Method for avoiding a collision of a vehicle and evaluation unit for carrying out the method

Abstract

The present invention relates to a method for avoiding a collision of a vehicle (2) with an object (8), comprising the method steps of determining a first variable (a) representative of the collision probability between the vehicle (2) and the object (8) on the basis of the trajectories of the vehicle (2) and the object (8), determining a second variable (b) representative of a time period until the probable collision between the vehicle (2) and the object (8), calculating a braking time (t b ) for the vehicle (2) as a function value of a function (f) with at least the first variable (a) and the second variable (b), comparison of the calculated braking time (t b ) with the current time (t a ) and automatic braking of the vehicle (2) when the calculated braking time (t b ) before the current time (t a ) or the current time (t a). Furthermore, the present invention relates to an evaluation unit (16) for a vehicle (2).

Description

State of the art

[0001] The present invention relates to a method for avoiding a collision between a vehicle and an object, in particular another road user. Furthermore, the present invention relates to an evaluation unit for a vehicle for implementing the method.

[0002] In the current state of the art, to avoid collisions between a vehicle and an object, usually another road user, automated interventions in the operation of the vehicle are increasingly being implemented, for example through automated emergency braking (AEB). The basis for determining the need for automated intervention is a situation analysis within the framework of a model of the vehicle's environment. The information required for this is determined or recorded in various ways, namely by the vehicle's environmental sensors, for example comprising a camera, radar, light-assisted object detection and distance measurement (LIDAR, "light detecting and ranging"), through communication between the vehicle and the traffic infrastructure, or through communication between the vehicle and other road users.

[0003] During the situation analysis based on the environment model, it is first determined whether there is any possibility of a collision between the vehicle and the object. This is done based on the determined trajectories of the vehicle and object. If the two trajectories intersect at a point or at least approach each other to within a predetermined limit, a collision possibility exists, and a collision probability is determined. Furthermore, a time period until the probable collision (TTC) is determined based on the provided information. If the determined collision probability exceeds a predetermined probability limit and / or the determined time period falls below a predetermined time limit, automatic emergency braking is triggered.

[0004] For example, CN 114906136 A describes a method for avoiding a collision between a vehicle and a pedestrian, whereby automatic emergency braking is triggered when the time until the probable collision falls below a predetermined time limit. Disclosure of the invention

[0005] With the present invention, an automatic emergency braking of a vehicle can be carried out at an optimal time in order to safely avoid, on the one hand, a collision of the vehicle with an object / road user and, on the other hand, a possibly unnecessary emergency braking, which in turn can cause an emergency situation, while also providing increased driving comfort.

[0006] According to the invention, a method for avoiding a collision between a vehicle and an object / road user having the features of patent claim 1, an evaluation unit for a vehicle having the features of patent claim 9 and a computer program having the features of patent claim 10 are therefore provided.

[0007] Accordingly, a method for avoiding a collision between a vehicle and an object is provided, comprising the following method steps: determining a first variable representative of the collision probability between the vehicle and the object, preferably a foreign vehicle, on the basis of the trajectories of the vehicle and the object, determining a second variable representative of a time period until the probable collision between the vehicle and the object, calculating a braking time for the vehicle as a function value of a function with at least the first variable and the second variable, comparing the calculated braking time with the current time, and automatically braking the vehicle if the calculated braking time is before the current time, thus in the past, or corresponds to the current time.

[0008] Furthermore, an evaluation unit, preferably an evaluation and control unit, for a vehicle, which is designed to carry out the aforementioned method, and a computer program are provided which, when executed on a processor, carries out the aforementioned method.

[0009] One idea of ​​the invention lies in overcoming rigid limits for the determined collision probability and the time until probable collision (TTC) by calculating a braking time from these variables, from which the necessity of automatic emergency braking (AEB) is derived. This achieves a sensible compromise between early emergency braking to avoid a collision and later emergency braking to avoid unjustified braking, whereby such a compromise can increase both safety and driving comfort. In particular, it is possible to calculate a braking time before or after the current time, whereby a calculated braking time in the future can rule out the initiation of measures to avoid automatic emergency braking, while a calculated braking time in the past can still rule out a collision.

[0010] Advantageous embodiments and further developments emerge from the further subclaims and from the description with reference to the figures.

[0011] According to a preferred embodiment of the invention, the vehicle issues a warning message if the calculated braking time is later than the current time and the difference between the calculated braking time and the current time falls below a predetermined limit. The warning message can preferably be any signal perceptible to the driver of the vehicle, which is particularly preferably output via a human-machine interface (HMI).

[0012] According to a further preferred embodiment of the invention, the function is designed such that the earlier the calculated braking time is, the larger the first variable and / or the smaller the second variable. This allows, for example, a relatively low collision probability to be offset by a relatively long time span until the probable collision, and vice versa. This allows unnecessary or premature automatic emergency braking to be avoided, which in turn can cause emergency situations.

[0013] According to a further preferred embodiment of the invention, at least two or more of the variables of the function are weighted differently to account for the significance of the individual variables and further optimize the timing of an emergency braking. It is preferred if the weighting values ​​for weighting the variables are determined empirically, preferably based on data from an accident database, and / or by a machine learning algorithm.

[0014] In order to further optimize the timing of emergency braking, the function in a further preferred embodiment of the invention includes at least one further determined variable.

[0015] According to a further preferred embodiment of the invention, the function includes a representative variable determined for the probable accident severity in the event of a collision, wherein the calculated braking time is preferably the earlier, the greater the representative variable determined for the probable accident severity.

[0016] According to a further preferred embodiment of the invention, the function includes a representative variable determined for the attention or activity of the driver of the vehicle, wherein the calculated braking time is preferably later, the greater the representative variable determined for the attention or activity of the driver of the vehicle is.

[0017] According to a further preferred embodiment of the invention, the function includes a representative variable determined for the accident risk due to braking of the vehicle, wherein the calculated braking time is preferably the earlier, the greater the representative variable determined for the accident risk due to braking of the vehicle is.

[0018] The invention will be explained in more detail below using an exemplary embodiment with reference to the accompanying drawings. They show: Fig. 1 a schematic representation of a vehicle with an embodiment of the evaluation unit according to the invention in a traffic environment and Fig. 2 a flowchart of an embodiment of the method according to the invention for implementation by the evaluation unit.

[0019] Fig. 1 shows a schematic representation of a vehicle with an embodiment of the evaluation unit according to the invention in a traffic environment.

[0020] The vehicle 2 moves in the direction of travel 4 within an environment with a traffic infrastructure 6. In a corresponding manner, an object 8, preferably another road user, moves in the environment in the direction of travel 10. The vehicle 2 has an environmental sensor system 12 for detecting information in the environment of the vehicle 2, which can include, for example, a camera, radar and light-based object recognition and distance measurement (LIDAR).

[0021] Furthermore, the vehicle 2 has a communication device 14 for receiving and transmitting information or data. The communication device 14 preferably enables a V2X ("vehicle-to-everything") exchange of information and data, for example, a V2V ("vehicle-to-vehicle") exchange between the vehicle 2 and the object 8, as well as a V2I ("vehicle-to-infrastructure") and I2V ("infrastructure-to-vehicle") exchange between the vehicle 2 and the traffic infrastructure 6.

[0022] The information or data from the environment sensor system 12 and the communication device 14 can be evaluated by an evaluation unit 16 of the vehicle 2, wherein the evaluation unit 16 is designed such that it can Fig. 2 described method. A corresponding computer program is provided, which is executed on a processor of the evaluation unit 16 in order to carry out the method according to Fig. 2. The evaluation unit 16 is operatively connected to a braking device 18 of the vehicle 2, so that the braking device 18 can be controlled by the evaluation unit 16 or an intermediate control device to achieve automatic braking or automatic emergency braking.

[0023] Fig. 2 shows a flowchart of an embodiment of the method according to the invention.

[0024] The method is carried out continuously or periodically by the evaluation unit 16, wherein in a method step 20 the environment or environment model is examined for the possibility of a collision. For this purpose, the trajectory of the vehicle 2 is compared with the trajectories of objects, preferably other road users. The trajectories are preferably determined using the MPP method ("most probable path"). Should the trajectories of the vehicle 2 and the object 8 overlap at a point or at least approach each other to within a predetermined distance limit, the possibility of a collision between the vehicle 2 and the object 8 is determined in method step 22. The method continues with the subsequent steps in which the object 8 is considered a possible collision opponent.

[0025] In the evaluation unit 16, variables are determined in process steps 24 to 32, which are used to calculate a braking time t b in method step 34. In the illustrated embodiment, the braking time t b as the function value of a function f with the variables a, b, c, d and e, i.e. t b = f(a, b, c, d, e).

[0026] In method step 24, the first variable a, which is representative of a collision probability between the vehicle 2 and the object 8, is determined on the basis of the trajectory of the vehicle 2 on the one hand and the trajectory of the object 8 on the other hand. The function f is designed such that the braking time t to be calculated bthe earlier it is, the larger the first variable a is. When determining the collision probability, the following criteria can be taken into account alternatively or additionally: If the trajectories do not overlap, taking into account the extent of the vehicles (e.g. the bounding boxes), reference can be made to the determined smallest distance between the trajectories (again taking into account the vehicle extent). The smaller this distance, the greater the first variable a and thus the collision probability. The trajectories can also be estimated either on the basis of the current dynamic status of the vehicle 2 and / or the object 8, for example on the basis of position, yaw angle, speed, acceleration and / or yaw rate, or they can be taken from the V2X exchange via the communication device 14.For example, a CAM message (cooperative awareness message) or an MCM message (maneuver coordination message) can serve as the basis for determining the trajectory of object 8. The dynamic status and the V2X data can also be combined (e.g., using data fusion) to improve the accuracy of trajectory determination.

[0027] The determination of the collision probability and thus of the first variable a also depends on the accuracy of the trajectory prediction. The accuracy of the trajectory can be represented, for example, by adding confidence intervals and / or standard deviations to the trajectory points or trajectory course. If the entire trajectory or individual trajectory points are inaccurate, the trajectory points become trajectory regions in which the vehicle or object can be located with a certain probability distribution. To calculate the collision probability of the trajectories, the distance calculation should be based on these trajectory regions rather than on the trajectory points. In general, the more accurate the trajectory, the more reliable the calculation of the collision probability becomes, and vice versa.

[0028] The accuracy of the trajectory prediction can be calculated depending on several factors. For example, the accuracy of the current dynamic status of the vehicle 2 and / or the object 8 can be taken into account when calculating the trajectory accuracy, i.e., a more accurate trajectory can be predicted based on a more accurate dynamic status. Additionally, the confidence levels of received V2X data can be considered when calculating the trajectory accuracy. The prediction horizon of the trajectory can also be considered, i.e., points of the trajectories further in the future have a higher degree of inaccuracy, so this could possibly be taken into account by reducing the accuracy of the trajectory prediction.The driving situation should also be taken into account, especially since a simple driving situation, for example on a highway with low traffic density, results in higher trajectory accuracy than is the case in more complex driving situations, such as in a city with high traffic density. The type of road user for whom the trajectory is predicted also influences the accuracy of the trajectory prediction. For example, the predicted trajectories can be sorted from highest to lowest accuracy depending on the road user type: train at a level crossing, tram, truck, car, motorcycle, bicycle, adult pedestrian, child. Finally, the driving style or movement pattern of the vehicle for which the trajectory is predicted influences the accuracy of the trajectory prediction.A detected aggressive driving style or an erratic movement pattern results in a lower accuracy of the trajectory prediction and vice versa.

[0029] The estimated trajectories of the potential collision opponent are preferably calculated as the most probable trajectories considering the available data. However, the collision probability may also depend on alternative trajectories that could also be taken by object 8. For example, it is possible that the estimated trajectories of vehicle 2 and object 8 have a high collision probability. However, if object 8 has sufficient time to brake comfortably and avoid the collision, the estimated collision probability will be lower. Alternative opponent trajectories can therefore preferably be estimated in the following way: Desired or alternative trajectories contained in a received MCM message can describe further possible trajectories for object 8 (with a lower probability than the planned trajectory).For example, the probability of each trajectory in the MCM message can be estimated as an inverse function of the trajectory cost. Furthermore, depending on the driving scenario, e.g., if object 8 has no driving priority, a braking trajectory can be estimated. Or, if object 8's lane is closed, a lane-change curve can be estimated. Furthermore, depending on the vehicle type, a restricted set of trajectories may be available. For example, some vehicle types have a restricted set of trajectories they can take. For example, a heavy truck has a reduced maximum acceleration, which limits its possible alternative trajectories; or a tram can only follow its tram tracks.

[0030] In method step 26, the second variable b, which is representative of the time period until the probable collision (TTC) between the vehicle 2 and the object 8, is determined. The function f is designed such that the braking time t to be calculated b the earlier it is, the smaller the second variable b is.

[0031] The second variable b is also determined or calculated based on the determined trajectories of vehicle 2 and object 8. Preferably, the time period between the time of the calculation and the time at which the trajectories first intersect at a point or have approached each other to such an extent that a predetermined distance limit or minimum distance is undershot is calculated.

[0032] As already explained above, the function f for calculating the braking time also includes the additional variables c, d, and e to determine the optimal time for automatic emergency braking. Thus, in method step 28, the third variable c, which is representative of the probable accident severity in the event of a collision between vehicle 2 and object 8, is also determined. The function f is designed such that the braking time t to be calculated b the earlier it is, the larger the third variable c is.

[0033] The severity of the accident is determined based on the expected degree of injury to the persons involved. The Abbreviated Injury Scale (AIS), developed by the Association for the Advancement of Automotive Medicine (AAAM), can preferably form the basis for determining the third variable c. In particular, the injury levels used in the AIS, from 1 (probability of death < 0.1%) to 6 (probability of death = 100%), are suitable parameters that can be used to determine the severity of the accident. The determination of the third variable c, and thus the severity of the accident, is preferably carried out using injury probability curves depending on the expected type of collision and / or the change in the speed of vehicle 2 during the potential accident.Alternatively or in addition, the PVERL scale commonly used in Germany can also be used, ranging from 1 (no injury) to 2 (slightly injured) and 3 (severely injured) to 4 (fatally injured). The higher the expected degree of injury and / or the probability of death, the higher the third variable c, i.e., the accident severity.

[0034] In method step 30, the fourth variable d is determined, which is representative of the attention or activity of the driver of the vehicle 2, and possibly also of the attention or activity of the driver of the object 8. The function f is designed in such a way that the braking time t to be calculated bThe later the braking time, the larger the fourth variable d is. The fourth variable d ensures that the driver's current ability to absorb and process information and, on this basis, to actively intervene in the course of the journey is taken into account when calculating the braking time t b is taken into account.

[0035] To determine the fourth variable d, the degree of actuation of the control elements operable by the driver, such as the accelerator pedal, brake pedal, or steering wheel, can preferably be determined. Thus, the exertion of pressure on the accelerator or brake pedal or a torque exerted on the steering wheel indicates increased activity and attention on the part of the driver. Alternatively or additionally, a driver monitoring system (DMS) can be provided in the vehicle 2, possibly also in the object 8, which provides information on the driver's attention as a basis for determining the fourth variable d. Such a driver monitoring system can, for example, comprise a camera, radar, or infrared sensors.

[0036] In method step 32, the fifth variable e, which is representative of the accident risk due to braking of vehicle 2, is determined. The function f is designed such that the braking time t to be calculatedb The earlier the time, the larger the fifth variable e is. Due to the fifth variable e, automatic braking of the vehicle could be initiated relatively early, but this allows more time for a more moderate braking maneuver, in which the determined accident risk can be reduced by the vehicle's own braking.

[0037] When determining the accident risk due to braking, a collision risk with a vehicle following vehicle 2 is preferably determined. The information required for this can preferably be recorded via the environment sensor system 12, but in principle also via V2V or V2X exchange, whereby in particular the distance to the following vehicle, the relative speed between vehicle 2 and the following vehicle and, if applicable, also the vehicle type or the vehicle weight of the following vehicle can be taken into account, especially since vehicle type and weight influence the braking distance of the following vehicle. A short distance, a high relative speed of the faster following vehicle and its weight thus increase the accident risk and hence the fifth variable e.

[0038] In process step 34, the braking time t bas the function value of the function f with the determined variables a, b, c, d and e. At least two of the variables a, b, c, d and e, preferably all of the variables mentioned, are weighted differently by assigning them different weighting values ​​g a , g b , g c , g d , and g e The weighting values ​​g a , g b , g c , g d , g e are preferably determined empirically, for example, based on data from an accident database, and / or by a machine learning algorithm. For example, the function could be designed as shown below: tb=f(a, b, c, d, e)=a*ga+b*gb+c*gc+d*gd+e*ge

[0039] The weighting values ​​are presented as weighting factors for example. Although the weighted variables are summed here, it is also possible to use a different arithmetic operation, such as multiplication.

[0040] The braking time t calculated in this way b is compared in process step 36 with the current time t a If the comparison leads to the result that the calculated braking time is after the current time t a and thus lies in the future, a warning message is only issued to the driver in process step 38 by the evaluation unit 16 of the vehicle 2 via a human-machine interface if the difference between the calculated braking time t b and the current time t afalls below a predetermined limit so that the driver is alerted to a potential collision and the probable collision can still be averted by human intervention if necessary.

[0041] However, if the comparison in process step 36 leads to the result that the calculated braking time t b before the current time t a and thus lies in the past or corresponds to it, an automatic braking or an automatic emergency braking takes place in method step 40 by controlling the braking device 18 on the basis of a corresponding signal from the evaluation unit 16. QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] CN 114906136 A

[0004]

Claims

[1] Method for avoiding a collision of a vehicle (2) with an object (8) with the method steps Determination of a first variable (a) representative of the collision probability between the vehicle (2) and the object (8) on the basis of the trajectories of the vehicle (2) and the object (8), Determination of a second variable (b) representative of a time period until the probable collision between vehicle (2) and object (8), Calculation of a braking time (t b ) for the vehicle (2) as a function value of a function (f) with at least the first variable (a) and the second variable (b), Comparison of the calculated braking time (t b ) with the current time (t a ) and automatic braking of the vehicle (2) when the calculated braking time (t b ) before the current time (t a ) or the current time (t a) corresponds. [2] Method according to claim 1, wherein the vehicle (2) issues a warning message when the calculated braking time (t b ) after the current time (t a ) and the difference between the calculated braking time (t b ) and current time (t a ) falls below a predetermined limit. [3] Method according to one of claims 1 or 2, in which the calculated braking time (t b ) the earlier it is, the larger the first variable (a) and / or the smaller the second variable (b) is. [4] Method according to one of the preceding claims, in which the function (f) includes at least one further determined variable (c; d; e) and / or at least two or more of the variables (a, b, c, d, e) of the function (f) are weighted differently. [5] Method according to claim 4, wherein the weighting values ​​(g a , g b , g c , g d , ge ) for weighting the variables (a, b, c, d, e) are determined empirically, preferably on the basis of data from an accident database, and / or by a machine learning algorithm. [6] Method according to one of claims 4 or 5, in which the function (f) includes a variable (c) representative of the probable severity of an accident in the event of a collision, the calculated braking time (t b ) preferably the earlier the time is, the greater the representative variable (c) determined for the probable accident severity is. [7] Method according to one of claims 4 to 6, wherein the function (f) includes a representative variable (d) determined for the attention or activity of the driver of the vehicle (2), wherein the calculated braking time (t b) preferably the later the time is, the greater the representative variable (d) determined for the attention or activity of the driver of the vehicle (2). [8] Method according to one of claims 4 to 7, in which the function (f) includes a variable (e) representative of the risk of an accident due to braking of the vehicle (2), the calculated braking time (t b ) preferably the earlier, the greater the representative variable (e) determined for the risk of an accident due to braking of the vehicle (2). [9] Evaluation unit for a vehicle (2) designed to carry out the method according to one of the preceding claims. [10] A computer program which, when executed on a processor, carries out the method according to any one of claims 1 to 8.

Images