the low detection accuracy of the millimeter-wave radar in front, and to improve the distance detection accuracy. The ultrasonic radar controller is used to collect the output signal of the ultrasonic radar and calculate the relative distance between the target and the radar, and calculate the position information of the target relative to the vehicle through the relationship between the installation position of the radar and the vehicle coordinate system.

The vehicle's front-view camera is used to cooperate with the front millimeter-wave radar to further identify the type of target, further judge the degree of danger of the target, and provide a basis for emergency collision avoidance. The information fusion of radar and vision can effectively improve the accuracy and effectiveness of the front environment perception.

In addition to the accurate detection of the relative distance of the target, the intelligent autonomous car following system also needs an accurate safety distance estimation model to estimate the warning distance and braking intervention distance, but the safety distance is not related to the dynamics of driver characteristics, vehicle braking performance and driving conditions. factors, and these factors are difficult to predict, even if predicted by the model, there is a lag. On the basis of the existing safety distance model, the present invention introduces a self-learning module, fully considers the braking performance of different vehicles and the personality differences between drivers, and the influencing factors such as driving road surface, braking performance of the front vehicle and driving state, etc. The safety warning distance model and the braking intervention distance model, as well as the hysteresis and failure of the distance estimation method with the measured value as the input of the model, propose a minimum safe vehicle distance closed-loop control model with the ultimate goal of controlling the minimum safe distance of the vehicle. During the collision avoidance process, the braking deceleration of the vehicle is constantly adjusted, and the safety distance is dynamically closed-loop controlled to ensure that after the relative speed of the vehicle and the vehicle in front is eliminated, the minimum safety distance remains in a constant range, avoiding too large or too small, It not only improves the accuracy of system distance estimation, reduces the risk of anti-collision, but also ensures driving comfort and driving efficiency.

The main control unit of the system communicates with the ultrasonic radar controller, front-view camera, and millimeter-wave radar through the CAN bus, obtains the detection data of each sensor, and establishes a corresponding relationship with the vehicle coordinate system through the installation position of each sensor and their respective coordinate systems. The specific position, relative speed and distance information of each sensor detection target relative to the vehicle is obtained, and provided to the control strategy algorithm for decision-making. The control strategy algorithm outputs early warning information based on the obtained target information combined with the vehicle speed and running state, and controls the alarm device to remind the driver or implement brake intervention by communicating with the brake-by-wire unit of the smart car.

In addition, there is a vehicle following detection radar at the rear of the vehicle, and side detection radars on both sides of the vehicle. The side detection radar and the vehicle following detection radar detect the rear information of the vehicle and output signals to the main control unit, and the main control unit outputs control signals to the vehicle. The rear display prompt unit and direction control unit. The car following detection radar can detect the speed and distance of the vehicle behind, and the prompt unit can be an LED display device, prompting that the vehicle is normally in the automatic cruise state, or displaying the braking force of the vehicle, and giving a warning to the driver behind. The direction control unit can be the steering wheel control device on the automatic driving unit, which can control the direction of the vehicle and is used to link with the automatic cruise. When the vehicle brakes, there may still be a collision, and it can be avoided by steering. Of course, it needs to pass through the side before turning The radar detects whether there are obstacles or vehicles on the side.

As shown in Figure 2, the control method of the minimum safe inter-vehicle distance is as follows:

1) Power on the system

2) The control unit, radar system, visual perception system and brake-by-wire unit perform fault self-check;

3) If there is a fault in the system, a fault alarm will be issued, otherwise execute 4;

4) The status of the vehicle in front is detected, and the vehicle in front is detected and identified through the millimeter wave radar sensor, ultrasonic radar sensor and visual perception system. If the vehicle in front slows down, go to step 5, otherwise continue to detect the vehicle in front in real time;

5) Estimate the warning distance D _w , and then judge, if the measured distance D is greater than or equal to D _w , repeat step 5, otherwise perform step 6;

6) Collision warning, while estimating the braking intervention distance D _z , and then judging, if the measured distance D is greater than or equal to D _z , then repeat step 6, otherwise perform step 7;

7) If the vehicle in front is stationary, the speed of the vehicle is greater than 0 or the speed of the vehicle is greater than the speed of the vehicle in front, perform braking intervention, and then perform step 8, otherwise perform step 6;

8) Estimate the safety distance D _h , if the measured distance DD _h is greater than or equal to 1 meter, the distance is too large, increase the braking deceleration of the vehicle, and perform step 7; if the measured distance D _h -D is greater than or equal to 1 meter, then the distance If it is too small, reduce the braking deceleration of the vehicle and go to step 7; otherwise, keep the braking deceleration of the vehicle and go to step 7.

If it is detected that there is a car behind or the speed of the car following the car is too fast or the distance is too close, the rear car can be prompted through the display unit. on autopilot.

In addition, collect the distance and speed of the vehicle in front, as well as the speed of the vehicle. If the deceleration of the vehicle in front is so large that the distance between the vehicle in front and the vehicle in front cannot meet the braking distance, then detect whether there are vehicles or obstacles on both sides of the vehicle. If there are obstacles on both sides If there are no obstacles or vehicles on one side, the direction of the vehicle will be adjusted to that side; if there are no obstacles or vehicles on both sides, the direction of the vehicle will be adjusted to the right side of the vehicle.

When the car was following the car normally, it suddenly found that the car in front slowed down urgently. The vehicle then adopts braking to decelerate and avoid collisions. The running state of the vehicle in front, that is, the final speed of the vehicle is unpredictable. Therefore, the estimation of the warning distance _Dw and the braking intervention distance _Dz is based on the assumption that the vehicle in front finally stops and the vehicle behind also stops. , then the safety distance estimation model under this working condition is:

In the formula, v _b is the current speed of the vehicle (m s ^-1 ); T _d is the driver’s response time (s) and coefficient, T _z is the braking coordination time (s); a _b and k are the vehicle’s Estimated value of braking deceleration (m·s ^-2 ) and coefficient; v _f is the current speed of the vehicle in front (m·s ^-1 ); a _f is the estimated value of braking deceleration of the vehicle in front (m·s ^-2 ); d is the minimum safe distance (m) between the vehicle in front and the vehicle in front after the vehicle in front and the vehicle in front stop.

Correct the parameter l through the self-learning module to the evaluation results of the vehicle's braking performance, and adjust the preset values of the vehicle's braking deceleration a and _b in the model to match the vehicle's braking performance. According to the evaluation result of the driver's response time by the self-learning module, the preset value of the driver's response time T _d in the model is adjusted. According to the evaluation result of the braking system coordination time by the self-learning module, the preset value of the driver's response time T _z in the model is adjusted. Adjust k through the road surface recognition module to make corrections, and adjust the preset values of the brake deceleration a ^b of the vehicle and the front vehicle in the model to match the current driving road surface. Through the adaptive correction of relevant parameters, the estimation of the safety distance is more reasonable, more in line with the actual application scenario, and the estimation is more accurate, ensuring that the vehicle does not collide while maintaining high driving efficiency, driving comfort and driving experience . Formulas (1)(2) are only applicable to estimate the starting warning distance and braking intervention distance. If the braking deceleration of the vehicle is not adjusted in real time during the deceleration process of the vehicle, the deviation between the minimum safe distance and the ideal distance may be too large or too small after the relative speed of the vehicle and the front vehicle is eliminated, and a high collision avoidance cannot be maintained. Probability and driving efficiency. Keep the minimum safe distance within the ideal range after the relative speed of the vehicle in front and the vehicle in front is eliminated. Aiming at the operating condition of the vehicle after brake intervention, a closed-loop approximation adaptive minimum safe distance estimation control model is proposed. After the vehicle starts to brake and decelerate, the factors that affect the value of d are mainly the braking distance of the front vehicle and the braking distance of the vehicle. Adjust the braking distance by adjusting the braking deceleration of the vehicle to ensure that d is in an ideal range Inside.

Assuming that D _h is the estimated value of the safe distance between the vehicle and the vehicle in front during the braking process of the vehicle, and D is the relative distance between the vehicle and the vehicle in front measured periodically by the radar, the minimum safe vehicle distance control model is:

In the formula, t is the radar ranging period (s), and d is the minimum safe distance between the vehicle and the vehicle in front after the relative speed of the vehicle and the vehicle in front is eliminated (m). Ideally, the relative distance D measured in each ranging period should be equal to D _h is close, keeping within a certain error range. If D is too large, d may be too large, and the vehicle may decelerate sharply, causing discomfort to drivers and passengers, and reducing road driving efficiency. If D is too small, d may be too small, increasing the risk of collision. In the model, the brake deceleration parameters of the front vehicle remain basically unchanged except for adjustments based on the road surface recognition module. Therefore, in each radar distance measurement period, the braking deceleration of the vehicle can be adjusted by adjusting the coefficient k so that the measured value D is close to the estimated value D _h . This estimation method does not need to consider the final running speed of the preceding vehicle. In the actual scene, the practicability is stronger, and the d value is kept within a reasonable range. a _b and a _f are updated according to the relative vehicle speed collected in real time by the millimeter-wave radar and the real-time vehicle speed of the vehicle to form a closed-loop estimation model, and the relative safety distance is adjusted by adjusting the braking deceleration parameter coefficient l of the vehicle in real time, so that the relative vehicle speed is eliminated After that, the relative distance d becomes more reasonable.

Among them, T _d generally ranges from 0.4s to 1.5s, with a typical value of 1s; T _z generally ranges from 200 to 400ms, with a typical value of 300ms; a _b generally ranges from 3.0 to 9.8m·s ^-2 , a typical value: car Braking deceleration; 5.5m·s ^-2 ; truck braking deceleration 3.6m·s ^-2 . The coefficients j, k, l, m are all preset to 1, and the range of variation is 0.5~1.5; the value range of d is 1~2m.

The present invention has been exemplarily described above in conjunction with the accompanying drawings. Obviously, the specific implementation of the present invention is not limited by the above methods, as long as various insubstantial improvements are adopted in the method concept and technical solutions of the present invention, or there is no improvement Directly applying the conception and technical solutions of the present invention to other occasions falls within the protection scope of the present invention.

Claims (6)

1. autonomous with car system, system is provided with the front anticollision millimetre-wave radar installed in vehicle front, and the front anticollision is in the least Metre wave radar outputs signal to main control unit, the main control unit output alarm signal to warning reminding list through CAN Unit, the main control unit exports brake signal to brake-by-wire unit, it is characterised in that：The front anticollision millimetre-wave radar Both sides are respectively provided with a ultrasonic radar, and the ultrasonic radar outputs signal to ultrasonic radar control unit, described super Sound radar control unit outputs signal to main control unit through CAN；

Vehicle front is provided with vehicle forward sight camera, and the vehicle forward sight camera outputs signal to main control list through CAN Unit；

Rear view of vehicle is provided with car detection radar, described to detect rear view of vehicle information with car detection radar and output signal to master control Unit processed, the main control unit outputs control signals to the display reminding unit of rear view of vehicle；

Vehicle both sides are provided with lateral detection radar, and the lateral detection radar simultaneously outputs signal to main control unit, the master control Unit processed outputs control signals to direction controlling unit；

Control method based on the autonomous minimum safe following distance with car system：

1) system electrification；

2) fault self-checking；

3) fault alarm is carried out if it there is failure, and closes system, system starts if in the absence of failure；

4) real-time detection front truck state, by millimetre-wave radar sensor, ultrasonic radar sensor and visually-perceptible system pair Front vehicles are detected and recognized, if front truck slows down, perform step 5), otherwise continue to examine front vehicles in real time Survey；

5) estimation early warning distance D_w, then judge, if whether measured distance D is more than or equal to D_w, then step 5 is repeated)；

6) anti-collision warning, while estimating brake regulation apart from D_z, then judge, if whether measured distance D is more than or equal to D_z, then Repeat step 6), otherwise perform step 7)；

If 7) front truck is static, this car speed is more than 0 or this car speed is more than front truck, then carry out brake regulation, then performs Step 8), otherwise perform step 6)；

8) safe distance D is carried out_hEstimation, if measured distance D-D_hMore than or equal to 1 meter, then apart from bigger than normal, increase the braking of this car and subtract Speed, performs step 7；If measured distance D_h- D is more than or equal to 1 meter, then apart from less than normal, reduce this car braking deceleration, performs Step 7；Otherwise this car braking deceleration keeps, and performs step 7.

2. the control method of the autonomous minimum safe following distance with car system according to claim 1, it is characterised in that：

$D_w=(v_f-v_b)\×T_d+(v_f-v_b)\×T_z+\frac{v_b^2}{2\×k\×a_b}-\frac{v_f^2}{2\×a_f}+d$

$D_z=(v_f-v_b)\×T_z+\frac{v_b^2}{2\×k\×a_b}-\frac{v_f^2}{2\×a_f}+d$

In formula, v_bIt is this car current vehicle speed ms^-1；T_dRespectively driver's response time s, T_zIt is brake coordination time s；a_bAnd k It is this car braking deceleration estimate ms^-2And coefficient；v_fIt is front truck current vehicle speed ms^-1；a_fFor front truck braking deceleration is estimated Evaluation ms^-2；D is the minimum safe distance m after front truck and Ben Che stoppings between this car and front truck.

3. the control method of the autonomous minimum safe following distance with car system according to claim 1, it is characterised in that：It is false If D_hIt is this car in this car braking procedure and front truck safe distance estimate, D is that this car of radar period measurement is relative with front truck Distance, minimum safe spacing Controlling model is：

$D_h=(v_b\×t-\frac{k\×a_b\×t^2}{2})-(v_f\×t-\frac{a_f\×t^2}{2})+d$

T is radar range finding cycle s, this car and front truck minimum safe distance m after d is Ben Che and front truck relative velocity is eliminated in formula.

4. the control method of the autonomous minimum safe following distance with car system according to claim 3, it is characterised in that： The range measurement cycle of each radar this car braking deceleration can be adjusted by regulation coefficient k so that measured value D values close to Estimate D_h。

5. the control method of the autonomous minimum safe following distance with car system according to claim 1, it is characterised in that：Adopt Following distance and speed before collection, and this car speed, if excessive the causing of front truck deceleration is difficult to meet braking distance with preceding following distance, Then detection vehicle both sides whether vehicle or barrier, if there are barrier or vehicle in both sides, not to direction active accommodation, if its Middle side clear or vehicle, then adjust this car direction to the side, if the equal clear in both sides or vehicle, to this car right side Adjust this car direction.

6. the control method of the autonomous minimum safe following distance with car system according to claim 5, it is characterised in that：When During active accommodation this car direction, the steering indicating light of institute's adjustment direction side is opened, while blowing a whistle.