EP3996965A1

EP3996965A1 - Method for making a decision for a vehicle, depending on its environment

Info

Publication number: EP3996965A1
Application number: EP20750314.5A
Authority: EP
Inventors: Ahmid El Hamdani
Original assignee: PSA Automobiles SA
Current assignee: PSA Automobiles SA
Priority date: 2019-07-11
Filing date: 2020-07-06
Publication date: 2022-05-18
Also published as: WO2021005296A1; FR3098480B1; FR3098480A1

Abstract

A decision-making method is implemented in a vehicle and comprises: - a first step (10-20) in which a traffic context of the vehicle is determined, - a second step (30-40) in which environmental data are analysed with a view to detecting objects, then, for each object, a type of mobility among first and second types corresponding to an absence of ability to move and an ability to move are determined, - a third step (50) in which, for each object, a position and a velocity vector are determined depending on the environmental data, and - a fourth step (60) in which decisions are made from among doing nothing, providing information, acting on the motion of the vehicle or protecting a passenger depending on the traffic context, type of mobility, position and velocity vector that are determined for each detected object.

Description

DESCRIPTION

TITLE: DECISION-MAKING PROCESS FOR A VEHICLE, DEPENDING ON ITS ENVIRONMENT

Technical field of the invention

The invention relates to land vehicles with at least partially automated driving, and more specifically to decision making in such vehicles.

By “land vehicle with at least partially automated (or autonomous) driving” is meant here a vehicle that can be driven on a road in an automated (partial or total (without intervention of its driver)) during an automated driving phase, or of manually (and therefore with the driver's intervention on the steering wheel (or handlebars) and / or pedals (or levers)) during a manual driving phase.

State of the art

Some at least partially automated (or autonomous) driving vehicles, generally of the automobile type, have environmental data that is representative of the environment around them. These environmental data may come from sensors that they embark and / or from other vehicles which are located near them and / or from communication stations which belong to the road infrastructures on which they travel.

Currently, the environmental data available to a vehicle with at least partially automated driving is analyzed internally by one or more driving assistance (or assistance) devices (for example of the ADAS type ("Advanced Driver Assistance System). »)), Each responsible for a single function. For example and without limitation, these functions can control the trajectory or the parking automated system or securing the vehicle or at least one passenger, or regulating the speed, or even detecting a potentially dangerous situation.

Generally, these driver assistance devices operate in parallel without sharing computing capacity (or CPU ("Central

Processing Unit ")), which increases the costs of the vehicle, or even does not allow some people to dispose of it given the high sum that their accumulation entails. Indeed, currently, each driving assistance (or assistance) uses a set of sensors in certain phases of life (parking, low speed, high speed, all speeds), and therefore requires the development and design of its own algorithms. In addition, it does not allow to be exhaustive in terms of analysis of the operational scene, and therefore certain situations

remain unanswered. In addition, this parallel operation makes centralized (or global) decision-making difficult and time-consuming, in particular to control the dynamics of the vehicle during phases of automated (or autonomous) driving, and therefore the time taken to take them. actually implementing it may turn out to be too long in relation to the maximum reaction time available in certain complex situations. In addition, in the absence of decision-making

centralized (and therefore in the case of purely parallel operation), we are forced to provide very low level arbitration.

Furthermore, most current decision-making is intended to make the vehicle and / or its passengers safe, but does not (or too little) inform other road users of their consequences, so that 'they do not allow other road users to get to (or be made) safe.

Finally, animals located in the environment of a vehicle are very rarely taken into account in individual or centralized (or global) decision-making, and therefore the implementation of these can sometimes injure or even kill them.

The aim of the invention is therefore in particular to improve the situation. Presentation of the invention

For this purpose, it proposes in particular a decision-making process, intended to be implemented in a vehicle with at least partially automated driving and having environmental data representative of a surrounding environment, and comprising:

- a first step in which a traffic (or operational) context of the vehicle is determined as a function of a current speed of this vehicle and / or of a current geographical position of this vehicle,

- a second step in which we analyze the data

environment in order to detect objects in the environment of the vehicle, then a type of mobility is determined for each object detected from among the first and second types corresponding respectively to an absence of mobility capacity and a mobility capacity,

- a third step in which a position and a speed vector relative to the vehicle are determined for each detected object based on the environmental data, and

a fourth step in which at least one decision is taken chosen from among doing nothing, informing, acting on a longitudinal dynamic and / or a lateral dynamic of the vehicle, and securing at least one passenger of the vehicle, depending on the context of circulation, type of mobility, position and speed vector determined for each object detected.

Thanks to the invention, each decision is now really centralized (or global) and therefore optimally adapted to the known environment of the vehicle since it results from taking into account all the environmental data available to the vehicle. the instant considered in this vehicle.

The decision-making method according to the invention can include other characteristics which can be taken separately or in combination, and in particular: - environmental data can be obtained from sensors which are on board the vehicle and / or at least one other vehicle located near the vehicle and / or a communication station belonging to a road infrastructure on which the vehicle;

- In its first step, the traffic context can be determined from among traffic in a parking lot, urban traffic, traffic on the road outside the built-up area, and traffic on the expressway;

- in its second step, if a second type of mobility associated with a detected object is determined, a mobility subtype can be determined for the latter from among a first subtype corresponding to a natural mobility of a being living (human or animal) and a second subtype corresponding to controlled mobility of a vehicle;

- In its third step, it is possible to determine for each object detected a position relative to the vehicle from a position to the right of the vehicle, a position to the left of the vehicle, a position in front of the vehicle and a position behind the vehicle;

- in its fourth step, in the event of a decision to inform, it is possible to inform a passenger of the vehicle and / or a passenger of at least one other vehicle located near the vehicle and / or at least one other vehicle located near the vehicle and / or a communication station belonging to a road infrastructure on which the vehicle runs and / or living beings located near the vehicle;

- in its fourth step, in the event of a decision to secure, it is possible to act on at least one first member of the vehicle directly participating in securing the passenger (s) of the vehicle and / or at least one second member of the vehicle participating in the safety of the vehicle.

The invention also proposes a computer program product comprising a set of instructions which, when it is executed by processing means, is suitable for implementing a decision-making method of the type of that presented above. to decide action (s) before be carried out by a vehicle with at least partially automated driving when it is traveling in a surrounding environment.

The invention also provides a decision-making device, intended to equip a vehicle with at least partially automated driving and having environmental data representative of a

surrounding environment, and comprising at least one processor and at least one memory arranged to perform the operations consisting of:

- determining a vehicle traffic context as a function of a current speed of this vehicle and / or a current geographical position of this vehicle,

- analyzing the environmental data in order to detect objects in the environment, then determining for each object detected a type of mobility among the first and second types corresponding respectively to an absence of mobility capacity and a mobility capacity,

- determining for each object detected a position and a speed vector relative to the vehicle based on the environmental data, and

- to take at least one decision chosen from among doing nothing, informing, acting on a longitudinal dynamic and / or a lateral dynamic of the vehicle, and to secure at least one passenger of the vehicle, depending on the traffic context, type of mobility , position and speed vector determined for each object detected.

The invention also proposes a vehicle, optionally of the automobile type, with at least partially automated driving, having environmental data representative of an environment surrounding it, and comprising a decision-making device of the type presented below. before.

Brief description of the figures Other characteristics and advantages of the invention will become apparent on examination of the detailed description below, and of the appended drawings, in which:

[Fig. 1] illustrates schematically and functionally an exemplary embodiment of a decision-making device implementing at least part of a decision-making process according to the invention, and

[Fig. 2] schematically illustrates an example of an algorithm implementing a decision-making method according to the invention.

Detailed description of the invention

The object of the invention is in particular to propose a decision-making method, and an associated decision-making device DP, intended to allow decision-making in a vehicle with at least partially automated driving, as a function of the environmental data. available to the latter.

In what follows, it is considered, by way of nonlimiting example, that the at least partially automated driving vehicle is of the automobile type. This is for example a car. But the invention is not limited to this type of vehicle. It relates in fact to any type of land vehicle that can travel on roads, with at least partially automated driving, and having environmental data. Thus, it also concerns utility vehicles, motorcycles, minibuses, coaches, road machinery, and construction or agricultural machinery.

It is recalled that the term “land vehicle with at least partially automated (or autonomous) driving” is understood here to mean a vehicle that can be driven on a road in an automated (partial or total (without intervention of its driver)) during a phase. driving

automated, or manually (and therefore with intervention of the driver on the steering wheel (or handlebars) and / or pedals (or levers)) during a manual driving phase. For example, driving automated (or autonomous) of such a vehicle may consist of steering it partially or totally or of providing any type of assistance to a natural person who drives it. This therefore covers any automated (or autonomous) driving, from level 1 to level 5 in the OICA scale.

(International Organization of Automobile Manufacturers).

It should be noted that for the invention to be implemented, it is not essential that the vehicle carries at least one passenger. Indeed, a fully automated driving vehicle can travel without

carry anybody.

As mentioned above, the invention notably provides a decision-making method intended to enable decision-making in an at least partially automated driving vehicle.

This decision-making method can be implemented at least partially by the decision-making device DP which comprises for this purpose, as illustrated without limitation in FIG. 1, at least one processor PR, for example a digital signal (or DSP (“Digital Signal Processor”)), and at least one MD memory, and therefore which can be produced in the form of a combination of circuits or electrical or electronic components (or “hardware”) and software modules (or “ software ”). The MD memory is RAM in order to store instructions for the implementation by the processor PR of at least part of the decision-making process. The processor PR can comprise integrated circuits (or printed), or else several integrated circuits (or printed) connected by wired or non-wired connections. By integrated (or printed) circuit is meant any type of device capable of performing at least one electrical or electronic operation.

In the example illustrated without limitation in FIG. 1, the decision-making device DP comprises its own computer CA because it is responsible for making centralized (or global) decisions. But it is not compulsory. Indeed, the decision-making device DP could form part of a computer performing at least one function other than its own in the vehicle. As illustrated without limitation in FIG. 2, the decision-making method, according to the invention, comprises first 10-20, second 30-40, third 50 and fourth 60 steps.

The first step 10-20 of the (decision-making) process begins when the vehicle has environmental data representative of the surrounding environment. The more this environment surrounds the vehicle, the greater the scope for decision-making and the better the quality of decision-making.

It should be noted that the environmental data dde can be obtained from sensors which are on board the vehicle concerned (and which produce them) and / or from at least one other vehicle located near the vehicle concerned and / or from a station. communication belonging to

the road infrastructure on which the vehicle concerned is traveling.

Any type of environmental data known to those skilled in the art is concerned here. Thus, it could be digital images or maps or relative positions of objects or relative speeds of objects or relative speed vectors of objects, for example.

Any sensor known to those skilled in the art and providing dde environment data can be used. Thus, it can be, for example, at least one camera (operating in the visible domain or in

infrared) observing at least part of the environment outside the vehicle, or at least one ultrasonic or radar or lidar sensor observing at least part of the environment outside the vehicle.

Some of the dde environmental data (available in a vehicle) may have been transmitted by waves to the vehicle by a communication station in the road infrastructure and / or by vehicles in its vicinity.

The obtaining of the environment data dde is performed by the decision-making device DP in the sub-step 10 of the first step 10-20 in the example algorithm of FIG. 2. This first step 10-20 continues with a sub-step 20 in which one (the processor PR and the memory MD) determines (s) a traffic context (or operational context) cdc of the vehicle as a function of the speed in progress. the latter and / or the current geographical position of this vehicle.

The current geographical position of the vehicle can, for example, be determined by a navigation aid device which is in the vehicle, either permanently or temporarily.

For example, the traffic context (or operational context) cdc can be determined by the processor PR and the memory MD among traffic in a parking lot, city traffic, off-road traffic, and freeway traffic.

Traffic in a car park may, for example, be characterized by a traffic speed of less than 10 km / h. Urban traffic can, for example, be characterized by a traffic speed of between 30 km / h and 50 km / h. Road traffic outside built-up areas can, for example, be characterized by a speed of between 70 km / h and 90 km / h. Motorway traffic can, for example, be characterized by a speed of between 100 km / h and 130 km / h. All the aforementioned values are purely illustrative and may vary according to the laws in force in the countries concerned.

It will be understood that in normal driving condition the current speed of the vehicle may be sufficient to determine the context of cdc traffic. But this is not always the case, especially in the event of a slowdown (or traffic jam). In the latter case, the current geographical position of the vehicle must be used in order to determine in a database storing traffic information the type of road (or location in the case of a parking lot) to which it corresponds.

In a second step 30-40 of the method, we (the processor PR and the memory MD) begin (s) by analyzing the environmental data dde (obtained in the first step 10-20), in order to detect objects in the environment of the vehicle.

This analysis is performed in the substep 30 of the second step 30-40 in the example algorithm of Figure 2.

The detection of an object can be done, for example, by recognition of shape or of characteristic dimensions or of characteristic speeds, or by the knowledge of a relative position pro or of a relative speed or of a relative speed vector vvro in relation to the vehicle concerned.

The second step 30-40 continues in a sub-step 40 in which one (the processor PR and the memory MD) determines (s) for each object detected a type of mobility tmo among a first type corresponding to an absence of capacity of mobility and a second type corresponding to a mobility capacity.

It will be understood that the first type of tmo mobility corresponds to any inanimate and therefore immobile object, such as for example an element of a road infrastructure, a panel, a post, a stud, a wall, a fence, or a plant. The second type of mobility tmo corresponds to any animated and / or mobile object, such as for example a living being (human or animal) or a controlled object (vehicle). Note that an object can be

temporarily stationary while belonging to the second type. It is the information obtained during the detection of an object that will make it possible to determine its type of mobility tmo (first or second).

In a third step 50 of the method, we (the processor PR and the memory MD) determine (s) for each object detected a position pro and a speed vector vvro with respect to the vehicle as a function of the environment data dde. It will be understood that the result of the analyzes carried out in sub-step 30 can be used here to determine the relative position pro and relative speed vector vvro of each object detected.

In a fourth step 60 of the method, we (the processor PR and the memory MD) take (take) at least one decision, chosen from among doing nothing, informing, acting on the longitudinal dynamics and / or the dynamics. side of the vehicle, and secure at least one passenger of the vehicle, as a function of the traffic context cdc, type of mobility tmo, relative position pro and relative speed vector vvro which have just been determined for each object detected.

Thus, each decision that is taken within the vehicle is now centralized (or global) and therefore optimally adapted to

the known environment of the vehicle since it results from taking into account all the environmental data dde which is available at the time considered in this vehicle. The DP decision-making system therefore constitutes an assistance system (in particular for driving) operating globally and no longer in parallel. This results in a significant reduction in the computing capacity (or CPU) on board the vehicle and therefore a significant reduction in the costs of the vehicle.

In addition, this centralized (and global) operation makes it possible to significantly reduce the time taken to make decisions, and therefore to significantly reduce the reaction time by making the latter compatible with many complex situations. In addition, this centralized (and global) operation avoids having to make very low level arbitrations. Finally, decision-making can now affect the vehicle as well as its passengers and other road users (including animals).

It will be noted that in the sub-step 40 of the second step 30-40, in the event of determination of a second type of mobility tmo associated with a detected object, we (the processor PR and the memory MD) little (fri) t determining for this object a stmo mobility subtype from among a first subtype which corresponds to a natural mobility of a living being (human or animal) and a second subtype which corresponds to a controlled mobility of a vehicle. This option allows even better characterization of the detected potentially mobile object, and therefore even better to make each decision in the presence of such an object in the environment.

It will also be noted that in the third step 50 one (the processor PR and the memory MD) can (come) t determine for each object detected a position relative to the vehicle (or relative position pro) from a position to the right of the vehicle, a position to the left of the vehicle, a position in front of the vehicle and a position behind the vehicle. This option

(proposing to use pro relative positions chosen in four sectors surrounding the vehicle) simplifies the calculations since we do not need to know precisely the relative position of an object in a frame of reference attached to the vehicle.

It will also be noted that in the fourth step 60, in the event of a decision to inform, one (the processor PR and the memory MD) can inform a passenger of the vehicle, and / or a passenger of at least one other vehicle located near the vehicle, and / or at least one other vehicle located near the vehicle, and / or a communication station belonging to

the road infrastructure on which the vehicle is traveling, and / or living beings located near the vehicle. It depends on the situation encountered, defined by the dde environment data.

The action consisting in informing a passenger of the vehicle can be done by any means known to those skilled in the art, and in particular by means of a message displayed on at least one screen on board the vehicle (for example that of the instrument panel. control panel or from the dashboard) and / or broadcast by at least one speaker installed in the vehicle, or a haptic function (via the sense of touch of the fingers (vibration, heat, texture, transient effect, in particular) ).

The action consisting in informing a passenger of another vehicle can be done by any means known to those skilled in the art, and in particular by means of a message transmitted by means of waves to this other vehicle and intended for be displayed and / or broadcast to his attention, or an intermittent activation of the lighting and / or signaling lights (flashes of headlights, hazard lights, in particular)), or even an activation of the horn (or horn).

The action consisting in informing another vehicle can be done by any means known to those skilled in the art, and in particular by means of a message transmitted by waves to this other vehicle and intended to be used in internal by its driver assistance system (s) or its decision-making system.

The action of informing a communication station can be done by any means known to those skilled in the art, and in particular by means of a message transmitted by waves to this communication station.

The action consisting in informing a living being can be done by any means known to those skilled in the art, and in particular by means of an intermittent activation of the lighting and / or signaling lights (calls of headlights, lights distress, in particular)), or activation of the audible warning device (or horn), or even the broadcast of a specific alert signal outside the vehicle.

It will also be noted that in the fourth step 60, in the event of a locking decision, it (the processor PR and the memory MD) can act on at least one first component of the vehicle which directly participates in a setting. safety of passenger (s) of the vehicle and / or at least one second component of the vehicle which participates in securing the vehicle.

In the case of a motor vehicle, a first component is fitted to its interior which can accommodate at least one passenger to be made safe. In this case, a first member may, for example, be a seat belt or an airbag (or “airbag”) protection device or a pedal or a steering wheel (with its steering column). In this case, we (the processor PR and the memory MD) can generate a first command intended, for example, to cause an action among a pre-tensioning of a seat belt in use by the driver. passenger, or a pre-conditioning of at least one air bag protection device (for example by an action on its pyrotechnic trigger) associated with the zone where the passenger is located, or a reorientation of the steering wheel and / or a depression of the steering column, or a displacement of at least one pedal.

Also in the case of a motor vehicle, a second member can, for example, be a braking system (when it is controllable electronically), a device involved in the connection to the ground (such as for example a controlled shock absorber or an electronic power-assisted steering of which at least one parameter is dynamically adaptable), a speed limitation device, or an opening (such as a window 'a side door). In this case, we (the processor PR and the memory MD) can generate a second command intended, for example, to cause an action among a reduction in the speed of the vehicle, or a preconditioning of the braking system. when this can be controlled electronically (for example by partial pressurization of elements), or a displacement of the vehicle towards a side of the road, or a specific configuration of dynamically adaptable parameters of organs

intervening in the ground connection (for example by making the controlled shock absorbers or the electric power steering stiffer), or total simultaneous closing of certain windows and of a possible sunroof and partial front windows associated with protective devices to air bag responsible for protecting the heads of front passengers.

It will also be noted that in the fourth step 60, in the event of a decision to act on the dynamics of the vehicle, we (the PR processor and the MD memory) can act on its longitudinal speed or on its speed.

transverse, or on its longitudinal and transverse speeds.

Thanks to the invention, it is now possible to inform living beings located inside and outside a vehicle with the use of the means available on board this vehicle, to act on the dynamics of a vehicle with the 'use the means available on board this vehicle, and protect living beings located inside and outside a vehicle with

the use of the means available on board this vehicle.

It will also be noted that the invention also provides a computer program product (or computer program) comprising a set of instructions which, when it is executed by processing means of electronic circuit (or hardware) type, such as for example the processor PR, is suitable for implementing the decision-making method described above to decide what action (s) should be taken by a vehicle when it is traveling in a surrounding environment.

It will also be noted, as illustrated without limitation in FIG. 1, that the decision-making device DP can also comprise, in

in addition to its random access memory MD and its processor PR, a mass memory MM, in particular for storing environmental data dde and road or navigation or warning information received by means of waves by the vehicle, and of data

intermediaries involved in all its calculations and processing. Furthermore, this decision-making device DP can also include an input interface IE for the reception of at least the environmental data dde and the road or navigation or warning information received by the vehicle in order to use them. in calculations or processing, possibly after having shaped and / or demodulated and / or amplified their contents, in a manner known per se, by means of a digital signal processor PR '. In addition, this decision-making device DP can also include an output interface IS, in particular for delivering orders, commands and messages intended to implement each decision taken.

Note also that one or more sub-steps of the first, second, third and fourth steps of the decision-making process can be performed by different components. Thus, the decision-making method can be implemented by a plurality of digital signal processors, RAM, mass memory, input interface, output interface.

Claims

1. A decision-making method for a vehicle with at least partially automated driving and having environmental data representative of an environment surrounding it, characterized in that it comprises i) a first step (10-20) in in which a traffic context of said vehicle is determined as a function of a current speed of said vehicle and / or of a current geographical position of said vehicle, ii) a second step (30-40) in which said data of said vehicle is analyzed environment in order to detect objects in said environment, then determining for each object detected a type of mobility among first and second types corresponding respectively to an absence of mobility capacity and a mobility capacity, iii) a third step (50) in which a position and a speed vector with respect to said vehicle are determined for each detected object as a function of said environmental data, and iv) a fourth step (60) in which one takes at least one decision chosen from among doing nothing, informing, acting on a longitudinal dynamic and / or a lateral dynamic of said vehicle, and securing at least one passenger of the vehicle, according to said traffic context, type of mobility, position and speed vector determined for each object detected.

2. Method according to claim 1, characterized in that one obtains said environmental data from sensors on board said vehicle and / or at least one other vehicle located near said vehicle and / or a station. communication belonging to a road infrastructure on which said vehicle travels.

3. Method according to claim 1 or 2, characterized in that in said first step (10-20) said traffic context is determined from among traffic in a parking lot, urban traffic, off-road traffic, and traffic. on the fast track.

4. Method according to one of claims 1 to 3, characterized in that in said second step (30-40), in the event of determination of a second type of mobility associated with a detected object, a mobility subtype is determined for the latter from among a first subtype corresponding to a natural mobility of a living being and a second subtype corresponding to a controlled mobility of a vehicle.

5. Method according to one of claims 1 to 4, characterized in that in said third step (50) is determined for each object detected a position relative to said vehicle from a position to the right of said vehicle, a position to the left of said vehicle. , a position in front of said vehicle and a position behind said vehicle.

6. Method according to one of claims 1 to 5, characterized in that in said fourth step (60), in the event of a decision to inform, a passenger of said vehicle and / or a passenger of at least one other is informed. vehicle located near said vehicle and / or at least one other vehicle located near said vehicle and / or a communication station belonging to a road infrastructure on which said vehicle and / or living beings located near said vehicle circulate.

7. Method according to one of claims 1 to 6, characterized in that in said fourth step (60), in the event of a decision to secure, one acts on at least a first member of said vehicle participating directly in a setting. safety of passenger (s) of said vehicle and / or at least a second member of said vehicle participating in securing said vehicle.

8. Computer program product comprising a set of instructions which, when it is executed by processing means, is suitable for implementing the decision-making method according to one of the preceding claims to decide on action. (s) to be performed by an at least partially automated driving vehicle when traveling in a surrounding environment.

9. Decision-making device (PD) for a vehicle with at least partially automated driving and having environmental data representative of an environment surrounding it, characterized in that it comprises at least one processor (PR) and at least one memory (MD) arranged to perform the operations of i) determining a traffic context of said vehicle as a function of a current speed of said vehicle and / or of a geographical position during said vehicle, ii) analyzing said environmental data in order to detect objects in said environment, then determining for each detected object a type of mobility among first and second types corresponding respectively to an absence of mobility capacity and a mobility capacity, iii) to determine for each object detected a position and a speed vector with respect to said vehicle as a function of said environmental data, and iv) to take at least one decision chosen from among doing nothing, informing, acting on longitudinal dynamics and / or lateral dynamics of said vehicle, and secure at least one passenger of the vehicle, depending on said traffic context, type of m obility, position and speed vector determined for each object detected.

10. Vehicle driving at least partially automated and having environmental data representative of an environment surrounding it, characterized in that it further comprises a decision-making device (PD) according to claim 9.