WO2022089989A1

WO2022089989A1 - Method and system for spatially avoiding collisions

Info

Publication number: WO2022089989A1
Application number: PCT/EP2021/078891
Authority: WO
Inventors: Nico Stock; David Schnitzler; Niklas Roth; Carsten Lenz
Original assignee: Zf Friedrichshafen Ag
Priority date: 2020-10-26
Filing date: 2021-10-19
Publication date: 2022-05-05
Also published as: DE102020213456A1

Abstract

The invention relates to a method for spatially avoiding collisions during a turning process of an ego-vehicle, the method comprising the following: a) acquiring, filtering and processing sensor data regarding the ego-vehicle and regarding an oncoming object; b) determining that, for the ego-vehicle, there is a turning situation with the oncoming object; c) analyzing and evaluating whether there is a critical turning situation by: i) analyzing the attention of the driver in the ego-vehicle; ii) detecting a turning maneuver of the ego-vehicle; and iii) evaluating the motion of the object; steps a) and/or b) being repeated if there is insufficient sensor data; and the following steps also being carried out if there is a critical turning situation: d) predicting the motion of the ego-vehicle and the motion of the object; e) carrying out an overlap analysis; steps a) to e) being repeated in full or in part if no collision course is detected; and the method also comprising the following if a collision course is detected: f) determining and analyzing a remaining distance for braking the ego-vehicle; g) initiating an automatic emergency-braking assistant if the remaining distance is greater than a necessary braking distance; or h) outputting a warning and triggering a prefill function if the remaining distance is less than the necessary braking distance.

Description

Method and system for spatial collision avoidance

field of invention

The present invention generally relates to collision situations between two vehicles or between vehicles and/or objects in a turning situation. In particular, methods and systems for avoiding a spatial collision based on a minimal braking effect on an ego vehicle are in the foreground.

Background of the Invention

Methods for detecting critical driving situations of motor vehicles in road traffic are known from the prior art, the emergency braking functions of which work with a maximum deceleration of the driving speed up to standstill when the ego vehicle is driving straight ahead.

US 10 486 707 B2 shows a system and a method for predicting a driver's intention at an intersection. At this time, it is estimated whether a driver of an own vehicle or a distant vehicle intends to turn left or right or to cross an intersection straight ahead before both reach the intersection. The estimation relies on a probabilistic model using a dynamic Bayesian network. External parameters at an intersection, such as the position and speed of the distant vehicle, and internal information on the operation of one's own vehicle are determined. The method then predicts a turning intent of the host vehicle and/or the remote vehicle at the intersection using the model based on the external parameters and the internal cues using the model.

However, the systems and methods known from the prior art for detecting critical driving situations of motor vehicles in road traffic show a range of disadvantages and difficulties. As a rule, the possibility of using known emergency brake assistants AEB (automatic emergency braking) very limited. In most cases, they are only able to react to vehicles driving ahead or parked vehicles. In addition to dangerous situations that are still not recognized or recognized too late, such automated braking systems do not intervene in turning scenarios, but only brake when the ego vehicle is driving straight ahead. Vehicle manufacturers are therefore increasingly interested in using emergency brake assistants in other traffic situations.

In addition, braking is often too early and/or too strong in order to avoid an inherent collision. As a result, all components involved in the braking process of the vehicle, such as brakes, brake pads, tires, etc., wear out prematurely. This not only results in increased maintenance costs for the vehicle in question, but possibly even a shorter product life cycle. False triggering of such methods and systems and increased or premature braking to a standstill also lead to avoidable increased fuel and energy consumption, which is not only based on an actuation of the braking system itself, but also on a subsequent braking process, again necessary acceleration of the vehicle to leave the dangerous situation. In addition, there is an uncomfortable driving behavior of the own vehicle for the driver and for possible other passengers due to a too abrupt, jerky and mass inertia-related reduction in the driving speed. Furthermore, such driving maneuvers, especially if they were carried out as a "false alarm" and thus possibly unnecessary, impair traffic both in the city and in the country. Traffic jams and possible rear-end collisions are the result.

object of the invention

It is the object of the present invention to overcome the above-described and other disadvantages of existing systems and methods for detecting critical driving situations. In particular, a prediction of the point of collision between an ego vehicle and an object in a turning situation and an avoidance of this collision are possible Deceleration intervention in the ego vehicle in the foreground, which leads to a standstill of the same.

Summary of the Invention

The above-mentioned object and other problems are solved by a method for spatial collision avoidance as part of a turning maneuver of an ego vehicle according to claims 1 or 2. The subclaims relate to preferred embodiments of the invention.

In particular, a method for spatial collision avoidance as part of a turning process of an ego vehicle is provided, which has the following: a) detecting, filtering and processing sensor data relating to the ego vehicle and an oncoming object; b) determining that the ego vehicle is in a turning situation with the oncoming object; c) analyzing and assessing whether a critical turning situation is present, by means of: i) analyzing the driver's attention in the ego vehicle; ii) detecting a turning maneuver of the ego vehicle; and iii) evaluating the movement of the object; wherein, if there is insufficient sensor data, steps a) and/or b) are repeated; and wherein, if there is a critical turning situation, the following steps are also carried out: d) predicting the respective movement of the ego vehicle and the object; e) performing an overlap analysis; wherein, if no collision course is detected, steps a) to e) are repeated in whole or in part; and wherein, if a collision course is detected, the method further comprises the following: f) determining and analyzing a remaining distance for braking the ego vehicle; and g) initiating an automatic emergency brake assistant if the remaining distance is greater than a necessary braking distance; or h) issuing a warning and initiating a prefill function if the remaining distance is less than the necessary braking distance.

In addition, a method for spatial collision avoidance as part of a turning process of an ego vehicle is provided, which has the following: a) detecting, filtering and processing sensor data relating to the ego vehicle and an oncoming object; b) determining that the ego vehicle is in a turning situation with the oncoming object; c) analyzing and assessing whether a critical turning situation is present, by means of: i) analyzing the driver's attention in the ego vehicle; ii) detecting a turning maneuver of the ego vehicle; and iii) evaluating the movement of the object; wherein, if there is insufficient sensor data, steps a) and/or b) are repeated; and wherein, if there is a critical turning situation, the following steps are also carried out: d) predicting the respective movement of the ego vehicle and the object; e) performing an overlap analysis; wherein, if no collision course is detected, steps a) to e) are repeated in whole or in part; and wherein, if a collision course is detected, the method further comprises the following: f) determining and analyzing a remaining distance for braking the ego vehicle; and g) initiating an automatic emergency brake assistant if the remaining distance is less than a necessary braking distance; or h) issuing a warning and initiating a prefill function if the remaining distance is greater than the necessary braking distance. According to a further exemplary embodiment, the sensor data recorded with regard to the ego vehicle and/or the object can be at least one of the following: speed and/or direction of movement of the ego vehicle, yaw rate of the ego vehicle, steering wheel angle of the ego vehicle, steering wheel angular velocity of the ego vehicle vehicle, speed and/or direction of movement of the object, longitudinal distance between the ego vehicle and the object, lateral distance between the ego vehicle and the object and an angle defined between a direction of movement of the object and a distance between a bumper center point of the ego vehicle and a point of the object that is frontally and centrally located, viewed from the ego vehicle.

According to a further exemplary embodiment, the object is an object, a vehicle with at least one wheel, or a person.

According to another embodiment, the object is a bumpered vehicle and the front-center point is a bumper center point.

According to a further exemplary embodiment, the sensor data can be processed vectorially.

According to a further exemplary embodiment, the prediction takes place in two-dimensional space.

According to a further exemplary embodiment, the prediction in two-dimensional space includes predicting a movement of the center point of the rear axle over time.

According to a further exemplary embodiment, the step of predicting has a step of precalculating travel directions and/or movement paths of the ego vehicle and the object. According to a further embodiment, the available route is in one-dimensional data.

According to a further exemplary embodiment, step i) of analyzing the driver's attention in the ego vehicle comprises analyzing a monitored brake and accelerator pedal interaction.

Furthermore, the present invention optionally provides at least one of the following steps in step ii) of detecting the turning maneuver of the ego vehicle: determining the start of the turning maneuver by monitoring the steering wheel speed and/or a lateral acceleration of the ego vehicle; determining whether the turning maneuver is smooth by monitoring the steering angle; determining a turn radius; and analyzing a monitored turn signal activity of the ego vehicle.

According to a further exemplary embodiment, step iii) of evaluating the state of the object includes assuming a constant straight-ahead movement of the object in the movement direction currently determined.

According to another embodiment, the step of determining and analyzing the distance available for braking the ego vehicle may include determining a desired remaining and collision-avoiding distance to the object.

According to another embodiment, determining the desired remaining and collision-avoiding distance can be done using a PTi term or a PT2 term.

It is to be regarded as advantageous within the meaning of the present invention that, in contrast to known methods and systems, an even later intervention in the driving behavior of an ego vehicle is possible, which reduces the number and risk of possible incorrect interventions. The risk of collisions with traffic behind is also reduced because the speed is only slightly reduced. This in turn results in a lower risk from the point of view of functional safety as well as possible cost savings through lower requirements for functional safety with the same system performance. Incorrect interventions are less abrupt for the driver, and therefore less frightening, and as a result result in less impairment of the driver's reaction behavior as the journey progresses.

Such a later possibility of intervention, which leads to a lower risk of incorrect interventions, is therefore particularly useful for low-cost systems. Due to sensor technology deficits that are often present, known low-cost systems, in comparison to high-end systems, only detect possible collision situations with a sufficient degree of delay or when the distance between the objects involved is (too) small. This is often due to a correspondingly large measurement tolerance of the installed sensors. Since it is essential to avoid triggering an Autonomous Emergency Braking or AEB assistant for no reason, i.e. a so-called “false negative” signal from an emergency brake assistant for vehicles, with known low-cost systems it had to be accepted that these if necessary, not trigger an AEB assistant or not in time because at the last possible triggering time at the given speeds there was not a sufficient probability of collision. However, if the emergency braking strategy of the present invention is taken into account, later triggering times are possible, at which even low-cost systems, appropriately retrofitted, predict a potential collision with sufficient probability.

In addition, the already mentioned advantages of the methods for spatial collision avoidance in the course of a turning maneuver of an ego vehicle of the present invention preferably build on one another at least in part or are dependent on one another in a synergetic manner. By intervening during a turning situation, it is possible to avoid a potential collision that would otherwise have resulted in a head-on collision between two vehicles or objects. In addition, the invention can be applied not only to oncoming vehicles but also to motorcycles, cyclists and pedestrians.

Brief description of the drawings

The invention and further details and advantages thereof are explained below using preferred exemplary embodiments with reference to the figures. Show it:

1A shows a plan view of a turning situation according to an exemplary embodiment of the present disclosure;

1B shows a further plan view of a turning situation according to an exemplary embodiment of the present disclosure;

FIG. 2 shows a flowchart of a method for avoiding a spatial collision avoidance as part of a turning process of an ego vehicle according to an exemplary embodiment of the present disclosure;

3 shows a minimum steering angle profile for a turning maneuver;

4 shows an exemplary state of the 2D prediction of an ego vehicle;

5 shows a further plan view of a turning situation according to an exemplary embodiment of the present disclosure.

Detailed description of the invention

In the present description, the terms top, bottom, right and left and similar information refer to the orientations or arrangements shown in the figures and only serve to describe the exemplary embodiments. These terms may indicate preferred arrangements but are not meant to be limiting. The present disclosure relates to various embodiments of methods for spatial collision avoidance during a turning operation. In this context, the following terms are used:

An intervention of an automatic emergency brake assistant in this context generally denotes a desired intervention or non-intervention of the automatic emergency brake assistant. Furthermore, a distinction is made between incorrect intervention and incorrect intervention. In the event of faulty intervention, a brake assistant intervenes too early or too late in terms of the optimization potential of the present disclosure. In the event of an incorrect intervention, a brake assistant incorrectly identifies a supposed collision situation and intervenes without any corresponding necessity.

The term “ego vehicle” is used below for the vehicle in which or for which the systems or methods described are to be used or applied. On the other hand there is a supposedly or potentially colliding object or a "target vehicle".

Furthermore, the present disclosure relates to “spatial” collision avoidance in turning situations. The term "spatial" collision avoidance is geared towards a complete braking of the ego vehicle in order to guarantee collision avoidance by spatially separating it from the target vehicle.

The term "predict" in the context of the present invention means predicting, calculating or estimating.

1A shows a plan view of a turning situation according to an exemplary embodiment of the present disclosure. In the case of the exemplary intersection situation presented here, an oncoming target vehicle in the opposite lane crosses the turning path of an ego vehicle while the ego vehicle is turning left. at Inattention on the part of the driver in the turning ego vehicle would often lead to an accident and corresponding considerable damage to property and persons. With the help of the automatic emergency brake assistant for spatial collision avoidance of the present disclosure, also referred to as junction turn assist, a braking strategy can be selected in a turning situation, which at best prevents or at least weakens an impending collision. The target vehicle shown as oncoming in FIG. 1A is an example of any potentially oncoming road user, such as a cyclist or a pedestrian.

Equally, it can also be an inanimate object without its own intention to move. In addition, the representation of the oncoming object on the road is exemplary. Rather, the object shown could be approaching on a footpath or cycle path or from an exit or an area opposite a T-junction.

FIG. 1B shows a further plan view of a turning situation according to an exemplary embodiment of the present disclosure. Analogously to FIG. 1A , an exemplary intersection situation is also shown here, in which an oncoming target vehicle in the opposite lane crosses the turning path of an ego vehicle while the ego vehicle is turning to the left. In addition, FIG. 1B shows various physical parameters that can be used for the methods and systems according to the invention for avoiding a spatial collision. In the coordinate system related to the illustrated ego vehicle, the following are shown as examples: the ego vehicle speed Vego, the yaw rate V'ego, the steering wheel angle aego as a steering angle request, which is usually communicated via a steering wheel, the steering wheel angle speed öego, the speed of the oncoming target -Vehicle or object Vobj and the longitudinal distance Ax and lateral distance Ay between the vehicles also shown as an example. In addition, there is the angle cp, which is spanned by the vectors Vobj and a distance between a bumper center point of the ego vehicle and a point of the object which is frontally and centrally located as viewed from the ego vehicle.

Optionally, this is the distance between the center points of the bumpers of two vehicles. The parameters mentioned are either measured directly with the help of sensors installed in the ego vehicle or calculated from the data obtained. They serve as a basis for the ego vehicle to recognize a potentially dangerous situation for the turning scenario described in good time and, if necessary, to initiate an emergency braking intervention. If, as an option, braking of the ego vehicle and/or oncoming object that has already been initiated is also measured or taken into account in the sense of negative acceleration, this may also result in a later triggering point in time for (additional) emergency braking, which is particularly beneficial for low-cost systems comes. Preferably, only a current speed Vobj of the oncoming object can be measured directly, for example via radar. In this case, a corresponding acceleration of the oncoming object could be derived, for example, from the time profile of the speed.

FIG. 2 shows a method (100) for spatial collision avoidance as part of a turning maneuver of an ego vehicle, as shown by way of example in FIGS. 1A and 2A, according to an exemplary embodiment of the present invention. The method (100) initially has a step (101) of acquiring, filtering and processing sensor data relating to the ego vehicle and an oncoming object. The respective states of the ego vehicle preferably have physical parameters of the ego vehicle and the object, but can also have information relating to the driver and the driving behavior of the ego vehicle, information relating to the movement behavior of the object and/or information relating to the environment. Based on the recorded, filtered and processed sensor data from step (101), the present situation is analyzed. First of all, this relates to determining that the ego vehicle is in a turning situation with the oncoming object. If such is the case, a further step (102) of analyzing and evaluating whether the present turning situation is also critical takes place. This is done by analyzing the driver's attention in the ego vehicle, by detecting a turning maneuver of the ego vehicle and by evaluating the movement of the object. If there is insufficient sensor data, further sensor data relating to the ego vehicle and the oncoming object is recorded, filtered and processed as required and/or it is determined again, also as required, that the ego vehicle is in a turning situation with the oncoming object . This is followed by a renewed analysis and evaluation as to whether the turning situation that is now present is critical.

If it has been determined that a critical turning situation is present, a step (103) of predicting the respective movement of the ego vehicle and of the object takes place. The corresponding movement prediction can optionally be designed as a prediction in two-dimensional space. The prediction in the two-dimensional space can also optionally include a prediction of a movement of the center point of the rear axle of the ego vehicle over time. Furthermore, the prediction can generally optionally include a step of precalculating travel directions and/or movement paths of the ego vehicle and the object.

According to the respectively predicted movements of the ego vehicle and the oncoming object, an overlap analysis is carried out in step (104). If it is detected here that there is no collision course, steps (101) to (103) are repeated in whole or in part as required, as shown in FIG. However, if an inherent collision course is detected in step (104), a step (105) then follows to determine and analyze a remaining distance for braking the ego vehicle. The determined remaining distance can optionally exist as an available distance in one-dimensional data.

According to the determined and analyzed remaining distance, there are now at least two variants for spatial collision avoidance as part of a turning maneuver:

Variant I

There follows either a step of initiating (106) an automatic emergency brake assistant if the remaining distance is greater than a necessary braking distance, or a step (107) of issuing a warning and initiating a prefill function if the remaining distance is less than the necessary braking distance is. In the context of the present disclosure, the named prefill function refers to the build-up of minimal pressure in the brake system for applying the brake discs in order to at least partially reduce a respective clearance of the brakes. The corresponding operating mode of the prefill function is optionally preset or can be set by the respective driver.

variant

This is followed either by a step of initiating (106) an automatic emergency brake assistant if the remaining distance is less than a necessary braking distance, or a step (107) of issuing a warning and initiating a prefill function if the remaining distance is greater than the necessary braking distance .

These variants mentioned in dealing with an emergency braking in a turning situation, depending on the system and the user, fundamentally allow the automatic emergency brake assistant according to the invention to intervene as late and/or as gently as possible. This reduces the risk for Incorrect interventions minimized. In addition, the necessary reduction in speed can be designed to be lower, so that the risk of potential subsequent collisions with traffic behind is reduced, which is considered the greatest risk for an automatic emergency braking system from the point of view of functional safety.

Basically, when considering whether a method according to Variant I or II should be chosen, different focuses are in the foreground. If, for example, there is a particular interest in the type of object colliding and if it is a vulnerable road user, i.e. a pedestrian, motorcyclist or cyclist, for example, the earliest possible intervention of the emergency brake assistant is usually preferred. However, if the focus is on considerations and considerations regarding the extent of a potential collision, it can be assumed that a front-to-front collision is generally preferable to a front-to-side collision, meaning in particular the side of an oncoming target vehicle is. This is true even if the collision energy would be higher in a front-to-front collision due to possibly delayed braking. In the latter case, use of the emergency brake assistant as late as possible is preferred. In addition, however, the desire may also be in the foreground not to allow an emergency brake assistant, which is basically present, to act automatically, in particular when a collision can no longer be avoided. This would sometimes be the case if the consequences for the occupants of the respective ego vehicle and/or the target vehicle were made even worse by using an emergency brake assistant shortly before the collision compared to an impact without using such an emergency brake assistant. A corresponding danger exists, for example, in the event of a potential side impact, specifically with the egooder of the target vehicle.

As already explained, the method (100) shown in FIG. vehicle up. This step optionally includes monitoring and analyzing driver inputs such as brake and accelerator pedal interaction.

As also already explained, the method (100) shown in FIG. 2 in step (102) of analyzing and evaluating whether the present turning situation is critical includes, among other things, detecting a turning maneuver of the ego vehicle. This detection process can optionally have at least one of the following steps: determining the start of the turning maneuver by monitoring the steering wheel speed and/or a lateral acceleration of the ego vehicle, determining whether the turning maneuver is steady, and, by monitoring the steering angle, determining a turning radius. Here, the following examination and detection of a potential turning request by the driver of the ego vehicle is optionally provided:

The steering wheel angle a ego , the steering wheel angular velocity _öego and the yaw rate V'ego are used for a corresponding examination and detection of the turning maneuver.

In order to determine the start of the turning maneuver correctly, it is first ensured by observing the steering wheel angular velocity äego that the ego vehicle was initially driving straight ahead. For this purpose, the steering wheel angular velocity öego is compared with a predetermined threshold value a _m in for a parameterizable time tmin_straight. The threshold value Omin is predetermined by the developer depending on the vehicle and the requirements for the response behavior of a system. In order to ensure that the vehicle is not moving in a dynamic limit range, the lateral acceleration a _y of the ego vehicle is also examined and compared to a further predetermined threshold value a _y _min. An exemplary dynamic limit range is present in those critical driving situations in which underlying models, such as a Single-track model, where applicable, no longer apply. The lateral acceleration a _y can be calculated with the help of the formula

can be determined from the yaw rate V'ego and the ego vehicle speed.

The potential turning maneuver is then evaluated using the yaw rate V'ego and the steering angle a _e go.

Furthermore, the current turning or curve radius r can be determined using the ego vehicle speed Vego and the yaw rate V'ego with the aid of the formula

be approximated. If this is greater than a further predetermined threshold value rmin, it can be assumed that the ego vehicle is currently cornering.

In addition, the steering wheel angle a _e go can be compared with a further predetermined threshold value amin(t), which corresponds to a required minimum angle profile a _min (t) as shown in FIG. This ensures that the ego vehicle is actually in a steady turning maneuver and that the risk of a possible incorrect intervention in an unwanted scenario is minimized.

The detection process described in step (102) can also optionally include analyzing a monitored indicator activity of the ego vehicle. Depending on the function requirement, this can be done using a parameter that can be set for this purpose. In an exemplary scenario with traffic on the right and corresponding turning maneuvers to the left, this means, for example, that an inactive left turn signal can do so, depending on the requirement can mean that no emergency braking is triggered, but only a warning is issued and a prefill function of the braking system is initiated. This is based on the following optional consideration with regard to a driver's actual intention to turn if no turn signal was activated: In particular, an incorrect intervention as well as an incorrect intervention within the meaning of the present disclosure should be avoided if an ego vehicle, for example, deliberately follows before a right-turn maneuver swerves to the left or drives around an obstacle.

As also already explained, the method (100) shown in FIG. 2 in step (102) of analyzing and evaluating whether the present turning situation is critical includes, among other things, evaluating the movement of the object. In the analysis of the object according to the invention, it is optionally checked what type of road user it is (pedestrians, drivers, cars/trucks, motorcyclists) in order to be able to react differently to different road users if necessary. Optionally, the direction of movement of the object (e.g. crossing, oncoming etc.) and/or the object speed Vobj according to the formula

Vp_min - Vobj - Vp_max (3), where the speeds V _p _min and V _p _max are predefined threshold values.

In addition, an absolute distance z between the ego vehicle and the object can optionally be calculated using the formula

are compared with the also predefined limit values z _p _min or z _p max , where Ax denotes the longitudinal distance and Ay denotes the lateral distance between the ego vehicle and the object. As already explained in connection with FIG. 2, the step (103) of predicting the respective movement of the ego vehicle and of the object takes place if a turning maneuver including a critical turning situation is present. The duration of such a predicted movement of the ego vehicle can optionally be set using a parameter tp_predict.

A model according to the invention for predicting the movement of the ego vehicle is based on the following assumptions:

- The speed Vego of the ego vehicle is constant.

- Yaw rate of V'egode's ego vehicle is constant.

- The rear axle of the ego vehicle rolls ideally on the ground so that there is no slippage. Consequently, as shown in FIG. 4 , the corresponding instantaneous center is at the level of the rear axle of the ego vehicle for each calculation time step.

In this context, FIG. 4 shows an initial state of an exemplary 2D prediction of the ego vehicle, where R denotes the turning radius of the ego vehicle, starting from the instantaneous center.

This allows the movement of the center point of the rear axle to be predicted with sufficient accuracy over time, and all other vehicle points can be calculated on the basis of this. To determine a potential collision, the vehicles and/or objects can be approximated by rectangles in a simplified manner. For the movement of the object, constant straight-ahead travel along the current direction of travel is optionally assumed.

Finally, in a collision analysis performed as an example in step (105), the distance between the rectangles is calculated. In the event that these intersect over time, it can be assumed that a collision will occur. With regard to the determination of the available braking distance, reference is now made to FIG. 5 as an example, which shows a further plan view of a turning situation according to an exemplary embodiment of the present disclosure. For this purpose, a distance b is defined which, as the desired remaining or remaining distance from the target, specifies that lateral offset from the object at which the ego vehicle should ideally come to a standstill and a collision is avoided. The determination of the desired remaining and collision-avoiding distance as part of a modeling of the ego vehicle braking system can optionally be carried out using a PTi element or a PT2 element. The latter shows a proportional transfer behavior with a 2nd order delay.

In addition, a parameterizable distance As is defined along the ego-vehicle path, which can compensate for any uncertainties caused by the sensors or any model assumptions made in advance in terms of a tolerance for prediction errors. The available braking distance then results from the calculated distance along the path of the ego vehicle _s ego until the distance b to the object is undershot according to the formula

Sego(distance object < b) - As (5).

Within the scope of the present disclosure, it should be noted that after a check as to whether a critical turning situation with an oncoming object is present, the latest possible braking intervention into the standstill is aimed for in order to avoid spatial collisions.

Optionally, both the oncoming object can be either a target vehicle, an object, or a pedestrian. Furthermore, all of the described methods 100 can optionally have a step for determining whether traffic is on the right or on the left. Furthermore, the methods according to the invention for spatial collision avoidance in the course of a turning maneuver of an ego vehicle are optionally designed in such a way that the automatic emergency brake assistant, which is also according to the invention, can be triggered coupled to an electronic stability program.

Furthermore, the present invention provides a computer program product that is stored on a computer-readable medium and has software including instructions that can be operated to execute the inventive method for spatial collision avoidance as part of a turning process of an ego vehicle.

This software can optionally be programmed to work with an Electronic Stability Program (ESP) system that selectively brakes individual wheels if a vehicle is about to skid. In this way, both oversteer and understeer of the vehicle are prevented in a spatial collision avoidance of the present invention.

In all of the exemplary embodiments described here, for the sake of simplicity, unless explicitly stated otherwise, collision situations in the plane spanned by a two-dimensional vector in each x and y direction of a Cartesian coordinate system to be regarded as known were spoken of. However, it is obvious to a person skilled in the art that most everyday collision situations likewise have a component in three-dimensional space once the geographic situation is no longer flat. A corresponding vectorial adaptability of the claimed method and system is therefore also part of the present disclosure.

Furthermore, various exemplary embodiments have been described which include an optional use of predefined threshold values. In this context, it should be noted that these threshold values depend on the respective system and the specific properties of an ego vehicle and depending on the exemplary use in a car or a truck can diverge greatly.

Furthermore, it is assumed to be known for the present invention that motor vehicles are equipped as standard with corresponding sensors (e.g. speedometer) and measuring or monitoring devices in order to measure and monitor speeds and/or accelerations on the own vehicle. If necessary, the concepts described here also provide for interfaces and adaptation options for receiving data and for data transmission to the outside (e.g. GPS, connectivity systems such as ConnectedDrive® etc.). In principle, the concepts and systems described here can be used equally well in low-cost systems and highly complex, networked vehicle systems.

The invention has been described using preferred embodiments, with the individual features of the described embodiments being able to be freely combined with one another and/or exchanged, provided they are compatible. Likewise, individual features of the described embodiments can be omitted if they are not absolutely necessary. Numerous modifications and configurations are possible and obvious to a person skilled in the art, without departing from the idea of the invention.

Claims

Method (100) for spatial collision avoidance as part of a turning process of an ego vehicle, the method (100) having the following: a) detecting, filtering and processing (101) sensor data relating to the ego vehicle and an oncoming object; b) determining (102) that the ego vehicle is in a turning situation with the oncoming object; c) analyzing and evaluating (102) whether a critical turning situation is present, by means of: i) analyzing the driver's attention in the ego vehicle; ii) detecting a turning maneuver of the ego vehicle; and iii) evaluating the movement of the object; wherein, if there is insufficient sensor data, steps a) and/or b) are repeated; and wherein, if there is a critical turning situation, the following steps are also carried out: d) predicting (103) the respective movement of the ego vehicle and the object; e) performing (104) an overlap analysis; wherein, if no collision course is detected, steps a) to e) are repeated in whole or in part; and wherein, if a collision course is detected, the method (100) further comprises the following: f) determining and analyzing (105) a remaining distance for braking the ego vehicle; and g) initiating (106) an automatic emergency brake assistant if the remaining distance is greater than a necessary braking distance; or h) issuing (107) a warning and initiating a prefill function if the remaining distance is less than the necessary braking distance.

22 Method (100) for spatial collision avoidance as part of a turning process of an ego vehicle, the method (100) having the following: a) detecting, filtering and processing (101) sensor data relating to the ego vehicle and an oncoming object; b) determining (102) that the ego vehicle is in a turning situation with the oncoming object; c) analyzing and evaluating (102) whether a critical turning situation is present, by means of: i) analyzing the driver's attention in the ego vehicle; ii) detecting a turning maneuver of the ego vehicle; and iii) evaluating the movement of the object; wherein, if there is insufficient sensor data, steps a) and/or b) are repeated; and wherein, if there is a critical turning situation, the following steps are also carried out: d) predicting (103) the respective movement of the ego vehicle and the object; e) performing (104) an overlap analysis; wherein, if no collision course is detected, steps a) to e) are repeated in whole or in part; and wherein, if a collision course is detected, the method (100) further comprises the following: f) determining and analyzing (105) a remaining distance for braking the ego vehicle; and g) initiating (106) an automatic emergency brake assistant if the remaining distance is less than a necessary braking distance; or h) issuing (107) a warning and initiating a prefill function if the remaining distance is greater than the necessary braking distance. The method (100) according to any one of the preceding claims, wherein the sensor data recorded with regard to the ego vehicle and/or the object are at least one of the following: Speed and/or direction of movement of the ego vehicle, yaw rate of the ego vehicle, steering wheel angle of the ego vehicle, steering wheel angular velocity of the ego vehicle, speed and/or direction of movement of the object, longitudinal distance between the ego vehicle and the object, lateral distance between the ego vehicle and the object and an angle spanned between a moving direction of the object and a distance between a bumper center point of the ego vehicle and a frontally central point of the object as viewed from the ego vehicle. The method (100) of any preceding claim, wherein the object is an object, a vehicle having at least one wheel, or a human. The method (100) of any preceding claim, wherein the object is a bumpered vehicle and wherein the front-center point is a bumper center point. The method (100) according to any one of the preceding claims, wherein the sensor data are processed vectorially. The method (100) according to any one of the preceding claims, wherein the predicting (103) occurs in two-dimensional space. The method (100) of claim 7, wherein predicting (103) in two-dimensional space comprises predicting movement of the center point of the rear axle over time. The method (100) according to any one of the preceding claims, wherein the predicting (103) has a step of precalculating travel directions and/or movement paths of the ego vehicle and the object. The method (100) of any preceding claim, wherein the available route is in one-dimensional data. The method (100) of any preceding claim, wherein step i) of analyzing driver attention in the ego vehicle comprises analyzing monitored brake and accelerator pedal interaction. The method (100) according to any one of the preceding claims, wherein step ii) of detecting the turning maneuver of the ego vehicle comprises at least one of the following steps:

determining the start of the turning maneuver by monitoring the steering wheel speed and/or a lateral acceleration of the ego vehicle;

determining whether the turning maneuver is smooth by monitoring the steering angle;

determining a turn radius; and analyzing a monitored turn signal activity of the ego vehicle. The method (100) according to any one of the preceding claims, wherein the step iii) of evaluating the state of the object comprises assuming constant straight-ahead movement of the object in the currently determined direction of movement. The method (100) of any preceding claim, wherein the step of determining and analyzing (105) the available distance for braking the ego vehicle comprises: determining a desired remaining and collision-avoiding distance to the object.

25 The method (100) of claim 14, wherein determining (105) the desired remaining and collision avoidance distance is performed using a PTi term or a PT2 term.

26