EP4188769A1

EP4188769A1 - Motor vehicle driver assistance method

Info

Publication number: EP4188769A1
Application number: EP21749614.0A
Authority: EP
Inventors: Keilatt Andriantavison; Thomas Heitzmann
Original assignee: Valeo Schalter und Sensoren GmbH
Current assignee: Valeo Schalter und Sensoren GmbH
Priority date: 2020-08-03
Filing date: 2021-07-27
Publication date: 2023-06-07
Also published as: WO2022028943A1; DE102020120401A1

Abstract

The invention relates to a driver assistance method for a motor vehicle of interest equipped with a driver assistance system (3) and travelling in a driving area, comprising the steps of: detecting, by the system (3), at least one third-party vehicle present in the environment of the vehicle of interest based on measurements delivered by at least one exteroceptive measurement sensor (2) installed on-board the vehicle of interest; predicting, by the system (3), a trajectory associated with the detected third-party vehicle; and determining, by the system (3), a safety area for the vehicle of interest, based on the projected trajectory and the current movement parameters of the vehicle of interest. Vehicle movement models dependent on a set of parameters representative of limitations of the kinematic performances associated with the vehicles are used in the prediction and determination steps. According to the invention, the value of at least one of the parameters of said set is dynamically modified by the system (3) depending on at least one additional piece of information transmitted to the system (3) by at least one external source (1, 2, 4) installed, or not, on-board the vehicle of interest.

Description

ASSISTANCE WITH THE DRIVING OF A MOTOR VEHICLE

Technical area

The present invention relates generally to the field of motor vehicles, and more specifically to assistance in driving a motor vehicle.

Technology background

[0002] In order to increase road safety, certain so-called semi-autonomous motor vehicles are equipped with partial automation systems or advanced driver assistance systems (known by the acronym ADAS in English terminology ), in particular systems performing, in place of the driver, the lateral control and/or the longitudinal control of the vehicle, or at the very least alerting the driver of a potentially dangerous situation to enable him to react in time. There are also plans to make motor vehicles completely autonomous, i.e. without a driver.

[0003] To enable an autonomous or semi-autonomous vehicle (hereinafter referred to as a "vehicle of interest") to detect dangerous situations and to react accordingly to avoid or reduce the risk of accidents, the assistance system to driving on board this vehicle must be able not only to detect all dynamic objects (hereinafter referred to as "third-party vehicles") present in the immediate environment of the vehicle, such as other motor vehicles (cars, trucks, motorcycles ), but also to predict the future movements of these third-party vehicles.

[0004] A known method for assisting the driving of a motor vehicle of interest moving in a driving zone thus generally comprises:

- a detection step during which a driving assistance system on board the motor vehicle of interest detects at least one third-party vehicle present in the environment of the motor vehicle of interest from measurements delivered by at least one exteroceptive measurement sensor on board the vehicle of interest; and - a step of prediction, by the driving assistance system, of a trajectory associated with said at least one detected third-party vehicle.

[0005] As described for example in the document entitled “A survey on motion prediction and risk assessment for intelligent vehicles” (Lefèvre et al., Robomech Journal 2014,1:1 http://www.robometechjournal.eom/content/ 1/1/1), the known trajectory prediction methods can be based on a motion model based on physics, that is to say a model considering that the future motion of a vehicle depends only on the laws physics and assumes that the vehicle does not change speed or direction.

[0006] Once the trajectories of the third-party vehicles have been predicted, the driving assistance system is able to determine a safety zone for the vehicle of interest, from the predicted trajectories and the current movement parameters. of the vehicle of interest.

[0007] The system described in the document entitled "Implementing the RSS Model on NHTSA Pre-Crash Scenarios", available at the following link https://www.mobileye.com/responsibility-sensitive-safety/rss_on_nhtsa.pdf, determines for example a safety zone for the vehicle of interest and the behavior to adopt in the event of violation thereof, by evaluating in particular a longitudinal safety distance and a lateral safety distance which must be maintained by the vehicle of interest with respect to to third-party vehicles moving around in its environment. To do this, the system uses, both in the trajectory prediction step and in the safety zone determination step, a set of parameters representative of the limitations of the kinematic performances associated with the vehicles, namely in particular:

- the maximum longitudinal acceleration of the vehicles;

- the maximum braking deceleration of vehicles;

- the maximum lateral acceleration of the vehicles; and

- the reaction time associated with the vehicles.

[0008] A disadvantage of this system lies in the fact that the above parameters are static, that is to say that they are fixed once and for all, and this, regardless of the vehicles present on the scene and the real environmental conditions associated with the driving scenario.

[0009] For example, it is considered empirically that the longitudinal acceleration of a vehicle, whatever it is, is equal to 3m/s ² and that the maximum braking deceleration of a vehicle is equal to -9m /s ² . The reaction time is linked to the average reaction time of a driver, and also fixed empirically, for example at 1 second.

[0010] This may result in cases in which the system risks over-evaluating the size of the safety zone, which may lead to an accident. It may also result in cases where the system may underestimate the size of the safety zone, and then trigger unnecessary lateral/longitudinal checks and/or an alert to the driver of the vehicle of interest.

Summary of the invention

[0011] The purpose of the present invention is to overcome the limitations of the prior art by proposing to adapt the value of at least one of the preceding parameters to the actual driving situations encountered.

Consequently, the subject of the present invention is a method for assisting the driving of a motor vehicle of interest moving in a driving zone comprising:

- a detection step during which a driving assistance system on board said motor vehicle of interest detects at least one third-party vehicle present in the environment of the motor vehicle of interest from measurements delivered by at least one exteroceptive measurement sensor on board the vehicle of interest;

- a step of prediction, by the driving assistance system, of a trajectory associated with said at least one detected third-party vehicle;

- a step of determining, by the driving assistance system, a safety zone for the vehicle of interest, from the predicted trajectory and the current movement parameters of said vehicle of interest, said step of predicting and said step of determining using vehicle motion models dependent on a set of parameters representative of limitations of the kinematic performances associated with the vehicles, characterized in that the value of at least one of the parameters of the said set used in the step of prediction and the step of determination is dynamically modified by the driver assistance system as a function of at least one additional piece of information transmitted to the driving assistance system by at least one external source whether or not on board the vehicle of interest.

Said set of parameters may comprise a maximum longitudinal acceleration of said vehicles, and/or a maximum braking deceleration of said vehicles, and/or a maximum lateral acceleration of said vehicles, and/or a reaction time associated with said vehicles.

[0014] Said at least one external source can be an exteroceptive sensor on board the vehicle of interest, for example a rain and/or light sensor.

[0015] Said at least one external source can also be a proprioceptive sensor on board the vehicle of interest.

[0016] Said at least one external source can also be a V2X communication module equipping an infrastructure located in the driving zone.

[0017] In one possible embodiment, the exteroceptive measurement sensor is a camera, or a radar or a lidar.

Brief description of the drawings

The invention will be better understood in view of the following description given with reference to the appended figures, in which:

- Figure 1 schematically illustrates, in top view, an example of a road scene serving to illustrate certain principles of the invention;

FIG. 2 schematically represents components of an example of an on-board system on a vehicle of interest capable of implementing a driving assistance method in accordance with the invention;

- Figure 3 illustrates an example of variation of a maximum braking deceleration as a function of the level of rain; - Figure 4 illustrates an embodiment of the invention in which the external source is a V2X type communicating terminal installed in a driving zone;

- Figure 5 shows steps that can be implemented according to one embodiment of a driving assistance method according to the invention.

Description of embodiment(s)

To fix ideas, the invention will now be described in the context of the non-limiting example of the road scene shown schematically in top view in Figure 1.

[0020] In this figure 1, a vehicle of interest Vi, having an advanced driving assistance system in accordance with the invention, moves in a rolling zone. Two other vehicles Vi and V2 also evolve in the environment of the vehicle of interest Vi. For simplicity, it is assumed, without limitation, that the driving zone corresponds to a portion of the motorway, and that all vehicles travel in the same direction (from left to right in figure 1), according to the French highway code ( overtaking on the left and speed limit of 130 km/h). In the nonlimiting example, the third-party vehicles Vi, V2 are motor vehicles. The nature of the third-party vehicles present in the environment of the vehicle of interest, however, has no impact on the principles of the present invention. In other words, a third-party vehicle can be any motorized vehicle (motor vehicle, motorcycle, truck, etc.), a semi-autonomous vehicle or an autonomous vehicle.

As shown schematically in Figure 2, the vehicle of interest Vi is conventionally equipped with several proprioceptive sensors (generally represented under the reference 1), such as a speed sensor, a steering wheel angle sensor and a GPS type navigation system, allowing it to have information on its current state (speed, acceleration, heading followed with respect to a reference (X,Y) linked to the vehicle of interest, current position with respect to an HD map board containing context-related information, such as speed limit regulations, type of road, etc.). [0022] The vehicle of interest Vi also comprises one or more exteroceptive sensors (generally represented under the reference 2), comprising at least one measurement sensor (for example image sensor, a Radar, a Lidar) allowing it to detect the third-party vehicles present in its environment, and optionally, the information relating to the geometry of the road scene (in particular the marking lines, the road signs, etc.). In the non-limiting embodiment shown, other exteroceptive sensors not used in the detection of third-party vehicles, such as a rain and/or light sensor, equip the vehicle of interest Vi.

[0023] A driver assistance system 3 is also on board the vehicle of interest Vi. As represented schematically in FIG. 2, this driving assistance system 3 conventionally comprises:

- a detection module 30 configured to detect the presence of third-party vehicles moving in the environment of the vehicle of interest Vi from measurements delivered by the exteroceptive measurement sensor(s) 2;

- a decision module 31 capable of determining a so-called safety zone for the vehicle of interest Vi; and

- a control module 34 capable here of generating commands allowing lateral and/or longitudinal control of the vehicle of interest according to the safety zone delivered at the output of the decision module 31 (as we have seen previously, the control module could be replaced or supplemented by a module for generating a visual and/or audible alert intended for the driver of the vehicle of interest).

The decision module 31 comprises a first so-called trajectory prediction sub-module 32, configured to predict the trajectory of each of the third-party vehicles detected in the environment of the vehicle of interest Vi by the detection module 30, such as than the vehicles Vi and V2 in the example of FIG. 1. This first sub-module 32 is more precisely capable of anticipating the trajectory of a detected third-party vehicle, according to its current speed and heading in the reference frame ( X,Y) associated with the vehicle of interest Vi, by applying a linear extrapolation on a motion model based on physics, and deducing therefrom a danger zone, such as the zones Ai and A2 represented on FIG. 1, symbolizing areas within which the vehicle of interest Vi must not penetrate.

The decision module 31 further comprises a second sub-module 33 called security evaluation, receiving on the one hand, the output of the first sub-module 32 (in this case the danger zones Ai and A2 associated with the third-party vehicles detected), and on the other hand, the information related to the current dynamics of the vehicle of interest Vi (in particular its current speed, acceleration and orientation in the reference frame (X, Y)) and delivered by the proprioceptive sensors 1 . From this information received, the sub-module 33 is able to evaluate the so-called security zone (such as the zone Ai in FIG. 1) for the vehicle of interest.

As indicated in the introduction, the calculations performed both by the first sub-module 32 for prediction and by the second sub-module 33 for security evaluation involve movement models based on physics, which models use a set of parameters representative of kinematic performance limitations associated with vehicles, including in particular:

- the maximum longitudinal acceleration a^ ¹ . cars ;

- the maximum braking deceleration p _max of the vehicles;

- the maximum lateral acceleration a^ ^l , _x of the vehicles; and

- the reaction time p associated with the vehicles.

[0027] For example, in the case (not shown) where the vehicle of interest Vi is preceded by a third vehicle traveling in the same direction and detected by the system 3, a minimum longitudinal safety distance d ¹ ™ separating the vehicle of interest of the third-party vehicle, and defining a longitudinal dimension of the safety zone for the vehicle of interest, can be calculated by applying the following relationship: in which : V/ and v _t are the current speeds of the vehicle of interest and of the third-party vehicle detected;

- Pmax is the maximum braking deceleration of the vehicles;

- Pmin is the minimum deceleration that the vehicle of interest must perform to avoid a collision.

Contrary to the known solutions which consist in using the same set of parameters { a^ ¹ ; _pmax ; a^ ^l _ax \ p } whose values are fixed once and for all and empirically, the present invention provides for the use of at least one external source to transmit to the driving assistance system 3 at least one additional piece of information used by the system 3 to dynamically modify the value of at least one of the parameters of the set.

In one possible embodiment, the external source is also embedded in the vehicle of interest Vi. It may for example be one of the exteroceptive sensors 2 of the vehicle of interest.

[0030] By way of non-limiting example, if one of the exteroceptive sensors 2 on board the vehicle of interest Vi is a sensor capable of detecting the presence of rain, it is possible to select the value assigned to the maximum deceleration braking of the assembly from two possible values stored in a database 35 of the system 3, depending on whether or not this sensor detects the presence of rain. We can for example decide that:

- in dry weather, the value to be assigned to the maximum braking deceleration

/ Pm Il aUx./ vehicles is - 9 m/s ² ; ^J

- but that if the rain sensor detects the presence of rain, the value to be assigned to the maximum braking deceleration p _max of the vehicles will be equal to - 4 m/s ² .

[0031] As a variant, if the sensor is able to measure rain levels, provision can be made to dynamically adapt the acceleration and deceleration values to braking as a function of the levels measured. FIG. 3 gives an example of variation of the values assigned to the maximum braking deceleration p _max of the vehicles as a function of the level of rain detected, these values being recorded beforehand in the database 35. According to another example, one of the exteroceptive sensors 2 of the vehicle of interest is a light sensor. However, it is understood that the reaction time p of any driver is generally higher when it is dark or at night than in broad daylight. In this case, provision can advantageously be made to modify the value to be assigned to the reaction time of the vehicles, in particular third-party vehicles, by selecting this value from among several values pre-stored in the database 35. It is possible, for example, to decide that:

- during the day or in good light conditions, the value to be assigned to the reaction time p at least of third-party vehicles is 1 second;

- but that if the sensor detects low light, or even night driving conditions, the value to be assigned to the reaction time p at least of the third-party vehicles will be higher, for example equal to 2 seconds.

Alternatively, the external source embedded in the vehicle of interest Vi can be one of the proprioceptive sensors 1 of the vehicle of interest. By way of example, the GPS-type navigation system of the vehicle of interest can be used to derive the exact time, and to deduce therefrom, like the luminosity sensor indicated previously, a value to be assigned to the time of reaction p, at least for third-party vehicles, as a function of the time supplied by this navigation system. Thanks to the GPS system, the vehicle of interest can also position itself very precisely on an on-board HD map, and derive at least one additional piece of information characteristic of the environment (for example the presence of a bridge, a tunnel, the type of road surface...). Depending on correspondence tables pre-stored in the database 35, the system 3 can in this case dynamically select a more suitable value for at least one of the parameters of the set. Typically, the type of road surface can have an advantageous effect on the values to be assigned to acceleration and deceleration under braking.

In another non-limiting example, the additional information provided to the driver assistance system 3 corresponds to a lane change maneuver performed or about to be performed by the vehicle of interest Vi. This information can be obtained by an exteroceptive sensor 2 of the vehicle of interest (for example an on-board camera detecting the crossing of a line of ground marking by the vehicle of interest) and/or by a proprioceptive sensor 1 (steering wheel angle sensor or detection of the on state of a turn signal of the vehicle of interest). In this case, it is possible to choose to assign to the maximum braking deceleration for the third-party vehicles a different value depending on whether the vehicle of interest is changing lanes or not. The precise value to be assigned in the event of a lane change maneuver can result either from a value obtained in a prior learning phase obtained online or offline (simulation of the behavior of third-party vehicles in terms of braking deceleration in different situations where the vehicle of interest is making a lane change).

[0035] In all the previous examples, the external source providing the additional information which will make it possible to dynamically adapt the value to be assigned to at least one of the parameters of the set of parameters {; Pmax>' ^a mix>'P} is ^a source on board the vehicle of interest Vi.

Of course, several additional pieces of information coming from several on-board sensors (proprioceptive and/or exteroceptive) can be combined to change the value of one or more of said parameters, without departing from the scope of the present invention.

In another embodiment, it is assumed that, in accordance with the non-limiting example shown in Figure 2, the vehicle of interest Vi is also a vehicle able to communicate, that is to say to transmit and receive information, with other vehicles and dedicated terminals equipping the road infrastructure, such as terminal 4 in FIG. 4. To do this, the system 3 comprises a V2X communication module 36 (meaning Vehicle to everything in Anglo-Saxon terminology). In this case, at least one of the additional pieces of information such as that described above (detection of rain level, time of day, detection of luminosity, etc.) can be acquired directly by the infrastructure 4 by equipping the latter suitable measurement sensors. In addition, the infrastructure 4 knows the environmental specificities linked to the zone in which this infrastructure is located. Thus, in the example of Figure 4, the terminal 4 is located at a crossing to which are assigned different driving rules (for example the presence of stop signs, priority rules), as well as different information that can have consequences on the assessment of the danger (type of road surface, presence of a building at the crossing affecting visibility, etc.).

From at least one of these additional pieces of information, the terminal 4 is able to dynamically select the value of at least one of the parameters of the set of parameters {a^ ¹ ; (3 _max ; a^ ^l _ax \ p }.

By, in the situation shown in Figure 4, the vehicle of interest Vi may not see the presence of the vehicle V3 arriving on the right lane early enough, due to the presence of the building located at the intersection. In this case, terminal 4 can choose to assign to the reaction time p of the vehicles a greater value (for example equal to 2 seconds) compared to the empirical value of one second generally used.

The values thus assigned to the set of parameters { ^a max >Pmax>' ^a m ^l ax> P } are then transmitted by a V2X transmission module (not shown) fitted to terminal 4 and can thus be received by the reception module 36 of the vehicle of interest Vi when the latter approaches the intersection. The system 3 for aiding the driving of the vehicle of interest can then advantageously decide to temporarily replace the values of the parameters available to it in the database 35 by the values which it receives from the terminal 4.

The steps of a method 100 for assisting the driving of the motor vehicle of interest Vi are summarized in FIG. 5:

- the step referenced 110 corresponds to a detection step during which the driving assistance system 3 of the motor vehicle of interest Vi detects at least one third-party vehicle present in its environment from measurements delivered by at least one exteroceptive measurement sensor 2;

- the driving assistance system 3 can then predict (step 120), the trajectory associated with said at least one detected third-party vehicle, and determine (step 130) a safety zone for the vehicle of interest Vi, from the predicted trajectory and current motion parameters of said vehicle of interest Vi, using vehicle motion models depending on a set of parameters representative of limitations of the kinematic performances associated with the vehicles;

- the value of at least one of the parameters of said set used in the prediction step 120 and the determination step 130 is dynamically modified (step 140) by the driving assistance system 3 as a function of at least additional information transmitted to the driving assistance system 3 by at least one external source whether onboard or not on the vehicle of interest Vi. The source external to the system 3 can be a proprioceptive sensor 1 or an exteroceptive sensor 2 of the vehicle of interest Vi, or a communication terminal 4 of the V2X type. Several of these sources can be combined to modify the value of several parameters of the set.

At the end of step 130, the system is able, if necessary, to alert the driver of a dangerous situation and/or to generate automatic longitudinal and/or lateral control commands for the vehicle of interest (steps not shown).

Claims

Claims Method (100) for assisting the driving of a motor vehicle of interest (Vi) moving in a driving zone comprising:

- a detection step (110) during which a driving assistance system (3) on board said motor vehicle of interest (Vi) detects at least one third-party vehicle (Vi, V2) present in the environment of the motor vehicle of interest (Vi) from measurements delivered by at least one exteroceptive measurement sensor (2) on board the vehicle of interest (Vi);

- a step (120) of prediction, by the driving assistance system (3), of a trajectory associated with said at least one detected third-party vehicle

(Vi, V2);

- a step (130) of determining, by the driving assistance system (3), a safety zone for the vehicle of interest (Vi), from the predicted trajectory and the current movement parameters of said vehicle of interest (Vi), said step (120) of predicting and said step (130) of determining using models of movement of vehicles depending on a set of parameters representative of limitations of the kinematic performances associated with the vehicles, characterized in that the value of at least one of the parameters of said set used in the prediction step and the determination step is dynamically modified (140) by the driving assistance system (3) as a function of at least additional information transmitted to the system

(3) driving assistance by at least one external source (1, 2, 4) whether or not on board the vehicle of interest (Vi). Method according to claim 1, characterized in that said set of parameters comprises a maximum longitudinal acceleration of said vehicles. Method according to any one of claims 1 or 2, characterized in that said set of parameters comprises a maximum braking deceleration of said vehicles.

4. Method according to any one of claims 1 to 3, characterized in that said set of parameters comprises a maximum lateral acceleration of said vehicles.

5. Method according to any one of claims 1 to 4, characterized in that said set of parameters comprises a reaction time associated with said vehicles.

6. Method according to any one of the preceding claims, characterized in that said at least one external source is an exteroceptive sensor (2) on board the vehicle of interest (Vi).

7. Method according to claim 6, characterized in that said exteroceptive sensor is a rain and/or light sensor.

8. Method according to any one of claims 1 to 5, characterized in that said at least one external source is a proprioceptive sensor (1) on board the vehicle of interest (Vi).

9. Method according to any one of claims 1 to 5, characterized in that said at least one external source is a V2X communication module equipping an infrastructure (4) located in the rolling area.

10. Method according to any one of the preceding claims, characterized in that the exteroceptive measurement sensor is a camera, or a radar or a Lidar.