An adaptive AEB control method for unmanned vehicles

Abstract

The present invention belongs to the technical field of active vehicle safety control, and specifically relates to an adaptive AEB control method for unmanned vehicles. The present invention employs a two-step screening logic to decouple the target objects in both the lateral and longitudinal directions. First, obstacles in the direction of the vehicle's course are screened from the targets in front of the vehicle based on the lateral distance between the vehicle and the target. A lateral collision time algorithm is then used to screen potential obstacles in the direction of the vehicle's course from the targets in front of the vehicle, thereby achieving accurate obstacle screening. Then, dangerous obstacles are identified based on the longitudinal collision time algorithm, and the braking intervention timing and control strategy are adaptively selected based on the dangerous interval of the dangerous obstacle's longitudinal collision time. This allows the vehicle to have better adaptability in complex scenarios and working conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle active safety control, and in particular relates to an adaptive AEB control method for an unmanned vehicle. Background Art

[0002] With the development of the automotive industry, driverless vehicles have become a hot topic in the field of automotive research. The safety of driverless vehicles is the primary issue to be considered in the research of driverless vehicles.

[0003] Automatic emergency braking (AEB), a key obstacle avoidance technology in the field of active safety technology for autonomous vehicles, is a crucial function in active safety. The AEB system consists of an environmental perception module, a control unit, and an actuator. The environmental perception module can typically be a visual sensor, lidar, millimeter-wave radar, or other sensor. The actuator is typically the vehicle's braking system, and may also include an alarm device. Its basic principle is to use the environmental perception module to collect environmental information around the autonomous vehicle, including information about surrounding vehicles, pedestrians, and other obstacles, and send this information to the control unit. Based on the information sent by the environmental perception module, the control unit determines whether the autonomous vehicle will collide with surrounding obstacles. If a collision is determined, it sends a corresponding braking target request to the vehicle's braking system, causing the vehicle's braking system to complete the braking process, or controls the alarm device to sound an alarm.

[0004] Research on AEB systems has attracted the attention of many researchers. Initially, the most common control strategy was rule-based design. This control strategy mainly depends on human exhaustive enumeration and design experience under various working conditions, and continuously improves the knowledge base based on technological advancements. It also requires multiple calibrations of working condition parameters to ensure accuracy. In recent years, commonly used algorithms include safe distance algorithms, safe time TTC algorithms, and a fusion algorithm of the two. Their research has shown good braking effects in specific scenarios with longitudinal obstacles. However, most studies only consider the situations in some scenarios and cannot adapt to some emergencies in complex and changing scenarios.

[0005] For example, a Chinese invention patent application with application publication number CN111332264A discloses an automatic emergency braking control method. This method uses first and second specific information obtained in two ways, combined with known vehicle data for analysis, to determine whether emergency braking should be performed. This method can avoid misidentification or missed identification of targets based on information obtained in a single way, which in turn leads to erroneous or missed operations. It effectively improves the accuracy and reliability of target identification, thereby enhancing driving safety and comfort and improving the user experience. However, this method can only meet specific working conditions. In complex and changing scenarios, the braking timing will be inaccurate, and it lacks good adaptability. Summary of the Invention

[0006] The purpose of the present invention is to provide an adaptive AEB control method for unmanned vehicles to solve the problem of poor adaptability of existing technical methods due to inaccurate braking timing in the face of changing and complex scenes.

[0007] To solve the above technical problems, the present invention provides an adaptive AEB control method for an unmanned vehicle, comprising the following steps:

[0008] 1) Obtain information about the vehicle and the target in front of it;

[0009] 2) Based on the information of the ego vehicle and the target object, the lateral distance between the ego vehicle and the target object is calculated. If the lateral distance between the ego vehicle and the target object is less than or equal to the ego vehicle's safety distance, the target object is determined to be an obstacle in the ego vehicle's direction. Based on the information of the ego vehicle and the target object, the lateral collision time TTC between the ego vehicle and the target object is calculated. y If the lateral distance between the vehicle and the target is greater than the set safety distance, and the lateral collision time TTC between the vehicle and the target is greater than the set safety distance, y Less than or equal to the lateral collision time threshold TTC yth , the target object is determined to be a potential obstacle to the vehicle's heading; obstacles to the vehicle's heading and potential obstacles to the vehicle's heading are collectively referred to as obstacles;

[0010] 3) Calculate the longitudinal collision time TTC between the vehicle and each obstacle based on the vehicle and obstacle information x , select the longitudinal collision time TTC x The shortest obstacle is considered a dangerous obstacle; if the longitudinal collision time TTC between the vehicle and the dangerous obstacle is x Less than or equal to the first longitudinal collision time threshold TTC xth1 , determine the longitudinal collision time TTC of the dangerous obstacle x The danger zone in which the vehicle is located, wherein a longitudinal collision time range constitutes a first-level danger zone, and the first-level danger zone corresponds to a braking strategy; the vehicle is braked according to the braking strategy corresponding to the danger zone;

[0011] The longitudinal direction is parallel to the lane line of the lane where the vehicle is located, and the transverse direction is perpendicular to the lane line.

[0012] The beneficial effects are as follows: The present invention designs a two-stage screening logic to decouple the target objects in the horizontal and vertical directions. First, based on the lateral distance between the vehicle and the target object, the vehicle's heading obstacles are screened from the target objects in front of the vehicle. Then, using the lateral collision time algorithm, potential vehicle heading obstacles are screened from the target objects in front of the vehicle as obstacles, thereby achieving accurate obstacle screening. Then, based on the longitudinal collision time algorithm, dangerous obstacles are found, and based on the dangerous interval of the dangerous obstacle's longitudinal collision time, the braking intervention timing and control strategy are adaptively selected, thereby making the vehicle more adaptable in the face of complex scene conditions. Moreover, through the adaptive calculation of the lateral and longitudinal collision time thresholds, the lateral and longitudinal decoupling of multiple targets can be achieved, with high adaptability. While meeting safety requirements, the multi-stage braking allows the driver to enjoy better comfort.

[0013] Furthermore, in step 2), if the lateral distance between the vehicle and the target object is greater than the vehicle safety distance, and the lateral collision time between the vehicle and the target object is greater than the lateral collision time threshold, the target object is determined to be a safe target object; in step 3), if the longitudinal collision time TTC between the vehicle and the dangerous obstacle is greater than the vehicle safety distance, the target object is determined to be a safe target object. x Greater than the first longitudinal collision time threshold TTC xth1 , and judge dangerous obstacles as safe targets.

[0014] Furthermore, the information of the vehicle and the target object includes the lateral velocity v 1y , target lateral velocity v 2y , the relative distance D between the vehicle and the target rel And the azimuth angle θ of the target object; in step 2), the following method is used to determine the lateral collision time TTC between the vehicle and the target object y :If the lateral relative speed v between the target object and the vehicle yrel =v 2y -v 1y ≥0, then the lateral collision time TTC y Set to a value greater than the set maximum value; otherwise, the lateral collision time TTC y for: B is the lateral width of the vehicle, d s The width is set to allow for safe passage of lateral redundancy.

[0015] Its beneficial effects are: yrel When ≥0, it means that the target object is moving away from the vehicle or moving parallel to the vehicle's heading direction. At this time, the vehicle is in a safe state and the lateral collision time TTC is set to y Set to a relatively large value to ensure TTC y ≥TTC yth , to avoid the occurrence of lateral braking by mistake. yrelWhen <0, the set horizontal redundant safety passage width d s Taken into account to calculate the lateral collision time TTC y , which avoids the situation where the car is unable to react in time in an emergency.

[0016] Furthermore, the following method is used to determine the lateral collision time threshold TTC: yth :

[0017] like and but

[0018] like and Then the lateral collision time TTC yth Set to a value less than the set minimum value;

[0019] Among them, D rel is the relative distance between the vehicle and the target; θ is the azimuth angle of the target; v 1x is the longitudinal velocity of the vehicle; v 2x is the longitudinal velocity of the target; a xmax is the maximum deceleration of the vehicle; d end The longitudinal redundant safety distance is set; v yrel =v 2y -v 1y The lateral relative speed between the target object and the vehicle. Its beneficial effect is: different lateral collision time thresholds TTC are used in different situations yth .exist When , it means that at this time, the longitudinal distance between the vehicle and the target in front is relatively safe, and the vehicle can be stopped by the maximum deceleration. The braking time T of the vehicle at the current speed with the maximum deceleration b1 Compare, if the comparison result is less than, it means that the vehicle has finished braking, and the target in front has not entered the lateral safety distance, so the lateral collision time threshold TTC is set at this time. yth A smaller value (for example, 0 or a smaller value close to 0) is used to avoid erroneous braking.

[0020] Furthermore, the lateral collision time threshold TTC is determined by the following method: yth :

[0021] like and Then TTC yth Set to a value less than the set minimum value;

[0022] like and TTC yth =TTC y ;

[0023] Among them, D rel is the relative distance between the vehicle and the obstacle; θ is the azimuth angle of the target; v1 is the longitudinal velocity of the vehicle; v2 is the longitudinal velocity of the target; a max is the maximum deceleration of the vehicle; d end The longitudinal redundant safety distance is set; v yrel =v 2y -v 1y is the lateral relative velocity between the target object and the vehicle; L1 and L2 are the lengths of the vehicle and the target object, respectively.

[0024] Its beneficial effect is: different lateral collision time thresholds TTC are used in different situations yth .exist When , it means that the vehicle cannot stop safely within the longitudinal distance of the target object at the maximum deceleration. and If the comparison result is greater than or equal to, it means that when the target object is within the lateral safety distance, the vehicle has already traveled beyond the collision point. At this time, the lateral collision time threshold TTC is set. yth A smaller value (e.g. 0 or a smaller value close to 0) is used to avoid false triggering of the brakes. If the comparison result is less than 0, it means that when the target in front enters the safe distance, the vehicle cannot exceed the collision point, nor can it stop within the longitudinal safe distance, so TTC is output. yth =TTC y .

[0025] Furthermore, the vehicle and obstacle information includes the vehicle longitudinal velocity v 1x , obstacle longitudinal velocity v 2x , the relative distance D between the vehicle and the obstacle rel , obstacle azimuth θ, vehicle longitudinal acceleration a 1x and the obstacle longitudinal acceleration a 2x ; In step 3), the following method is used to determine the longitudinal collision time TTC between the vehicle and the obstacle: x :If you are in any of the following situations: v 2x =v 1x And a 2x ≥a 1x , or v 2x >v 1x And a 2x ≥a 1x, then the longitudinal collision time TTC x Set to a value greater than the set value; otherwise, the longitudinal collision time TTC is determined according to the following formula x :

[0026]

[0027] The beneficial effect is that when v2 = v1 and a2 ≥ a1, or v2 > v1 and a2 ≥ a1, it means that the obstacle is moving in accordance with the vehicle and no collision will occur. Therefore, the longitudinal collision time TTC is x Set it to a larger value. In other cases, set the longitudinal collision time TTC according to the actual situation. x .

[0028] Furthermore, in step 3), three danger zones are divided by three-level longitudinal collision time thresholds, wherein the three-level longitudinal collision time thresholds are respectively the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 , and TTC xth1 >TTC xth2 >TTC xth3 , the three levels of danger areas are: Level 1 danger zone [0, TTC xth3 ]、Second level dangerous zone (TTC xth3 ,TTC xth2 ]、Third level dangerous zone (TTC xth2 ,TTC xth1 ].

[0029] Its beneficial effects are: dividing multiple danger zones according to the longitudinal collision time to classify dangerous obstacles, and adopting braking strategies in different danger zones to ensure driving comfort while avoiding safety accidents.

[0030] Furthermore, if the dangerous obstacle is in a stationary or uniformly moving state, the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 They are:

[0031]

[0032]

[0033]

[0034] Among them, v 1x is the longitudinal velocity of the vehicle; v2x is the longitudinal speed of the dangerous obstacle; t k is the set safety time; t2 is the deceleration growth time; d end The longitudinal redundant safety distance is set; Respectively represent the three-level braking deceleration a of the vehicle xcom , Secondary braking deceleration a xtra , first-level braking deceleration a xmax Braking, the difference in distance from the dangerous obstacle when braking to the safe moment, and

[0035] Its beneficial effect is that when the dangerous obstacle is at rest or in a uniform motion, the deceleration growth time, the set safety time, and the braking time of the relative motion between the vehicle and the dangerous obstacle are taken into account to accurately determine the first longitudinal collision time threshold TTC. xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 .

[0036] Furthermore, if the dangerous obstacle is in a variable speed motion state and a2≠a1, the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 They are:

[0037]

[0038]

[0039]

[0040] If the dangerous obstacle is in a variable speed motion state and a2 = a1, the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 They are:

[0041]

[0042]

[0043]

[0044] Where, v 1x is the longitudinal velocity of the vehicle; v 2x is the longitudinal speed of the dangerous obstacle; t kis the set safety time; t2 is the deceleration growth time; d end The longitudinal redundant safety distance is set; Respectively represent the three-level braking deceleration a of the vehicle xcom , Secondary braking deceleration a xtra , first-level braking deceleration a xmax Braking, the distance difference between the vehicle and the dangerous obstacle ahead when braking to the safe moment; any one of the three levels of deceleration is recorded as the target deceleration a xaim ;

[0045] a 2x <0, and a xaim <a 1x hour,

[0046]

[0047]

[0048]

[0049] When a2>0,

[0050]

[0051]

[0052]

[0053] Its beneficial effect is that when the dangerous obstacle is in a variable speed motion state, the deceleration growth time, the set safety time, and the braking time of the relative motion between the vehicle and the dangerous obstacle are taken into account to accurately determine the first longitudinal collision time threshold TTC. xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 .

[0054] Furthermore, if the vehicle is in a first-level dangerous zone, the seat belt needs to be tightened when the vehicle is braked; if the vehicle is in a second-level dangerous zone, a warning signal needs to be sent to the dangerous obstacle when the vehicle is braked; and when the vehicle is braked in step 3), the expected deceleration a of the vehicle is exp for:

[0055] a exp =(F t -F bexp -∑F) / m

[0056] Among them, F t is the driving force; F bexp is the expected braking force; ∑F is the rolling resistance Ff , slope resistance F α , air resistance F w and m is the curb mass of the vehicle.

[0057] The benefits are as follows: In the first danger zone, the situation is extremely dangerous, and the seat belt is tightened to improve driver safety. In the second danger zone, the situation is relatively dangerous, and a warning of dangerous obstacles is issued, giving the driver space and time to avoid them. Furthermore, the calculated expected deceleration must overcome rolling resistance, grade resistance, and air resistance, thus taking actual conditions into account and ensuring effective braking. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a flow chart of the self-adaptive AEB control method for an unmanned vehicle of the present invention;

[0059] Figure 2 This is a schematic diagram of the first front obstacle screening principle in the self-adaptive AEB control method for unmanned vehicles of the present invention;

[0060] Figure 3 This is a schematic diagram of the principle of screening dangerous obstacles and their degree of danger in the adaptive AEB control method for unmanned vehicles of the present invention;

[0061] Figure 4-1 This is a kinematic analysis diagram of an unmanned vehicle going uphill according to the present invention;

[0062] Figure 4-2 This is a kinematic analysis diagram of an unmanned vehicle when going downhill according to the present invention. DETAILED DESCRIPTION

[0063] The key points of the present invention include: First, in the face of complex scene conditions, the present invention decouples the target objects in front of the vehicle in the horizontal and vertical directions, and designs two screening logics. The first screening logic is to screen out the vehicle's heading obstacles and potential vehicle heading obstacles (collectively referred to as obstacles) from the target objects in front of the vehicle. The second screening logic screens out dangerous obstacles from the obstacles, and compares and judges the danger levels of dangerous obstacles, and adaptively selects the intervention timing and control strategy of AEB. Second, in the first screening logic, different lateral collision times TTC are adaptively determined for different working conditions. y and lateral collision time threshold TTC yth , in order to accurately identify the obstacles and potential obstacles in the direction of the vehicle, the second screening logic adaptively determines different longitudinal collision times TTC in the face of different working conditions x and longitudinal collision time threshold TTC xth, in order to accurately identify the danger level of dangerous obstacles, thus providing a good judgment benchmark for the timing of AEB intervention.

[0064] The present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0065] Embodiment of the adaptive AEB control method for unmanned vehicles:

[0066] The vehicle targeted by this embodiment (hereinafter referred to as the self-vehicle) is an unmanned vehicle. The overall process of the unmanned vehicle adaptive AEB control method of this embodiment is as follows: Figure 1 As shown, the specific process is as follows:

[0067] Step 1: Obtain information about the state of the vehicle and its external environment (mainly the state of the environment in front of the vehicle) through sensors installed on the vehicle.

[0068] The acquired vehicle information and vehicle external environment status information include: determining the real-time speed v1 of the vehicle based on the GPS / IMU positioning signal of the vehicle, including the longitudinal speed v 1x and the lateral velocity v 1y ; The real-time acceleration a1 of the vehicle is determined by the vehicle accelerometer installed on the vehicle, including the longitudinal acceleration a 1x The slope angle α of the road on which the vehicle is located is determined based on the gyroscope on the vehicle. The relative distance D between the vehicle and the target in front is determined based on the radar and visual sensor installed on the vehicle. rel The speed v2, acceleration a2 and azimuth angle θ of the target in front are determined by the algorithm of the sensor on the vehicle. The target speed includes the longitudinal speed v 2x and the lateral velocity v 2y , acceleration a2 includes longitudinal acceleration a 2y In this embodiment, the longitudinal direction is defined as the direction parallel to the lane markings of the lane in which the ego vehicle is located, and the transverse direction is defined as the direction perpendicular to the lane markings. Of course, if the vehicle is moving straight along the lane, rather than turning, changing lanes, or making a U-turn, then the heading direction of the ego vehicle is parallel to the lane markings.

[0069] Step 2: The first obstacle screening is to screen out the vehicle's heading obstacles and potential vehicle heading obstacles in front of the vehicle.

[0070] The principle diagram of the first front obstacle screening is as follows Figure 2 As shown, calculate the lateral distance D of the target object relative to the vehicle rel sinθ (if the vehicle is moving straight along the lane, the lateral distance between the target object and the vehicle is the lateral distance between the target object and the middle line of the target course). Considering the lateral width of the vehicle as B, the lateral redundant safety passing width d is set for vehicle driving safety. s, judge whether the lateral distance is less than the lateral safety distance of the vehicle

[0071] If the lateral distance is less than the vehicle's safe lateral distance range, This indicates that the target object is an obstacle in the vehicle's direction.

[0072] If the lateral distance is greater than or equal to the vehicle's safe lateral distance range, Then continue to calculate the lateral collision time TTC between the vehicle and the target object y And the set lateral collision time threshold TTC yth , and judge whether the target object is a potential obstacle to the vehicle: If TTC y ≤TTC yth , indicating that the target object is a potential obstacle to the vehicle; if TTC y >TTC yth , indicating that the target object is not a potential obstacle to the vehicle's course and is a safe target object.

[0073] Among them, the lateral relative speed v between the target object and the vehicle is defined as yrel , v yrel =v 2y -v 1y , v 2y Indicates the lateral velocity of the target, v 1y Indicates the lateral velocity of the vehicle. The target's velocity is positive when it is in the same direction as the vehicle's y-axis. The y-axis of the vehicle is the direction of the vehicle's lateral velocity. y and lateral collision time threshold TTC yth The method for determining is as follows:

[0074] 1) Transverse collision time TTC y When v yrel When ≥0, it means that the target moves away from the vehicle or moves parallel to the vehicle's heading, which is a safe state. At this time, the lateral collision time TTC y Directly set it to a larger value Y to avoid false triggering in the horizontal direction; when v yrel <0, lateral collision time TTC y It is calculated using the following formula:

[0075]

[0076] 2) Lateral collision time threshold TTC yth Lateral collision time threshold TTC ythThe setting of represents the likelihood that an object will enter the vehicle's direction of travel. A value that is too high increases the likelihood of false triggering, while a value that is too low may prevent the vehicle from reacting in an emergency. The lateral collision time threshold designed in this invention addresses the issue of whether an object ahead requires emergency braking before it enters the safe lateral distance.

[0077] ① If a xmax is the maximum longitudinal deceleration of the vehicle, d end It is the longitudinal redundant safety distance between the vehicle and the obstacle in front after braking, indicating that the longitudinal distance between the vehicle and the target object in front is relatively safe at this time. xmax You can stop the car. Then compare The braking time of the vehicle at the current speed and maximum deceleration The size of

[0078] like but

[0079] like This means that the vehicle has finished braking and the target in front has not entered the lateral safety distance, so TTC is set at this time. yth A smaller value, for example, 0 or a value close to 0.

[0080] ②If This means that the vehicle cannot safely stop within the longitudinal distance of the target object even at the maximum deceleration. and L1 and L2 are the lengths of the vehicle and the target object in front respectively:

[0081] like When the target object is within the lateral safety distance, the vehicle has already traveled beyond the collision point. At this time, TTC is set. yth A smaller value, for example, it can be set to 0 or a value close to 0;

[0082] like When the target object in front enters the safe distance, the vehicle cannot exceed the collision point, nor can it stop within the longitudinal safe distance. At this time, set TTC yth =TTC y .

[0083] Note: When v 1x ≤v 2x When the longitudinal situation is safe, TTC yth =0; when When the longitudinal distance meets the braking distance of the vehicle's maximum deceleration, the target in front has entered the lateral safety distance. At this time, set TTC yth A smaller value, for example, 0 or a value close to 0.

[0084] According to the above method, all obstacles and potential obstacles in the direction of the vehicle can be screened out and collectively referred to as obstacles. At this time, it is necessary to continue with step 3 to pass the longitudinal collision time TTC. x The algorithm further determines the safety of the vertical risk of the target object.

[0085] Step 3: Longitudinal collision time TTC x Algorithm, perform secondary classification on the vehicle heading obstacles and potential vehicle heading obstacles screened in step 2, first find the longitudinal collision time TTC x The shortest obstacle is a dangerous obstacle, and the dangerous obstacle is further judged to determine its degree of danger. For details, see the second front dangerous target screening principle diagram of the self-adaptive AEB control method for unmanned vehicles, such as Figure 3 shown.

[0086] First, we need to calculate the longitudinal collision time TTC of each obstacle x , select the longitudinal collision time TTC x The shortest obstacle is called a dangerous obstacle; and the time to pass the dangerous obstacle through the longitudinal collision TTC is calculated. x , and combined with the longitudinal collision time threshold TTC xth To judge the danger level of dangerous obstacles. Longitudinal collision time threshold TTC xth It includes three thresholds, namely the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 , and meet TTC xth1 >TTC xth2 >TTC xth3 , and thus classify this type of target through the three-level longitudinal collision time threshold: when TTC x >TTC xth1 When the dangerous obstacle is a safe target; when TTC xth2 <TTC x ≤TTC xth1 When the dangerous obstacle is a level 3 dangerous target; when TTC xth3 <TTC x ≤TTC xth2 When the dangerous obstacle is a secondary dangerous target; when TTC x ≤TTCxth3 , the dangerous obstacle is a first-level dangerous target.

[0087] Among them, the longitudinal collision time TTC x and longitudinal collision time threshold TTC xth The method for determining is as follows:

[0088] 1) Longitudinal collision time TTC x .

[0089] When faced with the following situations, ①v 2x =v 1x, a 2x =a 1x ;②v 2x =v 1x ,a 2x >a 1x ; ③v 2x >v 1x, a 2x >a 1x ; ④v 2x >v 1x, a 2x =a 1x ; This is a safe situation, the vehicle collision time TTC x Set to a larger value X to ensure that AEB is not triggered. The determination formula for other situations is as follows:

[0090]

[0091] Where, v 1x and v 2x are the longitudinal speeds of the vehicle and the dangerous obstacle respectively; a 1x and a 2x are the longitudinal accelerations of the ego vehicle and the dangerous obstacle, respectively. The longitudinal acceleration of the dangerous obstacle is positive when it is in the same direction as the ego vehicle.

[0092] 2) Longitudinal collision time threshold TTC xth Longitudinal collision time threshold TTC xth Related to the following data: vehicle data processing and algorithm calculation time t0, brake response time t1, deceleration growth time t2, peak road adhesion coefficient μ, longitudinal redundant safety distance d between the vehicle and the obstacle in front after braking end .

[0093] ① If the dangerous obstacle is stationary or moving at a constant speed:

[0094]

[0095]

[0096]

[0097] Where, Respectively represent the distance difference between the vehicle and the dangerous obstacle when braking to the safe moment with the third, second and first level braking deceleration, and

[0098] ② If the dangerous obstacle moves at a variable speed and a2≠a1:

[0099]

[0100]

[0101]

[0102] If the dangerous obstacle moves at a variable speed and a2 = a1:

[0103]

[0104]

[0105]

[0106] The target deceleration a is recorded as any one of the three deceleration levels. xaim , to determine whether the preceding vehicle stops before the self-vehicle, the formula is determined as follows:

[0107] ① When the dangerous obstacle slows down, that is, a 2x <0, and a xaim <a 1x hour:

[0108] If the vehicle stops before the dangerous obstacle, but

[0109] If the dangerous obstacle stops before the vehicle stops, but

[0110] If the vehicle and the dangerous obstacle stop at the same time, but

[0111] ②When the dangerous obstacle accelerates, that is, a 2x >0 o'clock:

[0112]

[0113]

[0114]

[0115] Step 4: Take appropriate control strategies for the dangerous obstacles identified above. x The algorithm outputs a reasonable expected deceleration for different levels of danger of the target object, and through the designed inverse longitudinal dynamics model, converts the expected deceleration into braking pressure on the vehicle, ultimately forming feedback control.

[0116] The distribution strategy of the expected deceleration is as follows: Analyze the three longitudinal collision time trigger thresholds TTC xth The significance of TTC xth1 The vehicle will enter the AEB level 3 braking control strategy and xth1 ~TTC xth2 Comfort braking is performed within the time period, and the acceleration and deceleration range is 2m / s according to the human body's comfort 2 ~-2m / s 2 , so during this period we will take a com =-2m / s 2 During this time period, the vehicle sends a warning signal to the target ahead, giving the target ahead space and time to avoid. xth2 The secondary braking comfort transition of the vehicle entering the AEB secondary braking strategy to the maximum braking force is expressed. According to the peak ground adhesion coefficient μ, the secondary braking deceleration is taken as TTC xth3 Expresses that the vehicle will take maximum deceleration to avoid collision and automatically tighten the seat belt to further improve safety. The maximum deceleration is a xmax =-μg.

[0117] According to the kinematic analysis diagram of the unmanned vehicle during braking, Figure 4-1 and Figure 4-2 As shown in Figure 2, the kinematic equation of the vehicle during braking is:

[0118] ma exp =F t -F bexp -∑F

[0119] Where a exp is the expected deceleration at the current moment; F t is the driving force; F exp is the expected braking force at the current moment; ∑F is the rolling resistance F f , slope resistance F α , air resistance Fw and; F α = mgcosα + μmgsinα, where m is the vehicle's curb mass, α is the instantaneous slope output by the vehicle's gyroscope, which is positive when going uphill and negative when going downhill, μ is the peak adhesion coefficient of the road at this time, and g is the acceleration due to gravity, which is 9.8 m / s 2 ; F f =mgf, where f is the rolling resistance coefficient; C D is the air resistance coefficient, A is the frontal area, and ρ is the air density.

[0120] The expected deceleration corresponds to the brake pressure in real time as follows:

[0121]

[0122] Where K b It is the proportional coefficient between braking force and braking pressure. This constant may vary depending on the vehicle model and can be obtained through simple experiments.

[0123] According to the adaptive AEB control method designed by the present invention, a test scenario is designed in a closed field. The vehicle is traveling at a constant speed of 10 m / s. There is a cyclist at a longitudinal distance of 20 m in front of the vehicle and a lateral distance of 8 m on the right side. The cyclist is traveling at a constant speed of 4 m / s in the lane of the vehicle. The lane width is 3.8 m, the width of the vehicle is 1.9 m, and the lateral redundant passing width d s =0.7m, longitudinal redundant safety distance d end =3m, peak adhesion coefficient μ=0.9, vehicle data processing and algorithm calculation time t0=0.4s, brake response time t1=0.15s, deceleration increase time t2=0.1s.

[0124] First, according to the first screening logic of the dangerous target ahead, at this time, the target does not belong to the obstacle in the direction of the vehicle. At this time, the lateral collision time TTC y =1.5875, lateral collision time threshold TTC yth =0.4544. When the lateral collision time TTC y =TTC yth When , the target object is screened as a potential obstacle in the direction of the vehicle. At this time, the longitudinal distance between the vehicle and the target object is 8.669m, and the lateral relative distance on the right is 3.4676m. At this time, the second forward dangerous target screening logic is entered, and the longitudinal collision time TTC x =0.8669, longitudinal collision time threshold TTC xth1 =3.45, TTC xth2 =2.0838, TTC xth3=1.5169, which triggers the maximum braking deceleration. After braking, the longitudinal distance between the vehicle and the target is 3m, and the braking time is 1.1338s. At this time, the target is on the left side of the vehicle at a lateral distance of 1.0676m, and the obstacle is avoided safely.

[0125] If only the lateral safety distance is considered, most studies consider the safety distance to be half the lane width, which is 1.9m in this scenario. If braking control is performed when the target object reaches this safety distance, the longitudinal distance between the vehicle and the target object is 4.75m. When braking with maximum deceleration, the braking distance is 5.669m. It is found that maximum deceleration cannot successfully stop the vehicle.

[0126] In summary, the adaptive AEB control method for unmanned vehicles of the present invention can be applied to unmanned vehicles under multiple working conditions. Through adaptive calculation of lateral and longitudinal collision time thresholds, it can achieve lateral and longitudinal decoupling of multiple target objects, has high adaptability, and can ensure safety while allowing passengers to enjoy better comfort through multi-level braking.

[0127] Furthermore, the above method can be implemented via a computer program. For example, an adaptive AEB control device for an unmanned vehicle can be designed, comprising a memory, a processor, and an internal bus. The processor and memory communicate and exchange data with each other via the internal bus. The memory includes at least one software functional module stored in the memory. The processor executes the software program and modules stored in the memory to perform various functional applications and process data, thereby implementing the adaptive AEB control method for an unmanned vehicle described in the method embodiment of the present invention. The processor can be a processing device such as a microprocessor (MCU) or a programmable logic device (FPGA). The memory can be any type of memory that stores information electrically, such as RAM and ROM; any type of memory that stores information magnetically, such as a hard disk, floppy disk, magnetic tape, magnetic core memory, bubble memory, or USB flash drive; any type of memory that stores information optically, such as a CD or DVD; and other types of memory, such as quantum memory and graphene memory.

Claims

1. An adaptive AEB control method for an unmanned vehicle, characterized in that: The steps include: 1) Obtain information about the vehicle and the target in front of it; 2) Based on the information of the ego vehicle and the target object, the lateral distance between the ego vehicle and the target object is calculated. If the lateral distance between the ego vehicle and the target object is less than or equal to the ego vehicle's safety distance, the target object is determined to be an obstacle in the ego vehicle's direction. Based on the information of the ego vehicle and the target object, the lateral collision time TTC between the ego vehicle and the target object is calculated. y If the lateral distance between the vehicle and the target is greater than the set safety distance, and the lateral collision time TTC between the vehicle and the target is greater than the set safety distance, y Less than or equal to the lateral collision time threshold TTC yth , the target object is determined to be a potential obstacle to the vehicle's heading; obstacles to the vehicle's heading and potential obstacles to the vehicle's heading are collectively referred to as obstacles; Among them, if and but like and Then the lateral collision time TTC yth Set to a value less than the set minimum value; Among them, D rel is the relative distance between the vehicle and the target; θ is the azimuth angle of the target; v 1x is the longitudinal velocity of the vehicle; v 2x is the longitudinal velocity of the target; a xmax is the maximum longitudinal deceleration of the vehicle; d end The longitudinal redundant safety distance is set; v yrel =v 2y -v 1y is the lateral relative speed between the target object and the vehicle; B is the lateral width of the vehicle, d s The horizontal redundant safety passage width is set, a 2x is the longitudinal acceleration of the obstacle, a2 is the acceleration of the target; 3) Calculate the longitudinal collision time TTC between the vehicle and each obstacle based on the vehicle and obstacle information x , select the longitudinal collision time TTC x The shortest obstacle is considered a dangerous obstacle; if the longitudinal collision time TTC between the vehicle and the dangerous obstacle is x Less than or equal to the first longitudinal collision time threshold TTC xth1 , determine the longitudinal collision time TTC of the dangerous obstacle x The danger zone in which the vehicle is located, wherein a longitudinal collision time range constitutes a first-level danger zone, and the first-level danger zone corresponds to a braking strategy; the vehicle is braked according to the braking strategy corresponding to the danger zone; The longitudinal direction is parallel to the lane line of the lane where the vehicle is located, and the transverse direction is perpendicular to the lane line.

2. The adaptive AEB control method for unmanned vehicles according to claim 1, characterized in that: In step 2), if the lateral distance between the vehicle and the target object is greater than the safety distance of the vehicle, and the lateral collision time between the vehicle and the target object is greater than the lateral collision time threshold, the target object is determined to be a safe target; in step 3), if the longitudinal collision time TTC between the vehicle and the dangerous obstacle is greater than the safety distance of the vehicle, the target object is determined to be a safe target. x Greater than the first longitudinal collision time threshold TTC xth1 , and judge dangerous obstacles as safe targets.

3. The adaptive AEB control method for unmanned vehicles according to claim 1, characterized in that: The information of the ego vehicle and the target object includes the lateral velocity v 1y , target lateral velocity v 2y , the relative distance D between the vehicle and the target rel and the target azimuth angle θ; In step 2), the following method is used to determine the lateral collision time TTC between the vehicle and the target object: y :If the lateral relative speed v between the target object and the vehicle yrel =v 2y -v 1y ≥0, then the lateral collision time TTC y Set to a value greater than the set maximum value; otherwise, the lateral collision time TTC y for:

4. The adaptive AEB control method for unmanned vehicles according to claim 1, characterized in that: The following method is used to determine the lateral collision time threshold TTC yth : like and Then TTC yth Set to a value less than the set minimum value; like and TTC yth =L y ; Among them, L1 and L2 are the lengths of the vehicle and the target object respectively.

5. The adaptive AEB control method for unmanned vehicles according to claim 1, characterized in that: Vehicle and obstacle information includes the longitudinal velocity v 1x , obstacle longitudinal velocity v 2x , the relative distance D between the vehicle and the obstacle rel , obstacle azimuth θ, vehicle longitudinal acceleration a 1x and the obstacle longitudinal acceleration a 2x ; In step 3), the following method is used to determine the longitudinal collision time TTC between the vehicle and the obstacle: x :If you are in any of the following situations: v 2x =v 1x And a 2x ≥a 1x , or v 2x >v 1x And a 2x ≥a 1x , then the longitudinal collision time TTC x Set to a value greater than the set maximum value; otherwise, the longitudinal collision time TTC is determined according to the following formula x :

6. The adaptive AEB control method for unmanned vehicles according to claim 1, characterized in that: In step 3), three danger zones are divided by three-level longitudinal collision time thresholds, where the three-level longitudinal collision time thresholds are the first longitudinal collision time threshold TTC, xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 , and TTC xth1 >TTC xth2 >TTC xth3 , the three levels of danger areas are: Level 1 danger zone [0, TTC xth3 ]、Second level dangerous zone (TTC xth3 ,TTC xth2 ]、Third level dangerous zone (TTC xth2 ,TTC xth1 ].

7. The method for adaptive AEB control of an unmanned vehicle according to claim 6, characterized in that: If the dangerous obstacle is at rest or in constant motion, the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 They are: Among them, t k is the set safety time; t2 is the deceleration growth time; Respectively represent the three-level braking deceleration a of the vehicle xcom , Secondary braking deceleration a xtra , first-level braking deceleration a xmax Braking, the difference in distance from the dangerous obstacle when braking to the safe moment, and 8. The method for adaptive AEB control of an unmanned vehicle according to claim 6, characterized in that: If the dangerous obstacle is in a variable speed motion state and a2≠a1, a1 is the acceleration of the vehicle, then the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 They are: If the dangerous obstacle is in a variable speed motion state and a2 = a1, the first longitudinal collision time threshold TTC xth1 , Second longitudinal collision time threshold TTC xth2 and the third longitudinal collision time threshold TTC xth3 They are: Where, t k is the set safety time; t2 is the deceleration growth time; Respectively represent the three-level braking deceleration a of the vehicle xcom , Secondary braking deceleration a xtra , first-level braking deceleration a xmax Braking, the distance difference between the vehicle and the dangerous obstacle ahead when braking to the safe moment; any one of the three levels of deceleration is recorded as the target deceleration a xaim ; a 2x <0, and a xaim 1x hour,​ When a2>0, 9. The method for adaptive AEB control of an unmanned vehicle according to claim 6, characterized in that: If the vehicle is in the first-level danger zone, the seat belt needs to be tightened when the vehicle is braked; if the vehicle is in the second-level danger zone, a warning signal needs to be sent to the dangerous obstacle when the vehicle is braked; and when the vehicle is braked in step 3), the expected deceleration a of the vehicle is exp for: a exp =(F t -F bexp -∑F) / m Among them, F t is the driving force; F bexp is the expected braking force; ∑F is the rolling resistance F f , slope resistance F α , air resistance F w and m is the curb mass of the vehicle.

Images