Night slope early warning method and device for automatic driving vehicle and electronic equipment

Abstract

The invention relates to a night slope early warning method and device for an automatic driving vehicle and electronic equipment, and the method comprises the steps: obtaining the detection data of a vehicle-mounted radar about the road condition in front of the vehicle; judging the road condition and whether the obstacle vehicle in front of the vehicle disappears or not according to the detection data; when the obstacle vehicle disappears, the current observation distance is determined, the relation between the current observation distance and the safe vehicle following distance is judged, and when the current observation distance is smaller than the safe vehicle following distance, a vehicle following early warning prompt is given. After the front obstacle vehicle disappears, the default front obstacle vehicle still exists, and the current observation distance between the front obstacle vehicle and the front obstacle vehicle continues to be calculated. And when the current observation distance is smaller than the safe vehicle following distance, giving a vehicle following early warning prompt, so that the automatic driving system controls the vehicle speed, and the current observation distance is always larger than or equal to the safe vehicle following distance. The problem that effective vehicle following early warning cannot be achieved in the night ramp scene through an existing vehicle following early warning method of the automatic driving vehicle is solved.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving, and in particular to a nighttime slope warning method, device and electronic equipment for an autonomous driving vehicle. Background Art

[0002] In order to improve the active safety of autonomous vehicles and reduce the probability of traffic accidents, various active safety-related technologies have begun to be applied in autonomous vehicles, among which vehicle-following collision warning has attracted much attention.

[0003] The premise of anti-collision is that the autonomous driving system can detect the vehicle in front in time, obtain the distance, speed and other parameters between the front and rear vehicles, analyze the gradient of distance change, and design a control algorithm to control the speed of the vehicle to maintain a safe distance between vehicles and achieve following safety. However, in actual situations, when an autonomous driving vehicle encounters a ramp, the radar can easily lose the vehicle target after the front vehicle goes uphill or downhill due to the detection angle. At night, under the influence of light reflection, the target features are easy to blur, and the use of image algorithms is prone to false detection and missed detection, which is not conducive to vehicle decision-making and planning. At present, the following vehicle warning and safety assurance methods are mainly aimed at horizontal roads. There is a lack of following vehicle warning methods for ramps. When encountering a ramp at night, the on-board sensor is affected and the detection is lost. The autonomous driving system will think that there is no obstacle vehicle in front and control the vehicle to accelerate to the maximum speed required by law; at this time, if the front vehicle starts to slow down, the distance between the front and rear vehicles will decrease rapidly. When the sensor finds the front vehicle again, the distance between the front and rear vehicles may not meet the driving safety requirements, thereby causing a collision accident.

[0004] There is currently no effective solution to the problem that the current vehicle following warning method for autonomous driving vehicles cannot achieve effective vehicle following warning in night-time slope scenarios. Summary of the invention

[0005] The present invention provides a nighttime slope warning method for an autonomous driving vehicle to solve the problem that the current vehicle following warning method for an autonomous driving vehicle cannot achieve effective vehicle following warning in a nighttime slope scenario.

[0006] The present invention provides a nighttime slope warning method for an autonomous driving vehicle, comprising:

[0007] Obtain detection data of the vehicle radar about the road conditions in front of the vehicle;

[0008] Determine the road condition and whether the obstacle vehicle in front of the vehicle has disappeared according to the detection data;

[0009] When the obstacle vehicle disappears, the current observation distance is determined, and the relationship between the current observation distance and the safe following distance is judged. When the current observation distance is less than the safe following distance, a following vehicle warning prompt is given;

[0010] The step of determining the current observation distance comprises:

[0011] When the road condition is a horizontal turning uphill, the current observation distance is determined by a first uphill sight distance model, and the first uphill sight distance model is:

[0012]

[0013] When the road condition is from uphill to level, the current observation distance is determined by a second uphill sight distance model, and the second uphill sight distance model is:

[0014] S v =2S r θ1|cscθ1|

[0015] When the road condition is a horizontal turning downhill, the current observation distance is determined by a first downhill sight distance model, and the first downhill sight distance model is:

[0016] S v =2S r θ1|cscθ1|

[0017] When the road condition is a downhill to horizontal road, the current observation distance is determined by a second downhill sight distance model, and the second downhill sight distance model is:

[0018]

[0019] Among them, S v Indicates the current observation distance, S r represents the maximum road distance detected by the vehicle-mounted radar, θ1 represents the detection centerline direction of the vehicle-mounted radar and S r The angle of the detection direction.

[0020] Compared with the related art, the present invention assumes that the obstacle vehicle ahead still exists after the obstacle vehicle ahead disappears, and continues to calculate the current observation distance between the two. When the current observation distance is less than the safe following distance, a following warning prompt is given, so that the automatic driving system controls the vehicle speed so that the current observation distance is always greater than or equal to the safe following distance, thereby ensuring driving safety. The present invention solves the problem that the current following warning method of the automatic driving vehicle cannot achieve effective following warning in the night ramp scene.

[0021] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1is a flow chart of a nighttime slope warning method for an autonomous driving vehicle provided in some embodiments of the present invention;

[0023] Figure 2 It is a calculation principle diagram of the first uphill sight distance model in the present invention;

[0024] Figure 3 It is a calculation principle diagram of the second uphill sight distance model in the present invention;

[0025] Figure 4 It is a calculation principle diagram of the first downhill sight distance model in the present invention;

[0026] Figure 5 It is a calculation principle diagram of the second downhill sight distance model in the present invention;

[0027] Figure 6 is a calculation principle diagram of the first uphill distance model in the present invention;

[0028] Figure 7 is a calculation principle diagram of the second uphill distance model in the present invention;

[0029] Figure 8 is a calculation principle diagram of the first downhill distance model in the present invention;

[0030] Fig. 9 It is a calculation principle diagram of the second downhill distance model in the present invention. DETAILED DESCRIPTION

[0031] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments.

[0032] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the general meaning understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the", "these" and the like in this application do not represent quantitative restrictions, and they can be singular or plural. The terms "include", "comprise", "have" and any variants thereof involved in this application are intended to cover non-exclusive inclusions; for example, a process, method and system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether directly or indirectly. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships. For example, "A and / or B" can mean: A exists alone, A and B exist at the same time, and B exists alone. Usually, the character " / " indicates that the objects associated with each other are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0033] In an embodiment of the present invention, a nighttime slope warning method for an autonomous driving vehicle is provided. Figure 1 is a flowchart of a nighttime slope warning method for an autonomous driving vehicle provided in some embodiments of the present invention, such as Figure 1 As shown, the process includes the following steps:

[0034] Step S110, obtaining detection data of the vehicle-mounted radar about the road condition in front of the vehicle.

[0035] Due to the poor light conditions at night, the use of image sensors to obtain the road conditions in front of the vehicle is prone to missed detection or false detection of image targets, making it difficult to accurately judge the road conditions in front. The present invention addresses the problem of poor light conditions at night by obtaining the road conditions in front of the vehicle through a vehicle-mounted radar. The vehicle-mounted radar detects the distance between the vehicle and various objects in front through the transmission and reception of radio waves. For example, within the detection range of the vehicle-mounted radar, the distance between the vehicle and the road surface, the vehicle in front, and pedestrians can be detected, and the above information can be obtained from the detection data of the vehicle-mounted radar.

[0036] Step S120, judging the road condition and whether the obstacle vehicle in front of the vehicle has disappeared based on the detection data.

[0037] When the obstacle vehicle disappears, step S130 is executed to determine the current observation distance, judge the relationship between the current observation distance and the safe following distance, and give a following vehicle warning prompt when the current observation distance is less than the safe following distance.

[0038] Among them, the disappearance of the obstacle vehicle in front of the vehicle refers to the disappearance of the obstacle vehicle from the detection range of the vehicle-mounted radar. This situation can be judged by the detection data of the vehicle-mounted radar. Usually, the moment the obstacle vehicle disappears, the distance value detected by the vehicle-mounted radar will suddenly change. For example, at the last moment, the detection data of the vehicle-mounted radar still contains the distance data between the vehicle and the obstacle vehicle in front (indicating that there is an obstacle vehicle in front), and at the next moment, the distance data between the vehicle and the obstacle vehicle in front in the detection data of the vehicle-mounted radar disappears or the value tends to infinity (these are two cases of sudden changes in distance values, which means that there is no obstacle vehicle in front). The change in value can be used to determine whether the obstacle vehicle in front has disappeared. The specific principle is that if the front vehicle is always in the radar detection range of the rear vehicle, the distance change between the front and rear vehicles is continuous, and the distance change range should conform to the kinematic law. Suppose the speed of the front vehicle is V f , the deceleration is a f , the speed of the rear vehicle is V e , the deceleration is a e , the distance gradient change value Δd at the next moment should satisfy:

[0039]

[0040] If the distance gradient change value Δd at a certain moment does not satisfy the above relationship, it means that the front vehicle disappears from the detection field of view of the rear vehicle at that moment.

[0041] When the obstacle vehicle in front disappears, the observation distance between the vehicle and the vehicle in front is calculated. Using the observation distance between the two as the early warning judgment standard, the safest autonomous driving can be achieved.

[0042] The steps for determining the current observation distance include:

[0043] Reference Figure 2 When the road condition is horizontal and turns uphill, the current observation distance is determined by the first uphill sight distance model, and the first uphill sight distance model is:

[0044]

[0045] Reference Figure 3 , when the road condition changes from uphill to level, the current observation distance is determined by the second uphill sight distance model, and the second uphill sight distance model is:

[0046] S v =2S r |θ1|csc|θ1|

[0047] Reference Figure 4 When the road condition is horizontal and turns downhill, the current observation distance is determined by the first downhill sight distance model, which is:

[0048] S v =2S r |θ1|csc|θ1|

[0049] Reference Figure 5 , when the road condition is downhill to horizontal, the current observation distance is determined by the second downhill sight distance model, and the second downhill sight distance model is:

[0050]

[0051] Among them, S v Indicates the current observation distance, S r represents the maximum road distance detected by the vehicle-mounted radar, θ1 represents the detection centerline direction of the vehicle-mounted radar and S r The angle of the detection direction, S r The detection direction refers to the farthest road position detected by the vehicle-mounted radar in that direction.

[0052] The above four situations are respectively for four road conditions. Horizontal to uphill means that the vehicle is on a horizontal road and the road ahead is uphill, uphill to horizontal means that the vehicle is on an uphill road and the road ahead is horizontal, horizontal to downhill means that the vehicle is on a horizontal road and the road ahead is downhill, downhill to horizontal means that the vehicle is on a downhill road and the road ahead is horizontal. It should be noted that the front mentioned above refers to the range detectable by the vehicle-mounted radar, not the infinitely extended front. Therefore, it will be explained later that the judgment of various road conditions is also based on the detection data of the vehicle-mounted radar.

[0053] It should be noted that in the first uphill sight distance model and the second downhill sight distance model, the road distance between the farthest road position and the vehicle is used as the observation distance between the vehicle and the vehicle in front. In the second uphill sight distance model and the first downhill sight distance model, considering the symmetry of the road surface, twice the road distance between the farthest road position and the vehicle in front is used as the observation distance between the vehicle and the vehicle in front. At this time, the actual distance between the vehicle and the vehicle in front must be greater than or equal to the observation distance.

[0054] In the prior art, when the obstacle vehicle in front disappears, it is usually judged that there is no obstacle vehicle in front, and no warning information will be given, and the automatic driving system will speed up the vehicle. In fact, the obstacle vehicle disappears from the detection field of view, it is very likely that it encounters a slope road condition, and it still exists in front of the vehicle. After the obstacle vehicle in front disappears, the present invention assumes that the obstacle vehicle in front still exists and continues to calculate the current observation distance between the two. When the current observation distance is less than the safe following distance, a following warning prompt is given, so that the automatic driving system controls the vehicle speed so that the current observation distance is always greater than or equal to the safe following distance, thereby ensuring driving safety. The present invention solves the problem that the current following warning method of the automatic driving vehicle cannot achieve effective following warning in the ramp scene at night.

[0055] In some of the embodiments, when the obstacle vehicle disappears, the road condition determination step includes:

[0056] When the vehicle is level and the azimuth of the maximum road distance detected by the on-board radar is greater than 0, it is determined that the road condition is level turning uphill.

[0057] When the vehicle is level and the azimuth of the maximum road distance detected by the on-board radar is less than 0, it is determined that the road condition is level turning downhill.

[0058] When the vehicle is facing upward, it is determined that the road condition is turning from uphill to level.

[0059] When the vehicle is facing downward, it is determined that the road condition is turning from downhill to level.

[0060] First, the state of the ego vehicle can be measured based on the on-board posture sensor (such as a gyroscope). When the ego vehicle is on a horizontal road and the front vehicle disappears, there are two situations: one is that the front is uphill, and the other is that the front is downhill. Both will cause the obstacle vehicle in front to suddenly disappear from the radar field of view. If the front is uphill, the road surface can be detected within the field of view of the on-board radar, and the road surface distance detected by the upper boundary of the field of view is the largest, that is, the azimuth of the maximum road surface distance (the sound wave conduction angle detected at this distance) is greater than 0. If the front is downhill, the on-board radar can only partially detect the road surface with the lower field of view, and then the azimuth of the maximum road surface distance is less than 0. When the ego vehicle is on an uphill and the front vehicle disappears, it usually means that the front is a horizontal road surface (it may also be a downhill, but it can be regarded as horizontal, and the sight distance model of the two is the same). When the ego vehicle is on a downhill and the front vehicle disappears, it usually means that the front is a horizontal road surface (it may also be an uphill, but it can be regarded as horizontal, and the sight distance model of the two is the same).

[0061] As mentioned above, it is mainly for the early warning processing steps when the obstacle vehicle in front suddenly disappears.

[0062] On the other hand, when the obstacle vehicle ahead is within the detection field of view, the existing technology usually directly uses the detected vehicle distance (usually the straight-line distance) as the actual vehicle distance, which is relatively accurate on horizontal roads, but inaccurate on sloped roads. For example, when the road surface turns from horizontal to uphill, the road surface is curved, and the vehicle distance is not equal to the straight-line distance between the two vehicles.

[0063] In some of the embodiments, when an obstacle vehicle exists, step S140 is executed to determine the current actual vehicle distance, judge the relationship between the current actual vehicle distance and the safe following distance, and give a following vehicle warning prompt when the current actual vehicle distance is less than the safe following distance.

[0064] The steps for determining the current actual vehicle distance include:

[0065] Reference Figure 6 When the road condition is horizontal and turns uphill, the current actual vehicle distance is determined by the first uphill distance model, and the first uphill distance model is:

[0066]

[0067] Reference Figure 7 When the road condition changes from uphill to level, the current actual vehicle distance is determined by the second uphill distance model, and the second uphill distance model is:

[0068] S d =2S c ·|θ2|csc|θ2|

[0069] Reference Figure 8 When the road condition is horizontal and turns downhill, the current actual vehicle distance is determined by the first downhill distance model, and the first downhill distance model is:

[0070] S d =2S c ·|θ2|csc|θ2|

[0071] Reference Fig. 9 When the road condition changes from downhill to level, the current actual vehicle distance is determined by the second downhill distance model, and the second downhill distance model is:

[0072]

[0073] Among them, S d Indicates the current actual vehicle distance, S c represents the straight-line distance between the vehicle and the obstacle vehicle detected by the vehicle-mounted radar, θ2 represents the detection centerline direction of the vehicle-mounted radar and S c The angle of the detection direction, S c The detection direction refers to the direction in which the vehicle-mounted radar detects the obstacle vehicle.

[0074] Through the above distance model, the actual distance between the front and rear vehicles can be accurately obtained.

[0075] When an obstacle vehicle exists, the road condition judgment steps include:

[0076] When the ego vehicle is level and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the on-board radar is greater than 0, it is determined that the road condition is level turning uphill.

[0077] When the ego vehicle is level and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the on-board radar is less than 0, it is determined that the road condition is level turning downhill.

[0078] When the ego vehicle is facing upward and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the on-board radar is less than the detection centerline azimuth, the road condition is uphill to level.

[0079] When the ego vehicle is facing downward and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the on-board radar is greater than the detection centerline azimuth, the road condition is downhill to level.

[0080] Specifically, if the obstacle vehicle in front is traveling in the same direction as the vehicle (for example, both are on a horizontal road or an uphill road, etc.), and the two are on the same straight line, the obstacle vehicle in front will be detected by the detection center line of the vehicle-mounted radar. With this as a reference, through the above steps, the relationship between the obstacle vehicle in front and the vehicle can be determined.

[0081] Correspondingly, when there is an obstacle vehicle, the road condition judgment step also includes:

[0082] When the ego vehicle is level and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the on-board radar is equal to 0, the road condition is determined to be continuously level.

[0083] When the ego vehicle is facing upward and the azimuth of the straight-line distance between the ego vehicle and the obstacle vehicle detected by the on-board radar is equal to the azimuth of the detection center line, it is determined that the road condition is a continuous uphill slope.

[0084] When the ego vehicle is facing downward and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the on-board radar is equal to the detection centerline azimuth, it is determined that the road condition is a continuous downhill slope.

[0085] The steps of determining the current actual vehicle distance also include:

[0086] When the road condition is continuously level, continuously uphill or continuously downhill, the straight-line distance between the vehicle and the obstacle vehicle detected by the on-board radar is determined as the current actual vehicle distance.

[0087] In the above road conditions, since the road surface is basically straight rather than curved, the straight-line distance between the vehicle and the obstacle vehicle detected by the on-board radar is the current actual vehicle distance.

[0088] In some of the embodiments, the safe following distance includes a first safe following distance and a second safe following distance, the first safe following distance is greater than the second safe following distance, and the following warning prompt includes a general following warning prompt and an emergency following warning prompt.

[0089] When the current observation distance is less than the safe following distance, the following vehicle warning prompts are given including:

[0090] When the current observation distance is less than the first safe following distance and greater than the second safe following distance, a general following warning prompt is given; when the current observation distance is less than the second safe following distance, an emergency following warning prompt is given.

[0091] And / or, when the current actual vehicle distance is less than the safe following distance, the following vehicle warning prompt is given including:

[0092] When the current actual vehicle distance is less than the first safe following distance and greater than the second safe following distance, a general following warning prompt is given; when the current actual vehicle distance is less than the second safe following distance, an emergency following warning prompt is given.

[0093] Among them, the calculation formula for the first safe following distance is:

[0094] S warning =V e *TTC warning

[0095] Among them, S warning Indicates the first safe following distance, V e Indicates the vehicle speed, TTC warning Indicates the preset safety time, which is usually between 2.5 and 3.5 seconds;

[0096] The calculation formula for the second safe following distance is:

[0097] S min =S e -S f +S p

[0098] Among them, S min Indicates the second safe following distance, S e Indicates the braking distance of the vehicle, S f Indicates the braking distance of the obstacle vehicle, S p Indicates the preset safety distance.

[0099] As described above, the nighttime slope warning method for an autonomous driving vehicle proposed in the present invention has been described in a relatively complete manner.

[0100] It should also be noted that when the obstacle vehicle in front of the vehicle disappears from the detection field of view, it may be caused by the obstacle vehicle turning. However, this situation can be excluded by the detection data. The most important feature is that when the obstacle vehicle turns and disappears, its last moment position in the detection field of view must be on the left or right side of the detection field of view, and when the obstacle vehicle disappears on the uphill or downhill slope, its last moment position in the detection field of view must be on the upper side or the upper side of the detection field of view.

[0101] Therefore, in some of the embodiments, judging whether the obstacle vehicle in front of the vehicle disappears refers to judging whether the obstacle vehicle in front of the vehicle disappears without turning, and the disappearance of the obstacle vehicle refers to the disappearance of the obstacle vehicle without turning. However, it should be emphasized that even if no distinction is made between the disappearance modes, that is, it is assumed that a sloped road condition occurs when the obstacle vehicle disappears, it will not affect the implementation of the method and the solution of its technical problems. At this time, the method only has certain usage defects, that is, in some cases, the turning of the obstacle vehicle is misjudged as the obstacle vehicle going uphill or downhill. Compared with the prior art that assumes the disappearance of the obstacle vehicle as the absence of the obstacle vehicle in front of the vehicle, there is still obvious progress.

[0102] The nighttime slope warning method for an autonomous driving vehicle provided by the present invention is supplemented as follows by a specific embodiment.

[0103] In a specific embodiment, a nighttime slope warning method for an autonomous driving vehicle includes: a scene perception module, a risk assessment module, a distance detection module, a warning module, and a safety assurance module.

[0104] The scene perception module analyzes the collected radar time series data, parses the gradient of distance parameter changes during the vehicle following process, and determines whether the leading vehicle is always in a detected state by comparing the preset gradient value.

[0105] Since the on-board visual sensor is prone to failure under night conditions, the vehicle in front is mainly detected by the on-board radar. The visual sensor fails due to the influence of nighttime lighting, and the vehicle in front can only be detected by the radar; usually the radar returns 32 target parameters, and according to the number of detected targets, it will output stable target parameters with speed and distance, and the rest are noise points; in order to extract accurate target parameters, the noise points are first filtered, and according to the high-frequency characteristics of the noise points, the high-frequency noise is filtered out by a low-pass filter, and the low-frequency effective target parameters are retained; then according to the need to follow the vehicle, the target in front is detected, and the sensor detection range in the horizontal direction is constrained to ±10°; the detection angle in the vertical direction is constrained to ≥0°. In the target detection described in the above, if there is a vehicle in front, the relative speed value in the returned target parameter tends to 0 after the vehicle is stabilized; if a slope is encountered, the vehicle in front disappears, and the detected target relative speed value is approximately the opposite value of the vehicle speed.

[0106] Radar detection conditions mainly include:

[0107] The detection conditions are divided into four types: horizontal condition, horizontal-ramp condition, ramp condition, and ramp-horizontal condition. A climbing process following a vehicle will successively experience horizontal condition, horizontal-ramp condition, ramp condition, and ramp-horizontal condition. The main purpose of analyzing the climbing process in stages is to obtain the accurate distance between the front and rear vehicles through precise modeling.

[0108] The horizontal working condition is that the front and rear vehicles are on a horizontal road. At this moment, the distance between the front and rear vehicles changes continuously, and the distance change range should conform to the kinematic law. Assume that the speed of the front vehicle is V f , the deceleration is a f , the speed of the rear vehicle is V e , the deceleration is a e , the distance gradient change value Δd at the next moment should satisfy:

[0109]

[0110] The horizontal-ramp working condition is that the front vehicle is driving on a ramp, and the rear vehicle has not gone uphill and is still on a horizontal road surface; if the slope is too large, and the radar detects that the vehicle has no tendency to move left or right at the previous moment, then the front vehicle is going uphill. The basis for judging this working condition is that the front vehicle leaves the radar detection field of view and the detection distance value suddenly increases, far exceeding Δd.

[0111] The ramp condition is that both the front and rear vehicles are driving on a ramp, the slope angle of the front vehicle relative to the rear vehicle is 0, the distance value detected by the radar is the actual distance between the front and rear vehicles, and the distance change satisfies the Δd change law. The basis for judging this condition is that the horizontal-ramp condition appears first, and then the front vehicle is detected again and satisfies the Δd change law.

[0112] The ramp-to-horizontal condition is that the front vehicle leaves the ramp and reaches the horizontal road surface at the top of the slope, while the rear vehicle is still on the slope. The basis for judging this condition is that, when the horizontal-to-ramp condition has occurred, the slope angle of the front vehicle relative to the rear vehicle becomes a negative value until the slope angle becomes 0, and both the front and rear vehicles return to the horizontal road surface.

[0113] The precise modeling described above extracts the road surface within the radar field of view for different detection conditions, and uses a mathematical expression to represent the distance obtained by the radar by analyzing geometric relationships.

[0114] The front vehicle is in a detected state, that is, the front vehicle is within the sensor detection field of view, and the distance value detected by the radar changes relatively stably. After the front vehicle disappears, the radar detection distance will suddenly increase due to the slope, and the increase value is significantly greater than Δd; in addition, because of the loss of the dynamic target, the relative speed detected by the radar will change suddenly, and the relative speed value of the detected static target is approximately equal to the inverse of the vehicle's speed.

[0115] According to the relative positions of the front and rear vehicles, the downhill process is the opposite of the uphill process, and successively goes through four stages: horizontal-horizontal condition (indicating that the front and rear vehicles are on level ground), horizontal-ramp condition (indicating that the rear vehicle is on level ground and the front vehicle is on a ramp), ramp condition (indicating that the front and rear vehicles are on a downhill ramp), and ramp-horizontal condition (indicating that the rear vehicle is on a ramp and the front vehicle returns to level ground).

[0116] To determine whether the leading vehicle is going uphill or downhill, it is mainly determined by analyzing the data returned by the radar sensor. If the leading vehicle is going uphill, the radar detects the vehicle first, and then obtains the farthest ground distance within the line of sight after the leading vehicle disappears. The azimuth is upward and greater than 0. Subsequently, various stages of the uphill process appear in sequence. In the slope condition, the front of the leading and rear vehicles have the same elevation angle and are relatively horizontal. The radar can always detect the vehicle in front. After the leading vehicle reaches the top of the slope, the vehicle suddenly disappears, and the azimuth of the farthest distance in the radar line of sight is downward and less than 0. When the waiting vehicle reaches the top of the slope, the leading vehicle appears in the radar field of view again. The position condition of the leading and rear vehicles is horizontal-horizontal, and the uphill process is completed.

[0117] Before going downhill, the front and rear vehicles are on the same level ground, and the front vehicle is always within the detection field of view of the radar; when going downhill, after the front vehicle disappears relative to the vehicle speed, the azimuth of the farthest distance of the radar line of sight is downward and less than 0; then it experiences the ramp condition and the ramp-horizontal condition respectively, and finally the horizontal-horizontal condition, the front and rear vehicles both leave the ramp, and the downhill process is completed.

[0118] The risk assessment module analyzes the possible motion behaviors that may threaten the safety of the rear vehicle after the front vehicle disappears according to the changes in the detected state of the front vehicle, and assesses the maximum risk faced by the driving.

[0119] Considering the failure of the camera detection function at night, the limited radar detection angle, the front vehicle entering the detection blind spot, and the movement changes of the front vehicle after entering the blind spot, it can be divided into three working conditions. The three working conditions are:

[0120] Working condition 1: The front vehicle accelerates after entering the blind spot; in the working condition 1, after the front vehicle accelerates into the blind spot, the rear vehicle maintains its original speed and the distance between the front and rear vehicles continues to increase. Based on the minimum safety theory, the increased vehicle distance does not pose a threat to driving safety; if the rear vehicle also accelerates after the front vehicle disappears, it is necessary to analyze the specific changes in the distance between the front and rear vehicles to evaluate whether the vehicle distance meets the driving safety requirements.

[0121] Working condition 2: The front vehicle enters the blind spot and drives at a constant speed; in the working condition 2, when the front vehicle enters the blind spot at a constant speed, the rear vehicle maintains a constant speed, the distance between the front and rear vehicles remains unchanged, the distance between the vehicles meets the safety requirements of following the vehicle, and driving safety is not threatened; if the rear vehicle accelerates because there is "no car" in front, the distance between the front and rear vehicles decreases rapidly, and based on the minimum safety theory, the vehicle safety will be seriously threatened.

[0122] Working condition three: the front vehicle decelerates after entering the blind spot; in the working condition three, the front vehicle decelerates and enters the detection blind spot, and the rear vehicle maintains the same speed, then the distance between the front and rear vehicles will gradually decrease; if the rear vehicle accelerates, the distance between the vehicles will continue to decrease; in both cases, when the front vehicle is detected again, the distance between the front and rear vehicles is completely unable to meet the safety of following the vehicle, and the probability of rear-end collision is extremely high.

[0123] Risk assessment is based on three working conditions. The minimum safety distance for following vehicles is calculated in real time according to the minimum safety distance theory, and the following safety is evaluated by comparing the actual vehicle distance. If the vehicle distance is less than the minimum safety distance, it is judged that the risk of following the vehicle is extremely high at this moment. The following warning distance is calculated by converting the warning collision time, and the actual following distance is compared. If the actual vehicle distance is less than the warning distance, it means that there is a risk in following the vehicle, and the system needs to adjust the vehicle speed in time to restore the vehicle distance to a safe value.

[0124] The distance detection module first analyzes the following distance after the front vehicle disappears based on the risk assessment result, establishes a following model according to the vehicle motion state and parameters, and calculates the vehicle-to-vehicle distance; at the same time, establishes a line-of-sight distance model according to the radar viewing angle, and calculates the maximum distance value that can be detected within the radar viewing angle;

[0125] The distance detection module targets four working conditions: horizontal, horizontal-ramp, ramp, and ramp-horizontal. The position relationship between the front and rear vehicles is different, and the distance detection acquisition methods are different. In the horizontal working condition, the radar can detect the front vehicle, and the radar output distance is the actual distance between the front and rear vehicles. In the horizontal-ramp working condition, it is necessary to establish a geometric model to calculate the distance between the front and rear vehicles based on the position relationship between the front and rear vehicles. In the ramp working condition, the front and rear vehicles are both on the ramp and the vehicles are in the same plane. The radar output distance is the actual distance between the front and rear vehicles. In the ramp-horizontal working condition, it is necessary to establish a geometric model to calculate the distance between the front and rear vehicles based on the position relationship between the front and rear vehicles. In the distance detection module, when the front vehicle enters the detection blind spot, a geometric model is established based on the sensor perspective to calculate the sight distance.

[0126] The early warning module grades the early warning information, and different grades output different degrees of early warning information.

[0127] The warning module classifies the warning information into general warning and emergency warning. The general warning is based on the warning collision time TTC. warning Multiply by the current vehicle speed V e Calculate the warning following distance S warning , calculated as follows:

[0128] S warning =V e *TTC warning

[0129] Among them, TTCwarning It is an empirical value, usually ranging from 2.5 to 3.5. When the warning following distance is greater than the actual following distance S, it means that the distance between the front and rear vehicles is too small, and the vehicle begins to issue a periodic sound warning to remind the driver to control the deceleration and increase the distance between the vehicles.

[0130] The emergency warning is based on the minimum safety distance theory to calculate the minimum following distance S under real-time working conditions. min , calculated as follows:

[0131] S min =S e -S f +S p

[0132] Among them, S e is the braking distance of the rear vehicle, S f is the braking distance of the vehicle ahead, calculated as follows:

[0133]

[0134] When the minimum safety distance is greater than the actual following distance, the vehicle is in a dangerous state and a collision is very likely to occur during the following process, and the system continues to issue a sound warning.

[0135] The safety assurance module plans and adjusts the driving speed by comparing the actual distance and warning distance when following a vehicle, so as to maintain a safe and reasonable following distance.

[0136] Safety assurance module: After the system issues a general warning, if the driver does not take effective measures to increase the distance between vehicles within 2 seconds, the system will take active safety measures and slow down by braking. The active safety measures after the general warning take into account the riding comfort and control the braking deceleration value to ≤4m / s while ensuring the safety of following the vehicle. 2 ;

[0137] Safety guarantee module: When the system finds that an emergency warning is about to occur during the following process, the system gives priority to driving safety and uses emergency braking to quickly increase the distance between vehicles to prevent collision accidents caused by the short distance between vehicles and the vehicles not having time to adjust their speed. Emergency braking means that the vehicle brakes at the maximum deceleration and the brake hydraulic pressure reaches the maximum value.

[0138] Figure 2This is a distance detection logic process diagram of a nighttime slope warning and active safety implementation method for automatic driving assistance of the present invention. First, the radar detection data is used to determine whether the road ahead is a ramp or a straight road. If the front vehicle goes uphill and disappears, the farthest distance azimuth detected by the radar is upward and greater than 0; then the horizontal-ramp condition, ramp condition, ramp-horizontal condition and horizontal-horizontal condition appear one by one, and the radar detection azimuth changes according to greater than 0, equal to 0, less than 0, and equal to 0 respectively, and the current working condition is determined according to the change of the angle in sequence; the distance value detected by the sensor in the straight road condition is equal to the actual vehicle distance between the front and rear vehicles; under the ramp condition, the radar detection distance cannot be used directly, and it is necessary to establish a geometric model based on the radar detection distance according to the uphill and downhill processes and the relative position between the front and rear vehicles to calculate the actual vehicle distance or line of sight.

[0139] The downhill process is opposite to the uphill process. During the downhill process, the radar detection angle will be less than 0, equal to 0, greater than 0 and equal to 0 respectively. The corresponding working conditions are horizontal-ramp condition, ramp condition, ramp-horizontal condition and horizontal-horizontal condition.

[0140] In an embodiment of the present invention, a nighttime slope warning device for an autonomous driving vehicle is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. The terms "module", "unit", "subunit", etc. used below may implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0141] The nighttime slope warning device for autonomous vehicles includes:

[0142] A data acquisition module is used to obtain detection data of the vehicle radar about the road conditions in front of the vehicle;

[0143] A state judgment module is used to judge the road condition and whether the obstacle vehicle in front of the vehicle has disappeared according to the detection data;

[0144] The early warning prompt module is used to determine the current observation distance when the obstacle vehicle disappears, judge the relationship between the current observation distance and the safe following distance, and give a following vehicle early warning prompt when the current observation distance is less than the safe following distance;

[0145] The step of determining the current observation distance comprises:

[0146] When the road condition is a horizontal turning uphill, the current observation distance is determined by a first uphill sight distance model, and the first uphill sight distance model is:

[0147]

[0148] When the road condition is from uphill to level, the current observation distance is determined by a second uphill sight distance model, and the second uphill sight distance model is:

[0149] S v =2S r |θ1|csc|θ1|

[0150] When the road condition is a horizontal turning downhill, the current observation distance is determined by a first downhill sight distance model, and the first downhill sight distance model is:

[0151] S v =2S r |θ1|csc|θ1|

[0152] When the road condition is a downhill to horizontal road, the current observation distance is determined by a second downhill sight distance model, and the second downhill sight distance model is:

[0153]

[0154] Among them, S v Indicates the current observation distance, S r represents the maximum road distance detected by the vehicle-mounted radar, θ1 represents the detection centerline direction of the vehicle-mounted radar and S r The angle of the detection direction.

[0155] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0156] In an embodiment of the present invention, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the night slope warning method for an autonomous driving vehicle provided by the present invention.

[0157] In an embodiment of the present invention, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the night slope warning method for an autonomous driving vehicle provided by the present invention are implemented.

[0158] It should be understood that the specific embodiments described herein are only used to explain the application, rather than to limit it. Based on the embodiments provided in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the protection scope of this application.

[0159] Obviously, the drawings are only some examples or embodiments of the present application. For ordinary technicians in the field, the present application can also be applied to other similar situations based on these drawings without creative work. In addition, it is understandable that although the work done in this development process may be complicated and lengthy, for ordinary technicians in the field, certain changes in design, manufacturing or production based on the technical content disclosed in this application are only conventional technical means and should not be regarded as insufficient content disclosed in this application.

Claims

1. A nighttime slope warning method for an autonomous driving vehicle, characterized in that: include: Obtain detection data of the vehicle radar about the road conditions in front of the vehicle; Determine the road condition and whether the obstacle vehicle in front of the vehicle has disappeared according to the detection data; When the obstacle vehicle disappears, determine the current observation distance, judge the relationship between the current observation distance and the safe following distance, and give a following vehicle warning prompt when the current observation distance is less than the safe following distance; The step of determining the current observation distance comprises: When the road condition is a horizontal turning uphill, the current observation distance is determined by a first uphill sight distance model, and the first uphill sight distance model is: When the road condition is from uphill to level, the current observation distance is determined by a second uphill sight distance model, and the second uphill sight distance model is: S v =2S r θ1|cscθ1| When the road condition is a horizontal turning downhill, the current observation distance is determined by a first downhill sight distance model, and the first downhill sight distance model is: S v =2S r |θ1|csc|θ1| When the road condition is a downhill to horizontal road, the current observation distance is determined by a second downhill sight distance model, and the second downhill sight distance model is: Among them, S v Indicates the current observation distance, S r represents the maximum road distance detected by the vehicle-mounted radar, θ1 represents the detection centerline direction of the vehicle-mounted radar and S r The angle of the detection direction.

2. The nighttime slope warning method for an autonomous driving vehicle according to claim 1, characterized in that: When the obstacle vehicle exists, determine the current actual vehicle distance, judge the relationship between the current actual vehicle distance and the safe following distance, and give a vehicle following warning prompt when the current actual vehicle distance is less than the safe following distance; The step of determining the current actual vehicle distance comprises: When the road condition is a horizontal turning uphill, the current actual vehicle distance is determined by a first uphill distance model, and the first uphill distance model is: When the road condition is an uphill to flat road, the current actual vehicle distance is determined by a second uphill distance model, and the second uphill distance model is: S d =2S c ·|θ2|csc|θ2| When the road condition is a horizontal turning downhill, the current actual vehicle distance is determined by a first downhill distance model, and the first downhill distance model is: S d =2S c ·|θ2|csc|θ2| When the road condition is a downhill to horizontal road, the current actual vehicle distance is determined by a second downhill distance model, and the second downhill distance model is: Among them, S d Indicates the current actual vehicle distance, S c represents the straight-line distance between the vehicle and the obstacle vehicle detected by the vehicle-mounted radar, θ2 represents the detection centerline direction of the vehicle-mounted radar and S c The angle of the detection direction.

3. The nighttime slope warning method for an autonomous driving vehicle according to claim 1, characterized in that: When the obstacle vehicle disappears, the road condition judgment step includes: When the vehicle is level and the azimuth of the maximum road distance detected by the vehicle-mounted radar is greater than 0, it is determined that the road condition is level turning uphill; When the vehicle is level and the azimuth angle of the maximum road surface distance detected by the vehicle-mounted radar is less than 0, the road condition is determined to be a level turning downhill; When the vehicle is facing upward, it is determined that the road condition is an uphill to horizontal condition; When the vehicle is facing downward, it is determined that the road condition is a downhill to level road.

4. The nighttime slope warning method for an autonomous driving vehicle according to claim 2, characterized in that: When the obstacle vehicle exists, the road condition judgment step includes: When the ego vehicle is level and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the vehicle-mounted radar is greater than 0, it is determined that the road condition is level turning uphill; When the ego vehicle is level and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the vehicle-mounted radar is less than 0, it is determined that the road condition is level turning downhill; When the vehicle is facing upward and the azimuth of the straight-line distance between the vehicle and the obstacle vehicle detected by the vehicle-mounted radar is less than the azimuth of the detection center line, the road condition is an uphill to horizontal road condition; When the vehicle is facing downward and the azimuth of the straight-line distance between the vehicle and the obstacle vehicle detected by the vehicle-mounted radar is greater than the azimuth of the detection center line, the road condition is downhill to level.

5. The nighttime slope warning method for an automatic driving vehicle according to claim 4, characterized in that: When the obstacle vehicle exists, the road condition judgment step further includes: When the ego vehicle is level and the straight-line distance azimuth between the ego vehicle and the obstacle vehicle detected by the vehicle-mounted radar is equal to 0, it is determined that the road condition is continuously level; When the vehicle is facing upward and the azimuth of the straight-line distance between the vehicle and the obstacle vehicle detected by the vehicle-mounted radar is equal to the azimuth of the detection center line, it is determined that the road condition is a continuous uphill slope; When the vehicle is facing downward and the straight-line distance azimuth between the vehicle and the obstacle vehicle detected by the vehicle-mounted radar is equal to the detection centerline azimuth, it is determined that the road condition is a continuous downhill slope; The step of determining the current actual vehicle distance also includes: When the road condition is continuously level, continuously uphill or continuously downhill, the straight-line distance between the vehicle and the obstacle vehicle detected by the vehicle-mounted radar is determined as the current actual vehicle distance.

6. The nighttime slope warning method for an autonomous driving vehicle according to claim 2, characterized in that: The safe following distance includes a first safe following distance and a second safe following distance, the first safe following distance is greater than the second safe following distance, and the following warning prompt includes a general following warning prompt and an emergency following warning prompt; When the current observation distance is less than the safe following distance, giving a following vehicle warning prompt includes: When the current observation distance is less than the first safe following distance and greater than the second safe following distance, giving the general following vehicle warning prompt; When the current observation distance is less than the second safe following distance, giving the emergency following warning prompt; And / or, when the current actual vehicle distance is less than the safe following distance, providing a following vehicle warning prompt includes: When the current actual vehicle distance is less than the first safe following distance and greater than the second safe following distance, giving the general following vehicle warning prompt; When the current actual vehicle distance is less than the second safe following distance, the emergency following warning prompt is given.

7. The nighttime slope warning method for an automatic driving vehicle according to claim 6, characterized in that: The calculation formula of the first safe following distance is: S warning =V e *TTC warning Among them, S warning Indicates the first safe following distance, V e Indicates the vehicle speed, TTC warning Indicates the preset safety time; The calculation formula of the second safe following distance is: S min =S e -S f +S p Among them, S min Indicates the second safe following distance, S e Indicates the braking distance of the vehicle, S f Indicates the braking distance of the obstacle vehicle, S p Indicates the preset safety distance.

8. A nighttime slope warning device for an autonomous driving vehicle, characterized in that: include: A data acquisition module is used to obtain detection data of the vehicle radar about the road conditions in front of the vehicle; A state judgment module is used to judge the road condition and whether the obstacle vehicle in front of the vehicle has disappeared according to the detection data; The early warning prompt module is used to determine the current observation distance when the obstacle vehicle disappears, judge the relationship between the current observation distance and the safe following distance, and give a following vehicle early warning prompt when the current observation distance is less than the safe following distance; The step of determining the current observation distance comprises: When the road condition is a horizontal turning uphill, the current observation distance is determined by a first uphill sight distance model, and the first uphill sight distance model is: When the road condition is from uphill to level, the current observation distance is determined by a second uphill sight distance model, and the second uphill sight distance model is: S v =2S r θ1|cscθ1| When the road condition is a horizontal turning downhill, the current observation distance is determined by a first downhill sight distance model, and the first downhill sight distance model is: S v =2S r θ1|cscθ1| When the road condition is a downhill to horizontal road, the current observation distance is determined by a second downhill sight distance model, and the second downhill sight distance model is: Among them, S v Indicates the current observation distance, S r represents the maximum road distance detected by the vehicle-mounted radar, θ1 represents the detection centerline direction of the vehicle-mounted radar and S r The angle of the detection direction.

9. An electronic device, comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor is configured to run the computer program to execute the nighttime slope warning method for an autonomous driving vehicle according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the nighttime slope warning method for an autonomous driving vehicle according to any one of claims 1 to 7 are implemented.

Images